Benumof - Thoracic Anesthesia

Anesthesia
For
Thoracic
Surgery
Second Edition
Jonathan L. Benumof, M.D.
Professor of Anesthesiology
University of California, San Diego, Medical Center
Department of Anesthesiology
San Diego, California
W.B. SAUNDERS COMPANY
A Division of Harcourt Brace & Company
Philadelphia London Toronto Montreal Sydney Tokyo
W.B. SAUNDERS COMPANY
A Division of Harcourt Brace Company
The Curtis Center
Independence Square West
Philadelphia, Pennsylvania 19106
Library of Congress Cataloging-in-Publication Data
Benumof, Jonathan
Anesthesia for thoracic surgery / Jonathan L. Benumof.2nd ed.
p. cm.
Includes bibliographical references and index.
ISBN 0-7216-4467-8
1. ChestSurgery. 2. Anesthesia. I. Title.
[DNLM: 1. Anesthesia. 2. Thoracic Surgery.
1994]
RD536.B46 1995
617. 9' 6754dc20
DNLM/DLC
WF 980 B478a
93-^2770
ANESTHESIA FOR THORACIC SURGERY, SECOND EDITION ISBN 0-7216-4467-8
Copyright 1995, 1987 by W.B. Saunders Company
All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any
means, electronic or mechanical, including photocopy, recording, or any information storage and
retrieval system, without permission in writing from the publisher.
Printed in the United States of America.
Last digit i s the print number: 9 8 7 6 5 4 3 2 1
Contents
CHAPTER 1
History of Anesthesia for Thoracic Surgery /
Introduction, 2
Where Are We Today? Brief Summary of Modern Anesthetic Practice for Thoracic Surgery, 2
How Did We Get There? Historical Evolution of Modern Anesthetic Practice for Thoracic Surgery, 4
The Future, 12
Summary, 13
CHAPTER 3
General Respiratory Physiology and Respiratory Function During Anesthesia 43
Respiratory Physiology, 44
Respiratory Function During Anesthesia, 94
CHAPTER 4
Special Respiratory Physiology of the Lateral Decubitus Position, the Open
Chest, and One-Lung Ventilation 123
Introduction, 124
Physiology of Spontaneous Ventilation With an Open Chest, 124
Physiology of the Lateral Decubitus Position and the Open Chest During Controlled Two-Lung
Ventilation: Distribution of Perfusion (Q) and Ventilation (V), 125
Physiology of One-Lung Ventilation, 131
CHAPTER 5
Preoperative Cardiopulmonary Evaluation 152
Introduction, 153
Tumors and Other Masses of the Lung and Bronchi, 153
Mediastinal Masses, 198
Pleural Disease/Effusion, 201
Pericardial Disease/Effusion, 203
CHAPTER 6
Preoperative Respiratory Preparation 211
Introduction, 212
Correlation of Respiratory Complications with Degree of Pre-Existing Lung Disease, 213
Correlation of Respiratory Complications With Site of Operation, 213
Proof That Preoperative Pulmonary Preparation Decreases Incidence of Postoperative Respiratory
Complications, 213
Preoperative Respiratory Preparation Maneuvers, 214
Mechanism of Preoperative Respiratory Preparation Benefit, 227
Premedication, 227
CHAPTER 7
Monitoring 232
Introduction, 233
Tier I: Essential Monitoring System, 234
Tier II: Special Intermittent and/or Continuous Monitoring, 252
Tier III: Advanced Monitoring Techniques, 256
xi
CHAPTER 2
Thoracic Anatomy 15
Introduction, 16
The Thoracic Wall, 16
The Airway and Lung, 21
The Mediastinum, 38
CHAPTER 8
Choice of Anesthetic Drugs and Techniques 300
Introduction, 301
Most Common and Important Cardiopulmonary Considerations for Patients Undergoing Thoracic
Surgery, 301
Choice of Anesthesia and Arterial Oxygenation During One-Lung Ventilation, 308
Recommended Anesthesia Induction and Maintenance Drugs and Techniques, 320
CHAPTER 9
Separation of the Two Lungs (Double-Lumen Tube and Bronchial
Blocker Intubation) 330
Introduction, 331
Indications for Separation of the Two Lungs, 331
Double-Lumen Tube Intubation, 334
Bronchial Blockers (With Single-Lumen Endotracheal Tubes), 370
Endobronchial Intubation With Single-Lumen Tubes, 383
CHAPTER 10
Routine Surgical Considerations that have Anesthetic Implications 390
Introduction, 391
Positioning the Patient, 391
Thoracic Incisions, 394
Common Major Elective Thoracic Operations, 402
CHAPTER 1 1
Conventional and Differential Lung Management of One-Lung Ventilation 406
Introduction, 407
One-Lung Ventilation Situation, 407
Conventional Management of One-Lung Ventilation, 408
Differential Lung Management of One-Lung Ventilation, 413
Recommended Combined Conventional and Differential Lung Management of One-Lung Ventilation,
426
CHAPTER 12
High-Frequency and High-Flow Apneic Ventilation During Thoracic Surgery ... 432
Introduction, 433
High-Frequency Ventilation, 433
Low- and High-Flow Apneic Ventilation, 445
CHAPTER 13
Anesthetic Considerations (Other Than Management of Ventilation) During
and at the End of Thoracic Surgery 453
Introduction, 454
Management of Bronchospasm, 454
Management of Blood Loss, 459
Treatment of Nonblood Loss, Deleterious Hemodynamic Changes, 472
Re-Expansion of Collapsed Lung During and at the End of Thoracic Surgery, 479
Transport of Patient, 480
CHAPTER 14
Anesthesia for Special Elective Diagnostic Procedures 491
Introduction, 492
Bronchoscopy, 492
Mediastinoscopy, 504
Thoracoscopy, 508
CHAPTER 15
Anesthesia for Special Elective Therapeutic Procedures 513
Introduction, 514
Laser Resection of Major Airway Obstructing Tumors, 514
Tracheal Resection, 530
Broncholithiasis, 541
Giant Bullous Emphysema and Air Cysts, 542
Pulmonary Resection in Patients After Pneumonectomy, 548
Unilateral Bronchopulmonary Lavage, 548
Pulmonary Arteriovenous Malformations, 554
Lung Transplantation, 556
Tumors at the Confluence of the Superior, Anterior, and Middle Mediastina, 567
Repair of Thoracic Aortic Aneurysms, 575
Contents
Thymectomy for Myasthenia Gravis, 575
One-Lung Anesthesia in Morbidly Obese Patients, 579
Thoracic Outlet Syndromes, 580
CHAPTER 16
Anesthesia for Esophageal Surgery 591
Preoperative Considerations, 592
Intraoperative Anesthetic Considerations, 604
Postoperative Considerations, 609
CHAPTER 17
Anesthesia for Emergency Thoracic Surgery 612
Introduction, 613
Massive Hemoptysis, 613
Thoracic Aortic Aneurysms and Dissections/Disruptions, 619
Bronchopleural Fistula, 626
Lung Abscesses and Empyema, 631
Chest Trauma, 633
Transvenous Pulmonary Embolectomy, 645
Emergency Room Thoracotomy in the Management of Trauma, 646
Removal of Tracheobronchial Tree Foreign Bodies, 648
CHAPTER 18
Anesthesia for Pediatric Thoracic Surgery 657
Introduction, 659
Special Problems Related to Premature and Newborn Infants, 659
Congenital Diaphragmatic Hernia, 669
Esophageal Atresia and Tracheoesophageal Fistula, 674
Ligation of Patent Ductus Arteriosus in Premature Infants, 681
Vascular Rings, 682
Congenital Parenchymal Lesions: Lobar Emphysema, Cysts, Sequestrations, and Cystic Adenomatoid
Malformations, 683
Thoracic Surgical Procedures Requiring One-Lung Ventilation, 686
Diagnostic Bronchoscopy, 688
Bronchography, 690
Asphyxiating Thoracic Dystrophy (Jeune's Syndrome), 691
CHAPTER 19
Early Serious Complications Specifically Related to Thoracic Surgery 696
Introduction, 697
Herniation of the Heart, 697
Pulmonary Torsion, 699
Major Hemorrhage, 699
Bronchial Disruption, 700
Respiratory Failure, 703
Unilateral Re-Expansion (Intraoperative)Pulmonary Edema, 707
Right-Sided Heart Failure, 708
Myocardial Ischemia/Infarction, 710
Arrhythmias, 710
Right-to-Left Shunting Across a Patent Foramen Ovale, 712
Massive Postpneumonectomy Malignant Pleural Effusion and Chylothorax, 713
Systemic Tumor Embolism, 714
Neural Injuries, 714
Complications of Intrathoracic Intercostal Nerve Blocks, 715
CHAPTER 20
Mechanical Ventilation and Weaning 720
Introduction, 721
Initial Ventilator Settings: Intermittent Mandatory Ventilation, 721
Goal #1: F,0
2
<0.5 and P
a
0
2
Acceptable, 724
Goal #2: F,0
2
<0.5, PEEP <10 cm H
2
0, and P
a
0
2
Acceptable, 733
Goal #3: F,0
2
<0.5, PEEP <10 cm H
2
0, IMV <1 Breath/Min, P
a
0
2
Acceptable, 737
Summary of Mechanical Ventilation and Weaning Procedure, 740
Other New Mechanical Ventilation and Weaning Modes, 741
Extubation of the Trachea, 745
Complications of Mechanical Respiratory Support, 748
CHAPTER 21
Management of Postoperative Pain 756
Introduction, 757
Anatomy and Physiology of Post-thoracotomy Pain, 757
Systemic Narcotic Administration, 758
Intercostal and Paravertebral Nerve Block, 760
XIV Contents
Interpleural Analgesia, 761
Epidural Local Anesthetic Administration, 763
Cryoanalgesia, 763
Transcutaneous Electric Nerve Stimulation, 765
Epidural Opioids, 765
INDEX 775
CHAPTER 1
History of Anesthesia for
Thoracic Surgery
I. Introduction
II. Where Are We Today? Brief
Summary of Modern Anesthetic
Practice for Thoracic Surgery
III. How Did We Get There? Historical
Evolution of Modern Anesthetic
Practice for Thoracic Surgery
A. Pre-1910: Basic Medical
Accomplishments Advancing All
Types of Surgery
B. 1900-1920: Extrathoracic Chest
Wall Procedures Were Performed to
Treat Infective Lung Disease
C. 1900-1920: Ventilatory Techniques
for Chest Wall Procedures Did Not
Include Endotracheal Intubation
D. 1920s: Beginning of Significant Use
of Endotracheal Intubation
E. 1940-1950: Decline in Chest Wall
Procedures to Treat Infective Lung
Disease
F. 1938-1950: Emergence of
Intrathoracic Lung and Esophageal
Resection for Malignancy
G. 1950s and 1960s: Development of
Double-Lumen Endotracheal Tubes
H. 1956: Introduction of Halogenated
Inhalation Anesthetic Drugs
I. 1970s: Increased Invasive and
Noninvasive Monitoring
J. 1970s: Introduction of PEEP and
CPAP Into Clinical Practice
K. 1975: Introduction of Fiberoptic
Bronchoscopy
L. 1980s: Use of Nondependent Lung
CPAP, Dependent Lung PEEP,
Differential Lung Ventilation, HFV
IV. The Future
V. Summary
History of Anesthesia for Thoracic Surgery
I. INTRODUCTION
Modern anesthesia practice for thoracic surgery
has a deep physiologic and pharmacologic basis,
is often technically demanding and complex, and
requires extensive monitoring. The know-how to
administer a modern anesthetic agent for thoracic
surgery was acquired mainly during the twentieth
century in irregular periods of advancement. A
good deal of this irregularity was caused by the
two world wars, especially World War II, which
created large populations of patients with chest
injuries. The urgent need to care for these patients
adequately greatly stimulated the growth of both
surgery and anesthesia, and, in particular, it stim-
ulated the development of the thoracic anesthesia
subspecialty (see section III).
The role and importance of each advancement
can be best understood if there is an appreciation
of what constitutes modern anesthesia practice for
thoracic surgery. Consequently, this chapter will
first briefly describe where we are today (by sum-
marizing a typical modern anesthetic for a typical
thoracic surgical procedure), and then the main
body of the chapter will trace how we got there
(by describing the historical evolution toward that
modern anesthetic practice).
II. WHERE ARE WE TODAY? BRIEF
SUMMARY OF MODERN
ANESTHETIC PRACTICE FOR
THORACIC SURGERY
Today, a patient with a disease that can be
treated by thoracic surgery (i.e., surgery on all
structures within the thorax but excluding cardiac
surgery or surgery requiring cardiopulmonary by-
pass) will usually have an accurate preoperative
anatomic diagnosis (see chapter 5), and all major
organ systems will have been prepared for anes-
thesia and surgery (see chapter 6). Consequently,
it is now unusual to have an alteration in diagnosis
or in treatment plan following the induction of
anesthesia or the incision for elective thoracic sur-
gery. Figure 1-1 summarizes the very common
and/or very important components of modern in-
traoperative anesthesia practice for thoracic sur-
gery.
1
2
Most patients are primarily anesthetized with a
combination of intravenous and halogenated drugs
(see chapter 8). The induction of anesthesia may
be accomplished rapidly and with hemodynamic
stability by the administration of short-acting intra-
venous anesthetic drugs (barbiturates, narcotics,
sedatives) and anesthetic adjuvants (muscle relax-
ants, lidocaine, vasoactive drugs). Following re-
laxation with a nondepolarizing drug and the mon-
itoring of relaxation with a neuromuscular
blockade monitor, a double-lumen endotracheal
tube or bronchial blocker system is inserted (see
chapter 9). Whenever necessary throughout the
procedure, the position of the double-lumen tube
or bronchial blocker may be very precisely and
accurately positioned with the aid of a pediatric
fiberoptic bronchoscope. The double-lumen tube
will be connected to a reliable volume-limited me-
chanical ventilator. Attached to the double-lumen
tube-connector apparatus are monitors that are
used to measure end-tidal and inspired oxygen,
carbon dioxide, nitrogen, and halogenated drug
concentrations. The end-tidal carbon dioxide con-
centration is extremely important because it allows
continuous feedback regarding the adequacy of
ventilation, including tracheal tube misplacement,
disconnection, and kinking/obstruction. The air-
way pressure and tidal volume to both lungs and
to each lung separately will be measured, allowing
the calculation of whole and independent lung
compliance (see chapter 7).
The double-lumen tube (and bronchial blocker
[to a lesser extent]) allows for differential and in-
dependent management of the operative and non-
operative lungs (see chapters 11 and 12). The op-
erative lung will most often simply be made
atelectatic (one-lung ventilation); however, if oxy-
genation is a problem during atelectasis of the
operative lung, the operative lung may be insuf-
flated with oxygen, statically distended with small
amounts of constant positive airway pressure
(CPAP), made to oscillate slightly with high-fre-
quency ventilation (HFV), or very occasionally ex-
panded with an intermittent positive-pressure
breath (IPPB). Each one of these operative lung
management modalities greatly facilitates the per-
formance of thoracic surgery in the opened hemi-
thorax, because the nonmoving operative lung al-
lows the surgeon to have a quiet operative field. In
addition, the operative lung may be optimally
managed by avoiding known causes of inhibition
of operative lung hypoxic pulmonary vasoconstric-
tion (HPV) (which diverts blood flow away from
Figure 1-1 Modern anesthesia practice for thoracic surgery. (IPPB = intermittent positive-pressure breath; PEEP = positive
end-expiratory pressure; HFV = high-frequency ventilation; SAP = systemic arterial pulse pressure; SVR = systemic vascular
resistance; ABG = arterial blood gases; CVP = central venous pressure; PAP = pulmonary arterial pressure; PAWP = pulmonary
artery wedge pressure; CO = cardiac output; PVR; pulmonary vascular resistance; VBG = venous blood gases; Q
s
/Q, = right to
left transpulmonary shunt; V0
2
= oxygen consumption; EKG = electrocardiogram; I & = intake and output; CPAP = constant
positive airway pressure; HPV = hypoxic pulmonary vasoconstriction.)
History of Anesthesia for Thoracic Surgery 3
Figure 1-1 See legend on opposite page
4 History of Anesthesia for Thoracic Surgery
the operative lung). Finally, if hypoxemia is a se-
vere and persistent problem, the operative lung
may be conventionally ventilated with IPPB (re-
turn to two-lung ventilation). The nonoperative
lung may be managed with conventional IPPB
alone or in conjunction with positive end-expira-
tory pressure (PEEP). Alternatively, the nonoper-
ative lung can be occasionally managed with HFV,
as in surgery of the major conducting airways or
the presence of a major bronchopleural fistula.
The patient's overall and specific organ well-
being will be followed by a variety of monitors
(see chapter 7). Routinely, the patient will have
the electrocardiogram (EKG) continuously dis-
played, with input (from intravenous fluids) and
output (from nasogastric and urinary catheters as
well as blood loss from the operative field) (I &
O) frequently tallied. The vast majority of patients
will have an indwelling arterial catheter, which
will continuously display the phasic systemic ar-
terial pulse pressure (SAP) and from which fre-
quent intermittent arterial blood gas (ABG) sam-
ples may be drawn. Additionally, a pulse oximeter
is routinely placed on a finger or toe and continu-
ously displays heart rate and systemic arterial ox-
ygen saturation. A central venous catheter (index
of right-heart filling pressure) will often be in-
serted to monitor intravascular volume status (cen-
tral venous pressure, or CVP). However, if the
overall cardiovascular status is questionable or at
high risk (i.e., the patient is at risk for myocardial
ischemia and/or failure), a pulmonary artery cath-
eter may be inserted, which will allow measure-
ment of pulmonary arterial pressure (PAP), pul-
monary artery wedge pressure (PAWP) (index of
left-heart filling pressure), frequent determination
of cardiac output (CO), pulmonary vascular resis-
tance (PVR), systemic vascular resistance (SVR),
and venous blood gases (VBG), which will allow
for calculation of right to left transpulmonary
shunt (Q
s
/Q
t
) and oxygen consumption (Vo
2
). If
the patient does not breathe adequately postopera-
tively, a sophisticated reliable mechanical ventila-
tor will provide respiratory support (see chapter
19). Finally, an epidural catheter is often placed
preoperatively or immediately postoperatively to
provide postoperative analgesia with epidural
opioids (see chapter 21). The routine or very com-
mon components of this modern anesthesia prac-
tice are enclosed in a box in Figure 1-1, with the
occasional, but still very important, new compo-
nents of anesthesia practice underlined.
III. HOW DID WE GET THERE?
HISTORICAL EVOLUTION OF
MODERN ANESTHETIC PRACTICE
FOR THORACIC SURGERY
It would be impossible to list all of the many
wonderful medical discoveries and accomplish-
ments that have contributed to the evolution of
modern anesthesia practice for thoracic surgery.
Consequently, this chapter will concentrate on
events that were especially important and relevant
to the development of anesthesia for thoracic sur-
gery as a subspecialty and will not describe, or
will simply just mention, events that were impor-
tant to the overall development of anesthesia as a
broad discipline.
3-7
Nevertheless, just as thoracic
surgery is definitely the child of general surgery,
8
administration of anesthesia for thoracic surgery is
the child of administration of anesthesia for gen-
eral surgery.
Figure 1-2 summarizes the evolution of modern
anesthesia practice for thoracic surgery. On the
x-axis of the figure is the passage of time in dec-
ades. The y-axis lists medical achievements ac-
cording to whether they were primarily anesthesia
related (positive y-axis) or nonanesthesia related
(negative y-axis). The height of the medical ac-
complishment above or below the x-axis indicates
its relative importance in the evolution of modern
thoracic anesthesia practice. This history/time
chart emphasizes change in conceptual approach
by listing achievements only when they were first
clinically established, although in almost all in-
stances the real "first" and background for the
achievement were provided by earlier workers,
often using animals, or in single isolated cases or
institutions (see the following text).
As can be seen from Figure 1-2, the sequential
ability to intubate the trachea, to administer con-
trolled IPPB under relatively light anesthesia (cu-
rare and halothane), and to separate the lungs in
order to manage the two lungs differentially dom-
inates the anesthesia-related evolution. The devel-
opment of antibiotics .and of ligation techniques
for individual structures for lung resection domi-
nates the non-anesthesia-related evolution. Most
of these major advances were made during World
War II.
A. Pre-1910: Basic Medical
Accomplishments Advancing All Types
of Surgery
Prior to 1910 many basic and important discov-
eries were made that permitted the initial overall
development of surgery as a discipline. Primarily
anesthesia-related medical accomplishments in-
cluded discovery of oxygen, carbon dioxide, and
inhalational, intravenous, local, and conduction
anesthesia. Primarily non-anesthesia-related med-
ical accomplishments included the administration
of intravenous fluids, including blood transfusion,
the use of antisepsis, asepsis, the discovery of x-
rays, and airway endoscopy. The names of a few
of the persons responsible for these medical ac-
Evolution of Modem Anesthesia Practice for Thoracic Surgery
Height of Bars = Relative importance of achievement
Figure 1-2 Evolution of modern anesthesia practice for thoracic surgery. (IV = intravenous; TB = tuberculosis; PEEP = positive end-expiratory pressure; CPAP = constant positive airway
pressure; CA = cancer; HFV = high-frequency ventilation.)
complishments, which are important to all aspects
of modern medicine, are Andreas Vesalius (1555,
described thoracic anatomy), William Harvey
(1628, described circulatory system), Joseph Pries-
tley (1772, discovered 0
2
and N
2
0), Horace Wells
(1844, introduced N
2
0 inhalation anesthesia), Wil-
liam Morton (1846, introduced ether inhalation an-
esthesia), Louis Pasteur (1860-1880, described
bacterial origin of infection), Joseph Lister (1867,
introduced antisepsis), Robert Koch (1878-1882,
introduced bacterial culture), Ernst von Bergmann
(1886, introduced asepsis), Alfred Kirstein (1895,
first used direct vision laryngoscope), Wilhelm
Conrad von Roentgen (1895, discovered x-rays),
Karl Landsteiner (1900, described ABO blood
groups), and Chevalier Jackson (1907, greatly ex-
tended endoscopy).
Although these advancements permitted the de-
velopment of surgery as a discipline, thoracic sur-
gery was still in a remedial stage and was confined
to extrathoracic chest wall procedures. Open-chest
intrathoracic surgery could not be performed be-
cause of the still unsolved "pneumothorax prob-
lem" (mediastinal shift and paradoxical respira-
tion; see chapter 4). The extrathoracic chest wall
procedures that were performed at this time (see
section B) were with ventilatory techniques that
did not include endotracheal intubation (see sec-
tion C).
B. 1900-1920: Extrathoracic Chest Wall
Procedures Were Performed to Treat
Infective Lung Disease
From 1900 to 1920, empyema and tuberculosis
were the main indications for thoracic surgery. The
former was treated by rib resection, decortication,
and drainage, and the latter primarily by lung col-
lapse. This was usually achieved by creating an
artificial pneumothorax, but a number of patients
required division of adhesions so that the lung
could retract, and some were selected for the more
aggressive treatment of phrenic nerve avulsion.
Thoracoplasty was introduced in 1912 by Morris-
ton Davies but was considered hazardous at first
and, thus, was performed infrequently. However,
as both experience and expertise were acquired
after World War I, the procedure greatly gained in
popularity. In addition, infrequent thoracotomies
were also performed for lung resection and the
removal of tumors, using a very quickly applied
total lesion snare or tourniquet technique. The sub-
sequent necrosis of the lesion required removal
several days later and was usually associated with
infection, hemorrhage, and/or air leak. Endotra-
cheal intubation, IPPB, and lung separation tech-
niques were not yet in use for anesthesia, and
because the surgery predominantly involved the
chest wall, they were not actually required.
C. 1900-1920: Ventilatory Techniques
for Chest Wall Procedures Did Not
Include Endotracheal Intubation
Around 1910 these types of chest wall opera-
tions were managed with two different ventilatory
techniques. The first involved intratracheal insuf-
flation anesthesia. This was first described by two
American physiologists, James Meltzer and John
Auer. They demonstrated in animals that if gas is
blown under pressure into the trachea through a
narrow catheter, and allowed to escape around it,
satisfactory gas exchange could be achieved.
Charles Elsberg expanded on Meltzer and Auer's
insufflation method for surgery in patients by
using Chevalier Jackson's laryngoscope (1907) to
deliver the respiratory gases (air-ether). This tech-
nique usually rendered the patient apneic and be-
came a very popular choice for chest surgery in
the United States in 1914 and for head and neck
surgery in Britain following its introduction there
by the ear, nose, and throat surgeon Robert Kelley.
Indeed, Elsberg was able to further promote the
technique when asked to write the chapter on an-
esthesia for thoracic surgery in the 1914 edition of
Gwathmay's textbook Anesthesia. It should be
noted that insufflation of gases during bronchos-
copy, now largely superseded by the injector tech-
nique of Sanders (1967), is a modern derivative of
this technique. Although laryngoscopy was a pre-
requisite for intratracheal insufflation, the value of
endotracheal intubation, which would have al-
lowed for the use of intermittent positive-pressure
ventilation to solve the open-chest "pneumothorax
problem," simply did not occur to these individu-
als.
The second ventilatory method for intrathoracic
operations was primarily used during World War I
(1914 to 1918) and consisted of administration of
nitrous oxide and oxygen via spontaneous respira-
tions through an airtight mask with a PEEP of 5 to
7 mm Hg (referred to as CPAP breathing today)
and intravenous morphine added for additional an-
algesia. Endotracheal intubation, although possible
(see next section), was not used for chest opera-
tions, even on casualties, because it required ex-
pertise not readily available, and since deeper an-
esthesia was required, cardiovascular collapse was
feared. The airtight mask CPAP ventilatory
method continued to be used a great deal for chest
surgery for another decade after the war. Individ-
uals responsible for the development of the nitrous
oxide/oxygen/morphine anesthesia technique and
apparatus included Charles Peter, J. A. Heidbrink,
Hisiorv of Anesthesia for Thoracic Surgery 1
James T. Gwathmay, Elmer I. McKesson, and
George W. Crile.
D. 1920s: Beginning of Significant Use
of Endotracheal Intubation
The history of endotracheal intubation is closely
related to the history of laryngoscopy. In the area
of laryngoscopy and endotracheal intubation (and.
perhaps, positive-pressure ventilation) there were
several '"firsts.*" but they were forgotten or dis-
carded as a result of contemporary criticism. Con-
sequently, the date of introduction of these new
techniques or apparatuses was often controversial.
Prior to the 1920s, attempts at laryngoscopy
and/or endotracheal intubation were sporadic and
short lived (with the exception of Chevalier Jack-
son's laryngoscope for throat and neck surgery).
Charles Kite first used an endotracheal tube in
resuscitation of drowning victims in 1788. Manuel
Garcia, a singing teacher, used indirect laryngos-
copy in 1855. Around 1880 George Fell, Joseph
O'Dwyer. and William Macewen practiced blind
oral intubation, and in 1893 Fell attached a bel-
lows system to the intubating cannulas to achieve
resuscitative lung expansion by positive pressure.
Rudolph Matas, a surgeon, applied some of the
Fell-O'Dwyer blind techniques and apparatus to
the treatment of traumatic pneumothorax, and a
few other surgeons at this time (1895 to 1900)
adopted the Fell-O'Dwyer blind intubation-bel-
lows apparatus for use in surgery (see introduction
of IPPB into thoracic anesthesia in section F). A
direct vision laryngoscope for endotracheal intu-
bation developed by Kirstein in 1895 was not used
by anyone in surgery.
During and after World War I (1916 to 1928)
Ivan W. Magill and Stanley Rowbotham were
prompted by the need to provide safe operating
conditions for maxillofacial surgery on injured
servicemen to develop the technique of endotra-
cheal intubation; this work pioneered contempo-
rary endotracheal anesthesia. At first, wide-bore
rubber tubes were passed by blind nasal and oral
intubation, a technique that did not require either
laryngoscopy or deep anesthesia. Nevertheless,
this technique was little used for chest wall proce-
dures with nitrous oxide/oxygen/morphine anes-
thesia during this period. In fact, despite his own
remarkable skill at both blind nasal and orotra-
cheal intubation. Magill expressed the following
view as late as 1928 in a communication to the
Anaesthetic Section of the Royal Society of Med-
icine: "Operations on the thorax, such as thoraco-
plasty and pulmonary decortication, give better re-
sults when nitrous oxide or ethylene and oxygen
are administered by means of a face-piece. The
condition of the patient after operation is ample
compensation for the difficulty sometimes entailed
in maintaining airtight apposition of the mask.
Moreover, an efficient apparatus provides for pos-
itive ventilation of the lungs without the necessity
for intubation."''
However, within 2 to 4 years of this statement.
Gale and Waters in the United States and Magill
in the United Kingdom, all of whom understood
the value of endotracheal intubation and were
aware of Chevalier Jackson's work with direct lar-
yngoscopy, began the natural marriage of direct
laryngoscopy with endotracheal intubation, and di-
rect orotracheal intubation soon became increas-
ingly used in the following years. This was a very
important step because it was prerequisite to the
use of intermittent positive-pressure ventilation,
which would prove, a decade later, to be the solu-
tion to the "pneumothorax problem."
The ability to perform laryngoscopy and endo-
tracheal intubation more routinely logically led to
the capability of separating the two lungs with
various endobronchial tubes and bronchial block-
ers during the 1930s (Tables 1-1 and 1-2). Al-
though these devices proved to be extremely val-
uable following 1940 (see section F). their use at
this particular time was still restricted to a few
experts in a few centers. Other than the ability to
separate the two lungs, the development of anes-
thesia for thoracic surgery between 1930 and 1940
was relatively insignificant. Following this rela-
tively static period and World War II. thoracic
surgery and anesthesia for thoracic surgery greatly-
changed; operations on the chest wall dramatically
decreased, and major operations within the thorax
dramatically increased.
E. 1940-1950: Decline in Chest Wall
Procedures to Treat Infective Lung
Disease
As with all infectious diseases, the development
o\~ antibiotics completely changed the clinical
course of infective chest diseases. The treatment
o( lung abscess, bronchiectasis, and empyema was
revolutionized in the post-World War II period by
the use of sulfonamides (introduced by May and
Baker in 1938 for the control of hemolytic strep-
tococcal infections) and the availability of penicil-
lin for civilian use (the antibiotic effects of peni-
cillin were first described by Alexander Fleming
in 1929). The discovery of streptomycin in 1943.
para-amino-salicylic acid in 1946. and isoniazid in
1952 had equally profound effects on the treatment
of tuberculosis. Because of the discovery of these
antibiotics, pulmonary tuberculosis and bronchiec-
tasis were treated surgically far less often, whereas
Table 1-1 DEVELOPMENT OF SINGLE-LUMEN ENDOBRONCHIAL TUBES'
Date Name Distinctive Characteristics
Bronchial Intubation
Technique!
1932 Gale and Waters
1936 Magill
1955 Macintosh and Leatherdale
1957 Gordon and Green
1958 Pallister
1958 Machray
Carinal cuff that only when inflated to give airtight seal
excludes other main bronchus; unstable
Right and left tube with large endobronchial cuff; right
tube has terminal wire spiral
Left endobronchial tube, angulated at carina; tracheal as
well as endobronchial cuff; incorporated tracheal
suction stem; good stability
Right endobronchial tube with tracheal and bronchial
cuffs; carinal hook; bronchial cuff has slot in lateral
wall
Left endobronchial tube with one tracheal and two
bronchial cuffs
Left endobronchial tube similar to the left-sided Magill
tube but has a shorter bronchial cuff
Blind
Intubating bronchoscope
Blind
Blind or intubating
bronchoscope
*Data taken from Lee and Atkinson.
3
tPrior to availability of fiberoptic bronchoscopy.
resection for malignant disease became more and
more common (see the following section), a trend
that has certainly continued to the present time.
6
F. 1938-1950: Emergence of
Intrathoracic Lung and Esophageal
Resection for Malignancy
Several events around 1938 resulted in an ex-
plosion of growth of intrathoracic surgery. These
were both non-anesthesia and anesthesia related.
The major non-anesthesia-related advance was the
development of techniques for dissecting and li-
gating individual structures at the hilum of the
lung. Prior to 1930, lung resection was only occa-
sionally performed using a quickly applied total
lesion snare or a tourniquet technique. The tech-
nique required two surgical stages: one to snare
the lesion, and then another several days later to
remove the necrotic tissue. Consequently, the tech-
nique was fraught with dangers of infection, he-
morrhage, and air leak, and it always resulted in
significant morbidity and mortality. The first suc-
Table 1-2 DEVELOPMENT OF ENDOBRONCHIAL BLOCKERS*
Date Name Distinctive Characteristics Bronchial Intubation Technique!
1938 Crafoord
1936 Magill
1943 Vernon
1953 Sturtzbecher
1954 Vellacott
1955 Macintosh and
Leatherdale
1958 Green
1981 Ginsberg
Simple Bronchial Blocker
Gauze pack inserted into affected bronchus
Rubber suction catheter with a small inflatable
bronchial cuff
Similar to Magill catheter but larger and with
cuff covered with gauze or woven nylon
Intubating bronchoscope under topical analgesia
Bronchoscope
Bronchoscope
Combined Endotracheal Tube Plus Bronchial Blocker
Essentially an endotracheal tube with an
incorporated suction catheter plus blocker cuff
Right-sided endobronchial tube with bronchial
cuff that blocks upper lobe orifice, plus
tracheal cuff, an orifice in left lateral wall
between cuffs
Cuffed endotracheal tube with an angulated
cuffed suction catheter blocker incorporated
for left bronchus blockade
Right-sided tube similar to Vellacott tube but
with carinal hook
Modern clear plastic endotracheal tube with
second smaller "kangaroo" lumen for balloon-
tipped bronchial blocker
Blind-blocker directed into diseased bronchus
with wire stylet
Blind
Blind or intubating bronchoscope
Blind
3
History of Anesthesia for Thoracic Surgery
cessful pneumonectomy using the snare technique
was carried out by Rudolf Nisson in Germany in
1931 for bronchiectasis. Cameron Haight and John
Alexander in 1932 performed the first successful
pneumonectomy in the western hemisphere using
the same technique in a patient with bronchiec-
tasis. In 1933 Evarts Graham performed a left
snare pneumonectomy on a physician with squa-
mous cell carcinoma, and the individual survived
30 years.
9
10
No patient had survived a total pneu-
monectomy for a malignant tumor of the lung be-
fore Dr. Graham's operation.
Although everyone recognized that Dr. Gra-
ham's success was a milestone, the development
of the individual-structure (bronchus, pulmonary
artery, pulmonary vein) ligation technique in 1929
and the 1930s was much more important because
it greatly reduced the incidence and risk of post-
operative bronchial leaks and tension pneumo-
thorax, pulmonary hemorrhage (from either artery
or vein), and infection from residual necrotic tis-
sue. The individual-structure ligation technique for
lobectomy was first extensively used by Harold
Brunn in 1929
11,12
and then subsequently by Ed-
ward Churchill starting in 1938.
13
W.F. Rienhoff
performed the first pneumonectomy using the in-
dividual ligation technique in the 1930s. However,
the missing surgical detail in allowing the individ-
ual structure ligation technique to work well was
the routine postoperative use of closed-chest tho-
racostomy drainage with intrapleural suction. The
first report of pleural drainage after lobectomy was
in. 1929"
12
and for lung abscess in 1936.
14
15
Closed-chest thoracostomy with pleural suction
would prove to be the single most important post-
operative factor enabling thoracic surgery to ad-
vance. By the end of World War II, intercostal
closed thoracostomy with intrapleural suction had
become the standard primary treatment for most
thoracic injuries.
16, l7
The work of these early sur-
geons in combining individual-structure ligation
with postoperative pleural drainage obviously laid
the groundwork for the performance of lung resec-
tion as it is done today.
9-18
Also during this decade,
esophageal resections evolved rapidly from the
successful use of esophagectomy without thoracot-
omy
19
to the successful one-stage transthoracic re-
section and reconstruction accomplished in Japan
in 1933
20
and introduced into the Western world
in 1938.
21
These early attempts at intrathoracic
procedures during the 1930s were aided by the
introduction of an endobronchial tube for one-lung
anesthesia by J.W. Gale and R.M. Waters in 1932
and by the introduction of an endobronchial tube,
bronchoscope, and bronchial blocker by Magill in
1936 (see Tables 1-1 and 1-2 for subsequent de-
velopment of endobronchial tubes and bronchial
blockers).
22
"
24
Aside from the primordial lung separation tech-
niques mentioned previously, the truly important
anesthesia-related advances that contributed to the
explosive growth in thoracic surgery around 1938
consisted of the development of IPPB for chest
surgery, which, combined with the introduction of
muscle relaxation with curare in 1942, led to the
ability to easily control ventilation in the intraoper-
ative period. Prior to this time, the primary diffi-
culty in performing intrathoracic procedures was
known as "the pneumothorax problem" (see
chapter 4). As soon as the chest of a spontaneously
breathing patient was opened, the lung in question
not only collapsed but also moved up and down
violently with each struggling breath (mediastinal
flap and paradoxical respirations). Within a few
minutes, the patient became cyanotic and hypoten-
sive, and, unless the chest was closed quickly, the
patient would die. These were extremely poor con-
ditions under which to make a diagnosis and to
undertake treatment. The single most important
intraoperative factor enabling thoracic surgery to
advance rapidly was the solution to this pneumo-
thorax problem by artificial rhythmic inflation of
the lungs (through controlled ventilation with
IPPB) after the patient had been rendered apneic
(by muscle relaxation).
The successful introduction of IPPB for the
management of intrathoracic operations first re-
quired the development of laryngoscopy, endotra-
cheal intubation, and adequate bellows or pump
machinery. Many early attempts to solve the
"pneumothorax problem" failed because they
lacked one of these essential components. In trying
to avoid atelectasis, Ernst Sauerbruch in Germany
during 1893 to 1904 did his early thoracotomies
in an airtight chamber with the pressure reduced 7
mm Hg below atmospheric pressure, while the pa-
tient's head and the anesthetist were outside in
atmospheric air (negative-pressure breathing). This
technique was followed in 1904 by a positive-
pressure breathing chamber of another German,
Ludolph Brauer. Brauer's method was essentially
the same as the CPAP method introduced by
Gregory et al in 1971
25
but used more complicated
machinery. The positive-pressure breathing
method of Brauer slowly replaced the negative-
pressure breathing cabinet of Sauerbruch. Sauer-
bruch, who was without a doubt the most domi-
nant influence in (European) thoracic surgery at
this time, was extremely resistant to the change
from negative-pressure to positive-pressure breath-
ing in spite of the obvious clinical failure of his
method of spontaneous negative-pressure breath-
ing to permit prolonged surgery. In fact, Sauer-
bruch's spontaneous negative-pressure breathing
method remained in widespread use until World
War II and indeed appeared as the recommended
method in an English textbook in 1937, of which
he was coauthor. However, at the end of his career,
Sauerbruch advocated Brauer's positive-pressure
breathing technique. Neither Sauerbruch's nor
Brauer's breathing chamber techniques involved
endotracheal intubation.
There was also considerable interest in the
United States in developing positive-pressure
breathing, and the development of positive-pres-
sure breathing was closely related to the develop-
ment of endotracheal intubation (see section D). In
the 1880s, George Fell popularized the use of bel-
lows for resuscitation (compression leading to pos-
itive pressure), and O'Dwyer designed a hooked
metal intubating cannula. In 1895 Theodore Tuf-
fier, a French surgeon, used a cuffed O'Dwyer
tube and rhythmic positive pressure with PEEP for
thoracic surgery. In 1898, Rudolph Matas clearly
recognized that the Fell-O'Dwyer apparatus could
revolutionize thoracic surgery and used their ap-
paratus in thoracic surgery. Indeed, in 1899 Matas
wrote, "The procedure that promises the most
benefit in preventing pulmonary collapse in oper-
ations on the chest is the artificial inflation of the
lung and the rhythmical maintenance of artificial
respiration by a tube in the glottis directly con-
nected with a bellows. Like other discoveries, it is
not only elementary in its simplicity, but the fun-
damental ideas involved in this important sugges-
tion have been lying idle before the eye of the
profession for years. It is curious that surgeons
should have failed to apply for so long a time the
suggestions of the physiological laboratory, where
the bellows and tracheal tubes have been in con-
stant use from the days of Magendie to the present
in practicing artificial respiration in animals."
7
In 1899 Matas's colleague P.W. Parham stated,
"so imbued am I with its [the apparatus] great
value that I believe no surgeon now would be
justified in attempting a thoracic resection without
having the Fell-O'Dwyer apparatus at hand. I be-
lieve it will revolutionize the field of surgery, mak-
ing possible operations in the chest that would
otherwise clearly be too hazardous."
7
It is strange
that pioneer thoracic surgeons like Tuffier, Matas,
and Parham neglected to try Kirstein's laryngo-
scope for endotracheal intubation and preferred in-
stead to depend entirely upon blind oral digital
methods that appeared uncertain and unhygienic
even to their colleagues. In 1905 Jane way and
Green from New York described a cuffed endotra-
cheal tube and a pump for experimental chest sur-
gery. In 1910, Dorrance from Philadelphia de-
scribed cuffed endotracheal tubes and a pump for
chest surgery. However, none of these early Amer-
ican "firsts" in intubation-bellows or pump sys-
tems attracted considerable attention.
The history of IPPB as we know and use it
today began in 1916 when Giertz, a former assis-
tant to Sauerbruch, conducted experiments in ani-
mals showing that rhythmic inflation of the lungs
was more effective than either negative-pressure
or continuous positive-pressure ventilation. In
1934, Guedel, using ether anesthesia, was the first
to routinely manually control respiration intraoper-
atively in intubated patients. In 1934, Frenckner, a
Swedish ear, nose, and throat surgeon, described
the first experimental model "Spiropulsater" ven-
tilator for chest operations. The Frenckner posi-
tive-pressure "Spiropulsater" respirator was first
used by Clarence Crafoord in surgery in Stock-
holm in 1938, and this actually was the dawn of
the modern era of mechanical IPPB. The con-
trolled IPPB ventilation concept was further pro-
moted by Waters and simplified by Guedel in this
country in the late 1930s and by Nos worthy in
England in 1941. All three advocated controlled
breathing by intermittent pressure on the reservoir
bag of a closed cyclopropane anesthesia circuit.
Although ether and cyclopropane, which were the
most widely used inhalation anesthetics at that
time, were safe and potent (and therefore permitted
controlled respiiration to a certain extent), they
were also highly explosive and therefore precluded
the use of electrocautery. Controlled positive-pres-
sure breathing became routinely possible during
entire intrathoracic procedures with the introduc-
tion of muscle relaxation with curare in 1942 by
H.R. Griffith. Curare had the great advantages of
inducing apnea, diminishing reflexes, and, most
importantly, allowing the unrestricted use of elec-
trocautery.
Thus, the late 1930s and early 1940s witnessed
the advent of manual and mechanical controlled
ventilation in paralyzed patients. Despite the fact
that Crafoord's results were enthusiastically ac-
cepted in Scandinavia, and despite the efforts of
Guedel, Waters, and Nosworthy and the introduc-
tion of muscle relaxation, a significant portion of
the British and American anesthesia community
continued to use the older methods of continuous
positive pressure by mask and intratracheal insuf-
flation during the 1940s.
The slow acceptance by American and British
anesthetists of intraoperative controlled IPPB
caused mechanical ventilators to first be used ex-
tensively and routinely outside the operating room.
The catastrophic Copenhagen polio epidemic of
1952 led to a crash production program of Eng-
strom volume ventilators. In 1955, Bjrk and Eng-
strom in Sweden first described the use of their
ventilator for postoperative respiratory care of
poor-risk thoracic surgery patients. The Jefferson
ventilator was developed on John Gibbons's tho-
racic surgery service in Philadelphia and was in-
troduced in 1957 as the first American ventilator
for controlled ventilation. The clear-cut efficacy of
mechanical ventilators outside of the operating
room soon convinced anesthetists to use these ma-
chines to control IPPB ventilation reliably in the
operating room, and operating room mechanical
ventilators became commonplace in the 1960s and
1970s.
G. 1950s and 1960s: Development of
The development of double-lumen tubes was a
response to the fast-growing capabilities in tho-
racic surgery, which now required faster, surer,
and simpler methods of separating the two lungs
and of causing the lung under operation to be
atelectatic. The Bjork and Carlens bronchospiro-
metric double-lumen tube was first used during
anesthesia in 1950. in the next two decades, sev-
eral different types of double-lumen tubes with
varying capabilities were introduced (Tabic 1-3).
Until the advent of fiberoptic bronchoscopy, the
method of placement of these tubes within the
tracheobronchial tree was essentially blind. Fiber-
optic bronchoscopy now allows for these tubes to
be placed and for the position of tubes to be
checked under direct vision repeatedly with ex-
treme precision and very low risk (see section K).
H. 1956: Introduction of Halogenated
Inhalation Anesthetic Drugs
The introduction of halothane in England in
1956 by M. Johnstone was the result of a search
for an inhalational anesthetic that would meet
many needs that remained unfulfilled by previ-
ously used drugs. It was nonflammable (which
ether and cyclopropane were not) and therefore
allowed the use of electrocautery, it allowed high
concentrations of oxygen to be used (which nitrous
oxide obviated); it was thought to be inert; it had
limited solubility in water and fat (as opposed to
ether and chloroform); it did not react with alkali
used for CO
:
absorption (as did trichloroethylene);
it was tolerated reasonably well by the heart: and
it produced a smooth induction and awakening. By
1959, the use of halothane was widespread in both
the United States and the rest of the world. This
signaled the beginning of the era of nonflammable,
inhalational anesthetics. Although halothane was
initially thought to have no toxicity, it now appears
that hepatic toxicity is a real, but very rare, entity.
Fortunately, other halogenated drugs have been
developed, such as isoflurane, which have all the
desirable properties of halothane and appear not to
cause any tissue toxicity.
I. 1970s: Increased Invasive and
Noninvasive Monitoring
The 1970s witnessed a phenomenal explosive
increase in monitoring patients intra- and postop-
eratively. Almost all of the monitors shown m
Figure 1-1 were introduced during this period.
The use of these monitors has allowed for the
diagnosis of major patient problems that heretofore
were impossible to make with certainty (such as
the diagnosis of cardiac failure by pulmonary ar-
Table 1-3 DEVELOPMENT OF DOUBLE-LUMEN ENDOBRONCHIAL TUBES*
Date Name Distinctive Characteristics
Bronchial
Intubation
Technique!
1950 Carions
1959 Bryce-SmitH
1960 Bryce-Smith and Salt
I960 White
1962 Robertshaw
1979 National Catheter Corporation
Double-lumen catheter with two inbuilt curves; tracheal and
a bronchial cult' for left main bronchus, carina! hook and
cross-sectional shapeoval in horizontal plane
Modification of the Carlens catheter with no carina! hook;
cross-sectional shapeoval in horizontal plane
Right-sided version of the Bryce-Smith tube, possessing slit
in endobronchial cult, no carinal hooks
Right-sided version of the Carlens catheter, possessing slit in
endobronchial cull and a carinal hook
Right and left double-lumen tubes; modification of the
Carlens catheter with a larger lumen and hence low
resistance to gas Hows; slotted right endobronchial cult;
no carinal hooks; cross-sectional shapeD-shaped in the
horizontal plane
Right and left Robertshaw-type disposable double-lumen
tube made of clear tissue implantable lastic with low-
pressure, high-volume cuffs
Blind
Blind
Blind
Blind
Blind
Blind
1
tery pressure monitoring). The monitors (oxygen-
ation and ventilation monitors, neuromuscular
blockade monitors) also refine the specificity of,
and can quantitate, many diagnoses that previously
were possible to make but without great precision
or accuracy.
J. 1970s: Introduction of PEEP and
CPAP Into Clinical Practice
PEEP was introduced into clinical practice al-
most 20 years ago
26
and has proved to be a rapid
and relatively high-benefit, low-risk method for
increasing the oxygenation capability of severely
diseased lungs. CPAP was introduced a few years
later to treat spontaneously breathing infants with
idiopathic respiratory distress syndrome.
25
Since
then, the use of CPAP and PEEP to treat patients
with respiratory disease can almost be character-
ized as routine because these are the primary ther-
apeutic mechanisms by which the inspired oxygen
concentration can be reduced below toxic levels.
K. 1975: Introduction of Fiberoptic
Bronchoscopy
Direct examination of the tracheobronchial tree
with flexible fiberoptic bronchoscopy has greatly
facilitated the diagnosis, staging, and management
of pulmonary neoplasms as well as many other
lung diseases.
27
With particular reference to tho-
racic anesthesia, the fiberoptic bronchoscope now
allows for placement of double-lumen and endo-
bronchial tubes and bronchial blockers with great
precision and accuracy and at low risk to the pa-
tient. Prior to the advent of the fiberoptic broncho-
scope, these instruments were placed blindly (see
Table 1-3). Consequently, their exact position was
often precarious, and they sometimes malfunc-
tioned, preventing collapse of the nondependent
lung and ventilation of the dependent lung. Today,
the fiberoptic bronchoscope, although expensive,
is considered to be a high-priority piece of equip-
ment for an individual practicing anesthesia for
thoracic surgery.
L. 1980s: Use of Nondependent Lung
CPAP, Dependent Lung PEEP,
Differential Lung Ventilation, HFV
The ability to place double-lumen tubes easily
and accurately and at low risk (in part because of
the development of the fiberoptic bronchoscope)
has led to the ability to manage the two different
lungs specifically according to their individual pa-
thology. Thus, the nondependent lung may be stat-
ically distended by low levels of CPAP (using
oxygen), which greatly enhances oxygenation dur-
ing one-lung ventilation. The dependent lung may
be ventilated in the conventional manner, but with
the addition of PEEP, thereby correcting a low
ventilation-to-perfusion situation that may exist
owing to dependent lung compression by the me-
diastinum, abdominal contents, and positioning ef-
fects. Use of these two ventilation modalities to-
gether is termed differential lung ventilation.
28
Similarly, in the intensive care unit, patients with
predominantly unilateral adult respiratory distress
syndrome have been treated with differential lung
ventilation.
29
HFV may be the treatment of choice
for some patients with major bronchopleural fistu-
las and for some patients having surgery on major
conducting airways. HFV is efficacious for major
bronchopleural fistulas because it involves low air-
way pressures, which minimize air leaks from the
fistulas. HFV may be efficacious for surgery on
major conducting airways because it requires that
only a small catheter pass through the surgeon's
operative field; consequently, the anastomosis of a
major conducting airway may be made much eas-
ier.
IV. THE FUTURE
The focus of today's research will determine to
a large extent the types of advances that will be
made in the future. Prevention of lung carcinoma
is probably the most important consideration, and
efforts to discourage smoking will certainly con-
tinue to increase and be vigorous. Most encourag-
ingly in this regard, lung cancer rates in White
men in the United States leveled off in the late
1970s and, in 1982 and 1983, decreased for the
first time ever in White American men.
30
The de-
crease in lung cancer rates in White American men
parallels the decrease in smoking rate in this one
segment of American society. In addition, lung
cancer prevention with diet and micronutrients has
raised considerable interest.
31-35
Because early detection of lung carcinoma by
radiologic and cytologic methods have limited, but
some,
36
potential (high cost, low yield), the future
of screening for lung carcinoma may lie in devel-
oping sensitive cellular markers. Indeed, the iden-
tification of oncolipids
37
38
and tumor antigens,
39
development of specific antisera to tumor antigens,
and radiolabeling of these antibodies may serve
not only for early detection but also for early ther-
apy.
40
Because the most important cause or failure
of resectional surgery to cure patients of thoracic
carcinoma is mediastinal and extrathoracic meta-
static disease, much effort will go into making the
diagnosis of metastatic disease, and we will likely
witness improvement in such noninvasive proce-
dures as computed tomography and magnetic res-
onance imaging. rachytherapy, the permanent or
temporary implantation of radioactive sources.
may otter a useful surgical approach in patients m
whom unresectable pulmonary or mediastinal ma-
lignancies are found at the time of thoracotomy or
m patients previously treated with other modalities
for whom limited therapeutic alternatives exist."
It appears likely that we will witness an increase
in the availability and use of continuous monitors
of oxygenation, ventilation, and acid-base balance.
In addition, it is likely that additional very short-
acting inhalational anesthetics, narcotics, seda-
tives, and relaxants will be available in the near
future.
Interest will continue to be high in multimodal
therapy (chemotherapy, irradiation, surgical exci-
sion) of carcinoma of both the lung and esopha-
gus.^
1:
A new, novel, promising multimodal ap-
proach to the treatment of lung cancer has been
the administration of immunotherapy (interleukin-
2 and tumor necrosis factor-alpha).
4S
Because of
progress with immunosuppression, lung transplan-
tation should begin to achieve a high degree of
long-term success in the next two to three decades.
Finally, lasers may be used to remove lesions
without sacrificing surrounding normal lung tissue.
prevent air leaks after resection, and prevent the
distortion of lung tissue caused by stapling."
1
V. SUMMARY
Anesthesia for thoracic surgery has gone
through a remarkable evolution over the last 50
years. We have moved from performing simple
chest wall surgery without a secure airway in
poorly monitored patients who were deeply anes-
thetized with inherently dangerous techniques to
performing extremely complicated intrathoracic
procedures, with firm control over each lung inde-
pendently, in very well-monitored patients who are
moderately anesthetized with inherently safe tech-
niques. Consequently, the present mortality rates
of a pneumonectomy and of a lobectomy are only
6 and 3 per cent, respectively.
4.
Conversely, with
non-small cell lung carcinoma and no hilar or
mediastinal node involvement, a 90 per cent 5-
year survival rate may now be expected following
definitive resection.
4S
These figures compare ex-
tremely favorably with a mortality of 52 per cent
for a lobectomy for bronchiectasis in 1922." Con-
sidering the fact that the first modern pneumonec-
tomy was performed only 50 years ago. thoracic
anesthesia represents a subspecialty that has had a
rapid and dramatic development.
REFERENCES
1. Benumof JL: Monitoring respiratory function during anes-
thesia. In Saidman LJ. Smith NT (eds): Monitoring During
Anesthesia. 2nd eel. Boston. Butterworth Publishers. 1983.
PP 35-78.
2. Alter) DD. Benumof JL: Anesthesia for thoracic surgery.
In Miller R (eel): Anesthesia. New York. Churchill Living-
stone. 1981. pp 925-980.
3. I.ee JA, Atkinson RS: The history of anesthesia. In Lee
JA, Atkinson RS (eds): A Synopsis o\' Anesthesia. 7th ed.
Baltimore. Williams & Wilkins Co, 1973. pp 1-24.
4. Lee J A. Atkinson RS: .Anesthesia for thoracic surgery. In
Lee JA. Atkinson RS (eds): A Synopsis of Anesthesia. 7th
ed. Baltimore. Williams & Wilkins Co. 1973. pp 599-633
5. Kaplan JA: Development of thoracic anesthesia. In Kaplan
JA (ed): Thoracic Anesthesia. New York. Churchill Liv-
ingstone, 1982, pp 3-8.
6. Gothard JWW. Branthwaithc VIA: History. In Gothard
JWW. Branthwaite MA (eds): Anesthesia for Thoracic
Surgerv. Oxford, Blackwell Scientific Publications. 1982.
PP 1-8.
7. Rendell-Baker L: History of anesthesia for thoracic sur-
gery. In Mushin WW (ed): Thoracic Anesthesia. 2nd ed.
Oxford. Blackwell Scientific Publications. 1963. pp 598-
661.
S. Nael AP: Pioneers on the road to thoracic surgery. J
Thorac Cardiovasc Surg 101:377-384, 1991.
(
* Brewei LA III: The first pneumonectomy. J Thorac Car
diovasc Surg 88:810-826, 1984.
10. Graham AE, Singer .1.1: Successful removal of the entire
lung for carcinoma of the bronchus. JAMA 101:1371-
1374. 1933.
11. Brunn H: Surgical principles underlying one-stage lobec-
tomy Arch Surg 18:490. 1929.
12 Meyer JA: Hugh Morriston Davies and lobectom) for can-
cer. 1912. Historical vignette. Ann Thorac Surg 46:472-
474. 1988.
13. Churchill I:. Belsey RHR: Segmental pneumonectomy in
bronchiectasis: Lingula segment o\' left upper lobe Ann
Surg 109:481-499. 1939.
14. Neuhol II. Touroff AS: Acute putrid abscess of the lung.
Principles o\ operative treatment. Surg Gynecol Obstet
63:353. 1936.
15. Wilkins LW Jr: Acute putrid abscess of the lung. Classics
in Thoracic Surgery. Ann Thorac Surg 44:560-561. 1987.
lb. Wagner RB. Slivko B: Highlights ol the historv of non-
penetrating chest trauma. Surg Clin North Am 69:1-14,
1989.
17. Bern FB: Earl) treatment of chest wounds by intrapleural
suction. Bull US Army Med Dept 8:21 I. 1948.
is. Skinner DB: Technical and scientific advances m general
thoracic surgery. Ann Thorac Surg 49:14-25. 1990.
19, Turner G: Excision o\ the thoracic oesophagus for carci-
noma with construction of an extra-thoracic gullet. Lancet
2:1315-1316. 1933.
20. Ohsawa T: Surgery of the esophagus. Arch b Jpn Surg
10:605-695. 1933.'
2 1. Adams W. Phemister DB: Carcinoma of the lower thoracic
esophagus. Report of a successful resection and esopha-
gogastrosiomy. .1 Thorac Surg 7:621-632. 1938.
22. Hillard EK. Thompson PW: Instruments used m thoracic
anaesthesia. In Mushin WW (ed): Thoracic Anaesthesia
2ml ed. Oxford. Blackwell Scientific Publications. I9%3.
pp. 267.
23. White CMJ: Evolution of endotracheal and endobronchial
intubation. Br J Anaesth 32:235. i960.
24. Roberishaw FL: Control of secretions and haemorrhages
in (he Lings during operation. In Gray TC. Nunn JF eds):
General Anaesthesia. 3rd ed. Vol 2. Butterworths, London.
1977. pp 217.
CHAPTER 2
Thoracic Anatomy
I. Introduction G. Collateral Ventilation
II. The Thoracic Wall H. Pulmonary Arteries and Veins
A. Bones and Cartilage I. The Pulmonary Capillaries
B. The Diaphragm 1. Alveolar Sheet Arrangement
C. The Intercostal Spaces 2. Ultrastructure of Pulmonary
D. The Relation of the Thoracic Wall to Capillary-Interstitial Space-
the Pleura and Lungs Alveolar Area
III. The Airway and Lung J. Lymphatic System
A. Trachea K. Bronchial Arteries and Veins
B. Main Bronchi IV. The Mediastinum
C. Hilar Anatomy A. Divisions of the Mediastinum
D. Lung Lobes and Fissures B. Mediastinal Relations of the
E. Bronchopulmonary Segments Trachea, Esophagus, Aorta, and
F. The Air Spaces Pulmonary Trunk
16 Thoracic Anatomy
I. INTRODUCTION II. THE THORACIC WALL
Understanding thoracic anatomy is important to
the thoracic anesthetist for several reasons. First,
knowledge of thoracic anatomy is important in
diagnosing and understanding the physiologic and
therapeutic implications of thoracic lesions. Sec-
ond, the anesthesiologist should understand how
surgery will proceed (patient position, incisions,
intrathoracic procedures) because surgical require-
ments often dictate anesthetic requirements. Third,
an appreciation of anatomy should allow the anes-
thesiologist to anticipate and to prepare for intra-
operative complications. Fourth, understanding
thoracic anatomy is an integral part of the safe and
successful use of regional anesthesia and pain re-
lief. This chapter first discusses the anatomy of the
thoracic wall, then the anatomy of the airway and
lung parenchyma, and finally the anatomy of the
mediastinum. Special pediatric anatomy is consid-
ered in chapter 18.
A. Bones and Cartilage
The thoracic bones and cartilages are covered
with muscles. The anterior muscles are the pecto-
ralis major and minor and serratus anterior (Fig.
2-1), and the lateral and posterior muscles are the
latissimus dorsi, rhomboid major and minor, ser-
ratus posterior, and trapezius (Fig. 2-2).
The thoracic bones and cartilages consist of the
sternum, 10 pairs of ribs and costal cartilages (to-
gether termed costae), two pairs of ribs without
cartilage, and 12 thoracic vertebrae and their inter-
vertebral disks (Figs. 2-3 and 2-4). Together these
bones and cartilages surround a cavity that is ren-
iform in cross section; the cavity has a narrow
thoracic inlet at the neck and a much larger tho-
racic outlet facing the abdominal cavity. The inlet
is surrounded by the manubrium of the sternum,
the first ribs, and the first thoracic vertebrae. The
Pectoralis major
Latissimus
dorsi
Serratus
anterior
Figure 2-1 The important anterior thoracic wall muscles are the pectoralis major and minor (which is underneath the pectoralis
major and is not shown) and the serratus anterior. The important lateral thoracic wall muscle is the latissimus dorsi. Corresponding
surface features are shown on the right. The fifth intercostal space is just below the inferior border of the pectoralis major.
Thoracic Anatomy 1 7
Trapezius
Levator
scapulae
Rhomboid
minor
Serratus post. sup.
Latissimus dorsi
Serratus post. inf.
Figure 2- 2 The important posterior thoracic wall muscles are the latissimus dorsi, rhomboid major and minor, serratus posterior
superior (post, sup.) and inferior (post, inf.), and trapezius. The corresponding surface markings are shown on the left.
Manubrium
Sternal angle
Body of sternum
5th Costal
Cartilage
Jugular notch
Clavicle
Scapula
Xiphoid process
Figure 2- 3 Anterior view of the bones and cartilages of the thorax. The ribs are numbered according to their vertebral origin.
Spinous process
Thoracic Vertebrae 1
Thoracic inlet
Manubrium
Body of Sternum
7th Costal
Cartilage
Figure 2- 4 Lateral view of the
bones and cartilages of the thorax.
The ribs are numbered according to
their vertebral origin.
18
Thoracic Anatomy 19
inlet is roofed by bilateral thickened endothoracic
fascia, called Sibson's fascia, and subjacent parie-
tal pleurae that project upward into the base of the
neck. The outlet is formed by the xiphoid process,
fused costal cartilages of ribs 7 to 10, the anterior
portions of the eleventh rib, the shaft of the twelfth
rib, and the body of the twelfth thoracic vertebra.
The anterior margin of the outlet is at the level of
the tenth thoracic, the lateral limits at the second
lumbar, and the posterior margin at the twelfth
thoracic vertebra. The outlet is therefore higher at
its anterior margin than at its posterior limit and
reaches its lowest level in the lateral aspect near
the midaxillary line. It is sealed off from the ab-
dominal cavity by the diaphragm.
The size and shape of the thorax are largely
determined by the ribs and the costal cartilages.
The relations of ribs and their costal cartilages to
the sternum and to each other vary at different
levels. The upper seven pairs of ribs articulate
directly with the sternum by way of costal carti-
lages and are therefore called "true" or vertebro-
sternal ribs. In contrast, the lower five pairs are
called "false" ribs because they do not articulate
with the sternum at all. Of the false ribs, three
pairs, the eighth, ninth, and tenth, are called the
vertebrocostal ribs because their associated carti-
lages articulate with immediately supradjacent car-
tilages. The remaining pairs, 11 and 12, terminate
in cartilaginous tips, which are embedded in the
muscles of the abdominal wall. Because the only
articulation of these latter two ribs is with the
vertebrae, they are called vertebral ribs. The flexi-
bility of the rib-vertebral and rib-sternal joints
permit the "bucket-handle" or "pump-handle"
movements of breathing.
B. The Diaphragm
The diaphragm is a dome-shaped musculofi-
brous septum that separates the thoracic from the
abdominal cavity: Its convex upper surface forms
the floor of the thoracic cavity, and its concave
undersurface forms the roof of the abdominal cav-
ity. Its peripheral part consists of muscular fibers
that originate from the circumference of the tho-
racic outlet and converge to be inserted into a large
central tendon. The central tendon is C-shaped and
concave toward the vertebrae (Fig. 2-5). The cen-
tral tendon is perforated a little to the right of the
midline and at the level of the lower border of the
eighth thoracic vertebra by the inferior vena cava.
The esophageal opening lies in the muscular part
of the diaphragm near the midline at the level of
the tenth thoracic vertebra. The aorta passes pos-
teriorly through the diaphragm as it lies on the
front of the body of the twelfth thoracic vertebra.
The abdominal surface of the diaphragm is closely
applied to the liver, the fundus of the stomach, the
spleen, and the right and left kidneys and suprare-
nal glands. The upper surface lies in relation to the
right and left lungs and pleura and to the heart and
pericardium.
Inferior vena cava
Esophagus
Left crus
Celiac
Trunk
Abdominal aorta
Figure 2-5 The diaphragm is viewed from below. The muscular part of the diaphragm consists of the peripheral part, a left and
right crus. The muscular part of the diaphragm inserts onto a C-shaped (concave posteriorly) central tendon. The diaphragm is
pierced by the inferior vena cava, the esophagus, and aorta.
20 Thoracic Anatomy
External
intercostal
muscle
Interior
intercostal
muscle
Vein
Artery
Nerve
Innermost
intercostal
muscle
Figure 2- 6 The muscle layers of the intercostal space, from
external to internal aspect, are the external intercostal, the in-
terior intercostal, and the innermost intercostal muscles. Be-
tween the interior and innermost intercostal muscle at the cau-
dad aspect of each rib is the intercostal vein, artery, and nerve;
the vein is most cephalad, and the nerve is most caudad.
C. The Intercostal Spaces
The intercostal spaces between the ribs are the
main surgical pathway to the thorax. The muscles
of the intercostal spaces themselves are arranged
in three layers corresponding to the three layers of
the lateral abdominal wall (Figs. 2-6 and 2-7).
From a surgical approach, the first layer of tissue
to be encountered within the intercostal space is
the external intercostal muscle, which passes
downward and forward from the lower border of
one rib to the upper border of the rib below and
extends from the tubercle of the rib posteriorly to
the neighborhood of the costochondral junction in
front. The next layer of muscle is called the inter-
nal intercostal muscle and extends from the ster-
Rectus muscle
& sheath
Anterior cutaneous
nerve
Peritoneum
9th Costal
cartilage
8th Costal
cartilage
Anterior branch
Mid-axillary line
Posterior
branch
Serratus anterior
muscle
Paravertebral gutter
Posterior
intercostal
membrane
Latissimus dorsi
Figure 2- 7 Distribution of the eighth intercostal nerve to show its relationships in both its thoracic and abdominal courses to the
muscle layers of the body wall. See text for full discussion of the anatomy of the intercostal space.
External oblique m.
Internal oblique m.
Transversus abdominis muscle
Lateral cutaneous nerve
External intercostal muscle
Internal intercostal muscle
Innermost intercostal muscle
8th Intercostal nerve
Erector spinae muscle
Posterior motor-
cutaneous
nerve
Trapezius muscle
Thoracic Anatomy 21
num to the angle of the rib posteriorly and then is
replaced by the posterior intercostal membrane.
The muscle fibers run obliquely downward and
backward. The innermost layer of muscle is called
the innermost intercostal and is often incomplete
with slips of muscle linked by membranous tissue.
The endothoracic fascia, which is equivalent to the
transversalis fascia of the abdominal wall, is no
more than a fine layer of areolar connective tissue
between the intercostal muscles and the parietal
pleura.
In each intercostal space lies a neurovascular
bundle comprising, from above downward, the in-
tercostal vein, the intercostal artery, and the inter-
costal nerve, protected externally by the costal
groove of the upper rib (see Fig. 2-6). Posteriorly,
this bundle lies between the pleura and the poste-
rior intercostal membrane, but at the angle of the
rib it passes between the internal and innermost
intercostal muscles (see Figs. 2-6 and 2-7). Each
intercostal nerve first branches off into a posterior
motor-cutaneous nerve, then into a lateral cuta-
neous nerve, which arises in the midaxillary line
and itself gives off an anterior and posterior
branch, and then finally into an anterior cutaneous
nerve (see Fig. 2-7). Obviously, when an intercos-
tal nerve block is performed posterior to the mid-
axillary line, it will produce anesthesia over most
of the course of the intercostal nerve (the anterior
80 per cent).
Because the major intercostal vessels and nerves
lie in close relation to the lower border of each rib,
incisions near this level are to be avoided. A pref-
erable site is along the upper margin of each rib.
Although accessory nerves and vessels may be
sectioned at this level, there is negligible loss of
sensation. It is equally important, however, to un-
derstand that the overlap of adjacent nerves is so
great that paralysis and complete anesthesia are
seldom produced within one intercostal space in-
cision unless its nerve and the nerves in the inter-
costal space above and below are all severed.
D. The Relation of the Thoracic Wall to
the Pleura and Lungs
For surgical purposes, the lungs and pleura may
be considered to be coextensive with their respec-
tive costal, mediastinal, and diaphragmatic sur-
faces separated only by a film of serous fluid.
Thus, the pleural borders can be considered to be
the lung borders and are formed by the continuity
of the surfaces of costal, mediastinal, and dia-
phragmatic pleurae.
The anterior, inferior, and posterior borders of
the pleurae may be related to the overlying struc-
tures of the thoracic cage (Figs. 2-8 and 2-9). The
anterior pleural borders of the pulmonary cupolae
are separated by visceral structures at the base of
the neck. As they descend medially behind the
sternum, they come to oppose one another at the
sternal angle. The right anterior border continues
downward, close to the midline. At the lower lim-
its of the body of the sternum, it diverges laterally
along the sixth or seventh costal cartilage to be-
come the inferior pleural border. The left anterior
pleural border may follow a similar course, but
more commonly it diverges laterally at the fourth
costal cartilage, lies at the lateral sternal margin at
the fifth, and still further laterally at the sixth car-
tilage, and then diverges laterally with increasing
severity at the seventh costal cartilage. The lateral
displacement of the left anteropleural border be-
tween the fourth and sixth costal cartilages forms
the cardiac notch.
The inferior borders of both pleural sacs diverge
laterally along the seventh costal cartilage and then
cross ribs 8, 9, and 10. Their lowest level is
reached about the middle of the eleventh rib in the
midaxillary line. From this point, they follow an
almost horizontal course, cutting across the twelfth
rib to meet the posterior border at the twelfth tho-
racic vertebra. The posterior pleural borders as-
cend alongside or in front of the bodies of the
thoracic vertebrae until they diverge superiorly
near the pulmonary cupola.
Knowledge of the surface relations of lobes and
fissures of the lungs is important in percussion,
auscultation, roentgen evaluation of the pulmonary
field, and the decision where to make the thoracic
incision (see Figs. 2-8 and 2-9). The horizontal
fissure of the right lung (which separates the right
upper from the right middle lobe) lies under the
fourth rib anteriorly and continues horizontally lat-
erally so that it cuts underneath the downsloping
fifth rib at the anterior axillary line. The oblique
fissure of the right lung (which separates the right
middle from the right lower lobe) lies under the
sixth rib. The oblique fissure of the left lung
(which separates the left upper lobe and lingula
from the left lower lobe) also lies under the sixth
rib.
III. THE AIRWAY AND LUNG
1 2 4 5
A. Trachea
In adults, the trachea extends from its attach-
ment to the lower end of the cricoid cartilage, at
the level of the sixth cervical vertebra, to its ter-
mination at the bronchial bifurcation at the level
of the fifth thoracic vertebra. The adult trachea is
approximately 6 inches long, of which 2 inches lie
above the suprasternal notch; this portion is some-
22 Thoracic Anatomy
Horizontal
fissure
Oblique
fissure
Oblique
fissure
Costodiaphragmatic
recess
Figure 28 Anterior view of the thorax showing the relationship of the pleura and lungs to the thoracic wall. The oblique fissure
runs nearly parallel to the sixth rib. The space between the inferior border of the lung at end exhalation and the inferior pleural
surface (dashed line) is called the costodiaphragmatic recess. The top of the lung is called the cupola. The ribs are numbered
according to their vertebral origin.
Thoracic Anatomy 23
Oblique
fissure
Cupola
Horizontal
fissure
Costodiaphragmatic
recess
Figure 2-9 Lateral view of the thorax showing the relationship of the pleura and lungs to the thoracic wall. The oblique fissure
runs nearly parallel to the sixth rib. The space between the inferior border of the lung at end exhalation and the inferior pleural
surface (dashed line) is the costodiaphragmatic recess. The top of the lung is termed the cupola. The ribs are numbered according
to their vertebral origin.
24 Thoracic Anatomy
what greater (nearly 3 inches) when the neck is
fully extended. The diameter of the trachea corre-
lates with the size of the subject, and a good work-
ing rule is that it has the same diameter as the
patient's index finger.
6
The patency of the trachea
is due to a series of 16 to 20 C-shaped cartilages
joined vertically by fibroelastic tissue and closed
posteriorly by the unstriped trachealis muscle (pos-
terior membrane). The cartilage at the tracheal bi-
furcation is the keel-shaped carina, which is seen
as a very obvious sagittal ridge when the trachea
is inspected bronchoscopically. A finding of the
sharp edge of the carina becoming flat usually
denotes enlargement of the hilar lymph nodes or
gross distortion of the pulmonary anatomy by fi-
brosis, tumor, or some other pathology.
The trachea lies exactly in the midline in the
cervical part of its course (but within the thorax it
is deviated slightly to the right by the arch of the
aorta). Tracheal relations to other organs in the
neck are shown in Figure 2-10. Anteriorly, the
trachea is bounded by muscles, pretracheal fascia,
and thyroid gland; laterally, by the pretracheal fas-
cia and the carotid sheaths; and posteriorly, by the
esophagus and vertebral bodies. The esophagus is
in the midline between the trachea and vertebral
bodies. Tracheal and esophageal relations in the
superior mediastinum are discussed in section IV
and are shown in Figure 2-11.
B. Main Bronchi
The adult trachea bifurcates at the level of the
fifth thoracic vertebra into right and left main-stem
bronchi. The right main bronchus is shorter, wider,
and more vertically placed than the left main-stem
bronchus. The right main-stem bronchus is shorter
than the left because it gives off its upper lobe
bronchus after a course of only 1.5 to 2.0 cm (but
may be shorter or even nonexistent [i.e., the right
upper lobe may arise from the trachea]),
7
whereas
the left main-stem bronchus gives off its upper
lobe bronchus after a course of 5.0 cm. The right
main-stem bronchus is wider than the left because
it supplies a larger lung, and it is more vertically
placed (at 25 to the vertical compared with 45
on the left) because the left bronchus has to extend
laterally behind the aortic arch to reach its lung
hilum. Obviously, inhaled foreign bodies or a
Pretracheal fascia
Thyroid gland
Trachea
Esophagus
Neck muscles
Pre-Vertebral fascia
Recurrent
Laryngeal
Nerve
sympathetic chain
Carotid sheath
(Containing Common carotid
artery, Internal jugular vein,
and Vagus nerve)
Figure 2-10 Transverse section of the neck through C-6 showing the fascial planes and the contents of the pretracheal fascia.
The trachea begins at this level.
Thoracic Anatomy 25
Thymus
Figure 2-11 Cross section of the thorax at the level of tracheal bifurcation (usually fifth thoracic vertebra). The pulmonary artery
(not shown) lies just caudad to this level.
D. Lung Lobes and Fissures
(see Figs. 2-8 and 2-9)
The right lung is composed of three lobesthe
upper, middle, and lowerand is the larger of the
two lungs. The left lung is composed of two lobes,
the upper and lower. Two fissures are present on
the right. The oblique fissure (major fissure) sepa-
rates the lower lobe from the upper and middle
lobes, and the horizontal fissure separates the up-
per from the middle lobe. On the left is the single
major oblique fissure dividing the upper and lower
lobes.
E. Bronchopulmonary Segments
Each lobe of the right and left lungs is subdi-
vided into several individual anatomic units, the
bronchopulmonary segments. The general pattern
is that of 18 segments10 in the right lung and 8
in the left lung. Each bronchopulmonary segment
has its own bronchus, artery, and vein and consists
of all of the lung that is distal to the third bifurca-
bronchial aspirating catheter is far more likely to
go to the wider and more vertically placed right
bronchus than the narrower and more obliquely
placed left main-stem bronchus.
C. Hilar Anatomy
From an anatomic surgical standpoint, knowl-
edge of hilar anatomy is extremely important
(Figs. 2-12 and 2-13). On the right side, the azy-
gos vein marks the uppermost aspect of the hilar
structures. From top to bottom on the right are the
right main-stem bronchus, the right main pulmo-
nary artery, and then the right superior pulmonary
vein. On the left side, the aortic arch marks the
superior aspect of the hilum, and from top to bot-
tom there is first the left main pulmonary artery,
the left main-stem bronchus, and then the left su-
perior pulmonary vein. In both hilar regions, the
most anterior structure is the superior pulmonary
vein, behind which is the main pulmonary artery,
and the most posterior structure is the main-stem
bronchus.
26 Thoracic Anatomy
Main stem
bronchus
Upper
lobe
Lower
lobe
Upper lobe a
Azygos v.
Lower
lobe a
Middle
lobe a.
Lower
lobe a.
Right
pulmonary a
Upper lobe
Superior
pulmonary v.
Middle lobe v.
nferior
pulmonary v.
A B C
Figure 2-12 Anatomy of the right hilum. A, Bronchi. B, Bronchi and pulmonary arteries. C, Bronchi, pulmonary arteries, and
veins. From top to bottom are azygos vein, bronchus, pulmonary artery, and superior pulmonary vein. From anterior to posterior
are the superior pulmonary vein, pulmonary artery, and bronchus (a. = artery, v. = vein).
tion. The bronchopulmonary segment names and
numbers as seen from the conducting airway view
are shown in Figure 2-14, and as seen from an
external lung view, they are shown in Figures 2-
15 and 2-16.
F. The Air Spaces
The tracheobronchial tree has 23 generations of
branches. The successive subdivisions of the bron-
chial tree are bronchi, bronchioles, respiratory
bronchioles, alveolar ducts, atria, alveolar sacs,
and alveoli (Fig. 2-17). From the second to the
eleventh generation, the airways are called bron-
chi, and from the twelfth to approximately the
seventeenth generation the airways are called
bronchioles. There is a change in name of the
airways at the eleventh generation because an im-
portant anatomic change occurs at this level: Car-
tilage disappears from the walls of these 1-mm
diameter air passages, and structural rigidity ceases
to be the principal factor in maintaining patency.
Fortunately, at this level the air passages leave
their fibrous sheath and become embedded directly
in the lung parenchyma. Elastic recoil of the alveo-
lar septa is then able to hold the air passages open
like the guy ropes of a bell tent. Similarly, the
patency of the alveolar duct is also maintained by
the retraction of the surrounding alveolar septa.
The caliber of airways and alveolar ducts beyond
the eleventh generation is, therefore, mainly influ-
enced by lung volume. The destruction of alveolar
septa in individuals with emphysema renders the
bronchioles and alveolar ducts more liable to clo-
Left
Upper division
Upper
lobe
Lower lobe
Upper lobe v.
Lower (lingular)
division
Lower
lobe a.
Upper lobe a.
Lingular a.
Superior pulmonary v,
Lingular
Inferior
pulmonary v.
A B C
Figure 2-13 Anatomy of the left hilum. A, Bronchi. B, Bronchi and pulmonary arteries. C, Bronchi, pulmonary arteries, and
veins. From top to bottom are aortic arch, pulmonary artery, bronchus, and superior pulmonary vein. From anterior to posterior are
superior pulmonary vein, pulmonary artery, and bronchus (a. = artery; v. = vein).
Thoracic Anatomy 27
Bronchopulmonary Segments
Figure 2- 14 The bronchopulmonary segments. There are 10 on the right and eight on the left. On the left, segments l and 2 are
combined into one apical posterior segment, and segments 7 and 8 are combined into one anterior medial segment (ANT =
anterior; LAT = lateral; MED = medial; POST = posterior).
Post.
Right Lung Bronchopulmonary Segments
Lateral View Medial View
Ant. Post.
Figure 2-15 The bronchopulmonary segments of the right lung as seen from lateral and medial views. The key of the numbering
is given in Figure 2-14 (ANT = anterior; POST = posterior).
28 Thoracic Anatont)
Left Lung Bronchopulmonary Segments
Lateral View Medial View
Ant. Post. Ant.
Figure 2-16 The bronchopulmonary segments of the left lung as seen from a lateral view and a medial view. The key of the
numbering is given in Figure 2-14 (ANT = anterior; POST = posterior).
sure. Each bronchiole with its subdivisions is
termed a primary lung lobule.
The respiratory bronchioles occur at the seven-
teenth generation. All generations above the respi-
ratory bronchioles are considered to be in the con-
ductive zone, and all generations below the
respiratory bronchioles are considered to be in the
transition and respiratory zones (Table 2-1).
The 23 Generations of the Bronchial Tree
(generation #)
CARTILAGE
PRESENT
NO CARTILAGE
PRESENT
Bronchi
(2-11)
Bronchiole
(12-17)
Respiratory
bronchiole
(17-18)
Alveolar ducts
(19-20)
Atrium
(21)
Alveolar sac
(22)
Alveolus
(23)
Figure 2-17 The successive sub-
divisions of the bronchial tree are
bronchi (2nd to Nth generation),
bronchioles ( 12th to 17th genera-
tion), respiratory bronchioles (I7th to
18th generation), alveolar ducts ( 19th
to 20th generation), atrium (21 st gen-
eration), alveolar sacs (22nd genera-
tion), and alveoli (23rd generation).
Thoracic Anatomy
29
Table 2-1 APPROXIMATE DISTRIBUTION OF TOTAL LUNG VOLUME IN MILLILITERS FOR AN
ADULT HUMAN LUNG AT THREE-FOURTHS TOTAL LUNG CAPACITY*
Compartments
Zones
(Generation) Air Channels Tissue Blood
Conducting Bronchi Walls Arteries Veins
(17 and above) 170 Septa 150 250
Fibers
Transition Respiratory bronchioles Lymph Arterioles Venules
(18-21) Alveolar ducts
1500 200 60 120
Respiratory Alveoli Barrier Capillaries
(22-23) 3150 150 140
"Total lung volume = 5700 ml
At each dichotomous division the diameter of
the airway decreases but at a rate that is less than
the increase in the number of airways, so that the
summed cross-sectional area at any given level
increases and the airway resistance decreases
down the airway tree. Reduction in diameter at
each division below the bronchiole level is less
than before, so that there is a very rapid increase
in the summed cross-sectional area with respect to
distance down the airway in the distal part of the
airway. In the adult human, the surface area of the
lungs normally in contact with the alveolar-gas
phase is of the order of 80 10
4
cm
2
, and that in
contact with the blood is estimated to be between
50 and 70 10
4
cm
2
.
8
The approximate distribu-
tion of total lung volume for the adult human lung
is shown in Table 2-l.
y
Aside from the columnar ciliated cells, the epi-
thelium of the proximal airway is dominated by
the mucus-secreting goblet cells and the distal air-
way by the surface tension-reducing material se-
creting Clara cells.
10
The surface tension-reducing
material produced by the Clara cell in the distal
bronchial tree is consistent in appearance with sur-
factant. Smokers have been demonstrated to have
a decrease in the number of distal Clara cells and
an increase in the number of distal mucus-produc-
ing goblets cells (normally predominantly located
in the proximal bronchial tree)." The hypothesis
has been advanced" that the disappearance of
Clara cells and their replacement by goblet cells
may result in the bronchioles becoming lined by a
film of mucus. Replacement of the surfactant layer
by a film of mucus would render the bronchioles
unstable, and they would close unduly easily and
open with greater difficulty, leading to functional
obstruction. The alveolar surfaces are lined with a
layer of water and a surface tension-reducing ma-
terial called surfactant (see chapter 3 for the phys-
iology of surfactant). The type II alveolar pneu-
mocytes are responsible for the secretion of
surfactant. The alveolar water/surfactant lining
layer is continuous with the tracheobronchial tree
surface lining layer.
The tracheobronchial tree and pulmonary com-
partment are ciliated down to the respiratory bron-
chioles. Cilia propel the airway fluids and particles
trapped in the airway fluids because the cyclic
movements that the cilia have are asymmetric.
There are two active parts in the ciliary beat cycle:
In the effective (or power) stroke of the cycle, the
cilium remains fully extended and moves through
an arc in a plane approximately perpendicular to
the cell surface, whereas in the recovery (or prep-
aratory) stroke, a bend is propagated along the
length of the cilium from base to tip, and the
cilium swings around near the surface of the cell
to reach the starting position for the next effective
stroke (Fig. 2-18).'" Normally, approximately 10
ml of mucus is expelled each day from the trachea
of a healthy adult human, which represents a layer
10 deep propelled continuously up the trachea
at about 0.2 mm/sec. It is unlikely that the inter-
action of the airflow in the respiratory airways
with the mucus makes a significant contribution to
mucus transport under normal circumstances, be-
Figure 2-18 The beat cycle of a
typical fluid propelling cilium con-
sists of two parts: an effective stroke
in a plane perpendicular to the sur-
face of the airway and a recovery
stroke in a plane parallel to the sur-
face of the airway. (Modified from
Sleigh MA, Blake JR. Liron N: The
propulsion of mucus by cilia. Am
Rev Respir Dis 137:726-741. 1988.)
Recovery
stroke
Effective
stroke
30 Thoracic Anatomy
cause airflow only reaches a maximal rate of about
10 m/sec in the narrowest part of the nose and falls
from 1 m/sec in the trachea to 10 mm/sec in the
respiratory bronchioles. However, during cough,
gases may be expelled from the mouth at 10 L/s
or more, implying gas velocities well in excess of
10 m/sec, even in narrow airways, and estimated
speeds in regions of airway compression range as
high as 250 m/sec, which is three quarters of the
speed of sound. If coughing can detach mucus
from the lining of the tracheobronchial tree, then
the airflows involved in coughing would be much
more effective than cilia in propelling fluid along
the airway walls, at least as far as the first through
the twelfth generations of airways are concerned.
12
G. Collateral Ventilation
There are several unusual bronchiolar and/or al-
veolar ventilation pathways, and these are collec-
tively termed collateral ventilation.
13
There are
four known pathways of collateral ventilation (Fig.
2-19). First, interalveolar communications (pores
of Kohn) exist in most species and may range from
8 to 50 per alveolus. In human lungs, the average
number of pores per alveolus ranges from 13 to
21,
14
and the number of pores is directly propor-
tional to age and the development of obstructive
lung disease. The average length of the major axes
is 7 to 19 . The distribution of the size of the
pores is uniform regardless of their location within
the alveolus and the size of the alveoli. Second,
there are respiratory bronchiole-to-alveolar com-
munications (channels of Lambert). The size of
these communications range from "practically
closed" to 30 in diameter.
12
Third, there are
connections between respiratory bronchioles to
terminal bronchioles from adjacent lung segments
(channels of Martin) in healthy dogs, in normal
humans, and in patients with lung disease.
13
I5
These collateral ventilation pathways are approxi-
mately 120 in diameter.
13 I5
Fourth, the func-
tional characteristics of interlobar collateral venti-
COLLATERAL VENTILATION PATHWAYS
Lobar
bronchus
Terminal conducting
bronchiole to respiratory
bronchiole communication
(channels of Martin)
Respiratory
bronchiole to alveolar
communications
(Bronchioloalveolar
channels of Lambert)
Alveolus
Interalveolar communications
(Pores of Kohn)
Figure 2-19 Four different types of collateral ventilation: lobe to lobe (interlobar), conducting bronchiole to respiratory bron-
chiole (interbronchiolar channels of Martin), respiratory bronchiole to aveolar (bronchioloalveolar channels of Lambert), and
alveolar to alveolar (interalveolar pores of Kohn). The left side shows the general level of subdivision of the bronchial tree where
the collateral communication takes place.
Thoracic Anatomy 31
lation through interlobar connections have recently
been described in dogs,
16
and these interlobar con-
nections have been observed in humans.
17
The role
of these four collateral ventilation pathways may
be to prevent or to minimize atelectasis and alveo-
lar hypoxia. Because the bronchioloalveolar, inter-
bronchiolar, and interlobar channels have a mus-
cular wall of their own, a regional control of
collateral ventilation is very possible. Figure 2-20
indicates how the level of airway obstruction de-
termines how the various collateral ventilation
pathways are utilized.
18
The physiology and deter-
minants of collateral ventilation are discussed in
chapter 3.
H. Pulmonary Arteries and Veins
19
The pulmonary artery and main-stem bronchus
of each lung are next to one another as they enter
the parenchyma of the lung. As these structures
enter the lung, they carry with them an invagina-
tion of the visceral pleura as a connective tissue
sheath. The pulmonary artery branches in similar
fashion to, and follows very closely, along the
dorsolateral surface, the bronchial tree. Distal to
the bronchiolar level (eleventh generation), the
penetration of pleural tissue stops, and both the
arteries and bronchi are attached directly to the
lung substance.
As in the bronchial tree, arterial branches de-
crease in diameter at each dichotomy down to the
precapillary branches, whereas the summed cross-
sectional area increases (see section F for surface
areas). Although the two trees are similar, the ar-
teries branch somewhat more profusely than the
bronchi do, especially peripherally, where they
give off branches to supply alveoli on respiratory
bronchioles as well as immediately adjacent al-
veoli of neighboring alveolar sacs. Although there
is a progressive reduction in arterial diameter from
the hilus to alveoli, the greatest fractional reduc-
tion occurs in the arterial segments that accom-
pany the respiratory bronchioles and alveolar
ducts.
8
These arterial segments, which contain two
to five circular smooth muscle layers, give rise to
pulmonary arterioles, which in turn contain one to
three circular layers of smooth muscle. The arteri-
oles arise at right angles and cross between the
alveolar sacs and alveoli for variable distances, as
precapillary branches, before dividing into alveolar
capillary nets that cross over the rounded surface
of the alveoli.
The pulmonary veins arise from capillaries at
alveolar duct junctions, as well as on respiratory
bronchioles, on the pleura, and in connective tissue
septa. Thus, the pulmonary veins drain three sets
of capillaries: alveolar, bronchial, and pleural.
Those veins that start within the acinus course
centrifugally to the periphery of the lobule. There
they join veins running between the lobules at
right angles. The interlobular veins are therefore
situated away from the conducting airways and
arteries. Unlike arteries, veins do not have a peri-
vascular space, so that veins within the lung are
connected directly to the lung substance and are
held open by elastic forces. The veins only come
together with the arteries and airways as these
three structures approach the hilum. Casts of arte-
ries and veins show that the venous system has a
larger volume than the arterial system (ratio 2:1;
see Table 2-1) and at any given level has a larger
cross-sectional area. The larger cross-sectional
area of the veins results in a very low resistance
system, which can still function with the low driv-
ing pressure available.
The Level of Airway Obstruction
Determines Which Collateral Ventilation
Pathways Are Utilized
1.6 mm 2.4 mm
4.8 mm
Figure 2-20 A model of collateral ventilation in dog lungs after airway occlusion by beads of different sizes: (A) 1.6 mm, (B)
2.4 mm, (C) 4.8 mm. This model suggests how collateral ventilation may become less effective the more proximal the obstruction.
(From Lee LN, Ueno O, Wagner PD, West JB: Pulmonary gas exchange after multiple airway occlusion by beads in the dog. Am
Rev Respir Dis 140:1216-1221, 1989. Used with permission.)
32 Thoracic Anatomy
The intimai lining of all vascular channels is
formed by endothelium, and the adventitia is
formed from vascular fibroblasts. The medial or
contractile coat is formed by the vascular smooth
muscle cell or one of its precursors, the interme-
diate cell or pericyte.
20
Along the muscular arteries
and veins, the completely muscular coat, with its
internal and external elastic lamina, disappears
well before the capillary bed (Fig. 2-21). Along
some pathways, it is no longer apparent in arteries
or veins smaller than 150 in diameter. Along
others, it is still apparent in arteries as small as 30
. In the capillary, the contractile cell is the
pericyte, but this cell is also found upstream and
downstream in vessels that, although larger, have
a wall resembling that of a capillary (nonmuseular
artery or vein). The vessels between the muscular
and nonmuseular sections have a partial muscular
structure (partially muscular artery or vein) in
which the contractile cells may be smooth muscle
or intermediate cells. The importance of the peri-
cyte and intermediate cells is that, in a variety of
injuries, they develop into cells that, by light mi-
croscopy, look like smooth muscle cells and be-
have as such in that they produce elastin.
8
The
change from pericyte or intermediate cell to a cell
that closely resembles a smooth muscle cell is a
striking change in phenotype. The response of the
pulmonary intermediate and pericyte cells is prob-
ably more important in disease than that of the
smooth muscle.
I. The Pulmonary Capillaries
7. Alveolar Sheet Arrangement
The pulmonary precapillaries and capillaries
usually arise at right angles from the pulmonary
arterioles. The pulmonary capillaries have been
classically described as long, branching tunnels
lying adjacent to the alveoli (Fig. 2-22.4).
2I
How-
ever, in reality, the capillaries spread out over the
alveolar surface in a complex, interconnecting pat-
tern, which is confined to or contained within the
two dimensions of the alveolar surface plane (Fig.
2-22/?).
2I
Because of these anatomic characteris-
tics, the geometry of the pulmonary capillaries has
been likened to the appearance of an underground
parking garage as viewed from within the garage
(Fig. 2-23).
2I
The top and bottom of the garage
are flat endothelial surfaces held together and sup-
MUSCULAR PARTIALLY
MUSCULAR
NON
MUSCULAR
Capillary
Figure 2-21 Diagrammatic representa-
tion of distal part of a pulmonary artery as
shown by light microscopy at the top and
bottom panels and by electronmicroscopy in
the middle panel. The complete coat of me-
dial muscle gives way to an incomplete coat
before disappearing in the walls of arteries
that are still larger than capillaries. The
smooth muscle cell (M) layer lies between
an internal and external elastic lamina. In the
nonmuseular part of a partially muscular ar-
tery, an "intermediate" cell (I) is seen: It is
surrounded by its basement membrane and
lies within a single elastic lamina. The prox-
imal part of the nonmuseular artery contains
a complete layer of these cells. Farther dis-
tally, the layer becomes incomplete, and a
solitary cell, the pericyte (P), is seen ablu-
minal to the endothelial cell (E). (From Reid
LM: Overview: The Third Grover Confer-
ence on the Pulmonary Circulation: The
control of cellular proliferation in the pul-
monary circulation. Am Rev Respir Dis
140:1490-1493, 1989. Used with permis-
sion.)
Thoracic Anatomy 33
Figure 2-22 A, A classical histologic view of a pulmonary capillary surrounded on each side by an alveolus; the capillary
appears as a long, branching tunnel. B, A histologic view of a pulmonary capillary bed cut tangential to the rounded surface of an
alveolus; the pulmonary capillaries (CAPS) are spread out in a complex, interconnecting pattern. Individual capillaries divide and
meet around connective tissue posts, which join the two possible endothelial plates together. See text for further explanation.
(Reproduced with permission from Fung YC: The microcirculation as seen by a red cell. Microvasc Res 10:246-264, 1975.)
ported by columns, or "posts," of connective tis-
sue (Figs. 2-22B and 2-23). An alveolus is on the
other side of each of the endothelial surfaces. The
connective tissue columns between the endothelial
plates have been called posts because of their
structural support function as well as "struts" be-
cause they require one alveolus on one side of the
capillary to expand if the alveolus on the other
side of the capillary collapses. The connective tis-
sue posts occupy the circular clear spaces between
the interlacing capillary network seen in Figure 2-
22B.
The endothelial surfaces form the boundaries of
the mainly two-dimensional capillary, and blood
flow through this two-dimensional world has been
called sheet flow.
22
In this model, red blood cells
must wind their way around and among the posts,
much as a car would negotiate the garage in
bumper-to-bumper traffic (see Fig. 2-23). The
sheet flow theory of the pulmonary microcircula-
tion is consistent with experimental data and, in
brief, disassociates resistance from flow, because
these two terms are normally related by Poi-
seuille's formula for fluid dynamics in a cylinder.
22
The precise flow pattern from post to post has been
modeled as a function of the perfusion pressure.
2
'
The sheet flow theory suggests that the capillar-
ies may occupy up to half the surface of an alveo-
lar wall and thereby expose an enormous capillary
surface area to the alveolar gas for gas exchange.
Previous estimates in humans have suggested that
alveolar blood vessels (primarily capillaries) have
a luminal surface area on the order of 80 m
2
.
24
However, these early calculations of surface area
assumed a more or less regular rounded shape of
capillaries having smooth luminal surfaces. It is
now evident that the capillary surfaces are not
smooth but are covered with irregular complex
projections.
25
The projections are approximately
300 nm in diameter and may reach 3000 nm in
length. Some projections come to blunt ends.
whereas others bud, branch, or reflect back on the
main body of the cell. The size and density of the
endothelial projection meshwork are such that an
eddy flow of cell-free plasma occurs along the
endothelial lining cell.
26
This feature has extremely
important implications for the exchange of metab-
olites between endothelium and blood (see the sec-
tion Pulmonary Metabolism and Synthesis in
chapter 3).
26
The endothelial cells also contain a
large population of plasmalemmal (pinocytotic)
vesicles (Fig. 2-24), many of which communicate
freely with vascular lumen (see figures in Pietra
27
and Szidon and colleagues
28
). These endothelial
vesicles and cavities function both as mass carriers
of fluid and solutes across the endothelium and as
34 Thoracic Anatomy
Figure 2- 23 Imaginary view of the inside of a pulmonary capillary as it might appear to a red blood cell. The view is similar to
that seen when one stands inside a modern underground parking garage. Endothelial plates form the top and bottom of the garage,
and the plates are held together and supported by connective tissue columns or "posts." On the other side of each of the endothelial
surfaces is an alveolus, which is not shown. (Reproduced with permission from Fung YC: The microcirculation as seen by a red
cell. Microvasc Res 10:246-264, 1975.)
generators of transendothelial channels by fusion
and fission with each other and with both endothe-
lial domains (vascular and tissular). Therefore, ow-
ing to the presence of endothelial vesicles and the
endothelial projections into the lumen of the cap-
illary, it appears that the true surface area of cap-
illary endothelial cells (at least for metabolism) is
much larger than previously estimated.
Figure 2- 24 Electron photomicrograph of an alveolar cap-
illary membrane. (EN = endothelium; EP = epithelium; CAP
= capillary; ALV = alveolus; PV = plasmalemma vesicles
[pointed to by thin, dark arrows]; IS = interstitial space [indi-
cated by bars]; BM = basement membrane [indicated by mid-
dle-sized dark arrows]; Junc't = loose endothelial junction
[indicated by large open and dark arrows]). (Modified with
permission from Pietra GG: The basis of pulmonary edema
with emphasis on ultrastructure. In Thurlbeck WM [ed]: The
Lung: Structure, Function and Disease. Baltimore, Williams &
Wilkins Co., 1978, pp. 215-234.) See text for explanation and
function of the plasmalemma vesicles. See Figure 2-25 for a
schematic of the arrangement of endothelial cells, loose junc-
tions, basement membranes, interstitial space, and epithelial
cells shown in this figure.
Thoracic Anatomy 35
2. Ultrastructure of Pulmonary
Capillary-Interstitial Space-Alveolar
Area
A cross section through any capillary channel
shown in Figure 2-23 would reveal a vascular
channel lined by endothelium that usually contains
one or, at most, two or three red blood cells.
27
A
schematic of the ultrastructural appearance of an
alveolar septum is shown in Figure 2-25.
29
Capil-
lary blood is separated from alveolar gas by a
series of anatomic layers: capillary endothelium,
endothelial basement membrane, interstitial space,
epithelial basement membrane, and alveolar epi-
thelium (of the type I pneumocyte).
On one side of the alveolar septum (the thick,
upper [see Fig. 2-25], fluid- and gas-exchanging
side), the epithelial and endothelial basement
membranes are separated by a space of variable
thickness containing connective tissue fibrils, elas-
tic fibers, fibroblasts, and macrophages. This con-
nective tissue is the "backbone" of the lung pa-
renchyma and forms a continuum with the
connective tissue sheaths around the conducting
airways and blood vessels. Thus, the pericapillary
perialveolar interstitial space is continuous with
the interstitial tissue space that surrounds terminal
bronchioles and vessels, and both spaces constitute
the connective tissue space of the lung. There are
no lymphatics in the interstitial space of the alveo-
lar septum. Instead, lymphatic capillaries first ap-
pear in the interstitial space surrounding terminal
bronchioles, small arteries, and veins.
The opposite side of the alveolar septum (the
thin, down [see Fig. 2-25], gas-exchanging only
side) contains only fused epithelial and endothelial
basement membranes. The interstitial space is thus
greatly restricted on this side owing to fusion of
the basement membranes. Interstitial fluid cannot
separate the endothelial and epithelial cells from
one another, and as a result the space and distance
barrier to fluid movement from the capillary to
alveolar compartment is reduced and is composed
only of the two cell linings with their associated
basement membranes.
30
31
Between the individual endothelial and epithe-
lial cells are holes or junctions that provide a po-
tential pathway for fluid to move from the intra-
vascular space to the interstitial space and finally
from the interstitial space to the alveolar space.
The junctions between endothelial cells are rela-
tively large and are therefore termed "loose"; the
ALV
EPI
BM
BM
ENDO
RBC
ENDO
BM
EPI
ALV
Figure 2-25 Schematic summary of the ultrastructure appearance of the pulmonary capillary. (RBC = red blood cell; ENDO
= endothelium; BM = basement membrane; IS = interstitial space; EPI = epithelium; LJ = loose junction; TJ = tight junction;
ALV = alveolus.) The upper side of the capillary has the endothelial and epithelial basement membranes separated by an interstitial
space, whereas the lower side of the capillary contains only fused endothelial and epithelial basement membranes. The dashed
arrows indicate a potential pathway for fluid to move from the intravascular space to the interstitial space (through loose junctions
in the endothelium) and from the interstitial space to the alveolar space (through tight junctions in the epithelium). (Modified with
permission from Fishman AP: Pulmonary edema: The water-exchanging function of the lung. Circulation 46:390-408, 1972, and
by permission of the American Heart Association, Inc.)
36 Thoracic Anatomy
junctions between epithelial cells are relatively
small and are therefore termed "tight." Pulmo-
nary capillary permeability is a direct function of
and essentially equivalent to the size of the holes
in the endothelial and epithelial linings (see chap-
ter 3).
J. Lymphatic System
32
A superficial lymphatic plexus drains the vis-
ceral pleura, and a deep plexus, lying alongside
the pulmonary vessels, drains the bronchi but does
not reach beyond the alveolar ducts into the more
distal spaces. Consequently, these microscopic
vessels have been called "juxta-alveolar lymphatic
capillaries." Both of these lymphatic plexuses
drain into progressively more proximal lymph
node stations (Fig. 2-26). The stations, in se-
quence, consist of pulmonary lymph nodes, the
bronchopulmonary lymph nodes placed at the
points of bifurcation of the larger bronchi, tracheo-
bronchial lymph nodes, right and left paratracheal
(mediastinal) lymph nodes, and scalene and deep
cervical lymph nodes. In an attempt to better de-
fine prognosis with lung cancer, the lymph node
stations have been numbered 1 through 13, with 1
being the highest mediastinal and 13 being the
most peripheral or segmental node station.
33
The
stations of the lymph chain on either side drain
into right and left lymphatic vessels. Most often,
the right and left lymphatic vessels open directly
and independently into the junction between the
internal jugular and subclavian veins on either
side. However, the right vessel may drain into the
main right lymphatic duct, and the left vessel may
empty into the thoracic duct.
Malignant disease in the right lung spreads
mainly up the lymph node chain on the same
side.
32
The superior tracheobronchial nodes, being
the regional lymph node station, are readily in-
volved. From there, the lymphatic pathway leads
up to the paratracheal and finally to the right sca-
lene or inferior deep cervical nodes. In contrast,
3. Superior
tracheobronchia
lymph nodes
Right
bronchus
Right upper
lobe bronchus
Right middle
lobe bronchus
1. Pulmonary lymph
nodes
5. Scalene nodes and
deep cervical nodes
4. Tracheal lymph
nodes
Trachea
Left bronchus
Left upper
lobe bronchus:
superior part
lingular
part
Left lower
lobe bronchus
3. Inferior
tracheobronchial
lymph nodes
Figure 2-26 The trachea, bronchi, and associated lymph nodes. Lymph (lows from l (pulmonary lymph nodes) to 2 (broncho-
pulmonary lymph nodes) to 3 (inferior and superior tracheobronchial lymph nodes) to 4 (tracheal lymph nodes) to 5 (scalene nodes
and deep cervical nodes).
Right lower lobe bronchus
Bronchopulmonary
lymph nodes
Thoracic Anatomy 37
when malignant disease involves the left lung,
contralateral lymphatic spread occurs frequently in
addition to proximal ipsilateral spread.
32
In addi-
tion, malignant lesions of the left lower lobe can
also spread subdiaphragmatically through para-
aortic nodes to the abdomen. The malignant dis-
semination of left lower lobe tumors by both con-
tralateral and subdiaphragmatic spread are a rea-
son, at least in part, why left lower lobe bronchial
carcinoma has the worst prognosis of all the lobar
carcinomas.
32
Unfortunately, any lung segment can drain di-
rectly to the mediastinal nodes. In an anatomic
study on 260 adult cadavers, direct lymph path-
ways from lung segments to the mediastinal nodes
were observed in 25 per cent of lungs.
34
Clinical
observations support these anatomic observations
that a given region can drain directly to the me-
diastinal lymph nodes. Systematic excision of me-
diastinal lymphatic chains has shown the presence
of mediastinal lymph nodes infiltrated with tumor
in many tumors of the NO type (i.e., without any
involvement or disease of intrapulmonary, hilar, or
mediastinal lymph nodes). In addition, it is not
rare for tumors considered to be as NO to recur in
the mediastinum. Furthermore, the anatomic stud-
ies demonstrate in rare cases (0.5-1 per cent) that
right lymph drainage can go directly to the right
jugular-subclavian venous junction and to the tho-
racic duct and that each lung can drain to the
contralateral side.
34
K. Bronchial Arteries and Veins
3
'
The bronchial arterial system arises from the
systemic circulation and accounts for approxi-
mately 1 per cent of the cardiac output. The ori-
gins of the bronchial arteries are variable and in-
clude the aorta, intercostal arteries, and.
occasionally, the subclavian or innominate arteries
(Fig. 2-27). On the right, the major source is from
the first or, at times, the second right aortic inter-
costal artery. Distally on the right as well as on
the left, the other bronchial arteries take their ori-
gin directly from the aorta, the level of origin
varying between the third and the eighth thoracic
vertebrae. The bronchial arteries to either side en-
ter the hilus of the lung and form a communicating
arc around the main bronchus. From here, the
main arterial divisions radiate along the longitudi-
nal axis of the major bronchi in close apposition
to the bronchial wall. The vessels follow the
course of the bronchus and divide, as do the bron-
chi. Discrete bronchial arterial branches can be
identified as far as third- to fifth-order bronchi.
Networks of intercommunicating vessels are often
present on the bronchial walls. Throughout their
course, the bronchial arteries give off the vasa
vasorum of the pulmonary arteries. The arterial
oxygen tension in the vasa vasorum may be re-
sponsible, in part, for determining pulmonary ar-
terial tone with stimulation of hypoxic pulmonary
vasoconstriction at moderate hypoxemia and of
hypoxic pulmonary vasodilation at profound hy-
poxemia.
36
The bronchial veins form two distinct systems.
The superficial bronchial veins, which constitute a
relatively small system, drain the main and lobar
bronchi and empty into the azygos vein on the
right side and into the hemiazygos and mediastinal
veins on the left side (see Fig. 2-27). The deep
venous effluent of segmental and distal bronchi,
which constitutes a relatively large system, emp-
THE BRONCHIAL CIRCULATION
Artery
Figure 2- 27 Schematic diagram of the bronchial circulation. The outflow system is predominantly the deep (pulmonary vein)
one (Ba = bronchial arteries).
38 Thoracic Anatomy
ties into the pulmonary veins. It has been esti-
mated that two thirds of the bronchial blood sup-
ply empty into the pulmonary veins (the deep
system), and one third empties into the bronchial
vein-azygos system (the superficial system). Be-
cause the deep bronchial circulation by-passes the
lungs and empties into the left heart without being
oxygenated, it is a normally present source of
right-to-left shunting (approximately 0.7 per cent
of the cardiac output).
Because all the blood returning from the tra-
cheobronchial tree empties directly into the heart
or one of the central veins, it might be expected
that, provided a drug were absorbed speedily, in-
stillation into the trachea would lead to rapid at-
tainment of high concentrations in the heart. In-
deed, several drugs are effective when given by
the intratracheal route: Naloxone, diazepam, lido-
caine, adrenaline, and atropine are notable exam-
ples.
37
The blood flow through the already present (pre-
formed) bronchial circulation suddenly increases
in acute lung disease, and there is also brisk local
development of new vessels in certain subacute
and chronic pulmonary and bronchial disease con-
ditions. The mechanism of the acute increase in
bronchial artery-to-pulmonary vein blood flow
during acute inflammation
38
is probably due to the
release of vasoactive mediators, such as the metab-
olites of archidonic acids.
39
40
The bronchial artery
to pulmonary vein anastomoses are a protective
mechanism against desaturation of peripheral
blood. Because an inflamed part of the lung can-
not ventilate satisfactorily, the bronchopulmonary
shunts thus permit saturated bronchial blood to
flow into the inflamed part of the lung; the bron-
chopulmonary shunts can disappear after treatment
of the inflammatory process.
41
The subacute and chronic conditions that cause
new vessel growth include prolonged infections,
cavitation, neoplastic vessel growth, and atelec-
tasis of the lung. The proliferation of the bronchial
circulation under the previously mentioned patho-
logic conditions is often responsible for repeated,
sometimes serious, episodes of hemoptysis. Of
course, right-to-left shunting will be increased
anytime there is increased bronchial blood flow.
Several other connections between the pulmo-
nary and bronchial systems have been demon-
strated in humans and in guinea pigs. Pulmonary
artery-to-bronchial artery, capillary-to-capillary,
and vein-to-vein as well as the bronchial artery-
to-pulmonary vein connections exist. Thus, almost
any part of the lung may potentially be supplied
by blood from either artery and be drained by
either set of veins. The bronchial artery-to-pul-
monary vein connections are very important, par-
ticularly in such lesions as pulmonary stenosis and
pulmonary embolism. Under these two circum-
stances, bronchial collaterals supply tissues distal
to blocked arteries and help to maintain a normal
ventilation-perfusion ratio.
Study of Figure 2-27 makes the physiologic
determinants of the amount of bronchial circula-
tion blood flow obvious. The directly proportional
determinants are the upstream systemic (bronchial)
artery pressure,
42
degree of midstream alveolar hy-
poxia,
43-45
and hypercapnia.
45
The inversely pro-
portional determinants are the downstream pul-
monary vein/left atrial pressure,
42
downstream
azygos vein pressure (i.e., central venous and right
atrial pressure)increased azygos vein pressure
decreases bronchial vein flow and increases anas-
tomotic vein flow
46
and the midstream alveolar/
airway (positive end-expiratory pressure) pres-
sure.
46-48
Thus, it is not surprising that lung infarc-
tion (i.e., ischemic death of pulmonary paren-
chyma) is rarely a complication of pulmonary
emboli except in patients with cardiopulmonary
disease (i.e., low upstream pressure and high
downstream pressure). In these patients, an inade-
quate response of the systemic-to-pulmonary
blood flow from the bronchial circulation in the
region distal to the embolic obstruction has been
suggested as the cause of the ischemic damage.
49
The effect of hypoxia and hypercapnia is thought
to be due either to a direct effect on the bronchial
vessel smooth muscle (similar to the response of
other systemic arteries)
43
or to one or more vaso-
dilating prostaglandins.
44
45
IV. THE MEDIASTINUM
1 24
A. Divisions of the Mediastinum
The mediastinum is a wide organ-filled septum
between the two pleural sacs. It is divided, some-
what arbitrarily for the purposes of description,
into an upper and lower part by a plane that ex-
tends from the sternal angle (the angle of Louis)
across the upper level of the pericardium to the
lower border of the fourth thoracic vertebra (Fig.
2-28). The upper part is named the superior me-
diastinum, and the lower part is subdivided into
three partsthe anterior mediastinum in front of
the pericardium, the middle mediastinum contain-
ing the heart and pericardium, and the posterior
mediastinum behind the pericardium (see Fig. 2-
28).
The superior mediastinum is bounded above by
the thoracic inlet, below by the plane of the supe-
rior limit of the pericardium, anteriorly by the ma-
nubrium, posteriorly by the upper four thoracic
vertebrae, and laterally by the mediastinal pleura
of the two lungs. It contains the aortic arch, the
Thoracic Anatomy 39
Suprasternal
notch
Angle of
Louis
Anterior
mediastinum
Xiphoid
Superior
mediastinum
Posterior
mediastinum
Middle
mediastinum
Figure 2- 28 The mediastinum is divided into a superior mediastinum and inferior mediastinum. The inferior mediastinum is
divided into an anterior, a middle, and a posterior mediastinum. See text for further explanation.
innominate artery, the thoracic portions of the left
common carotid and left subclavian arteries, the
innominate veins, the upper halves of the superior
vena cava and trachea and esophagus, the thoracic
duct, the remains of the thymus, the recurrent la-
ryngeal and phrenic nerves, and some lymph
nodes. The anterior mediastinum is bounded ante-
riorly by the body of the sternum and posteriorly
by the parietal pericardium, and it extends down-
ward as far as the diaphragm. The anterior medias-
tinum does not contain any vital structures. The
middle mediastinum is the broadest part of the
interpleural mediastinal septum and contains the
heart, which is enclosed in the pericardium, the
ascending aorta, the lower half of the superior
vena cava with the azygos vein opening into it, the
pulmonary artery dividing into its two branches,
the right and left pulmonary veins, and the phrenic
nerves. The posterior mediastinum is an irregularly
shaped mass running parallel with the vertebral
column, and because of the slope of the dia-
phragm, it extends caudally beyond the pericar-
dium. It is bounded anteriorly by the pericardium
and more caudally by the diaphragm, posteriorly
by the vertebral column from the lower border of
the fourth to the twelfth thoracic vertebrae, and on
either side by the mediastinal pleura. It contains
the thoracic part of the distending aorta, the azygos
and hemiazygos veins, the bifurcation of the tra-
chea and the two bronchi, the esophagus, the tho-
racic duct, and many large lymph nodes.
B. Mediastinal Relations of the Trachea,
Esophagus, Aorta, and Pulmonary Trunk
(for Cervical Tracheal and Esophageal
Relations see section III)
From the thoracic inlet to the tracheal bifurca-
tion, the thoracic esophagus remains in intimate
relationship with the posterior wall of the trachea
and the prevertebral fascia (see Figs. 2-10, 2-11
and 2-29). Just above the tracheal bifurcation, the
esophagus is to the right of the aorta. This ana-
tomic positioning can cause a notch indentation in
the esophageal left lateral wall on the barium swal-
low radiograph. Immediately below this notch the
esophagus crosses both the bifurcation of the tra-
chea and the left main-stem bronchus. From there
down it passes over the posterior surface of the
subcarinal lymph nodes and then descends over
the pericardium of the left atrium. From below the
40 Thoracic Anatomy
2nd Costal cartilage Internal mammary
artery and veins
Left phrenic nerve
Left vagus nerve
Aortic arch
Left recurrent
laryngeal nerve Esophagus
Figure 2-29 The thoracic trachea and esophagus and their relation to other mediastinal structures ai the level of the fourth
thoracic vertebra. See Figure 2-11 for a cross section at a slight!) lower level and Figure 2-10 for a cross section at a somewhat
higher level.
Superior
vena cava
Phrenic Nerve
Azygos vein
Right vagus
nerve
Trachea
Thoracic Anatomy 41
Right
common carotid
Right subclavian
artery
Innominate artery
Ligamentum
arteriosum
Superior
vena cava
Left common
carotid artery
Left subclavian
artery
Left bronchus
Pulmonary trunk
Heart
Figure 2-30 The relations of the great vessels to the trachea. See text for full explanation.
aortic arch, the esophagus lies to the right of and
then in front of the descending thoracic aorta, the
transition occurring at the level of the eighth tho-
racic vertebra. Posteriorly, the thoracic esophagus
follows the curvature of the spine and remains in
close contact with the vertebral bodies. From the
eighth thoracic vertebra downward the esophagus
moves anteriorly away from the spine and passes
through the esophageal hiatus of the diaphragm in
front of the aorta.
Figure 2-30 shows the relationship of the aorta,
the pulmonary arteries, the tracheal bifurcation,
and the esophagus to each other. The pulmonary
trunk, which arises from the right ventricle, lies at
first slightly in front of the ascending aorta and
then to its left side before dividing into the left
and right pulmonary arteries. The left pulmonary
artery is connected to the concavity of the arch of
the aorta by the ligamentum arteriosum, which
represents the obliterated ductus arteriosus of the
fetus. The ascending aorta becomes the arch of the
aorta at the level of the manubrial-sternal joint, so
that the arch lies entirely within the superior me-
diastinum. The arch passes backward and to the
left and gives off in turn the innominate artery
(which divides behind the sternal-clavicular joint
into the right subclavian and right common carotid
arteries) and the left common carotid and left sub-
clavian arteries. The innominate and left common
carotid arteries are located at first in front of the
trachea and then past to its right and left sides,
respectively.
REFERENCES
1. Ellis H. McLarty M: Anatomy for Anaesthetists. 2nd ed.
Philadelphia. F. A. Davis Co, 1968.
2. Gray H, Goss CM: Gray's Anatomy. 27th ed. Philadelphia.
Lea and Febiger. 1959.
3. Blevins CE: Anatomy of the thorax and pleura. In Shields
TW (ed): General Thoracic Surgery. Philadelphia. Lea and
Febiger. 1983, pp 43-60.
4. Moffat DB: Anatomical aspects of thoracic anesthesia. In
Mushin WW (cd): Thoracic Anesthesia. Oxford, Blackwell
Scientific Publications. 1963, pp 143175.
5. Shields TW: Surgical anatomy of the lungs. In Shields TW
(cd): General Thoracic Surgery. 2nd ed. Philadelphia. Lea
and Febiger. 1983. pp 61-71.
6. Fukuoka RH, Kelly JW. Franklin CM: Correlation between
ETT size, distal digit diameter and the penlington formu-
lae. Anesth Analg 72:S85. 199L
CHAPTER 3
General Respiratory Physiology
and Respiratory Function
During Anesthesia
I. Respiratory Physiology
A. Introduction
B. Normal (Gravity-Determined)
Distribution of Perfusion, Ventilation,
and the Ventilation-Perfusion Ratio
1. Distribution of Pulmonary
Perfusion
2. Distribution of Ventilation
3. Distribution of the Ventilation-
Perfusion Ratio
4. Pulmonary Interstitial Fluid
Kinetics (Zone 4)
C. Other (Nongravitational) Important
Determinants of Pulmonary Vascular
Resistance and Blood Flow
Distribution
1. Cardiac Output
2. Alveolar Hypoxia
3. Lung Volume
4. Alternate (Nonalveolar) Pathways
of Blood Flow Through the Lung
D. Other (Nongravitational) Important
Determinants of Pulmonary
Compliance, Resistance, and Lung
Volume
1. Pulmonary Compliance
2. Airway Resistance
3. Different Regional Lung Time
Constants
4. Work of Breathing
5. Lung Volumes, Functional
Residual Capacity, and Closing
Capacity
a. Lung Volumes and Functional
Residual Capacity
b. Airway Closure and Closing
Capacity
c. Relationship Between
Functional Residual Capacity
and Closing Capacity
E. Oxygen and Carbon Dioxide
Transport
1. Alveolar and Dead-Space
Ventilation and Alveolar Gas
Tensions
2. Oxygen Transport
a. Overview of Oxygen
Transport
b. Oxygen-Hemoglobin
Dissociation Curve
c. Effect of (VQ, on P
a
0
2
d. Effect of Q, and Vo
2
on C
a
0
2
e. Fick Principle
f. Relationship Between Oxygen
Delivery (Do
2
) and Vo
2
3. Carbon Dioxide Transport
4. Bohr and Haldane Effects
F. Lung-Heart Interactions
1. Spontaneous Ventilation
2. Positive-Pressure Ventilation and
PEEP
G. Pulmonary Vascular Reflexes
H. Pulmonary Metabolism and
Synthesis
1. Metabolism of Endogenous
Substances
2. Synthesis and Release of
Endogenous Substances
3. Handling of Exogenous Drugs
I. Pulmonary Host Defenses
J. Other Special Lung Functions
1. Reservoir for Left Ventricle
2. Circulatory Filtrative Function
3. Nutrition
4. Administration of Drugs Via the
Respiratory Tract for Systemic
Effect
II. Respiratory Function During
Anesthesia
A. Introduction
B. Effect of Anesthetic Depth on
Respiratory Pattern
C. Effect of Anesthetic Depth on
Spontaneous Minute Ventilation
D. Effect of Pre-Existing Respiratory
Dysfunction on the Respiratory
Effects of Anesthesia
E. Effect of Special Intraoperative
Conditions on the Respiratory Effects
of Anesthesia
F. Mechanisms of Hypoxemia During
Anesthesia
1. Malfunction of Equipment
a. Mechanical Failure of
Anesthesia Apparatus to
Deliver Oxygen to the Patient
b. Mechanical Failure of
Endotracheal Tube: Main-
Stem Bronchus Intubation
44 General Respiratory Physiology and Respiratory Function During Anesthesia
2. Hypoventilation (Decreased Tidal
Volume)
3. Hyperventilation
4. Decrease in Functional Residual
Capacity
a. Supine Position
b. Induction of General
AnesthesiaChange in
Thoracic Cage Muscle Tone
c. Paralysis
d. Light or Inadequate
Anesthesia and Active
Expiration
e. Increased Airway Resistance
f. Supine Position, Immobility,
and Excessive Intravenous
Fluid Administration
g. High Inspired Oxygen
Concentration and Absorption
Atelectasis
h. Surgical Position
i. Ventilation History (Rapid,
Shallow Breathing)
j. Decreased Removal of
Secretions (Decreased
Mucociliary Flow)
I. RESPIRATORY PHYSIOLOGY
A. Introduction
Understanding normal respiratory physiology is
prerequisite to understanding mechanisms of im-
paired gas exchange during any type of anesthesia
and surgery. Toward this end, the normal (gravity-
determined) distribution of perfusion and ventila-
tion, the major nongravitational determinants of
resistance to perfusion and ventilation, the trans-
port of the respiratory gases, special heart-lung
interactions, and the pulmonary reflexes and spe-
cial functions (defense and metabolic) are pre-
sented first in this chapter. These processes and
concepts are then utilized in the discussion of the
general mechanisms of impaired gas exchange
during anesthesia and surgery.
B. Normal (Gravity-Determined)
Distribution of Perfusion, Ventilation,
and the Ventilation-Perfusion Ratio
1. Distribution of Pulmonary Perfusion
Contraction of the right ventricle imparts kinetic
energy to the blood in the main pulmonary artery.
Most of the kinetic energy in the main pulmonary
artery is dissipated in climbing a vertical hydro-
static gradient, and the absolute pressure in the
pulmonary artery (P
pa
) decreases l cm H
2
0 per
centimeter of vertical distance up the lung (Fig.
5. Decreased Cardiac Output and
Increased Oxygen Consumption
6. Inhibition of Hypoxic Pulmonary
Vasoconstriction
7. Paralysis
8. Right-to-Left Interatrial Shunting
9. Differential Diagnosis of Cyanosis
10. Involvement of Mechanisms of
Hypoxemia in Specific Diseases
G. Mechanisms of the Hypercapnia and
Hypocapnia During Anesthesia
1. Hypercapnia
a. Hypoventilation
b. Increased Dead-Space
Ventilation
c. Increased Carbon Dioxide
Production
d. Inadvertent Switching Off of a
Carbon Dioxide Absorber
2. Hypocapnia
H. Physiologic Effects of Abnormalities
in the Respiratory Gases
1. Hypoxia
2. Hyperoxia (Oxygen Toxicity)
3. Hypercapnia
4. Hypocapnia
3-1). At some great enough height above the
heart, P
pa
becomes zero (atmospheric), and still
higher in the lung the P
pa
becomes negative.
1
In
this region alveolar pressure (P
A
) then exceeds P
pa
and pulmonary venous pressure (P ) (which is
very negative at this vertical height). Since the
pressure outside the vessels is greater than the
pressure inside the vessels, the vessels in this re-
gion of the lung are collapsed and there is no
blood flow (zone 1,
> P
pa
> P
pv
). Since there
is no blood flow, no gas exchange is possible, and
the region functions as alveolar dead space or
"wasted" ventilation. Little or no zone 1 exists in
the lung under normal conditions,
2
but the amount
of zone 1 lung may be greatly increased if P
pu
is
reduced, as in oligemic shock, or if P
A
is increased,
as in positive-pressure ventilation.
Further down the lung absolute P
pa
becomes
positive and blood flow will begin when P
pa
ex-
ceeds
(zone 2, P
pa
>
> P
pv
). At this vertical
level in the lung,
exceeds P
pv
, and blood flow
is determined by the mean P
pa
P
A
difference
rather than the more conventional P
pa
P
pv
differ-
ence (see the following).
3
The zone 2 blood flow-
alveolar pressure relationship has the same physi-
cal characteristics as a river waterfall flowing over
a dam (Fig. 3-2). The height of the upstream river
(before reaching the dam) is equivalent to P
pa
, and
the height of the dam is equivalent to P
A
. The rate
of water flow over the dam is only proportional to
the difference between the height of the upstream
river and the dam ( P
p a
- P
A
), and it does not matter
how far below the dam the downstream river bed
General Respiratory Physiology and Respiratory Function During Anesthesia
T h e Four Zones of the Lung
Figure 3-1 The distribution of blood flow in the upright lung. In zone I. alveolar pressure (P
A
j exceeds pulmonary artery
pressure (P
p
J. and no flow occurs because the intra-alveolar vessels are collapsed by the compressing alveolar pressure. In zone 2.
arterial pressure exceeds alveolar pressure, but alveolar pressure exceeds pulmonary venous pressure (P
p
J. Flow in zone 2 is
determined by the arterial-alveolar pressure difference (
.,-
) and has been likened to an upstream river waterfall over a dam (see
Fig. 3-2). Because P
pa
increases down zone 2 and
remains constant, the perfusion pressure increases, and flow steadily increases
down the zone. In zone 3. pulmonary venous pressure exceeds alveolar pressure, and flow is determined by the arterial-venous
pressure difference (P
pa
P
P
J, which is constant down this portion of the lung. However, the transmural pressure across the wall of
the vessel increases down this zone so that the caliber of the vessels increases (resistance decreases), and therefore flow increases.
Finally, in zone 4, pulmonary interstitial pressure becomes positive and exceeds both pulmonary venous pressure and alveolar
pressure (see Fig. 3-3). Consequently, flow in zone 4 is determined by the arterial-interstitial fluid pressure difference (P
pa
-P
lsl
.).
(Redrawn by permission from West JB: Ventilation/Blood Flow and Gas Exchange. 4th ed. Oxford, Blackwell Scientific Publica-
tions, 1985.)
(PpJ is. This phenomenon has various names, in-
cluding the "waterfall/' "Starling resistor,"
"weir," and "sluice" effect. Since mean in-
creases down this region of the lung but mean

is relatively constant, the mean driving pressure
( P
p a
-P
A
) increases linearly, and therefore mean
blood flow increases linearly.
However, it should be noted that respiration and
pulmonary blood flow are cyclic phenomena.
Therefore, absolute instantaneous P
pu
, P
pv
, and P
A
are changing all the time, and the relationships
among P
pa
, P
pv
, and
are dynamically determined

by the phase lags between the cardiac and respira-
tory cycles. In other words, the upstream and
downstream blood flows actually approach and
leave, respectively, the respiratory dam as a wave,
with the crest being equal to systolic pressure and
the trough being equal to diastolic pressure; the
respiratory dam is actually going up and down in
a manner dependent on whether positive- or nega-
tive-pressure ventilation is being used (Fig. 3-25).
Consequently, a given point in zone 2 may ac-
tually be in either a zone 1 or 3 condition at a
given moment depending on whether the patient is
in respiratory systole or diastole or cardiac systole
or diastole.
A simple numerical example further illustrates
this point (the example assumes the alveolar pres-
sure is constant). In a man of ordinary height, the
distance between the top (apex) and the bottom
(base) of the lung is about 30 cm. If the main
pulmonary trunk is midway between the apex and
the base in the erect position, there is a column of
blood 15 cm high between the pulmonary trunk
and arterioles in the apex and a similar column of
blood between it and arterioles in the base. A col-
Zone 2 Waterfall Phenomenon
Static Mean Pressure Condition
Dynamic Instantaneous Pressure Condition
Respiratory
Dam
Figure 32 See legend on opposite page
Chest Wall
Figure 3- 4 This schematic diagram of the lung within the
chest wall shows the tendency of the lung to assume a globular
shape because of the lung's viscoelastic nature. The tendency
of the top of the lung to collapse inward creates a relatively
negative pressure at the apex of the lung, and the tendency of
the bottom of the lung to spread outward creates a relatively
positive pressure at the base of the lung. Thus, pleural pressure
increases by 0.25 cm H
2
0 per centimeter of lung dependency.
(Modified with permission from Benumof JL: Respiratory
physiology and respiratory function during anesthesia. In
Miller RD (cd): Anesthesia. 2nd ed. New York, Churchill Liv-
ingstone, 1983. chapter 32.)
apical alveoli (there is an approximately fourfold
alveolar volume difference).
12
If the regional dif-
ferences in alveolar volume are translated over to
a pressure-volume curve for normal lung (Fig.
3-5), the dependent small alveoli are on the mid-
portion and the nondependent large alveoli are on
the upper portion of the S-shaped pressure-volume
curve. Since the different regional slopes of the
composite curve are equal to the different regional
lung compliances, dependent alveoli are relatively
compliant (steep slope), and nondependent alveoli
are relatively noncompliant (flat slope). Thus, the
majority of the tidal volume is preferentially dis-
tributed to dependent alveoli because they expand
more per unit pressure change than nondependent
alveoli.
3. Distribution of the Ventilation-
Perfusion Ratio
Figure 3-6 shows that both blood flow and ven-
tilation (both on the left-hand vertical axis) in-
crease linearly with distance down the normal up-
right lung (horizontal axis, reverse polarity).
1
'
Since blood flow increases from a very low value
and increases more rapidly than ventilation with
distance down the lung, the ventilation-perfusion
(V
A
/Q) ratio (right-hand vertical axis) decreases
rapidly at first and then more slowly.
The V
A
/Q ratio best expresses the amount o\'
ventilation relative to perfusion in any given lung
region. Thus, alveoli at the base of the lung are
somewhat overperfused in relation to their venti-
lation (V
A
/Q > 1). Figure 3-7 shows the calcu-
lated ventilation (V
A
) and blood flow (Q) in liters
per minute, V
A
/Q ratio, and the alveolar Po
2
. and
Pco
2
in mm Hg for horizontal slices from the top
(7 per cent of lung volume), middle (11 per cent
of lung volume), and bottom (13 per cent of lung
volume) of the lung.
14
It can be seen that the P
A
0
:
increases by more than 40 mm Hg from 89 mm
Hg at the base to 132 mm Hg at the apex, while
the Pco
:
decreases by 14 mm Hg from 42 mm Hg
at the bottom to 28 mm Hg at the top. Thus, in
Transpulmonary Pressure
cm H
2
0
Figure 3-5 Pleural pressure increases 0.25 cm H
2
0 every centimeter down the lung. The increase in pleural pressure causes a
fourfold decrease in alveolar volume. The caliber of the air passages also decreases as lung volume decreases. When regional
alveolar volume is translated over to a regional transpulmonary pressure-alveolar volume curve, small alveoli are on a steep (large
slope) portion of the curve, and large alveoli are on a flat (small slope) portion of the curve. Because the regional slope equals
regional compliance, the dependent small alveoli normally receive the largest share of the tidal volume. Over the normal tidal
volume range (lung volume increases by 500 ml from 2500 [normal FRC] to 3000 ml), the pressure-volume relationship is linear.
Lung volume values in this diagram relate to the upright position. (Modified with permission from Benumof JL: Respiratory
physiology and respiratory function during anesthesia. In Miller RD (ed): Anesthesia. 2nd ed. New York, Churchill Livingstone.
1983, chapter 32.)
Figure 3- 6 Distribution of ventilation and blood flow (left-hand vertical axis) and the ventilation-perfusion ratio (right-hand
vertical axis) in a normal upright lung. Both blood flow and ventilation are expressed in L/min per cent alveolar volume and have
been drawn as smoothed out linear functions of vertical height. The closed circles mark the ventilation-perfusion ratios of horizontal
lung slices (three of which are shown in Fig. 3-7). A cardiac output of 6 L/min and a total minute ventilation of 5.1 L/min were
assumed. (Reproduced with permission from West JB: Ventilation/Blood Flow and Gas Exchange. 4th ed. Oxford, Blackwell
Scientific Publications, 1985.)
keeping with the regional V
A
/Q, the bottom of the
lung is relatively hypoxic and hypercarbic com-
pared with the top of the lung.
Wagner and colleagues
15
described a method of
determining the continuous distribution of V
A
/Q
ratios within the lung based on the pattern of elim-
ination of a series of intravenously infused inert
gases. Gases of differing solubility are dissolved
in physiologic saline solution and infused into a
peripheral vein until a steady state is achieved (20
min). Toward the end of the infusion period, sam-
ples of arterial and mixed expired gas are col-
lected, and total ventilation and cardiac output are
measured. For each gas the ratio of arterial to
mixed venous concentration (retention) and the ra-
tio of expired to mixed venous concentration (ex-
cretion) are calculated, and retention-solubility and
excretion-solubility curves are drawn. The reten-
tion/excretion-solubility curves can be regarded as
a "fingerprint" of the particular distribution of
ventilation-perfusion ratios that give rise to it.
Figure 3-8A shows the type of distribution
found in young normal subjects breathing air in
the semirecumbent position.
16
The distributions of
both ventilation and blood flow are relatively nar-
row. The upper and lower 95 per cent limits shown
(vertical interrupted lines) correspond to V
A
/Q ra-
tios of 0.3 and 2.1, respectively. Notice these
young normal subjects had no blood flow perfus-
ing areas with very low V
A
/Q ratios nor did they
have any blood flow to unventilated or shunted
areas (V
A
/Q = 0) or unperfused areas (V
A
/Q
= oo). Figure 3-8 also shows alveolar Po
2
and
Pco
2
in respiratory units having different V
A
/Q
ratios. It can be seen that, within the 95 per cent
range of V
A
/Q ratios (0.3 to 2.1), the Po
2
ranges
from 60 to 123 mm Hg, whereas the corresponding
Pco
:
range is 44 to 33 mm Hg.
4. Pulmonary Interstitial Fluid Kinetics
(Zone 4)
Pulmonary edema is defined as the abnormal
accumulation of interstitial fluid in the lung. The
Figure 3-7 The ventilation-perfusion ratio (V
A
/Q) and the regional composition of alveolar gas. Values for the regional flow
(Q), ventilation (V
A
), P
02
, and P
ro2
are derived from Figure 3-6. P
N
, has been obtained by what remains from the total gas pressure
(which, including water vapor, equals 760 mm Hg). The volumes (vol [per cent]) of the three lung slices are also shown. Compared
with the top of the lung, the bottom of the lung has a low ventilation-perfusion ratio and is relatively hypoxic and hypercarbic.
(Reproduced with permission from West JB: Regional differences in gas exchange in the lung of erect man. J Appl Physiol 17893
1962.)
Figure 3-8 A, The average distribution of ventilation-perfusion ratios in young, semirecumbent normal subjects. The 95 per cent
range covers ventilation-perfusion ratios from 0.3 to 2.1 (between dashed lines). The corresponding variations of P
O2
and P
CO2:
in
the alveolar gas can be seen in B. (Reproduced with permission from West J13: Blood How to the lung and gas exchange.
Anesthesiology 41:124, 1974.)
abnormal accumulation of pulmonary interstitial
fluid can have a profound effect on the distribution
of pulmonary blood flow, ventilation, the ventila-
tion-perfusion relationship, and the efficiency of
gas exchange. This section reviews the mecha-
nisms of how pulmonary interstitial fluid is
formed, stored, and cleared.
The concepts of a continuous connective tissue
sheath-alveolar septum interstitial space and of a
negative interstitial space pressure gradient are
prerequisite to understanding interstitial fluid ki-
netics (Fig. 3-9). After entering the lung paren-
chyma, both the bronchi and arteries run within a
connective tissue sheath that is formed by an in-
vagination of the pleura at the hilum and that ends
at the level of the bronchioles (Fig. 3-9A). Thus,
there is a potential perivascular and peribronchial
space, respectively, between the arteries and the
bronchi and the connective tissue sheath. The neg-
ative pressure in the pulmonary connective tissue
sheath exerts a radial outward traction force on the
sheath. The radial traction creates a negative pres-
sure within the sheath, and the negative pressure
is transmitted to the bronchi and arteries, which
tends to hold them open and increase their diame-
ters (Fig. 3-9).
3
The alveolar septum interstitial
space is the space between the capillaries and al-
veoli (or, more precisely, the space between the
endothelial and epithelial basement membranes)
and is continuous with the interstitial tissue space
that surrounds the larger arteries and bronchi (Fig.
3-9A) (also see chapter 2). Studies indicate that
the alveolar interstitial pressure is also uniquely
negative but not as much as the negative interstitial
space pressure around the larger arteries and bron-
chi.
17
is a capillary filtration coefficient expressed in
ml/min/mm Hg/100 g. The filtration coefficient is
the product of the effective capillary surface area
in a given mass of tissue and the permeability per
unit surface area of the capillary wall to filter the
fluid. Under normal circumstances, and at a verti-
cal height in the lung that is at the junction of
zones 2 and 3, the intravascular colloid osmotic
pressure (about 26 mm Hg) acts to keep water in
the capillary lumen, and working against this
force, the pulmonary capillary hydrostatic pressure
(about 10 mm Hg) acts to force water across the
loose endothelial junctions into the interstitial
space. If these were the only operative forces, the
interstitial space, and consequently the alveolar
surfaces, would be constantly dry, and there would
be no lymph flow. In fact, alveolar surfaces are
moist, and lymphatic flow from the interstitial
compartment is constant (approximately 500
ml/day). This can be explained in part by the
(8 mm Hg) and in part by the negative
(approximately 8 mm Hg). Negative (subat-
mospheric) interstitial space pressure would pro-
mote, by suction, a slow loss of fluid across the
endothelial holes.
18
Recently, much has been learned about pulmo-
nary interstitial edema caused by a very negative
Indeed, extremely negative pleural (and
perivascular hydrostatic) pressure, such as might
occur in a vigorously spontaneously breathing pa-
tient with an obstructed airway (upper airway
masses [tumors, hematoma, abscess, edema], la-
ryngospasm [most common], strangulation, infec-
tious processes [epiglottis, pharyngitis, croup], and
vocal cord paralysis),
19-23
rapid re-expansion of
lung, and application of very negative pleural pres-
sure during thoracentesis,
24
can cause pulmonary
interstitial edema (Table 3-1). Other factors pos-
sibly involved in the generation of pulmonary
edema after obstructed breathing may include re-
lease of the obstruction to exhalation (loss of auto-
positive end-expiratory pressure [PEEP]) (which
may account for the often sudden appearance of
pulmonary edema after relief of the obstruction),
23
hypoxia-induced increases in P
i ns i de
caused by pre-
and postcapillary constriction,
21
25
26
increased cap-
illary permeability and myocardial depression, pul-
monary vascular engorgement caused by increased
Table 3-1 CAUSES OF VERY NEGATIVE
PULMONARY INTERSTITIAL
FLUID PRESSURE
PULMONARY EDEMA
1. Vigorous spontaneous ventilation against an obstructed
airway
a. Upper airway mass (tumor, hematoma, abscess, foreign
body, etc.)
b. Laryngospasm
c. Infection, inflammation, edema
d. Vocal cord paralysis
e. Strangulation
2. Rapid re-expansion of lung
3. Vigorous pleural suctioning (thoracentesis, chest tube)
Continuous Connective Tissue Sheath-
Alveolar Septum Interstitial Space
CT Sheath Containing the
Extra-Alveolar
Interstitial Space
Alveolar Septum
Containing the
ntra-Alveolar Interstitial
Space
Interstitial Fluid Moves From Alveolar Septum
To Connective Tissue Space Via Three Mechanisms
2. Valves
1. Negative Pressure
Gradient
3. Arterial Pulsations
Connective Tissue Sheath
Loose
Junctions
Alveolar Septum
Continuous Interstitial Space
k
Figure 3- 9 A, Schematic diagram of the concept of a continuous connective tissue (CT) sheath-alveolar septum interstitial space.
The entry of the main-stem bronchi and pulmonary artery into the lung parenchyma invaginates the pleura at the hilum, forming a
surrounding connective tissue sheath (heavy black line). The connective tissue sheath ends at the level of the bronchioles. The
space between the pulmonary arteries and bronchi and the connective tissue sheath constitutes the extra-alveolar interstitial space.
The connective tissue sheath-extra-alveolar interstitial space is continuous with the alveolar septum interstitial space. The alveolar
septum interstitial space is contained within the endothelial and epithelial basement membranes of the capillaries and alveoli,
respectively. B, Schematic diagram showing how interstitial fluid moves from the alveolar septum interstitial space to the connective
tissue interstitial space. The mechanisms are via a negative pressure gradient ("sump"), the presence of one-way valves in the
lymphatics, and the massaging action of arterial pulsations. CAP = capillary.
Lymphatic Fluid Kinetics
rounding terminal bronchioles and small arteries.
Interstitial fluid is normally removed from the al-
veolar interstitial space into the lymphatics by a
"sump" (pressure gradient) mechanism, which is
caused by the presence of the more negative pres-
sure surrounding the larger arteries and
bronchi.
33
34
Increases in filtration rate cause fluid
to collect in the terminal lymphatics and the peri-
vascular spaces. The resulting increase in fluid
pressure and volume in terminal lymph vessels
causes a decrease in lymph vessel resistance and
an increase in lymph flow.
35
The sump mechanism
is aided by the presence of valves in the lymph
vessels. In addition, because the lymphatics run in
the same sheath as the pulmonary arteries, they are
exposed to the massaging action of the arterial
pulsations. The differential negative pressure, the
lymphatic valves, and the arterial pulsations all
help to propel the lymph proximally toward the
hilum and central venous circulation depot (see
Fig. 3-9B).
Any situation that impairs lymph flow may en-
hance the formation of pulmonary edema. The
lymphatics that drain the lung empty into the su-
perior vena cava. In most patients, superior vena
cava pressure is approximately equal to the central
venous pressure (CVP) so that the CVP represents
the outflow pressure against which the lung lym-
phatics must pump; consequently, numerous in-
vestigators documented that lung lymph flow rate
decreases if the CVP is elevated (see Fig. 3-10).
30
Indeed, lung lymph flow is essentially stopped by
a CVP of 20 cm H
2
0,
35
a value not uncommonly
observed clinically in the intensive care unit. Thus,
CVP elevations enhance the formation*of pulmo-
nary edema and pleural effusion
36
by impairing the
ability of the lymphatics to remove interstitial
fluid/
When flow of interstitial fluid through the en-
dothelial holes is excessive and cannot be cleared
adequately by lymphatics, fluid will accumulate in
the interstitial connective tissue compartment
around the large vessels and airways, forming per-
ibronchial and periarteriolar edema fluid cuffs, re-
spectively, increased pulmonary vascular resis-
tance and airway resistance, respectively, and a
Figure 3-10 The balance of forces across the pulmonary endothelial membrane is commonly altered in disease. See text for full
explanation.
C. Other (Nongravitational) Important
Determinants of Pulmonary Vascular
Resistance and Blood Flow Distribution
1. Cardiac Output
The passive effect of changes in cardiac output
on the pulmonary circulation are as follows.
40
As
cardiac output increases, pulmonary vascular pres-
sures increase (Fig. 3-11). Since the pulmonary
vasculature is distensible, an increase in pulmo-
nary artery pressure increases the radius of the
HIGH -.HIGH
HIGH
Resistance (PVR)
The Pulmonary Circulation and Vasomotion
Figure 3-12 The pulmonary circulation and vasomotion. Relationships among flow (Q), pressure (P), and resistance (R) and the
pulmonary circulation in the adult respiratory distress syndrome (ARDS) {A) and during deliberate hypotension ( j, BP) (B). Active
vasoconstriction is present in the pulmonary circulation whenever cardiac output decreases and pressure remains constant or
increases; these findings are common in ARDS. Active vasodilation is present in the pulmonary circulation whenever cardiac output
increases and pressure remains constant or decreases: these findings are often present with deliberate hypotension. (Modified with
permission from Benumof JL: The pulmonary circulation. In Kaplan JA (ed): Thoracic Anesthesia. New York, Churchill Living-
stone, 1983, chapter 7.)
greater effect than Pv0
2
because oxygen uptake is
from the alveolar space to the blood in the small
pulmonary arteries.
43
However, in those regions of
the lung that are atelectatic, the stimulus for HPV
is only v0
2
.
There are two major theories as to how alveo-
lar hypoxia may cause pulmonary vasoconstric-
tion.
44
"
48
First, alveolar hypoxia may cause the
release of vasoconstrictor substance(s) into the
pulmonary interstitial compartment where the sub-
stance^) may then cause vasoconstriction. In the
past 10 years, many vasoactive substances have
been proposed as the mediators of HPV (e.g., leu-
kotrienes, prostaglandins, catecholamines, seroto-
nin, histamine, angiotensin, and bradykinin), but
none has been proved to be involved primarily in
the process. Most recently, nitric oxide (endothe-
lium-derived relaxing factor) has been proposed to
have a pivotal role in modulating pulmonary vas-
cular resistance.
49
Although the precise mediator of HPV is not
known, it is certain that the prostaglandin products
of arachidonic acid metabolism can inhibit the
HPV response, and it is very possible that the
leukotriene products of arachidonic acid metabo-
lism mediate, or are at least required for, the HPV
response. The general scheme of arachidonic acid
metabolism is shown in Figure 3-13. On an appro-
priate stimulus, such as alveolar hypoxia, phos-
pholipase A
2
converts the phospholipid in the cell
membranes to arachidonic acid (this reaction is
possible in all 40 lung cell types).
50
The released
arachidonic acid can be metabolized in two ways.
First, the enzyme cyclo-oxygenase can convert
arachidonic acid to prostaglandins; the major pros-
taglandin is prostaglandin I
2
(prostacyclin).
51 52
Prostacyclin is a potent pulmonary vasodilator that
can abolish HPV.
51 52
(However, prostaglandin F
2
a
is a pulmonary vasoconstrictor.) The cyclo-oxy-
genase pathways can also produce thromboxane;
this product is thought not to be important with
regard to HPV.
53 54
Second, the enzyme lipoxygen-
ase can convert arachidonic acid to leukotrienes.
All the leukotrienes are potent pulmonary vaso-
constrictors and can enhance HPV; indeed, the
leukotrienes have received considerable attention
as the mediator of HPV.
54
The amount of pulmo-
nary vasoconstriction caused by hypoxia is regu-
lated by a balance between leukotriene agonist and
prostaglandin antagonist effects.
There are many experimental data that support
and are consistent with this relationship between
the products of arachidonic acid metabolism and
HPV (see Fig. 3-13). Blockage of cyclo-oxygen-
Figure 3-13 The leukotriene products of arachidonic acid metabolism can increase hypoxic pulmonary vasoconstriction (HPV),
and the prostaglandin products of arachidonic acid metabolism can decrease HPV. The amount of HPV is determined by a balance
between these agonist and antagonist effects of the leukotrienes and prostaglandins, respectively.
and results in gas exchange in a region of lung not
normally utilized (i.e., zone 1). Thus, with a low
F,0
2
, the P
a
o
2
is greater, and the alveolar-arterial
0
2
tension difference and dead space/tidal volume
ratio are less than would be expected or predicted
on the basis of a normal (sea level) distribution of
ventilation and blood flow.
85
High-altitude pulmo-
nary hypertension is an important component in
the development of mountain sickness subacutely
(hours to days) and cor pulmonale chronically
(weeks).
86
In fact, there is now good evidence that
in patients with chronic obstructive pulmonary dis-
ease, even nocturnal episodes of arterial oxygen
desaturation (caused by episodic hypoventilation)
are accompanied by elevations in P
pa
and may ac-
count for or lead to sustained pulmonary hyperten-
sion and cor pulmonale.
87
Second, hypoventilation (low V
A
/Q), atelectasis,
or nitrogen ventilation of any region of the lung
(one lung, lobe, lobule) generally causes a diver-
sion of blood flow away from the hypoxic to the
nonhypoxic lung (40 to 50 per cent, 50 to 60 per
cent, 60 to 70 per cent, respectively) (Fig. 3-
14).
88 w
The regional vasoconstriction and blood
flow diversion are of great importance in minimiz-
ing transpulmonary shunt and normalizing re-
gional V/Q ratios during disease of one lung, one-
lung anesthesia (see chapter 4), and inadvertent
intubation of a main-stem bronchus. In regard to
one-lung anesthesia in the lateral decubitus posi-
Regi onal HPV
Figure 3-14 Schematic drawing of regional hypoxic pul-
monary vasoconstriction (HPV); one-lung ventilation is a com-
mon clinical example of regional HPV. HPV in the hypoxic
atelectatic lung causes a redistribution of blood Mow away from
the hypoxic lung to the normoxic lung, thereby diminishing the
amount of shunt flow (Q./Q,) that can occur through the hy-
poxic lung. Inhibition of hypoxic lung HPV causes an increase
in the amount of shunt flow through the hypoxic lung, thereby
decreasing Pao
:
.
tion, it is important to note that the strength of the
HPV response in humans is sufficient to overcome
significant vertical hydrostatic gradients.
90
Third, in patients with chronic obstructive pul-
monary disease, asthma, pneumonia, and mitral
stenosis who do not have bronchospasm, adminis-
tration of pulmonary vasodilator drugs such as is-
oproterenol, sodium nitroprusside, and nitroglyc-
erin causes a decrease in P
a
0
2
and pulmonary
vascular resistance and an increase in right-to-left
transpulmonary shunt.
91
The mechanism for these
changes is thought to be deleterious inhibition of
pre-existing and, in some of the lesions, geograph-
ically widespread HPV without a concomitant and
beneficial bronchodilation.
91
In accordance with
the latter two lines of evidence (one-lung or re-
gional hypoxia and vasodilator drug effects on
whole-lung or generalized disease), HPV is
thought to divert blood flow away from hypoxic
regions of the lung, thereby serving as an autoreg-
ulatory mechanism that protects P
a
0
2
by favorably
adjusting regional V
A
/Q ratios. Factors that inhibit
regional HPV are extensively discussed under
physiology and one-lung ventilation in chapter 4.
3. Lung Volume
The functional residual capacity (FRC) is the
volume of lung that exists at the end of a normal
exhalation after a normal tidal volume and when
there is no muscle activity or pressure difference
between alveoli and atmosphere. Total pulmonary
vascular resistance is increased when lung volume
is either increased or decreased from FRC (Fig. 3-
15).
92-94
The increase in total pulmonary vascular
Figure 3-15 An asymmetrical U-shaped curve relates total
pulmonary vascular resistance (PVR) to lung volume. The
trough of the curve occurs when lung volume equals func-
tional residual capacity (FRC). Total pulmonary resistance is
the sum of resistance in small vessels (increased by increas-
ing lung volume) and in large vessels (increased by decreas-
ing lung volume). The end point for increasing lung volume
(toward total lung capacity [TLC]) is the creation of zone l
conditions, and the end point for decreasing lung volume
(toward residual volume [RV]) is the creation of low V/Q
and atelectatic (atel) areas that have hypoxic pulmonary vaso-
constriction (HPV). The curve represents a composite of data
from Simmons and colleagues,
42
Burton and Patel,"
1
and Wit-
tenberger and associates."
4
(Modified with permission from
Benumof JL: Respiratory physiology and respiratory function
during anesthesia. In Miller RD (ed): Anesthesia. 2nd ed.
New York, Churchill Livingstone, 1983, chapter 32.)
resistance above FRC is due to alveolar compres-
sion of small intra-alveolar vessels, which results
in an increase in small-vessel pulmonary vascular
resistance (i.e., creation of zone 1 or 2).
95 96
As a
relatively small mitigating or counterbalancing ef-
fect to the small-vessel compression, the large ex-
tra-alveolar vessels may be expanded by the in-
creased negativity of the perivascular pressure at
high FRC. The increase in total pulmonary vascu-
lar resistance below FRC is due to an increase in
pulmonary vascular resistance of large extra-al-
veolar vessels. The increase in large-vessel pul-
monary vascular resistance was previously thought
to be due to a mechanical tortuosity or kinking of
these vessels. However, small or grossly atelectatic
lungs are hypoxic, and it has recently been shown
that the mechanism of increased large-vessel pul-
monary vascular resistance in these lungs is due
entirely to hypoxic pulmonary vasoconstriction.
97
This conclusion has been found to be true whether
the chest is open or closed and whether ventilation
is by positive pressure or is spontaneous.
98
The relationship between lung volume and pul-
monary vascular resistance can determine the dis-
tribution of pulmonary blood flow within a given
single lung if there is a vertical gradient of alveolar
volume (Fig. 3-16). At total lung capacity, all
alveoli are large, and the small intra-alveolar ves-
sels are homogeneously compressed; there are no
low V/Q or atelectatic areas, and there is no large
extra-alveolar vessel vasoconstriction (HPV).
Since P
pa
and P
pv
increase down the upright lung,
blood flow simply increases down the upright lung
at total lung capacity (Fig. 3-16A) as it does at
FRC (Fig. 3-16/?). Near residual volume depen-
62 General Respiratory Physiology and Respiratory Function Dui ig Anesthesia
General Respiratory Physiology and Respirator} Function During Anesthesia 63
dent alveoli are much smaller than nondependent
alveoli, and dependent lung may contain hypoxic
low V/Q and atelectatic areas. Consequently, at
residual volume, the large extra-alveolar vessels in
the dependent lung may be constricted (because of
HPV), and blood flow either is uniform
5 6
or de-
creases slightly" down the upright lung, even
though P
pa
and P
pv
are increasing (Fig. 3-16C).
4. Alternate (Nonalveolar) Pathways of
Blood Flow Through the Lung
Figure 3-17 shows all the possible pathways for
blood to travel from the right side to the left side
of the heart without being oxygenated and the
pathologic conditions during which blood flow
through these shunt pathways is significantly in-
creased. Blood flow through poorly ventilated al-
veoli (low V/Q regions at F,0
2
less than 0.3 have
a right-to-left shunt effect on oxygenation) and
through nonventilated alveoli (atelectatic or con-
solidated regions) (V/Q = 0 at all F,0
2
) is the
source of right-to-left shunt. Low V/Q and atelec-
tatic lung units occur in conditions in which the
FRC is less than the closing capacity (CC) of the
lung (see section on lung volume presented later).
There are several right-to-left heart blood flow
pathways that do not pass by or involve alveoli at
all. The bronchial and pleural circulations origi-
nate from systemic arteries and empty into the left
side of the heart without being oxygenated, consti-
tuting the 1 to 3 per cent true right-to-left shunt
normally present. With chronic bronchitis, the
bronchial circulation may carry 10 per cent of the
cardiac output, and with pleuritis the pleural cir-
culation may carry 5 per cent of the cardiac output.
Consequently, there may be as much as a 10 per
cent and 5 per cent obligatory right-to-left shunt
present, respectively, under these conditions. Intra-
pulmonary arteriovenous anastomoses are nor-
mally closed, but in the face of acute pulmonary
hypertension, such as may be caused by a pulmo-
nary embolus, or cirrhosis,
100
they may open and
cause a direct increase in right-to-left shunt. The
foramen ovale is patent in 20 to 30 per cent of
individuals
101
but normally remains functionally
closed because left atrial pressure exceeds right
atrial pressure. However, any condition that results
in right atrial pressure being greater than left atrial
pressure may produce a right-to-left shunt with
resultant hypoxemia and possible paradoxical
embolization. Such conditions include the use of
high levels of PEEP, pulmonary embolization, pul-
monary hypertension, chronic obstructive pulmo-
nary disease, pulmonary valvular stenosis, conges-
tive heart failure, and postpneumonectomy states
(see chapter 19 and Figure 19-3). Indeed, even
such common events as mechanical ventilation
102
and breath holding and reaction to the presence of
an endotracheal tube during the excitement phase
of emergence from anesthesia
103 l04
have caused
right-to-left shunting across a patent foramen ovale
and severe arterial desaturation (thereby creating
the potential for paradoxical embolization)."*
4
Esophageal to mediastinal to bronchial to pulmo-
nary vein pathways have been described and may
explain in part the hypoxemia associated with por-
tal hypertension and cirrhosis. There are no known
conditions that selectively increase thebesian chan-
nel blood flow (thebesian vessels nourish the left-
heart myocardium and originate and empty into
the left heart).
Pulmonary Shunt Pathways
PLEURITIS
CIRRHOSIS
NONE
BRONCHITIS
PULMONARY
EMBOLISM
General Respiratory' Physiology and Respiratory Function During Anesthesia 65
ically, depending on whether a peak or plateau
inspiratory pressure gradient (respectively) is used
for the C
T
calculation.
During a positive- or negative-pressure inspira-
tion of sufficient duration, the transthoracic pres-
sure gradient first increases to a peak value and
then decreases to a somewhat lower plateau value
(Fig. 3-18). The peak transthoracic pressure value
is due to pressure required to overcome both elas-
tic (lung and chest wall) and airway resistance
forces (see the following section, Airway Resis-
tance). The transthoracic pressure decreases to a
plateau value following the peak value because,
with time, gas redistributes from stiff alveoli
(which expand only a little and therefore have only
a very short inspiratory period) into more com-
pliant alveoli (which expand a great deal and
therefore have a long inspiratory period). Because
the gas redistributes into more compliant alveoli,
less pressure is required to contain the same
amount of gas, and this explains why the pressure
decreases (see Fig. 3-18). Total thoracic compli-
ance (C
T
) may be expressed as either a dynamic or
static compliance. In practical terms, dynamic
compliance is the volume change divided by the
TIME (sec)
Figure 3-18 Relationship between tidal volume and airway
pressure in a patient receiving mechanical ventilation. The in-
spiratory plateau in the volume tracing is achieved by tempo-
rary occlusion of the expiratory tubing. During this period,
airway pressure falls from a peak of 20 cm H
:
0 to a plateau
pressure of 10 cm H,0. Total thoracic compliance is calculated
as volume delivered/airway pressure. If the peak airway pres-
sure is used, then dynamic compliance is calculated, whereas
if the plateau pressure is used, static compliance is calculated.
(PEEP = positive end-expiratory pressure.)
peak inspiratory transthoracic pressure; static com-
pliance is the volume change divided by the pla-
teau inspiratory transthoracic pressure. Therefore,
static C
T
is usually greater than dynamic C
T
be-
cause the former calculation uses a smaller denom-
inator (lower pressure) than the latter calculation.
However, if the patient is receiving PEEP, this
must be first subtracted from the plateau pressure
before calculating total thoracic compliance (i.e.,
compliance = volume delivered/(plateau pressure
- PEEP).
Total thoracic compliance is also helpful as a
predictor of weaning outcome, with the attraction
that it is less dependent on patient cooperation than
other indices. The likelihood of a successful wean-
ing outcome is uncommon in patients with a com-
pliance of less than 25 ml/cm H
2
0, because the
associated increase in the work of breathing makes
it difficult to sustain a level of ventilation adequate
for satisfactory gas exchange.
The pressure in an alveolus (P
al veo
|
ar
) deserves
special comment. The alveoli are lined with a layer
of liquid. The lining of a curved surface (sphere or
cylinder, as are the alveoli, bronchioles, bronchi)
with liquid creates a surface tension that tends to
make the surface area that is exposed to the atmos-
phere as small as possible. Simply stated, water
molecules crowd much closer together on the sur-
face of a curved layer of water than elsewhere in
the fluid. As lung or alveoli size decreases, the
degree of curvature and the retractive surface ten-
sion force increase.
According to Laplace's equation, the pressure in
an alveolus (P, in dyn/cm
2
) is above ambient pres-
sure by an amount depending on the surface ten-
sion of the lining liquid (T, in dyn/cm) and the
radius of curvature of the alveolus (R, in cm). This
is expressed in the following:
(4)
= 2T/R
Although surface tension contributes to the elas-
tic resistance and retractive forces of the lung, two
difficulties must be resolved. The first problem is
that the pressure inside small alveoli should be
higher than that for large alveoli, a conclusion that
stems directly from Laplace's equation (small R in
the denominator). From this reasoning, one would
expect a progressive discharge of each small al-
veolus into a larger one, until eventually only one
gigantic alveolus would be left (Fig. 3-19). The
second problem concerns the relationship between
lung volume and the transpulmonary pressure gra-
dient (P
a Kc ol a r
= Ppieurai) Theoretically, the retrac-
tive forces of the lung should increase as the lung
volume decreases. If this were true, lung volume
should decrease in a vicious cycle, with the ten-
dency to collapse increasing progressively as the
lung volume diminished.
These two problems are resolved by the fact that
the surface tension of the fluid lining the alveoli is
variable and decreases as its surface area is re-
duced. The surface tension of alveolar fluid can
reach levels that are well below the normal range
for body fluids such as water and plasma. When
an alveolus decreases in size, the surface tension
of the lining fluid falls to a greater extent than the
corresponding reduction of radius, so that the
transmural pressure gradient ( = 2T/R) diminishes.
This explains why small alveoli do not discharge
their contents into large alveoli (Fig. 3-19B) and
why the elastic recoil of small alveoli is less than
that of large alveoli.
The substance responsible for the reduction (and
variability) of alveolar surface tension is secreted
by the intra-alveolar type II pneumocyte and is a
lipoprotein called surfactant that floats as a 50-A
thick film on the surface of the alveolar-lining
fluid. When the surface film is reduced in area and
the concentration of surfactant at the surface is
increased, there is an increased surface-reducing
pressure that counteracts the surface tension of the
fluid lining the alveoli.
2. Airway Resistance
In order for air to flow into the lungs, a pressure
gradient must also be developed to overcome the
nonelastic airway resistance of the lungs to air-
flow. The relationship between the pressure gra-
dient () and the rate of airflow (V) is known as
airway resistance (R) and is expressed as follows:
The pressure gradient () along the airway de-
pends on the caliber of the airway and the rate and
pattern of airflow. There are three main patterns of
airflow; laminar flow occurs when the gas passes
down parallel-sided tubes at less than a certain
critical velocity. With laminar flow, the pressure
drop down the tube is proportional to the flow rate
and may be calculated from the equation derived
by Poiseuille: = V X 8 L X u/rr
4
X 980,
where = pressure drop (in cm H
2
0), V =
volume flow rate (in ml/sec), L = length of tube
(in cm), r = radius of tube (in cm), and u =
viscosity (in poise).
When flow exceeds the critical velocity, it be-
comes turbulent. The significant feature of turbu-
lent flow is that the pressure drop along the airway
is no longer directly proportional to flow rate but
is proportional to the square of flow rate according
to the equation = V
2
pfL/4ir
2
r\ where is a gas
or fluid density term and f is a friction factor that
depends on the roughness of the tube wall.
l(,?
Thus,
with increases in turbulent flow (and/or orifice
flow, see immediately below) increases much
more than V, and therefore R increases (see equa-
tion 5).
Orifice flow occurs at severe constrictions, such
as a nearly closed larynx. In these situations, the
pressure drop is also proportional to the square of
flow rate, but density replaces viscosity as the im-
portant factor in the numerator. This explains why
low-density gas such as helium diminishes the re-
sistance to flow (by threefold compared with air)
in severe obstruction of the upper airway.
Since the total cross-sectional area of the air-
ways increases as branching occurs, the velocity
of airflow decreases; laminar flow is therefore
chiefly confined to the airways below the main
bronchi. Orifice flow occurs at the larynx, and flow
in the trachea is turbulent during most of the res-
piratory cycle. Viewing the preceding five equa-
tions, one can see that many factors obviously may
affect the pressure drop down the airways during
respiration. However, variations in diameter of the
smaller bronchi and bronchioles are particularly
critical (bronchoconstriction may convert laminar
flow to turbulent flow), and the pressure drop
along the airways may become much more related
to flow rate.
3. Different Regional Lung Time
Constants
So far, the compliance and airway resistance
properties of the chest have been discussed sepa-
rately. In the following analysis, the pressure at
the mouth is assumed to increase suddenly to a
fixed positive value (Fig. 3-20)
106
that overcomes
both elastic and airway resistance and is main-
tained at this value during inflation of the lungs.
As shown in Figure 3-20, the pressure gradient
required to overcome nonelastic airway resistance
is the difference between the fixed mouth pressure
and the instantaneous height of the dashed line and
is proportional to the flow rate during most of the
respiratory cycle. Thus, the pressure gradient re-
quired to overcome nonelastic airway resistance is
maximal initially but then decreases exponentially
(Fig. 3-20/4, hatched lines). The rate of filling,
therefore, also declines in an approximately expo-
nential manner. The remainder of the pressure gra-
dient overcomes the elastic resistance (the instan-
taneous height of the dashed line in Fig. 3-20)
and is proportional to the change in lung volume.
Thus, the pressure gradient required to overcome
elastic resistance is minimal initially but then in-
creases exponentially (as does lung volume). Al-
veolar filling ceases (lung volume remains con-
stant) when the pressure resulting from the
retractive elastic forces balances the applied
(mouth) pressure (Fig. 3-20, dashed line).
Since there is only a finite time available for
alveolar filling, and alveolar filling occurs in an
exponential manner, the degree of filling is ob-
viously dependent on the duration of the inspira-
tion. The rapidity of change in an exponential
curve can be described by its time constant tau ().
Tau (T) is the time required to complete 63 per
cent of an exponentially changing function if the
total time allowed for the function change is un-
limited (2 = 87 per cent, 3 = 95 per cent, and
4 = 99 per cent). For lung inflation, = C
T
X
R; normally, C
T
= 0.1 L/cm H,0, R - 2.0 cm
H
2
0/L/sec, and = 0.2 sec and 3 = 0.6 sec.
Applying this equation to individual alveolar
units, the time taken to fill such a unit clearly
increases as airway resistance increases. The time
to fill an alveolar unit also increases as compliance
increases, since a greater volume of air will be
transferred into a more compliant alveolus before
the retractive force equals the applied pressure.
The compliance of individual alveoli differs from
the top to the bottom of the lung, and the resis-
tance of individual airways will vary widely de-
pending on their length and caliber. Therefore, a
wide variety of time constants for inflation exists
throughout the lung.
4. Work of Breathing
The pressure-volume characteristics of the lung
also determine the work of breathing. Since
(6)
Work = Force X Distance
and
Force = Pressure X Area
and
Distance = Volume/Area
Work = (Pressure X Area) (Volume/Area)
= Pressure X Volume
and ventilatory work may be analyzed by plotting
pressure against volume.
107
In the presence of in-
Figure 3-20 Artificial ventilation by intermittent application of a constant pressure (square wave). Expiration is passive. The
pressure required to overcome airway resistance (hatched lines, A) and airflow rate (V of equation 5; see panel C), which are
proportional to one another, decreases exponentially. The pressure required to overcome elastic resistance (height of dashed line,
panel A) and lung volume (see panel B), which are proportional to one another, increases exponentially. Values shown are typical
for an anesthetized supine paralyzed patient: total dynamic compliance, 50 ml/cm H
:
0; pulmonary resistance, 3 cm H
:
0/L/sec;
apparatus resistance, 7 cm ,/L/sec; total resistance, 10 cm H
2
0/L/sec; time constant, 0.5 sec. (Reprinted by permission of the
publisher, from Nunn JF: Applied Respiratory Physiology. 2nd ed. London, Butterworths [Publishers] Ltd, 1977.)
creased airway resistance or decreased lung com-
pliance, increased transpulmonary pressure is re-
quired to achieve a given V
T
with a consequent
increase in the work of breathing. The metabolic
cost of the work of breathing at rest constitutes
only 1 to 3 per cent of the total 0
2
consumption in
healthy subjects, but it is increased considerably
(up to 50 per cent) in patients with pulmonary
disease.
Two different pressure-volume diagrams are
shown in Figure 3-21. During normal inspiration
(left graph), transpulmonary pressure increases
from 0 to 5 cm H
2
0 while 500 ml of air is drawn
into the lung. Potential energy is stored by the lung
during inspiration and expended during expiration;
as a consequence, the entire expiratory cycle is
passive. The hatched area plus the triangular area
ABC represents pressure multiplied by volume
and is the work of breathing. Line AB is a section
of the pressure-volume curve of Figure 3-5. The
triangular area ABC is the work required to over-
come elastic forces (C
T
), whereas the hatched area
is the work required to overcome airflow or fric-
tional resistance (R). The graph on the right shows
an anesthetized patient with diffuse obstructive
disease resulting from the accumulation of mucous
secretions. There is a marked increase in both the
elastic (triangle AB'C) and airway (hatched area)
resistive components of respiratory work. During
expiration, only 250 ml of air leaves the lungs
during the passive phase when intrathoracic pres-
sure reaches the equilibrium value of 0 cm H
2
0.
Active effort-producing work is required to force
out the remaining 250 ml of air, and intrathoracic
pressure actually becomes positive.
For a constant minute volume, the work done
against elastic resistance is increased when breath-
ing is deep and slow. On the other hand, the work
done against airflow resistance is increased when
breathing is rapid and shallow. If the two compo-
Transpulmonary Pressure (cm H
2
0)
Figure 3-21 Lung volume is plotted against transpulmonary pressure in a pressure-volume diagram for an awake (normal) and
an anesthetized patient. The lung compliance of the awake patient (slope of line AB = 100 ml/cm H
:
0) equals that shown for the
small dependent alveoli in Figure 3-5. The lung compliance of the anesthetized patient (slope of line AB' = 50 ml/cm H
;
0) equals
that shown for the medium midlung alveoli in Figure 3-5 and for the anesthetized patient in Figure 3-20. The total area within the
oval and triangles has the dimensions of pressure multiplied by volume and represents the total work of breathing. The hatched
area to the right of lines AB and AB' represents active inspiratory work necessary to overcome resistance to airflow during
inspiration (INSP). The hatched area to the left of the triangle AB'C represents active expiratory work necessary to overcome
resistance to airflow during expiration (EXP). Expiration is passive in the normal subject because sufficient potential energy is
stored during inspiration to produce expiratory airflow. The fraction of total inspiratory work necessary to overcome elastic
resistance is shown by the triangles ABC and AB'C. The anesthetized patient has a decreased compliance and increased elastic
resistance work (triangle AB'C) compared with the normal patient's compliance and elastic resistance work (triangle ABC). The
anesthetized patient used as an example in this figure has an increased airway resistance to both inspiratory and expiratory work.
(Reproduced with permission from Benumof JL: Respiratory physiology and respiratory function during anesthesia. In Miller RD
(ed): Anesthesia. 2nd ed. Churchill Livingstone, New York, 1983, chapter 32.)
nents are summed and the total work is plotted
against the respiratory frequency, there is an opti-
mal respiratory frequency at which the total work
of breathing is minimal (Fig. 3-22).
mx
In patients
with diseased lungs, when elastic resistance is high
(pulmonary fibrosis, pulmonary edema, infants),
the optimum frequency is increased, and rapid
shallow breaths are favored. When airway resis-
tance is high (asthma, obstructive lung disease),
the optimum frequency is decreased and slow,
deep breaths are favored.
109
However, these
breathing patterns may be altered by such factors
as inflammatory response, release of mediators,
and emotional, physiologic, and/or psychological
stress.""
5. Lung Volumes, Functional Residual
Capacity, and Closing Capacity
a. LUNG VOLUMES AND FUNCTIONAL
RESIDUAL CAPACITY
The FRC is defined as the volume of gas in the
lung that exists at the end of a normal expiration
when there is no airflow and alveolar pressure
equals the ambient pressure. Under these condi-
tions, expansive chest wall elastic forces are ex-
actly balanced by retractive lung tissue elastic
forces (Fig. 3-23).' "
The expiratory reserve volume is part of the
FRC and is that additional gas beyond the end-
tidal volume that can be consciously exhaled and
results in the minimum volume of lung possible,
known as the residual volume. Thus, the FRC
equals the residual volume plus the expiratory re-
serve volume (Fig. 3-24). With regard to the other
lung volumes shown in Figure 3-24, tidal volume,
vital capacity, inspiratory capacity, inspiratory re-
serve volume, and expiratory reserve volume can
all be measured by simple spirometry. Total lung
volume, FRC, and residual volume all contain a
fraction (the residual volume) that cannot be meas-
ured by simple spirometry. However, if one of
these three volumes is measured, the others can be
easily derived because the other lung volumes,
which relate these three volumes to one another,
can be measured by simple spirometry.
FRC can be measured by one of three tech-
niques. The first method is to wash the N
2
out of
the lungs by several minutes of 0
2
breathing with
measurement of the total quantity of N
2
elimi-
nated. Thus, if 2 L of N-, are eliminated and the
Normal
Increased elastic
resistance
Increased air flow
resistance
Respiratory f requency (breat hs per mi n ut e )
Figure 3-22 The diagrams show the work done against elastic and airflow resistance separately and are combined to indicate the
total work of breathing at different respiratory frequencies. The total work of breathing has a minimum value at about 15 breaths
per minute under normal circumstances. For the same minute volume, minimum work is performed at higher frequencies with stiff
(less compliant) lungs and at lower frequencies when the airflow resistance is increased. (Reprinted by permission of the publisher,
from Nunn JF: Applied Respiratory Physiology. 2nd ed. London, Butterworths [Publishers] Ltd, 1977.)
initial alveolar N
2
concentration was 80 per cent,
it follows that the initial volume of the lung was
2.5 L. The second method uses the wash-in of a
tracer gas such as helium. If 50 ml of helium is
introduced into the lungs and the helium concen-
tration is then found to be 1 per cent, it follows
that the volume of the lung is 5 L. The third
method of measurement of FRC uses Boyle's law,
namely, PV = K, where = pressure, V =
volume, and = a constant. The subject is con-
fined within a gas-tight box (plethysmograph) so
that changes in the volume of the body may be
readily determined as a change in pressure within
the box. Disparity between FRC as measured in
Figure 3-23 A, The resting state of normal lungs when they are removed from the chest cavity;
i.e., elastic recoil causes total collapse. B, The resting state of a normal chest wall and diaphragm
when the thoracic apex is open to the atmosphere and the thoracic contents are removed. C, The
lung volume that exists at the end of expirationthe functional residual capacity. At functional
residual capacity, the elastic forces of lung and chest walls are equal and in opposite directions.
The pleural surfaces link these two opposing forces. (Reproduced with permission from Shapiro
A, Harrison RA, Trout CA: Clinical Application of Respiratory Care. 2nd ed. Copyright 1979
by Year Book Medical Publishers, Inc., Chicago.)
Figure 3-24 The dynamic lung volumes that can be measured by simple spirometry are the tidal volume, inspiratory reserve
volume, expiratory reserve volume, inspiratory capacity, and vital capacity. The static lung volumes are the residual volume,
functional residual capacity, and total lung capacity. The static lung volumes cannot be measured by observation of a spirometer
trace and require separate methods of measurement. (Reproduced with permission from Benumof JL: Respiratory physiology and
respiratory function during anesthesia. In Miller RD (ed): Anesthesia. 2nd ed. New York, Churchill Livingstone, 1983, chapter 32.)
the body plethysmograph and by the helium
method is often used as a way of detecting large,
nonventilating air-trapped blebs."
2
Obviously,
there are difficulties in the application of the body
plethysmograph to anesthetized patients.
b. AIRWAY CLOSURE AND CLOSING CAPACITY
As discussed in the section Distribution of Ven-
tilation, pleural pressure increases from the top to
the bottom of the lung and determines regional
alveolar size, compliance, and ventilation. Of even
greater importance to the anesthesiologist is the
recognition that these gradients in pleural pressure
may lead to airway closure and collapse of alveoli.
(1) AIRWAY CLOSURE IN PATIENTS WITH NOR-
MAL LUNGS. Figure 3-25A illustrates the normal
resting end-expiratory (FRC) position of the lung-
chest wall combination. The distending transpul-
monary and the intrathoracic air passage transmu-
ral pressure gradients are 5 cm H
;
0, and the air-
ways remain patent. During the middle of a normal
inspiration (Fig. 3-255), there is an increase in the
transmural pressure gradient (to 6.8 cm H
;
0),
which encourages distention of intrathoracic air
passages. During the middle of a normal expira-
tion (Fig. 3-25C), expiration is passive; alveolar
pressure is caused only by the elastic recoil of the
lung (2 cm ,), and there is a decrease (to 5.2
cm H
2
0) but still a favorable (distending) intralu-
minal transmural pressure gradient.
During the middle of a severe forced expiration
(Fig. 3-25D), pleural pressure increases far above
atmospheric pressure and is communicated to the
alveoli, which have a pressure higher still owing
to the elastic recoil of the alveolar septa (an addi-
tional 2 cm H
2
0). At high gas-flow rates, the pres-
sure drop down the air passage is increased, and
there will be a point at which intraluminal pressure
equals either surrounding parenchymal or pleural
pressure; that point is termed the equal pressure
point (EPP). If the EPP occurs in small intratho-
racic air passages (distal to the eleventh generation
the airways have no cartilage and are called bron-
chioles), they may be held open at that particular
point by the tethering effect of the elastic recoil of
the immediately adjacent or surrounding lung pa-
renchyma. If the EPP occurs in large extrathoracic
air passages (proximal to the eleventh generation
the airways are called bronchi), they may be held
open at that particular point by their cartilage.
Downstream of the EPP (in either small or large
airways), the transmural pressure gradient is re-
versed ( - 6 c m H
2
0) and will result in airway
closure. Thus, the patency of airways distal to the
eleventh generation is a function of lung volume,
and the patency of airways proximal to the elev-
enth generation is a function of intrathoracic
(pleural) pressure. In extrathoracic bronchi with
cartilage, the posterior membranous sheath appears
to give first by invaginating into the lumen."
3
If
lung volume was abnormally decreased (e.g., due
to splinting) and expiration was still forced, the
caliber of the airways would be relatively reduced
at all times, causing the EPP and point of collapse
to move progressively from larger to smaller air
passages (closer to the alveolus).
End
Expiration
NORMAL LUNG
Middle
Inspiration
NORMAL LUNG
Middle Passive
Expiration
NORMAL LUNG
Middle Forced
Expiration
NORMAL LUNG
Mild Forced
Expiration
EMPHYSEMA
Mild Forced
Expiration
EMPHYSEMA
LARYNX PARTLY
CLOSED
Figure 3-25 Pressure gradients across the airways. The airways consist of a thin-walled intrathoracic portion (near the alveoli)
and a more rigid (cartilaginous) intrathoracic and extrathoracic portion. During expiration the pressure from elastic recoil is assumed
to be +2 cm H
:
0 in normal lungs (A-D) and + 1 cm H
2
0 in abnormal lungs (E and F). The total pressure inside the alveolus is
pleural pressure plus the elastic recoil. The arrows indicate direction of airflow. (EPP = equal pressure point.) See text for
explanation. (Modified with permission from Benumof JL: Respiratory physiology and respiratory function during anesthesia. In
Miller RD (ed): Anesthesia. 2nd ed. New York. Churchill Livingstone, 1983, chapter 32.)
In patients with normal lungs, airway closure
may still occur even if exhalation is not forced,
provided residual volume is approached closely
enough. Even in patients with normal lungs, as
lung volume decreases during expiration toward
residual volume, the small airways (0.5 to 0.9 mm
in diameter) will show a progressive tendency to
close, whereas larger airways still remain
patent."
4
"
5
Airway closure occurs first in the de-
pendent-lung regions (as has been directly ob-
served by computed tomography),"
6
"
8
since the
distending transpulmonary pressure is less and the
volume change during expiration is greater. The
airway closure is most likely to occur in the depen-
dent regions of the lung whether the patient is in
the supine or lateral decubitus position, and
whether ventilation is spontaneous or by positive-
pressure ventilation."
6-
"
8
(2) AIRWAY CLOSURE IN PATIENTS WITH AB-
NORMAL LUNGS. Airway closure occurs with
milder active expiration, lower gas-flow rates, and
higher lung volumes and occurs closer to the al-
veolus in patients with emphysema, bronchitis,
asthma, and pulmonary interstitial edema. In all
four of these conditions, airway resistance is in-
creased, causing a larger pressure decrease from
the alveoli to the larger bronchi, thereby creating
the potential for negative intrathoracic transmural
pressure gradients and narrowed and collapsed air-
ways. In addition, the structural integrity of the
conducting airways may be diminished because of
inflammation and scarring and, therefore, may
close more readily for any given lung volume or
transluminal pressure gradient.
In emphysema, the elastic recoil of the lung is
reduced (to 1 cm H
2
0 in Fig. 3-25E),
t n e a
'
r
passages are poorly supported by the lung paren-
chyma, the point of airway resistance is close to
the alveolus, and the transmural pressure gradient
can become negative quickly. Therefore, during
only a mild forced expiration in an emphysema-
tous patient, the EPP and the point of collapse are
near the alveolus (Fig. 3-25E). Use of pursed-lip
or grunting expiration (the equivalent of partly
closing the larynx during expiration}, PEEP, and a
continuous positive airway pressure in an emphy-
sematous patient restores a favorable (distending)
intrathoracic air pressure transmural gradient (Fig.
3-25F). In bronchitis, the airways are structurally
weakened and may close when only a small nega-
tive transmural pressure gradient is present (as
with a mild forced expiration). In asthma, the mid-
dle-sized airways are narrowed by bronchospasm,
and if expiration is forced, they are further nar-
rowed by a negative transmural pressure gradient.
Finally, with pulmonary interstitial edema, peri-
alveolar interstitial edema compresses alveoli and
acutely decreases FRC; the peribronchial edema
fluid cuffs (within the connective tissue sheaths
around the larger arteries and bronchi) compress
the bronchi and acutely increase closing vol-
ume."
9
-
121
(3) THE MEASUREMENT OF CLOSING CAPAC-
ITY. CC is a sensitive test of early small-airway
disease and is performed by having the patient
exhale to residual volume (Fig. 3-26). ' " An inha-
lation from residual volume toward total lung ca-
pacity is begun, and at the beginning of the inha-
lation a bolus of tracer gas (
,33
Xe, helium) is
injected into the inspired gas. During the initial
part of this inhalation from residual volume, the
first gas to enter the alveolus is the dead-space gas
and the tracer bolus. The tracer gas will only enter
alveoli that are already open (presumably, the ap-
ices of the lung; hatched lines, Fig. 3-26) and does
not enter alveoli that are already closed (presuma-
bly the bases of the lung; no hatched lines, Fig. 3-
26). As the inhalation continues, apical alveoli
complete filling, and basilar alveoli begin to open
and to fill, but with gas that does not contain any
tracer gas.
A differential tracer gas concentration is thus
established; the gas in the apices has a higher
tracer concentration (Fig. 3-26, hatched lines) than
that in the bases (Fig. 3-26, no hatched lines). As
the subject exhales, and the diaphragm ascends, a
point is reached at which the small airways just
above the diaphragm start to close, limiting airflow
from these areas. The airflow now comes more
from the upper lung fields, where the alveolar gas
has a much higher tracer concentration, thus re-
sulting in a sudden increase in the tracer gas con-
centration toward the end of exhalation (phase IV).
The closing volume (CV) is the difference be-
tween the onset of phase IV and residual volume;
since it represents part of a vital capacity maneu-
ver, it is expressed as a per cent of the vital lung
capacity. The CV plus the residual volume is
known as the CC and is expressed as a per cent of
total lung capacity. Smoking, obesity, aging, and
the supine position increase the CC.
123
In healthy
individuals at a mean age of 44 years, CC = FRC
in the supine position, and at a mean age of 66
years CC = FRC in the upright position.
124
c. RELATI ONSHI P BETWEEN FUNCTI ONAL
RESIDUAL CAPACI TY AND CLOSI NG CAPACITY
The relationship between FRC and CC is far
more important than consideration of the FRC or
CC alone because it is the relationship between the
two that determines whether a given respiratory
unit is normal or atelectatic or has a low V/Q ratio.
The relationship between FRC and CC is as fol-
lows: When the volume of lung at which some
airways close is greater than the whole of the tidal
volume, then lung volume never increases enough
Total Lung
Capacity
Residual
Volume
Closing
Capacity
Spirogram
nhalation
of bolus
of
133
Xe
Total lung
FRC
Residual
Volume
Closing
Capacity
Concentration
of
133
Xe
of mouth
TLC
Phase IV
FRC CC RV
Vital Capacity
Figure 3-26 Measurement of closing capacity by the use of a tracer gas such as xenon 133 (' "Xe). The bolus of tracer gas is
inhaled near residual volume and, because of airway closure in the dependent lung, is distributed only to those nondependent
alveoli whose air passages are still open (shown hatched in diagram). During expiration, the concentration of the tracer gas becomes
constant after the dead space is washed out. This plateau (phase III) gives way to a rising concentration of tracer gas (phase IV)
when there is once again closure of the dependent airways because the only contribution made to the expired gas is by the high
' "Xe concentration nondependent alveoli. (Reprinted by permission of the publisher, from Nunn JF: Applied Respiratory Physiol-
ogy. 2nd ed. London, Butterworths [Publishers] Ltd, 1977.)
during tidal inspiration to open any of these air-
ways. Thus, these airways stay closed during all
of the tidal respiration. Airways that are closed all
of the time are equivalent to atelectasis (Fig. 3-
27). If the closing volume of some airways lies
within the tidal volume, then as lung volume in-
creases during inspiration, some previously closed
airways will open for a short period of time until
lung volume once again recedes below the closing
volume of these airways. Since these opening and
closing airways are open for a shorter period of
time than normal airways, they have less chance
or time to participate in fresh gas exchange, which
is a circumstance equivalent to a low ventilation-
perfusion region. If the closing volume of the lung
is below the whole of tidal respiration, then no
airways are closed at any time during tidal respi-
ration; this is a normal circumstance. Anything
that decreases FRC relative to CC or increases CC
relative to FRC will convert normal areas to low
V/Q and atelectatic areas. Development of low
V/Q and atelectatic areas will cause hypoxemia.
The theory of the FRC-CC-,, relationship
has been quantitatively confirmed in a study in
which patients with varying initial FRC-CC rela-
tionships were moved from the seated to the su-
pine position (thereby decreasing FRC).
125
In this
study,
125
subjects were divided into four groups. In
group 1, FRC exceeded CV in both the supine and
erect positions. In group 2, FRC exceeded CV
only in the seated position. In group 3, CV oc-
curred within the tidal volume in the seated posi-
tion and exceeded the tidal volume in the supine
position. In group 4, CV occurred above the tidal
volume in both positions.
Group 1 subjects, whose tidal breathing was
above the level of the CV in both postures, showed
improved gas exchange in the supine position.
This was explained by the presumed improvement
in overall uniformity of V
A
/Q and by the observed
increase in cardiac output. In group 2 subjects,
FRC was greater than CV in the erect position and
less than CV in the supine position. The P(A-a)o,
and shunt increased upon changing posture from
erect to supine (presumably owing to conversion
of normal units to low V/Q and atelectatic units).
In group 3 subjects, CV occurred within the tidal
breathing range in the erect position, and the erect
P(A-a)o
2
and shunt were similar to those of group
2 subjects when supine, while a change to the
supine position resulted in a further increase in
P(A-a)o
2
and shunt fraction (presumably owing to
Functional Residual Capacity (FRC) to
Closing Capacity (CC) Relationship
Time
Figure 3-27 The relationship between the functional residual capacity (FRC) (which is the per cent of total lung capacity that
exists at the end of the exhalation), shown by the level of each trough of the sine wave tidal volume, and the closing capacity (CC)
of the lung (three different closing capacities are indicated by the three different straight lines). The abscissa is time. See text for
explanation of why the three different functional residual capacity to closing capacity relationships depicted result in normal or low
ventilation-perfusion ratios (V
A
/Q) or atelectasis. (Modified with permission from Benumof JL: The pulmonary circulation. In
Kaplan JA (ed): Thoracic Anesthesia. New York, Churchill Livingstone. 1983, chapter 7.)
76
conversion of normal units to low V/Q and atelec-
tatic units). In group 4 subjects, CV exceeded the
whole of tidal respiration in the upright position.
Changing to the supine position caused a signifi-
cant increase in shunt and cardiac output, resulting
in no significant change in P(A-a)o
2
. Thus, the
change in cardiac output (and perhaps more uni-
form lung perfusion) counteracted further airway
closure.
The reason mechanical intermittent positive-
pressure breathing (IPPB) may be efficacious is
that it can take a previously spontaneously breath-
ing patient with a low ventilation-perfusion rela-
tionship (where the CC is greater than FRC but
still within the tidal volume, as depicted in Fig. 3-
2&4) and increase the amount of inspiratory time
that some previously closed (at end-exhalation)
airways spend in fresh gas exchange and thereby
increase the ventilation-perfusion relationship
(Fig. 3-285). However, if PEEP is added to the
IPPB, the PEEP increases FRC above or to a lung
volume greater than CC and thereby restores a
normal FRC-to-CC relationship so that no airways
are closed at any time during the tidal respiration,
as depicted in Figure 3-28C (IPPB + PEEP).
Thus, anesthesia-induced crescent-shaped densities
in the dependent regions of patients' lungs have
not been reversed with IPPB alone but have been
reversed with IPPB + PEEP (5 to 10 cm H.O)."
6
E. Oxygen and Carbon Dioxide
Transport
1. Alveolar and Dead-Space Ventilation
and Alveolar Gas Tensions
In patients with normal lungs, approximately
two thirds of each breath reach perfused alveoli
and thereby take part in gas exchange. This consti-
tutes the effective or alveolar ventilation. The re-
maining one third of each breath takes no part in
gas exchange and is therefore termed the total (or
effective or physiologic) dead-space ventilation.
The total dead-space ventilation may be divided
into two components: a volume of gas that venti-
lates the conducting airways (the anatomic dead-
space ventilation) and a volume of gas that venti-
Figure 3-28 The functional residual capacity to closing capacity relationship during spontaneous ventilation, intermittent
positive-pressure breathing (IPPB), and IPPB and positive end-expiratory pressure (IPPB). See text for explanation of the effect of
the two ventilatory maneuvers (IPPB and PEEP) on the functional residual capacity to closing capacity relationship. (TLC = total
lung capacity.) (Reproduced with permission from Benumof JL: Respiratory physiology and respiratory function during anesthesia.
In Miller RD (ed): Anesthesia. 2nd ed. Churchill Livingstone, New York, 1983, chapter 32.)
2. Oxygen Transport
a. OVERVIEW OF OXYGEN TRANSPORT
The principal function of the heart and lungs is
supporting oxygen delivery to, and C0
2
removal
from, the tissues in accordance with metabolic re-
quirements while maintaining arterial blood oxy-
gen and carbon dioxide partial pressures within a
narrow range. The respiratory and cardiovascular
systems are series-linked to accomplish this func-
tion over a wide range of metabolic requirements,
which may increase 30-fold from rest to heavy
exercise. The steps in this functional linkage are
as follows: (1) ventilation and distribution of ven-
tilation with respect to perfusion, (2) diffusion of
oxygen into blood, (3) chemical reaction of oxy-
gen with hemoglobin, (4) cardiac output of arterial
blood, and (5) distribution of blood to tissues and
release of oxygen (Table 3-4). The system is sel-
dom stressed except at exercise, and the earliest
symptoms of cardiac or respiratory diseases are
often seen only during exercise.
The maximum functional capacity of each link
can be determined independently. Table 3-4 lists
these measured functional capacities for healthy,
young men.
126
I27
The maximum capacity of each
step to deliver oxygen without desaturating arterial
blood and without causing tissue hypoxia can be
estimated from these functional capacities if 15
g/dL hemoglobin is assumed. The other assump-
tions are that physiologic dead-space ratio is 0.25
and that alveolar oxygen tension does not fall be-
low 110 mm Hg. Because theoretical maximum
oxygen transport at the ventilatory step or the dif-
fusion and chemical reaction steps (about 6 L/min
in normal humans at sea level) exceeds the oxygen
transportable by maximum cardiac output and dis-
tribution steps, the limit to oxygen transport is the
cardiovascular system. Respiratory diseases would
not be expected to limit maximum oxygen trans-
port until their functional capacities were reduced
nearly 40 to 50 per cent. Therefore, for healthy,
well-perfused tissues, oxygen uptake is determined
primarily by metabolic need rather than by oxygen
*Controversy exists over the magnitude of this number.
Originally 1.34 had been used,
121
* but with the determination of
the molecular weight of hemoglobin (64,458) the theoretical
value of 1.39 has become popular.
129
Following extensive hu-
man studies, Gregory observed in 1974 that the applicable
value was 1.306 ml/g per cent in human adults.
130
Most of the
literature still, however, utilizes 1.39.
5 L/min and C
a
0
2
= 20.4 ml O
2
/0.l L, then the
arterial point corresponds to 1020 ml/min going to
the periphery, and the venous point corresponds to
760 ml/min returning to the lungs, with Vo
2
=
260 ml/min.
The oxy-Hb curve can also relate the 0
2
actually
available to the tissues (leftmost y-axis in Fig. 3-
33) as a function of Po
2
. Of the 1000 ml/min of
0
2
normally going to the periphery, 200 ml/min of
0
2
cannot be extracted because it would lower the
Po
2
below the level (rectangular dashed line in
Fig. 3-33) at which organs such as the brain can
survive; the 0
:
available to the tissues is therefore
0.8 L/min. This is approximately three to four
times the normal resting Vo
2
. When Q, = 5
L/min, and the arterial saturation is less than 40
per cent, the total flow of 0
2
to the periphery is
reduced to 400 ml/min so that the available 0
2
is
now 200 ml/min and 0
2
supply just equals 0
2
demand. Consequently, with low arterial satura-
tion, tissue demand can be met only by an increase
in cardiac output or, in the longer term, by an
increase in Hb concentration.
The position of the oxy-Hb curve is best de-
scribed by the Po
2
at which Hb is 50 per cent
saturated (the P
M)
). The normal adult P
S()
, which is
the point on the left side and steep portion of the
oxy-Hb curve in Figure 3-33, is 26.7 mm Hg.
The effect of a shift in the position of the oxy-
Hb curve on hemoglobin saturation depends
greatly on the Po
2
. In the region of the normal
P
a
0
2
(75 to 100 mm Hg), the curve is relatively
horizontal, so that shifts of the curve have little
effect on saturation. In the region of the mixed
venous Po
:
where the curve is relatively steep, a
shift of the curve leads to a much greater differ-
ence in saturation. A P
so
of less than (<) 27 mm
Hg describes a left-shifted oxy-Hb curve, which
means that at any given Po
2
, Hb has a higher
affinity for 0
2
and is therefore more saturated than
normal. This may require a higher tissue perfusion
than normal to produce the normal amount of 0
2
unloading. The causes of a left-shifted oxy-Hb
curve are alkalosis (metabolic and respiratory,
called the Bohr effect), hypothermia, abnormal and
fetal hemoglobin, CO-Hb, methemoglobin, and
decreased RBC 2,3-diphosphoglycerate (2,3-DPG)
content (which may occur with transfusion of old
acid-citrate-dextrose stored blood; storage of blood
in citrate-phosphate-dextrose minimizes changes
in 2.3-DPG with time).
A PJO of greater than (>) 27 mm Hg describes
a right-shifted oxy-Hb curve, which means that at
any given Po
2
, Hb has a low affinity for 0
2
and is
less saturated than normal. This may allow a lower
tissue perfusion than normal to produce the normal
amount of 0
2
unloading. The causes of a right-
shifted oxy-Hb curve are acidosis (metabolic and
respiratory, called the Bohr effect), hyperthermia,
abnormal Hb, inhalation anesthetics (see later dis-
cussion), and increased RBC 2,3-DPG content.
Abnormalities in acid-base balance result in al-
teration of 2,3-DPG metabolism to shift the oxy-
hemoglobin dissociation curve to its normal posi-
tion. This "compensatory" change in 2,3-DPG
requires between 24 and 48 hours. Thus, with
acute acid-base abnormalities, oxygen affinity and
the position of the oxy-Hb curve change. How-
ever, with more prolonged acid-base changes, the
altered levels of 2,3-DPG shift the oxy-Hb curve
and, therefore, oxygen affinity back toward nor-
mal.
Many inhalation anesthetics have been shown to
shift the oxyhemoglobin dissociation curve to the
right.'"
I32
Isoflurane shifts the P
x
to the right 2.6
0.07 mm Hg at a vapor pressure of approxi-
mately l MAC (minimum alveolar concentration
for adequate anesthesia in 50 per cent of subjects)
(1.25 per cent).'" On the other hand, high-dose
fentanyl, morphine, and merperidine do not alter
the position of the curve.
c. EFFECT OF Q
s
/Q
t
ON P
a
0
2
Figure 3-34'
w
shows the relationship between
F|0
2
and P.,0
2
for a family of right-to-left transpul-
monary shunts (QJQ
t
); the calculations assume a
constant and normal cardiac output and P
a
C0
2
.
With no QJQ
t
, a linear increase in F,0
2
results in
a linear increase in P
a
0
2
(solid straight line). As
shunt is increased, the Q
s
/Q, lines relating F,0
2
to
P.,0
2
become progressively flatter.
113
With a shunt
of 50 per cent of the cardiac output, an increase in
F,0
2
results in almost no increase in P
a
0
2
. Thus, it
is obvious that the solution to the problem of hy-
poxemia secondary to a large shunt is not simply
increasing the F,0
2
but rather causing a reduction
in shunt (fiberoptic bronchoscopy, PEEP, patient
positioning, antibiotics, suctioning, and diuretics).
d. EFFECT OF Q
t
AND Vo
2
ON C
a
0
2
In addition to an increased Q
s
/Q,, the C
a
0
2
is
decreased by a decreased Q, (for a constant Vo
2
)
and by an increased Vo
2
(for a constant Q,). In
either case (decreased Q
t
or increased Vo
2
), along
with a constant right-to-left shunt, the tissues must
extract more 0
:
from the blood per unit blood
volume, and, therefore, the 0
2
content of mixed
venous blood (C
%
0
2
) must primarily decrease (Fig.
3-35). When the blood with lower Q0
2
passes
through whatever shunt that exists in the lung and
remains unchanged in its oxygen composition, it
must inevitably mix with oxygenated end-pulmo-
82 General Respirator} Physiology and Respiratory Function During Anesthesia
Figure 3-34 Effect of changes in inspired oxygen concentration on arterial Po
3
for various right to left transpulmonary shunts.
Cardiac output (Q,), hemoglobin (Hb), oxygen consumption (Vo
2
), and arteriovenous oxygen content differences (C[a v]0
:
) were
assumed to be normal. (Modified from Nunn JF: Applied Respiratory Physiology. London, Butterworths [Publishers] Ltd. 1977.
Reprinted by permission of the publisher, Butterworths.)
Figure 3-35 Effect of a decrease in cardiac output or an increase in oxygen consumption on mixed venous and arterial oxygen
contents. Mixed venous blood (v) perfuses either ventilated alveolar (ALV 0
:
) capillaries and becomes oxygenated end-pulmonary
capillary blood (c') or perfuses whatever true shunt pathways exist and remains the same in composition (desaturated). These two
pathways must ultimately join together to form mixed arterial (a) blood. If the cardiac output (Q,) decreases and/or the oxygen
consumption (Vo
:
) increases, the tissues must extract more oxygen per unit volume of blood than under normal conditions. Thus,
the primary effect of a decrease in Q, or an increase in Vo, is a decrease in mixed venous oxygen content. The mixed venous blood
with a decreased oxygen content must flow through the shunt pathway as before (which may remain constant in size) and lower the
arterial content of oxygen. Thus, the secondary effect of a decrease in Q, or an increase in Vo, is a decrease in arterial oxygen
content. (Reproduced with permission from Benumof JL: Respiratory physiology and respiratory function during anesthesia. In
Miller RD (ed): Anesthesia. 2nd ed. New York, Churchill Livingstone, 1983, chapter 32.)
Figure 3-36 The equivalent circuit of the pulmonary circulation in a patient with a 50 per cent right-to-left shunt. Oxygen
content is measured in milliliters per deciliter of blood (vol %). A decrease in cardiac output (Q,) or an increase in O, consumption
(Vo
2
) can cause a decrease in mixed venous oxygen content (from 15 vol % to 10 vol % in this example), which in turn will cause
a decrease in the arterial content of oxygen (from 17.5 vol % to 15.0 vol %). In this 50 per cent shunt example, the decrease in
mixed venous oxygen content was twice the decrease in arterial oxygen content. (Reproduced with permission from Benumof JL:
Respiratory physiology and respiratory function during anesthesia. In Miller RD (ed): Anesthesia. 2nd ed. New York, Churchill
Livingstone, 1983, chapter 32.)
Figure 3-37 Effects of changes in cardiac output (Q) on the 0
:
content of end-pulmonary capillary, arterial, and mixed venous
blood for a family of different transpulmonary right-to-left shunts. The magnitude of the right-to-left shunt is indicated by the
various numbered per cent symbols for arterial (solid line) and mixed venous (dashed line) blood; the oxygen content of end-
capillary blood is unaffected by the degree of shunting. Note that a decrease in Q results in a greater decrease in the arterial content
of 0\, the larger the shunt. (Modified with permission from Kelman GF, Nunn JF, Prys-Roberts C, et al: The influence of the
cardiac output on arterial oxygenation: A theoretical study. Br J Anaesth 39:450, 1967.)
Figure 3-38 Change in arterial oxygen saturation as shunt
is introduced and then cardiac output decreased. The fraction
of inspired oxygen is 0.33. (Reproduced with permission from
Westbrook JL, Sykes MK: Preoperative arterial hypoxemia.
The interaction between intrapulmonary shunt and cardiac out-
put. Anaesthesia 47:307-310, 1982.)
Figure 3-39 Relationship between systemic oxygen delivery and systemic oxygen consumption in control dogs (control = solid
line) and after administration of Escherichia coli endotoxin (endotoxin = dotted line). Endotoxin increased 0
:
demand by tissues,
but decreased the 0
:
extraction ratio (VoVDo,) from 78 to 54 per cent. (Modified with permission from Nelson DP. Samsel RW.
Wood LDH. et al: Pathological supply dependence of systemic and intestinal O, uptake during cndotoxemia. J Appl Physiol
64:2410. 1988.)
drated to carbonic acid (H
2
CO,) and as bicarbonate
(HCO_,~). In the erythrocyte, C0
2
combines with
hemoglobin as carbaminohemoglobin (Hb-C0
2
).
In plasma, C0
2
exists both in physical solution
and as H
2
C0
3
:
In this equation, is the solubility coefficient of
C0
2
in plasma (0.03 mmol/L/mm Hg at 37C).
However, the major fraction of C0
2
produced
passes into the erythrocyte. As in plasma, CO
:
combines with water to produce carbonic acid.
However, unlike in plasma where the reaction is
slow and most of the equilibrium is to the left, the
reaction in th erythrocyte is catalyzed by the en-
zyme carbonic anhydrase. This zinc-containing en-
zyme moves the reaction to the right at a rate 1000
times faster than that in plasma. Furthermore,
nearly 99.9 per cent of the carbonic acid disso-
ciates to the bicarbonate and hydrogen ions:
capnia and acidosis shift the curve to the right, and
hypocapnia and alkalosis shift the curve to the left.
The Haldane effect is the shift in the relationship
of Pco
2
to total C0
2
(i.e., the C0
2
dissociation
curve) caused by altered levels of oxygen. Low
Po
2
shifts the C0
2
dissociation curve to the left so
that the blood is able to pick up more C0
2
.
F. Lung-Heart Interactions
1. Spontaneous Ventilation
Spontaneous inspiration generates a decrease in
pleural pressure. Normal inspiration at rest is as-
sociated with an increased right ventricular (RV)
stroke volume and decreased left ventricular (LV)
stroke volume.'
43
l 44
The increase in RV stroke
volume is believed to be the result of a greater
augmentation in RV preload than in RV afterload
(increased transmural pulmonary artery pres-
sure).
143
However, in some situations (acute
asthma), the effect of an increase in RV afterload
may exceed the effect of an increase in RV pre-
load, and RV output may decrease.
145
The reduc-
tion in LV stroke volume is responsible for the
inspiratory fall in systemic blood pressure, also
known as pulsus paradoxus. Although most data
suggest the inspiratory reduction in LV stroke vol-
ume results from a reduction in LV preload,
146
l 47
some studies give evidence that the mechanism
responsible for the reduction in LV stroke volume
during normal inspiration is an increase in LV
afterload.
148,149
The reduction in preload could be
caused by pulmonary venous pooling
150-152
and/or
ventricular interdependence.
153
2. Positive-Pressure Ventilation and
PEEP
Positive-pressure inspiration reduces LV and
RV stroke volumes.
154
I55
RV stroke volume is
reduced primarily because of a decreased pressure
gradient driving venous return to the RV but also
because RV afterload is increased. The mecha-
nisms responsible for reducing LV stroke volume
during positive-pressure inspiration include de-
creased LV compliance (as a result of increased
pleural, pericardial, and lung parenchymal pres-
sure), altered LV geometry and filling (ventricular
interdependence),
153
and reduced RV output.
154
155
G. Pulmonary Vascular Reflexes
Cardiopulmonary receptors, whose afferent fi-
bers course in the cardiac sympathetic nerves, have
been described in both of the right-sided cardiac
It has become clear that the pulmonary efferent
sympathetic nerves are part of an extensive control
system that can be stimulated centrally and reflex-
ively to modify pulmonary vascular tone.
164-166
The
effect of sympathetic stimulation is predominantly
due to alpha-adrenergic receptor stimulation (alpha
receptors predominate both numerically and func-
tionally in the pulmonary circulation)
167
and results
in diminished distensibility (stiffening) of the pul-
monary arterial tree.
166
The effects of pulmonary
sympathetic nerve stimulation on the pulmonary
vascular resistance (as opposed to distensibility)
are not large, but in special circumstances they can
be important. For example, distention of the main
pulmonary artery by balloon inflation in the con-
scious dog reflexively produces constriction of
pulmonary arterioles and possibly venules (pres-
sure distal to the balloon increased from 21/6 to
43/14 mm Hg), and this is due to excitation of
receptors located in the pulmonary artery or pos-
sibly the right side of the heart, or both.
163
For
further example, when the influence of the me-
chanical effect of an increase in cardiac output
caused by sympathetic activation is prevented or
reduced, then the pulmonary vasoconstrictor re-
sponse to the sympathetic activation is even more
apparent.
168
169
Finally, several pharmacologic
studies suggested that vasodilatation in the lungs
can be mediated by stimulation of -adrenocep-
tors,
170-173
although vasodilatation is hard to docu-
ment because of the low vascular tone that is nor-
mally present.
Although the effects of pulmonary sympathetic
nerve stimulation are not ordinarily quantitatively
large, they are vital for homeostasis during stress
and exercise and for maintaining a precise balance
between right and left ventricular outputs. An in-
crease in right ventricular output causes an in-
crease in pulmonary sympathetic nerve activity.
The resultant increase in stiffness of the large pul-
monary arteries increases the rate of transmission
of a pulse wave of a given right ventricular beat to
a level sufficiently rapid to cause an increase in
the stroke volume of the subsequent left ventricu-
lar beat.
174
Without this type of automatic adjustment, the
increase in heart rate and cardiac output during
exercise or excitement would alter the synchrony
of the two ventricles and upset the balance in ven-
tricular outputs. In canine studies, an abrupt onset
of exercise (treadmill) without previous warning
produces an increase in right ventricular stroke
volume, which precedes an increase in left ventric-
ular stroke volume by several beats.
175
Because of
this asynchrony between the two ventricles, pul-
monary blood volume has to increase (see section
J). However, when dogs are trained to anticipate
the start of exercise, and sympathetic nervous ac-
tivity is presumably heightened, the ventricular
outputs change exactly in phase. Direct measure-
ments in humans are consistent with these obser-
vations in dogs, since it has been shown that the
pulmonary blood volume remains virtually un-
changed during exercise.
176
In addition, it is now
clear that during experimental excitation of the
hypothalmic integrative area, which is necessary
in order to produce or stimulate the defense reac-
tion, the pulmonary sympathetic nerves are stimu-
lated to produce moderate pulmonary vasocon-
striction, with presumably similar secondary
effects on balancing ventricular outputs.
164
On the
other hand, there is virtually no effect of vagal
stimulation on the pulmonary circulation.
Finally, vagal affrents from the lung cause a
number of important reflexes. Pulmonary venous
congestion (caused by mitral valve obstruction
sufficient to increase left atrial pressure by 5 mm
Hg) causes a significant increase in respiratory
rate.
177
Lung expansion in the resting tidal volume
range causes a brief tachycardia known as respira-
tory arrhythmia,
178
which arises from the vagal
afferent myelinated fibers responsible for the Her-
ing-Breuer reflex, which regulates depth of tidal
volume. Greater expansion of the lungs (beyond
the tidal volume) results in a proportional cardio-
vascular depression with reflex bradycardia, de-
creases in ventricular contractile state, systemic
vasodilation, and hypotension; these depressor re-
flexes are mediated through unmyelinated afferent
fibers that also lie in the vagal nerves
179-182
and
may explain, in part, why lung hyperexpansion
(e.g., PEEP, acute asthma) results in bradycardia
and hypotension.
183
H. Pulmonary Metabolism and
Synthesis
1. Metabolism of Endogenous
Substances
It is now well established that the pulmonary
circulation has three fundamentally and physiolog-
ically important pharmacokinetic functions.
I84
-'
86
First, the pulmonary circulation can inactivate or
remove vasoactive endogenous substances from
the venous blood (5-hydroxytryptamine [5-HT],
bradykinin, norepinephrine [NE], prostaglandins
[PGE,, PGE,, and PGFJ), whereas other endoge-
nous substances, often very closely related to each
other as well as to these substances, are allowed
free passage (epinephrine, angiotensin II, oxytocin,
and vasopressin) (see Table 3-6). This function
seems to be appropriate for the lung because, in
contrast to all other organs, it receives the total
Table 3-6 HANDLING OF BIOLOGICALLY
ACTIVE MATERIALS IN THE
PULMONARY CAPILLARY BED*
A. Metabolized at endothelial surface without uptake from
plasma
1. Bradykinin inactivated
2. Adenine nucleotides inactivated
3. Angiotensin 1 activated
4. Enkephalin inactivated
B. Metabolized intracellular^ after uptake from plasma
1. Serotonin
2. Norepinephrine
3. Prostaglandins and F
C. Inactivated by simple pharmacokinetic binding
1. Atrial natriuretic peptide
D. Unaffected by traversing lungs
1. Epinephrine 6. Dopamine
2. Isoproterenol 7. Vasopressin
3. Histamine 8. Acetylcholinet
4. Prostaglandins A and 9. Substance
I,
5. Angiotensin II
E. Synthesized within lung and released into blood
1. Prostaglandins and F
2. Prostacyclin
3. Endothelium-derived relaxing factor (EDRF)+
4. Hormones
F. Discharged from intrapulmonary stores into blood
1. Histamine
2. Prostaglandins
3. Slow-reacting substance of anaphylaxis
4. Kallikreins
5. Eosinophil leukocyte chemotactic factor of anaphylaxis
*Based on data from Bakhle and Vane,
11
*
4
Bakhle,'*
5
and
Camus and Jeanin.
IK6
tNormally very little or none reaches the lungs owing to
peripheral tissue and blood cholinesterase hydrolysis. However.
90 per cent of what little acetylcholine does reach the lungs is
unaffected by passage through the pulmonary circulation.
Local messenger of pulmonary vascular smooth muscle
relaxation.
See Table 3-7.
venous return and is therefore in an ideal position
to regulate the concentration of vasoactive sub-
stances in pulmonary capillary blood before they
reach the arterial circulation and have profound
systemic effects. Consequently, the lung can be
thought of as a biochemical or metabolic filter.
The removal of some substances but not of oth-
ers has led to the classification of vasoactive hor-
mones as local or as circulating, depending upon
whether they were removed by the lungs. A local
hormone is released at or near the target cells, has
its local effect, and is inactivated before reaching
the arterial circulation. The inactivation of the hor-
mone occurs either immediately in the tissues,
within a few seconds in the venous blood, or
within a few more seconds in the pulmonary cir-
culation. It is now apparent that enzymatic degra-
dation (5-HT, NE) and uptake processes (prosta-
glandins) play a part in the pulmonary inactivation
processes for these substances. Defects in this met-
abolic function of the lung are definitely impli-
cated in the causation of some clinical conditions.
For example, potentiation of the cardiovascular ef-
fects of NE produced by some drugs (e.g., cocaine,
tricyclic antidepressants, some steroids, certain
antihypertensive drugs) may be due to inhibition
of pulmonary removal of NE in addition to inter-
ference with uptake and storage of NE in periph-
eral tissues.
The presence of luminal endothelial projections
and plasmalemmal cavities, with histochemical
verification of the presence of enzymes at these
endothelial locations, invites the functional ana-
tomic picture of enzymes in the luminal membrane
of the endothelial cell being washed and bathed
with substrates that are in a continuously flowing
liquid phase.
187-190
The enzyme-substrate interface
is enormous at the level of the capillary bed
(greater than 100 m
2
; approximately half the en-
dothelial mass of the organism is located in the
lung!),
186
a condition that favors efficient metabo-
lism by relatively small amounts of enzyme. The
presence of intracellular and membrane cavities
very likely explains how the metabolic products of
some of the substrates (adenine nucleotides and
prostaglandins) are returned to the circulation with
no apparent delay or uptake by tissue.
The lungs have a number of ways in which they
may influence the quantity of endothelial enzymes
exposed to circulating substrates. The number of
capillaries open at any given time and therefore
the amount of endothelial surface exposed to blood
flow are complex functions of right-to-left shunt-
ing, pulmonary venous pressure, posture, exercise
(cardiac output), depth of ventilation, the compo-
sition of inhalants, and several other factors, not
the least of which is the structural integrity of the
lungs themselves. It might be expected that factors
that significantly increase or decrease mean transit
time of blood through the lungs (e.g., cardiac out-
put, viscosity) could affect the amount of endothe-
lial enzyme-substrate interaction. Changes in in-
spired oxygen tension have produced somewhat
variable and undramatic changes in the amount of
lung modification of most vasoactive substances,
with angiotensin I conversion being an excep-
tion.
m
-
194
On the other hand, a circulating hormone is
released into the venous blood and then distributed
through the arterial circulation without loss of ac-
tivity on passage through the lungs. By definition,
it is not possible for the lung to malfunction in its
handling of a circulating hormone. At present,
there are no known conditions in which the lung
begins to inactivate previously unmetabolized hor-
mones.
2. Synthesis and Release of
Endogenous Substances
The second important pharmacokinetic function
of the pulmonary circulation is an endocrine one:
The lung is an organ that can contribute vasoactive
endogenous hormones to the circulation as well as
remove them. It is possible that any substance the
lung is capable of synthesizing and/or storing may,
under certain conditions, be released into the pul-
monary circulation.
195-197
Table 3-7 shows that
neoplastic lung disease may produce almost any
biologically active polypeptide and the associated
characteristic clinical syndrome caused by the
polypeptide.
198
Mechanical stimulation of the lung,
such as stroking of the lung surface or retraction
during surgery, may cause release of various vaso-
constrictor and vasodilator substances, respec-
tively.
199
2(,()
Hyper- or hypoinflation of the lung
may cause release of prostaglandins.
199 2QI 202
Chemical stimulation of the lung, such as by al-
veolar hypoxia, also causes release of prostaglan-
dins, which may then serve to normalize V/Q re-
lationships.
203-205
In addition, anaphylaxis has been shown to be
one of the conditions in which biologically active
substances (histamine, slow-reacting substance of
anaphylaxis [SRS-A], PGE,, PGE
2
, and PGF
2
, and
possibly bradykinin) are released from the
lungs.
206
207
Indeed, in humans the lung seems to
be the major shock organ of anaphylaxis and the
source of mediators. Some of the mediators re-
leased during anaphylaxis are also released by
damage to lung tissue, such as is caused by over-
inflation, pulmonary embolism, sepsis, and physi-
cal manipulation. The release of these substances,
in acute lung injury states, into the systemic cir-
culation creates the general cardiovascular depres-
sion that is typically present.
208-210
Finally, biolog-
Table 3-7 POLYPEPTIDE HORMONE SECRETION IN PULMONARY DISEASE:
RESULTANT SYNDROMES AND COMMONLY ASSOCIATED LESIONS
Lung Lesion Hormones Syndrome
Oat cell carcinoma, adenoma
Oat cell carcinoma, adenoma,
tuberculosis, pneumonia,
aspergillosis
Squamous cell carcinoma,
adenocarcinoma, and large-
cell undifferentiated
carcinoma
Large-cell anaplastic carcinoma
Squamous, large-, or oat cell
carcinoma
Squamous cell carcinoma,
bronchial adenoma, and oat
cell carcinoma
Mesenchymal cell tumors
Fibrosarcoma
Anaplastic cell carcinoma
Anaplastic cell carcinoma
ACTH
ADH (arginine vasopressin)
PTH or related peptide
Gonadotropins
VIP or related peptide
Growth hormone, serotonin,
kinins (PGs and other)
Insulin-like peptide
Glucagon or related peptide
Prolactin
Combinations o( above
Hypokalemic alkalosis, edema, Cushing's
syndrome
Hyponatremia (SIADH)
Hypercalcemia
Gynecomastia (adults), precocious puberty
(children)
Watery diarrhea or no symptoms
Hypertrophic osteoarthropathy, "carcinoid"
Hypoglycemia
Hyperglycemia
Galactorrhea (or no symptoms)
Multiple syndromes
Abbreviations: ACTH = adrenocorticotrophic hormone; ADH = antidiuretic hormone; SIADH = syndrome of inappropriate
secretion of antidiuretic hormone; PTH = parathyroid hormone. PCs = prostaglandins; VIP = vasoactive intestinal polypeptide.
ically active substances are also released from the
lung when various pharmacologic agents are in-
jected or infused into the pulmonary circulation
(see next section). Table 3-6 summarizes many of
these considerations and the way in which the pul-
monary capillary bed handles biologically active
materials.
There are some newer candidates for "pulmo-
nary vasoactive hormone" status. In the past few
years, the endothelium has been identified as the
source of three more vasoactive mediators: plate-
let-activating factor (PAF), endothelium-derived
relaxing factor (EDRF), and endothelin. The phos-
pholipid, PAF, has a wide spectrum of mostly
proinflammatory activities.
2
"
: | 2
The vasodilator
substance EDRF
213-213
has now been identified as
nitric oxide.
216
This substance is formed by the
endothelial cells; its significance in the pulmonary
circulation remains to be established. One possibil-
ity is that hypoxic vasoconstriction could reflect a
lack of vasodilator nitric oxide, but this seems to
be unlikely in view of recent results showing the
potentiation of hypoxic vasoconstriction by inhib-
itors of the effects of EDRF.
217 : i s
Of particular
interest are the reports of the alteration of endothe-
lium-dependent relaxation in vascular smooth
muscle strips by halothane, enflurane, and isoflur-
ane.
217 2I9 22
but the relevance of these results to
responses in vivo is uncertain. Third, the most
recently described endothelial cell product is a 21-
amino acid peptide, endothelin. This peptide is a
potent vasoconstrictor
221
and bronchoconstrictor.
222
As implied with EDRF, it is clear that, for all these
pulmonary vasoactive hormones, the major prob-
lem is to make physiologic sense out of the con-
flicting activities of the endothelial cell products,
both for the underlying smooth muscle and for the
blood cells in the lumen.
3. Handling of Exogenous Drugs
The major difference between the fate of endog-
enous and exogenous (drug) substrates in the lung
is that the majority of the exogenous substrate that
is removed by passage through the pulmonary cir-
culation is not metabolized, but is simply bound,
more or less reversibly, to some component of
lung tissue.
184-186
Uptake of drugs into the lung
may be desired, when a therapeutic effect has to
be exerted in the lung, as, for example, relaxation
of airway smooth muscle, inhibition of angioten-
sin-converting enzyme, and management of infec-
tious and neoplastic lung diseases. In other in-
stances, however, penetration of drugs in the lung
is unwanted because either the major activity of
the drug concerns distant organs or the drug in-
duces an untoward effect in the lung.
A great number of drugs of diverse pharmaco-
logic activity accumulate in the lung. The most
reliable predictor of lung uptake is the amphiphilic
structure of a drug; amphiphilics are characterized
by the presence on the molecule of a vast hydro-
phobic area, connected to a terminal amino group
by a short aliphatic carbon side chain.
186 223
This
feature is shared by numerous drugs used in clini-
cal practice (e.g.. amiodarone. chlorimipramine.
chlorphentermine, chlorpromazine, imipramine,
methadone, morphine, propranolol, quinine, and
verapamil).
The localization of drugs after uptake also dif-
fers from that of endogenous compounds. Whereas
the latter are processed in endothelial cells, the
former mainly exit the pulmonary vascular com-
partment.
223
A fraction of the drug probably binds
to endothelial cell membranes, as suggested by the
decreased pulmonary 5-HT uptake that follows ad-
ministration of certain amphiphilics. However,
most of the drug binds to epithelial alveolar cells
and to the alveolar macrophage. Generally, amphi-
philics enter the lung very well, but their efflux is
usually both slow and incomplete, leaving a pool
with a prolonged efflux half-life in excess of sev-
eral hours; this has been reported for methadone,
imipramine, and amiodarone. Thus, significant
amounts of these drugs will persist in the lung,
where they may produce toxic reactions like amio-
darone pneumonitis.
Once accumulated, drugs affect the normal met-
abolic functions of lung endothelium.
224
A first
type of drug interaction with lung metabolism may
be seen as a side effect of a drug's major activity.
For example, tricyclic antidepressants can mark-
edly reduce in vitro and in vivo uptake and metab-
olism of circulating monoamines (i.e., the process-
ing of norepinephrine and serotonin), while
monoamine oxidase inhibitors depress their metab-
olism only. Therefore, following pretreatment with
a tricyclic drug, enhanced systemic hypertensive
response to intravenously administered norepi-
nephrine has been observed, presumably because
of impaired pulmonary norepinephrine clearance.
A second example of a drug/pulmonary function
interaction is drug interference with the pulmonary
generation of biologically active eicosanoids
(prostaglandins [PG];, thromboxanes, and leuko-
trienes [LT]). The lung is one of the major sites of
production of the eicosanoids, especially following
immunologic challenge. The beneficial effects of
corticosteroids in asthma have been linked, in part,
to inhibition of the synthesis of both PG- and LT-
type eicosanoids.
Drug-to-drug interactions are a third aspect of
pharmacokinetic events in the pulmonary circula-
tion: Both inhibition of uptake and displacement
of previously bound drugs have been observed in
vitro and in vivo. In general, amphiphilic com-
pounds mutually inhibit their uptake, presumably
because they share and thus compete for similar
binding sites in the lung. This issue is important
because displacement may lead to unexpected re-
lease of a previously bound drug, whereas inhibi-
tion of uptake may increase systemic availability
and systemic effects of a drug. Although readily
demonstrable, displacements are presently an in-
triguing but clinically insignificant phenomenon
because the displacements are usually transient
and of small magnitude. An interesting idea to take
advantage of displacement has been to dislodge a
pneumotoxicant from the lung.
186
Attempts to dis-
place paraquat taken up in the lung have been done
by subsequent administration of chlorpromazine.
Unfortunately, although an accelerated efflux of
paraquat could be readily demonstrated, there was
no reduction in lung toxicity of paraquat.
The pharmacokinetic functions that have been
mentioned here refer to a normal pulmonary cir-
culation in lung. Progressively, the idea has
emerged that some of the aforementioned sub-
strates could be used to assess the metabolic state
of lung endothelium in disease (i.e., develop a
biochemical index of lung injury). This hypothesis
is attractive because it is hoped that lung damage
in certain conditions that lead to the respiratory
distress syndrome may be detected earlier than is
currently possible by conventional means, like
blood gas tension analysis and chest X-rays. For
example, different groups, including Junod's, have
shown that following cardiopulmonary bypass
(CPB), there is no change in extraction of the
endogenous substrates 5-HT and PGE,, but there
is a reduction in propranolol extraction lasting up
to 5 days postoperatively.
225
This decrease is not
the consequence of CPB itself because it was also
observed after surgical procedures not requiring
CPB. This transient decrease in propranolol ex-
traction is believed to result from postoperative
atelectasis because continuous positive airway
pressure, which increases lung volume and recruits
lung vessels, rapidly normalizes lung propranolol
extraction.
225
I. Pulmonary Host Defenses
226
Protective mechanisms spaced along the entire
respiratory tract largely prevent noxious sub-
stances and particles inspired or aspirated into the
lungs from injuring the mucosa and the delicate
alveolar surface (Fig. 3-40). The upper airway
passages are very important for gross filtering of
air and for sneezing and coughing. A mucosal
surface covers the airway from the upper trachea
down to the respiratory bronchioles, and it consists
of a pseudostratified, columnar, ciliated, mucus-
secreting epithelium. The cilia, of course, propel
foreign material proximally (see Fig. 2-18), and
the mucus contains immunologic components in-
cluding immunoglobulins. Lymphocytes and
plasma cells are distributed in the submucosa, and
a few macrophages are present along the surface
of the conducting airways.
The junction between the respiratory bron-
Respiratory Defense Mechanisms
Figure 3-40 Alveolar unit enlarged to illustrate local de-
fenses that can combat a bacterium entering it. See text for
explanation of the dynamics of respiratory defense mecha-
nisms. = bacterium; II = type 2 pneumocyte; AM =
alveolar macrophage; LYM = lymphocyte; PMN = polymor-
phonuclear leukocytes. (Modified with permission from Rey-
nolds HY: Pulmonary host defenses: State of the art. Chest
95:223S-230S, 1989.)
chioles and the conducting airways is important in
several ways: ciliated epithelium and prominent
secretory cells (goblet cells) are lost and a type 1
and type 2 pneumocyte layer develops, different
sources of blood supply exist, and lymphatic chan-
nels terminate. Gas flow in the alveoli ceases and
molecules of air are propelled by brownian mo-
tion. With the loss of mucociliary activity and of
effective clearance by coughing, the body in its
evolutionary wisdom has developed a different
group of host defenses (see Fig. 3-40) to deal with
the particles or bacteria that may have eluded aero-
dynamic filtration mechanisms in the proximal air-
ways. If a bacterium (B) has escaped the aerody-
namic filtration defenses in the upper respiratory
tract, several adversaries await it; four defense
mechanisms are thought to exist. First, the bacte-
rium is coated (or opsonized) by the alveolar lin-
ing fluid containing surfactant and immunoglobu-
lins after being pushed against the sides of the
alveolus (1 in Fig. 3-40), and if complement were
also available, the microorganism could be lysed
(2 in Fig. 3^t0). Second, the scavenger and roving
alveolar macrophage would phagocytose it (3 in
Fig. 3^0), especially if the bacterium is opsonized
and could optimally be attached to the macro-
phage's membrane receptors. Under these circum-
stances, the microbe's fate would be an intracellu-
lar macrophage death. Third, because certain
microbes are difficult for the macrophage to con-
tain and destroy after ingestion, unless the macro-
phage is energized or activated with certain exog-
enous mediators produced by lymphocytes (called
lymphokines), a signal could go to the lympho-
cytes to join in and help. In this situation, a num-
ber of monokines might be produced by special
subsets of T-helper cells. Finally, depending on
the balance between macrophage defense versus
microbial virulence and inoculum size, it may be
imperative to recruit some mobile polymorphonu-
clear leukocytes (PMNs) from the vasculature
space for phagocytic firepower and the inflamma-
tory response (4 in Fig. 3-40). A number of fac-
tors may initiate the influx of PMNs into the al-
veoli, including chemotaxins that are produced by
macrophages.
The ultimate mobilization of lung defenses is to
generate an inflammatory response (pneumonitis)
with an influx of additional phagocytic help in the
form of PMNs and other immunologic factors
from the plasma (complement, IgM, protease in-
hibitors). What signals of distress must emanate
from the alveolar surface to set this response in
motion are multiple, and what triggers the macro-
phage to recruit help is uncertain. Special viru-
lence or a large inoculum of microbes seems likely
to elicit the response. A number of mediators (e.g.,
philogist complement fragments such as C5a, che-
motaxins arising from alveolar macrophages, or
metabolic products of the kinin system with che-
moattractant activity to cause the directed migra-
tion of PMNs from the vasculature) have been
identified. Reviews of this subject give more
details.
227
228
J. Other Special Lung Functions
229 231
1. Reservoir for Left Ventricle
The pulmonary vessels contain about 750 ml of
blood, of which more than half is in readily disten-
sible veins. Since these veins are an extension of
the left atrium, they constitute a blood reservoir
that supplies blood to fill the left ventricle and to
maintain its output, even when the right ventricu-
lar pump falls behind for a few beats. Indeed,
under experimental conditions the extra-alveolar
pulmonary veins can actually be observed chang-
ing in volume with each heartbeat, making up tem-
porary differences between the outputs of the right
and left ventricles. Thus, the left ventricular stroke
volume (output per beat) has been shown to re-
main unchanged for several beats, even when the
pulmonary artery is completely blocked by a bal-
loon. The.larger the left atrium and the more dis-
tended the pulmonary veins, the more capacious
will be the reservoir.
2. Circulatory Filtrative Function
The pulmonary vessels act as a filter to trap and
prevent emboli from reaching and blocking sys-
temic arteries, arterioles, or capillaries. Although a
major amount of pulmonary thromboembolism can
be lethal, the lungs can still perform their gas-
exchange function in the presence of a moderately
reduced number of pulmonary arteries, whereas
obstruction of a systemic artery is much more
likely to lead to tissue necrosis. Thus, concurrent
blockage of some pulmonary vessels is far better
tolerated than single blockage of most systemic
vessels.
3. Nutrition
Blood flow through the alveolar capillaries ap-
pears to be essential for the nutrition of the alveoli
and alveolar ducts. After unilateral pulmonary ar-
tery occlusion, many alveoli become hemorrhagic
and collapse within 1 to 3 days, although they
recover later and appear almost normal after some
months. The early damage implies that the pul-
monary circulation has a nutritive function and
that compensation by the bronchial circulation re-
quires several weeks to develop fully.
4. Administration of Drugs Via the
Respiratory Tract for Systemic Effect
Several drugs are effective when given by the
intratracheal route: naloxone, diazepam, lidocaine,
adrenaline, and atropine are notable exam-
ples.
n2
~
236
Redding et al.
237
introduced the proce-
dure as a means of drug administration in cardiac
arrest; in dogs, intratracheal adrenaline (1 mg in
10 ml of water) was as effective in restoring the
circulation after a hypoxic arrest as was the same
dose of intravenous or intracardiac adrenaline.
It is sometimes assumed that when a drug is
given into the respiratory tract, it is absorbed via
the alveolar capillary membrane,
238
but this route
applies only to the smallest particles or droplets
(6-0.1 or less in diameter). Larger particles or
droplets are absorbed via the mucosa of the tra-
cheobronchial tree.
239
Because all the blood return-
ing from the tracheobronchial tree empties either
directly into the heart (deep intrapulmonary bron-
chial veins drain to the pulmonary veins)
240
or one
of the central veins (superficial extrapulmonary
veins drain to systemic veins),
240
it might be ex-
pected that, provided a drug were absorbed
speedily, its instillation into the trachea would lead
to rapid attainment of high concentrations in the
heart. Bray et al.
236
confirmed this suggestion in a
study in anesthetized patients. They showed not
only that atropine, 0.6 mg, given via an endotra-
cheal tube produced the same degree of tachycar-
dia as it did when given into a vein on the back of
the hand, but also that the onset of the tachycardia
after endotracheal administration was significantly
faster than with the intravenous route (45.5 vs.
95.5 sec) if not as long lasting (4.4 vs. 8.8 min).
However, if adrenaline is given this way, it is not
as effective as it is by the intravenous route. In
dogs, an intratracheal dose of adrenaline 10 times
that given intravenously was needed to produce
the same increase in blood pressure,
241
although in
another study the drug had a more sustained effect
than when given intravenously.
242
In an emergency, it is sometimes impossible to
obtain venous access, and the intratracheal admin-
istration of drugs provides an alternative that is
usually readily available. The Resuscitation Coun-
cil (United Kingdom) and the American Heart As-
sociation both recommend intratracheal adminis-
tration of adrenaline and atropine but, because of
the absence of definitive data in humans, fail to
agree on the appropriate dose: The American rec-
ommendation is that the same dose of any drug
should be given intratracheally as would be used
intravenously,
243
whereas the Resuscitation Coun-
cil
244
recommends that the intratracheal dose
should be twice the intravenous dose. Thus, the
intratracheal route appears to be a useful method
of giving drugs in an emergency, although it seems
to be much less effective during cardiac massage
than it does if a degree of normal circulation per-
sists.
245
If drugs are to be given through an endo-
tracheal tube, they should be administered in a
suitably large volume of saline (not distilled
water)
246
(i.e., for an adult 10-20 ml).
237
The di-
luted drug is probably best delivered direct to the
respiratory mucosa; intratracheal instillation
should be followed by five forcible hyperinflations
of the lung to ensure distribution throughout the
respiratory tract and thereby enhance absorp-
tion.
236
247
II. RESPIRATORY FUNCTION
DURING ANESTHESIA
A. Introduction
Arterial oxygenation is impaired in most pa-
tients during anesthesia with either spontaneous or
controlled ventilation.
248-253
In otherwise normal
patients, it is generally accepted that the impair-
ment of arterial oxygenation during anesthesia is
more severe in the elderly,
254
255
the obese,
256
and
smokers.
257
In various studies of healthy young to
middle-aged patients, venous admixture (shunt)
has been found to average 10 per cent and the
scatter in V
A
/Q ratios to be small to moderate,
255,258
whereas in patients with more marked deteriora-
tion in preoperative pulmonary function, general
anesthesia causes considerable widening of V
A
/Q
distribution and large increases in both low V
A
/Q
(0.005 < V
A
/Q < 0.1) (underventilated) regions
and shunt.
254
257
259
The magnitude of shunt corre-
lates very closely with the degree of atelecta-
sis.
254
259
In addition to these generalizations concerning
respiratory function during anesthesia, the effect
of a given anesthetic on respiratory function will
depend on the depth of general anesthesia, the
patient's preoperative respiratory condition, and
the presence of special intraoperative anesthetic
and surgical conditions.
B. Effect of Anesthetic Depth on
Respiratory Pattern
The respiratory pattern is altered by the induc-
tion and deepening of anesthesia.
260
When the
depth of anesthesia is inadequate (less than MAC),
the respiratory pattern may vary from excessive
hyperventilation and vocalization to breath hold-
ing. As anesthetic depth approaches or equals
MAC (light anesthesia), irregular respiration pro-
gresses to a more regular pattern, which is associ-
ated with a larger than normal tidal volume. How-
ever, during light, but deepening anesthesia, the
approach to a more regular respiratory pattern may
be interrupted by a pause at the end of inspiration
(a sort of "hitch" in the inspiration), followed by
a relatively prolonged and active expiration in
which the patient seems to exhale forcefully rather
than passively. As anesthesia deepens to moderate
levels, respiration becomes faster, more regular,
but more shallow. The respiratory pattern is a sine
wave losing the inspiratory hitch and lengthened
expiratory pause. There is little or no inspiratory
or expiratory pause, and the inspiratory and expi-
ratory periods are equivalent. Intercostal muscle
activity is still present, and there is normal move-
ment of the thoracic cage with lifting of the chest
during inspiration. The respiratory rate is generally
slower and the tidal volume larger with nitrous
oxide-narcotic anesthesia compared with anesthe-
sia with halogenated drugs. During deep anesthe-
sia with halogenated drugs, increasing depression
of respiration is manifested by even more rapid,
shallow breathing (panting). On the other hand,
with deep nitrous oxide-narcotic anesthesia, res-
pirations become slower but may remain deep.
With very deep anesthesia with all drugs, respira-
tions are jerky or gasping in character and irregular
in pattern. This results from loss of active intercos-
tal muscle contribution to inspiration. As a result,
a "rocking boat" movement occurs in which there
is an out-of-phase depression of the chest wall
during inspiration, a flaring of the lower chest mar-
gins, and a billowing of the abdomen. The reason
for this type of movement is that inspiration is
dependent solely on diaphragmatic effort. Inde-
pendent of anesthetic depth, similar chest move-
ments may be simulated by upper and lower air-
way obstruction and by partial paralysis.
C. Effect of Anesthetic Depth on
Spontaneous Minute Ventilation
Despite the variable changes in respiratory pat-
tern and rate as anesthesia deepens, overall spon-
taneous minute ventilation progressively de-
creases. The normal awake response to breathing
C0
2
(the x-axis in Fig. 3-41 shows increasing end-
tidal concentration of C0
2
) causes a linear increase
in minute ventilation (y-axis in Fig. 3-41). In Fig-
ure 3-41 the slope of the line relating minute ven-
tilation to end-tidal C0
2
concentration is 2
L/min/mm Hg. Figure 3-41 also shows that in-
creasing halothane concentration displaces the
end-tidal C0
2
concentration Pco
2
-ventilation re-
sponse curve progressively to the right (meaning
that at any C0
2
concentration ventilation is less
than before), decreases the slope of the curve, and
shifts the apneic threshold to a higher end-tidal
C0
2
concentration level.
261
Similar alterations are
observed with halogenated anesthetics and narcot-
ics (Fig. 3^42). Figure 3-30 shows that decreases
in minute ventilation will cause increases in P
a
C0
2
and decreases in P
a
0
2
. In healthy, unstimulated
spontaneously breathing male volunteers, 1 MAC
halothane, isoflurane, and enflurane causes a
P
a
C0
2
of approximately 46, 48, and 62 mm Hg,
respectively.
D. Effect of Pre-Existing Respiratory
Dysfunction on the Respiratory Effects
of Anesthesia
Among the patients whom anesthesiologists are
frequently required to care for are (1) those with
acute chest disease (pulmonary infection, atelec-
tasis) or systemic diseases (sepsis, cardiac and
renal failure, or multiple trauma) who require
emergency operations; (2) heavy smokers with
subtle pathologic airway and parenchymal condi-
tions and hyperreactive airways; (3) those with
End-expiratory P
C0/
Figure 3-41 In conscious controls (heavy solid line) increasing end-expiratory Pco
:
increases pulmonary minute volume. The
dashed line is an extrapolation of the CO, response curve to zero ventilation and represents the apneic threshold. An increase in
anesthetic (halothane) concentration (end-expiratory concentration) progressively diminishes the slope of the CO
:
response curve
and shifts the apneic threshold to a higher Pco
2
. The heavy line interrupted by dots shows the decrease in minute ventilation and
the increase in Pco, that occurs with increasing depth of anesthesia. (Modified with permission from Munson ES, Larson CP Jr,
Babad AA, et al: The effects of halothane, fluoroxene and cyclopropane on ventilation: A comparative study in man. Anesthesiology
27:716, 1966.)
classic emphysematous and bronchitic problems;
(4) obese people prone to decreases in FRC during
anesthesia;
262
263
(5) patients with chest deformi-
ties; and (6) elderly patients.
Figure 342 The slope of Pco
:
-ventilation response curves
as a function of multiples of the minimal alveolar concentration
(MAC) required for anesthesia for different anesthetics. (Re-
produced from Nunn JF: Applied Respiratory Physiology. 3rd
ed. London, Butterworths, 1987, with the permission of the
publisher.)
The nature and magnitude of these pre-existing
respiratory conditions will determine, in part, the
effect of a given standard anesthetic on respiratory
function. For example, in Figure 3-43 the FRC-
CC relationship is depicted for normal, obese,
bronchitic, and emphysematous patients. In the
normal patient, FRC exceeds CC by approximately
1 L. In the latter three respiratory conditions, CC
is 0.5 to 0.75 L less than FRC. If anesthesia causes
a 1-L decrease in FRC, then the normal patient
will have no change in the qualitative relationship
between FRC and CC. In the patient with special
respiratory conditions, a 1-L decrease in FRC will
cause CC to exceed FRC and change the previ-
ously marginally normal FRC-CC relationship to
either a grossly low V/Q relationship or an atelec-
tatic FRC-CC relationship. Similarly, patients with
chronic bronchitis, who have copious airway se-
cretions, may suffer more from an anesthetic-
induced decrease in mucous velocity flow than
other patients. Finally, if an anesthetic drug inhib-
its HPV, the drug may increase shunting more in
patients with pre-existing HPV than in those with-
out pre-existing HPV. Thus, the effect of a stan-
dard anesthetic can be expected to produce varying
degrees of respiratory change among patients who
have different degrees of pre-existing respiratory
dysfunction.
Figure 3- 43 The lung volume (ordinate) at which the tidal volume is breathed decreases (by 1 L) from the awake state to the
anesthetized state. The functional residual capacity, which is the volume of lung existing at the end of the tidal volume, therefore
also decreases (by 1 L) from the awake to the anesthetized state. In normal, obese, bronchitic, and emphysematous patients, the
awake functional residual capacity considerably exceeds the closing capacity (CC). In obese, bronchitic, and emphysematous
patients, the anesthetized state causes functional residual capacity to be less than closing capacity. In the normal patient, anesthesia
causes the functional residual capacity to equal the closing capacity. (Reproduced with permission from Benumof JL: Respiratory
physiology and respiratory function during anesthesia. In Miller RD (ed): Anesthesia. 2nd ed. Churchill Livingstone, New York,
1983, chapter 32.)
E. Effect of Special Intraoperative
Conditions on the Respiratory Effects
of Anesthesia
Some special intraoperative conditions (such as
surgical position, massive blood loss, and surgical
retraction on the lung) may cause impaired gas
exchange. For example, some of the surgical posi-
tions (e.g., the lithotomy, jack-knife, and kidney
rest positions) and surgical exposure requirements
may decrease cardiac output, cause hypoventila-
tion in a spontaneously breathing patient, and re-
duce the FRC. The respiratory depressant effects
of any anesthetic will be magnified by the type
and severity of pre-existing respiratory dysfunc-
tion as well as by the number and severity of
special intraoperative conditions that can embar-
rass respiratory function.
F. Mechanisms of Hypoxemia During
Anesthesia
1. Malfunction of Equipment
a. MECHANICAL FAILURE OF ANESTHESIA
APPARATUS TO DELIVER OXYGEN TO THE
PATIENT
Hypoxemia caused by mechanical failure of the
0
2
supply system or the anesthesia machine is a
recognized hazard of anesthesia. Disconnection of
the patient from the oxygen supply system (usually
at the juncture of the endotracheal tube elbow con-
nector) is by far the most common cause of me-
chanical failure to deliver oxygen to the patient.
Other reported causes of 0
2
supply failure during
anesthesia include an empty or depleted 0
2
cylin-
der, substitution of a non-0
2
cylinder at the 0
2
yoke because of absence or failure of the pin in-
dex, an erroneously filled 0
2
cylinder, insufficient
opening of the 0
2
cylinder (which hinders a free
flow of gas as pressure decreases), failure of gas
pressure in a piped 0
2
system, faulty locking of
the piped 0
2
system to the anesthesia machine,
inadvertent switching of the Schrader adapters on
piped lines, crossing of piped lines during con-
struction, failure of a reducing valve or gas mani-
fold, inadvertent disturbance of the setting of the
0
2
flowmeter instead of the coarse flowmeter, frac-
tured or sticking flowmeters, transposition of rota-
meter tubes, erroneous filling of a liquid 0
2
reser-
voir with N
2
, and fresh gas line disconnection from
machine to in-line hosing.
264-267
Monitoring the in-
spired oxygen concentration with an in-line F,0
2
analyzer and airway pressure should detect most
of these causes of failure to deliver oxygen to the
patient.
264
"
267
b. MECHANICAL FAILURE OF ENDOTRACHEAL
TUBE: MAIN-STEM BRONCHUS INTUBATION
Esophageal intubation results in almost no ven-
tilation. Virtually all other mechanical problems
(except disconnect and ruptured cuff) with endo-
tracheal tubes (such as kinking, secretion block-
age, and herniated cuffs) cause an increase in air-
way resistance and may result in hypoventilation.
Intubation of a main-stem bronchus results in the
absence of ventilation of the contralateral lung.
Although potentially minimized by HPV, some
perfusion to the contralateral lung will always re-
main, and shunting will increase while P
a
0
2
will
decrease. A tube previously well positioned in the
trachea may enter a bronchus after the patient, or
the head of a patient, is turned or moved into a
new position.
268
Flexion of the head causes caudad
movements,
268
and extension of the head causes
cephalad movement of an endotracheal tube.
268
A high incidence of main-stem bronchus intu-
bation following institution of a 30-degree Tren-
delenburg position has been reported.
269
Cephalad
shift of the carina during the Trendelenburg posi-
tion caused the previously "fixed" endotracheal
tube to become located in a main-stem bronchus.
A main-stem bronchial intubation may obstruct the
ipsilateral upper lobe in addition to the contralat-
eral lung.
270271
Infrequently, the right upper bron-
chus, or one of its segmental bronchi, branches
from the lateral wall of the trachea and may be
occluded by a properly positioned intratracheal en-
dotracheal tube (see chapter 9 for complete discus-
sion of this problem with reference to double-lu-
men tubes).
2. Hypoventilation (Decreased Tidal
Volume)
Patients under general anesthesia may have a
reduced spontaneous tidal volume for two reasons.
First, it may be more difficult to breathe during
general anesthesia because of increased airway re-
sistance and decreased lung compliance. Airway
resistance may be increased because of reduced
FRC, endotracheal intubation, the presence of ex-
ternal breathing apparatus and circuitry, and pos-
sible airway obstruction in nonintubated patients
(see Fig. 3-46).
272
~
274
Lung compliance reduced
owing to some (or all) of the factors that can
decrease FRC (see Decrease in Functional Resid-
ual Capacity).
275
Second, the patient may be less
willing to breathe spontaneously (decreased chem-
ical control of breathing) during general anesthe-
sia, which usually includes premedication, intra-
venous induction, and maintenance with N
2
0 and
a halogenated drug (see Fig. 3-41).
276
There are two ways in which a decreased tidal
volume may cause hypoxemia. First, shallow
breathing may promote atelectasis and cause a de-
crease in FRC. The decrease in FRC likely results
from a progressive increase in surface tension and
can be blocked by increases in tidal volume (see
Ventilation History).
277-280
Second, decreased min-
ute ventilation decreases the overall V
A
/Q ratio of
the lung, which will decrease P
a
0
2
(see Fig. 3-31).
This is likely to occur with spontaneous ventilation
during moderate to deep levels of anesthesia in
which the chemical control of breathing is signifi-
cantly altered.
3. Hyperventilation
Hypocapnic alkalosis (hyperventilation) may re-
sult in a decreased P
a
0
2
via several mechanisms.
These mechanisms are decreased cardiac out-
put
136
137
and increased oxygen consumption
281
282
(see Decreased Cardiac Output and Increased Ox-
ygen Consumption), a left-shifted oxygen-hemo-
globin dissociation curve (see Oxygen-Hemoglo-
bin Dissociation Curve), decreased HPV
283
(see
Inhibition of Hypoxic Pulmonary Vasoconstric-
tion), and/or increased airway resistance and de-
creased compliance
284
(see Increased Airway Re-
sistance).
4. Decrease in Functional Residual
Capacity
Induction of general anesthesia is consistently
accompanied by a significant (15 to 20 per cent)
decrease in FRC,
116
l25
253
285
which usually causes
a decrease in compliance
253
275
and is associated
with compression atelectasis.
116
ll7
253
The maxi-
mum decrease in FRC appears to occur within the
first few minutes of anesthesia."
6
286-288
In the ab-
sence of any other complicating factor, the FRC
does not seem to decrease progressively during
anesthesia. During anesthesia, the appearance of
compression atelectasis and the reduction in FRC
are of the same order of magnitude whether venti-
lation is spontaneous or controlled or whether an-
esthesia is intravenous or inhalation."
8
Con-
versely, in awake patients, FRC is only slightly
reduced during controlled ventilation.
288
The re-
duction of FRC continues into the postoperative
period.
289
For individual patients, the reduction in
FRC correlates well with an increase in alveolar-
arterial Po
2
gradient during anesthesia with spon-
taneous breathing,
290
during anesthesia with artifi-
cial ventilation.
287
and in the postoperative pe-
riod.
289
The reduced FRC may be restored to
normal or above normal by the application of
PEEP,
116 291
and 10 cm H
2
0 PEEP eliminates or
reduces the compression atelectasis."
6
253
In the
following discussion, all possible causes of re-
duced FRC are considered.
a. SUPINE POSITION
Anesthesia and surgery are usually performed
with the patient in the supine position. When the
patient is changed from the upright to the supine
position, FRC decreases by 0.5 to 1.0 L"
6
125
253
285.292
De c a u s e 0
f
a
4-cm cephalad displacement of
the diaphragm by the abdominal viscera (Fig. 3-
44). Pulmonary vascular congestion may also con-
tribute to the decrease in FRC in the supine posi-
tion, particularly in patients who preoperatively
experienced orthopnea.
b. INDUCTION OF GENERAL ANESTHESIA-
CHANGE IN THORACIC CAGE MUSCLE TONE
At the end of a normal (awake state) exhalation,
there is slight tension in the inspiratory muscles
and no tension in the expiratory muscles. Thus, at
the end of a normal exhalation, there is a force
tending to maintain lung volume and no force de-
creasing lung volume (Fig. 3-45). After the induc-
tion of general anesthesia, there is a loss of the
inspiratory tone
293
and an appearance of end-expi-
ratory tone in the abdominal expiratory muscles at
the end of exhalation. Relaxation of inspiratory
muscles inserted onto the rib cage could be re-
sponsible for part of the decrease in FRC that
occurs with anesthesia.
293
294
The gain of end-ex-
piratory tone in the abdominal expiratory muscles
increases intra-abdominal pressure, forces the dia-
phragm cephalad, and decreases FRC (see Fig. 3-
AA \ 253, 286, 295
Thus, following the induction of general anes-
thesia, there is loss of a force tending to maintain
lung volume and gain of a force tending to de-
crease lung volume. Indeed, Innovar (droperidol
and fentanyl citrate) may increase tone in expira-
tory muscles to such an extent that the reduction
in FRC with Innovar anesthesia alone is greater
than that with Innovar plus paralysis induced by
succinylcholine.
296
With emphysema, exhalation may be accompa-
nied by pursing the lips or grunting (partially
closed larynx). The emphysematous patient ex-
hales in either of these ways because both these
maneuvers cause an expiratory retard that pro-
duces PEEP in the intrathoracic air passage and
decreases the possibility of airway closure and a
decrease in FRC (Fig. 3-25F). Endotracheal intu-
bation bypasses the lips and glottis and may abol-
ish normally present pursed-lip or grunting exha-
lation and in this way contributes to airway closure
and a loss in FRC in some spontaneously breath-
ing patients.
Progressive Cephalad Displacement
of the Diaphragm
Initial Upright
Position of
Diaphragm
Surgical Position
and Displacement
Supine
Position
Induction of
Anesthesia
Paralysis
Figure 3-44 Anesthesia and surgery may cause a progressive cephalad displacement of the diaphragm. The sequence of events
involves assuming the supine position, induction of anesthesia, causation of paralysis, the assumption of several surgical positions,
and displacement by retractors and packs. The cephalad displacement of the diaphragm results in a decreased functional residual
capacity ( [ FRC). (P
ab
= pressure of the abdominal contents.)
Spontaneous Ventilation: End Exhalation
Present:
Absent:
Force Increasing
Lung Volume
Force Decreasing
Lung Volume
Present: Present:
Absent: Absent:
Larger Force Decreasing
Lung Volume
Force Increasing
Lung Volume
Force Decreasing
Lung Volume
Force Increasing
Lung Volume
Figure 3-45 Schematic diagram of forces on the chest wall during the awake, anesthetized, and inadequately anesthetized
conditions. In the awake state, inspiratory tone is a force increasing lung volume, and there is no force decreasing lung volume. In
the anesthetized condition, loss of normally present inspiratory tone results in the absence of a force increasing lung volume,
whereas gain of expiratory tone results in the gain of a force decreasing lung volume. In inadequate anesthesia, forceful exhalations
result in a larger force tending to decrease lung volume, and the continued loss in inspiratory tone results in a continued absence of
a force tending to increase lung volume.
c. PARALYSIS
In the upright subject, the FRC and the position
of the diaphragm are determined by the balance
between the lung elastic recoil pulling the dia-
phragm cephalad and the weight of the abdominal
contents pulling it caudad.
297
There is no transdia-
phragmatic pressure gradient.
The situation is more complex in the supine
position. The diaphragm separates two compart-
ments of markedly different hydrostatic gradients.
On the thoracic side, pressure increases by approx-
imately 0.25 cm H
2
0/cm lung height
10
" and on
the abdominal side by 1.0 cm H
2
0/cm abdominal
height.
297
This means that in horizontal postures
progressively higher transdiaphragmatic pressures
must be generated toward dependent parts of the
diaphragm to keep the abdominal contents out of
the thorax. In the unparalyzed patient, this tension
is developed either by passive stretch and shape
changes of the diaphragm (causing an increased
contractile force) or by neurally mediated active
tension. With acute muscle paralysis, neither of
these two mechanisms can be operative, and a shift
of the diaphragm to a more cephalad position oc-
curs (Fig. 3-44).
253
298
The latter position must
express the true balance of forces on the dia-
phragm, unmodified by any passive or active mus-
cle activity.
The cephalad shift in the FRC position of the
diaphragm due to expiratory muscle tone during
general anesthesia is equal to the shift observed
during paralysis (awake or anesthetized
patients).
286
2
" The equal shift suggests that the
pressure on the diaphragm caused by an increase
in expiratory muscle tone during general anesthe-
sia is equal to the pressure on the diaphragm
caused by the weight of the abdominal contents
during paralysis. It is quite probable that the mag-
nitude of these changes in FRC due to paralysis is
also dependent on the body habitus.
d. LIGHT OR INADEQUATE ANESTHESIA AND
ACTIVE EXPIRATION
The induction of general anesthesia can result in
increased expiratory muscle tone,
295
but the in-
creased expiratory muscle tone is not coordinated
and does not contribute to the exhaled volume of
gas. In contrast, spontaneous ventilation during
light general anesthesia usually results in a coor-
dinated and moderately forceful active exhalation
and larger exhaled volumes (Fig. 3-45). Exces-
sively inadequate anesthesia (relative to a given
stimulus) results in very forceful active exhalation,
which may produce exhaled volumes of gas equal
to an awake expiratory vital capacity.
As during an awake expiratory vital capacity
maneuver, a forced expiration during anesthesia
raises the intrathoracic and alveolar pressures con-
siderably above atmospheric pressure (Fig. 3-25).
This results in a rapid outflow of gas, and since
part of the expiratory resistance lies in the smaller
air passages, a pressure drop will occur between
the alveoli and the main bronchi. Under these cir-
cumstances, the intrathoracic pressure rises consid-
erably above the pressure within the main bronchi.
Collapse will occur if this reversed pressure gra-
dient is sufficiently high to overcome the tethering
effect of the surrounding parenchyma on the small
intrathoracic bronchioles or the structural rigidity
of cartilage in the large extrathoracic bronchi.
Such collapse occurs in the normal subject during
a maximal forced expiration, and it is responsible
for the associated wheeze in both awake and an-
esthetized patients.
300
In the paralyzed anesthetized patient, the use of
a subatmospheric expiratory pressure phase is
analogous to a forced expiration in the conscious
subject; the "negative phase" may set up the same
adverse pressure gradients that can cause airway
closure, gas trapping, and a decrease in FRC. An
excessively rapidly descending bellows of a venti-
lator during expiration has caused a subatmo-
spheric expiratory pressure and has resulted in
wheezing.
301
e. INCREASED AIRWAY RESISTANCE
The overall reduction in all components of lung
volume during anesthesia results in a reduced cal-
iber of airway, which increases airway resistance
and any tendency toward airway collapse (Fig. 3-
46). The relationship between airway resistance
and lung volume is well established (Fig. 3-47).
The decreases in FRC caused by the supine posi-
tion (about 0.8 L)
292
and by the induction of anes-
thesia (about 0.4 L) are often sufficient to explain
the increased resistance seen in the healthy, anes-
thetized patient.
272
In addition to this expected increase in airway
resistance in anesthetized patients, there are a
number of additional special potential sites of in-
creased airway resistance. These consist of the en-
dotracheal tube (if present), the upper and lower
airway passages, and the external anesthesia appa-
ratus. Endotracheal intubation reduces the size of
the trachea, usually by 30 to 50 per cent (see Fig.
3-46). However, the resistance to breathing
through an endotracheal tube in vivo may be
higher than in vitro measurements predict because
of secretions, head or neck position, and tube de-
formation, which may cause kinking and/or in-
creased airflow turbulence.
273
Pharyngeal obstruc-
tion, which can be considered to be a normal
feature of unconsciousness, is most common. A
minor degree of this type of obstruction occurs in
snoring. Laryngospasm and obstructed endotra-
1 02 General Respiratory Physiology and Respiratory Function During Anesthesia
Upright Awake Supine Anesthetized
Increased Airway Resistance
Figure 346 The anesthetized patient in the supine position has an increased airway resistance because of decreased functional
residual capacity (FRC), decreased caliber of the airways, endotracheal (ET) intubation, and connection of the endotracheal tube to
external breathing apparatus and circuitry.
cheal tubes (secretions, kinking, herniated cuffs)
are not rare and may be life-threatening.
Respiratory apparatus often causes resistance
that is considerably higher than the resistance in
the normal human respiratory tract (see Fig. 3-
46).
| : :
When a number of resistors such as those
shown in Figure 346 are joined in a series to
form an anesthetic gas circuit, they generally pro-
duce a larger resistance (as with resistances in an
electrical circuit). The increase in resistance asso-
ciated with commonly used breathing circuits and
endotracheal tubes may impose an additional work
of breathing that is two to three times normal.
274
f. SUPINE POSITION, IMMOBILITY, AND
EXCESSIVE INTRAVENOUS FLUID
ADMINISTRATION
Patients undergoing anesthesia and surgery are
often kept supine and immobile for long periods
of time. Thus, some of the lung may be continually
dependent and below the left atrium and therefore
in a zone 3 or 4 condition. Being in a dependent
position, the lung is predisposed to fluid accumu-
lation. Coupled with excessive fluid administra-
tion, conditions sufficient to promote transudation
of fluid into the lung are present and will result in
Figure 3-47 Airway resistance is an increasing hyperbolic func-
tion of decreasing lung volume. Functional residual capacity (FRC)
decreases in changing from the upright to supine position. (Re-
printed with modification by permission of the publisher from
Nunn JF: Applied Respiratory Physiology. 2nd ed. London. Butter-
worths (Publishers] Ltd, 1977.)
pulmonary edema and a decreased FRC. Figure 3-
48 shows that when mongrel dogs are placed in a
lateral decubitus position and are anesthetized for
several hours (bottom horizontal axis), expansion
of the extracellular space with fluid (top horizontal
axis) causes the Po
2
(left-hand axis) of blood
draining the dependent lung (closed circles) to de-
crease precipitously to mixed venous levels (no 0
2
uptake).
302
Blood draining the nondependent lung
maintains its Po
2
for a period of time but in the
face of the extracellular fluid expansion also suf-
fers a decline in its Po
2
after 5 hours. Transpul-
monary shunt (right-hand axis) progressively in-
creases. If the animals were turned every hour (and
received the same fluid challenge), only the depen-
dent lung, at the end of each hour period, suffered
a decrease in oxygenation. If the animals were
turned every half hour and received the same fluid
challenge, neither lung suffered a decrease in ox-
ygenation. In patients undergoing pulmonary re-
section (who therefore have, or will have, a re-
stricted pulmonary vascular bed) in the lateral
decubitus position who receive excessive intrave-
Figure 3-48 Mongrel dogs anesthetized with pentobarbital (bottom axis), placed in a lateral decubitus position, and subjected to
progressive extracellular fluid expansion (top axis) have a marked decrease in the Po
2
(left vertical axis) of blood draining the
dependent lung (solid circles) and a smaller, much slower decrease in Po
2
of blood draining the nondependent lung (open circles).
The pulmonary arteriovenous shunt (right ventrical axis) rises progressively (triangles). (Modified with permission from Ray JF,
Yost L, Moallem S, et al: Immobility, hypoxemia, and pulmonary arteriovenous shunting. Arch Surg 109:537, 1974. Copyright
1974, American Medical Association.)
increase in the thickness of the layer of mucus
proceeding distally from the cuff. Another possi-
bility is that mechanical distention of the trachea
by the endotracheal tube cuff initiated a neuro-
genic reflex are that altered mucous secretions or
frequency of ciliary beating.
Other investigators showed that when all of the
preceding factors are controlled, halothane revers-
ibly and progressively decreases, but does not
stop, mucous flow over an inspired concentration
of 1 to 3 MAC.
317
The halothane-induced depres-
sion of mucociliary clearance was likely due to
depression of the ciliary beat, an effect that caused
slow clearance of mucus from the distal and pe-
ripheral airways. In support of this hypothesis is
the fact that cilia are morphologically similar
throughout the animal kingdom, and in clinical
dosages, inhaled anesthetics, including halothane,
have been found to cause reversible depression of
the ciliary beat of protozoa.
318
5. Decreased Cardiac Output and
Increased Oxygen Consumption
A decreased cardiac output (Q
t
) in the presence
of a constant 0
2
consumption (Vo
2
), or an in-
creased Vo^ in the presence of a constant Q or a
decreased Q, and an increased Vo
2
must all result
in a lower mixed venous 0
2
content (C
0
2
). The
lowered Q0
2
will then flow through whichever
shunt pathways exist, mix with the oxygenated
end-pulmonary capillary blood, and lower the 0
2
content of the arterial blood (C
a
0
2
) (Figs. 3-35, 3-
36, and 3-37). Decreased Q, may occur with myo-
cardial failure and hypovolemia; the specific
causes of these two conditions is beyond the scope
of this chapter. Increased Vo
2
may occur with ex-
cessive sympathetic nervous system stimulation,
hyperthermia, or shivering and can further contrib-
ute to impaired oxygenation of arterial blood.
319
6. Inhibition of Hypoxic Pulmonary
Vasoconstriction
Decreased regional alveolar Po
2
causes regional
pulmonary vasoconstriction, which diverts blood
flow away from hypoxic regions of the lung to
better ventilated normoxic regions of the lung. The
diversion of blood flow minimizes venous admix-
ture from the underventilated or nonventilated
lung regions. Inhibition of regional HPV might
impair arterial oxygenation by permitting in-
creased venous admixture from hypoxic or atelec-
tatic areas of the lung (Fig. 3-14). Chapters 4 and
9 discuss in detail the determinants (physiologic
variables and anesthetic drugs, respectively) of the
amount of HPV during one-lung ventilation. The
next paragraph briefly summarizes most of these
regional HPV determinants.
Since the pulmonary circulation is poorly en-
dowed with smooth muscle, any condition that
increases the pressure against which the vessels
must constrict (i.e., the P
pa
) will decrease HPV.
There are numerous clinical conditions that can
increase P
n
, and therefore decrease HPV. Mitral
pa
stenosis,
320
volume overload,
320
low (but above
room air) F,0
2
in nondiseased lung,
321
a progres-
sive increase in the amount of diseased lung
thromboembolism,
321
hypothermia,
322
and vasoac-
tive drugs can all increase P
pa
. Direct vasodilating
drugs (such as isoproterenol, nitroglycerin, and
nitroprusside),
91
323
inhaled anesthetics,
324
325
and
hypocapnia
283
can directly decrease HPV. The se-
lective application of PEEP to only the nondis-
eased lung can selectively increase nondiseased
lung pulmonary vascular resistance and divert
blood flow back into the diseased lung.
326
7. Paralysis
In the supine position, the weight of the abdom-
inal contents pressing against the diaphragm is
greatest in the dependent or posterior part of the
diaphragm and least in the nondependent or ante-
rior part of the diaphragm (Fig. 3-44). In the
awake patient breathing spontaneously, the active
tension in the diaphragm is capable of overcoming
the weight of the abdominal contents, and the dia-
phragm moves the most in the posterior portion
(because the posterior diaphragm is stretched
higher into the chest it has the smallest radius of
curvature and therefore it contracts most effec-
tively) and least in the anterior portion. This is a
healthy circumstance because the greatest amount
of ventilation occurs where there is the most per-
fusion (posteriorly or dependently), and the least
amount of ventilation occurs where there is the
least perfusion (anteriorly or nondependently).
During paralysis and positive-pressure breathing,
the passive diaphragm is displaced by the positive
pressure preferentially in the anterior nondepen-
dent portion (where there is the least resistance to
diaphragmatic movement) and is displaced mini-
mally in the posterior dependent portion (where
there is the most resistance to diaphragmatic
movement). This is an unhealthy circumstance be-
cause the greatest amount of ventilation now oc-
curs where there is the least perfusion, and the
least amount of ventilation now occurs where there
is the most perfusion.
299
However, data from the prone-suspended posi-
tion cannot be explained in the same way.
327
328
Prone, the dorsal, nondependent regions of the dia-
phragm exhibit greater displacement during both
spontaneous breathing and mechanical ventilation.
Paralysis did not have an impact on dependent
diaphragm motion nearly as much in the prone-
suspended as in the supine position largely be-
cause the dependent zone was not moving too well
to start with when subjects were prone. It is pos-
sible that the costal diaphragm (i.e., the dependent
part when prone) simply cannot load compensate
as effectively as does the crural diaphragm (i.e.,
the dependent part when supine). Therefore, al-
though dependent zone displacement of the dia-
phragm will always be greater during spontaneous
breathing as opposed to mechanical ventilation,
the magnitude of the change in pattern with paral-
ysis will vary with body position.
8. Right-to-Left Interatrial Shunting
Acute arterial hypoxemia from a transient right-
to-left shunt through a patent foramen ovale has
been described, particularly during emergence
from anesthesia.
329,33
However, unless there is a
real-time technique of imagining the cardiac cham-
bers (color flow Doppler mapping), it is difficult
to document an acute and transient right-to-left
intracardiac shunt as a cause of arterial hypoxemia.
Nevertheless, right-to-left shunting through a pat-
ent foramen ovale has been described in virtually
every conceivable clinical situation that afterloads
the right heart and increases right atrial pressure
(see, e.g., Section I.C.4.).
9. Differential Diagnosis of Cyanosis
It is generally accepted that cyanosis is apparent
in the skin and mucous membranes when there is
5 g'dl
-1
of deoxyhemoglobin present. Cyanosis
may be peripheral or central. Peripheral cyanosis
is due to reduced flow in the extremities, with
subsequent desaturation. It is characteristically
seen in the nail beds and the nose. The arterial
blood will be red. Central cyanosis is observed in
the mucous membranes of the mouth, the lips, and
the conjunctiva. The arterial blood will not be red.
The differential diagnosis of cyanosis is given
in Table 3-8. The most common cause of intra-
operative cyanosis is hypoxemia and can be deter-
mined on-line by S
P
0
2
. Polycythemia can be deter-
mined with reference to the patient's preoperative
hemoglobin. The third, and much rarer cause, is a
low-affinity hemoglobin. These hemoglobin mu-
tants have oxygen dissociation curves shifted to
the right (e.g., Hb Kansas has a P
50
of 70). This
form of cyanosis should be noticeable preopera-
tively.
Methemoglobin is the hemoglobin molecule
with the heme moiety in the oxidized or ferric
Table 3-8 DIFFERENTIAL DIAGNOSIS OF
CYANOSIS
1. Deoxyhemoglobin
A. Hypoxemia
B. Polycythemia
C. Hemoglobin with increased P
50
2. Methemoglobinemia
A. Congenital
1. Enzyme deficiency
2. Cytochrome b
5
deficiency
3. Hemoglobin M
. Acquired
1. Toxic (drugs, local anesthetics, amyl nitrite,
nitroglycerin, phenacetin, sulfur drugs)
3. Sulfhemoglobinemia
4. Peripheral (stagnant blood flow)
state. Normally, less than 0.15 g-dl
-1
of hemoglo-
bin is in this form, and when methemoglobin lev-
els rise to 1.5 to 2.0 g*dl~' cyanosis can be de-
tected clinically.
331
Patients are often described as
having a grayish hue, and the arterial blood is
chocolate brown in color. Symptoms do not usu-
ally occur until levels are greater than 4 g-dl
-1
.
Patients who are symptomatic with methemoglo-
binemia can be treated with intravenous methylene
blue. The initial dose should be 1 mg*kg
_1
over 5
min. If no effect is seen in 30 to 60 min, a repeat
dose of 2 mg'kg
-1
may be given.
331
However, met-
hemoglobinemia patients who have a deficiency of
glucose-6 phosphate dehydrogenase are unaffected
by methylene blue.
Sulfhemoglobinemia is the final cause of cy-
anosis. It is a green-pigmented molecule with a
sulfur atom incorporated in the porphyrin ring. It
results from irreversible oxidation of hemoglobin
by drugs or chemicals. Patients with sulfhemoglo-
binemia can be profoundly cyanosed but have
minimal dyspnea. This is because cyanosis can be
detected at a blood level of 0.5 g*dl~ ', lower than
methemoglobin (1.5-2.0 g'dl
-1
) or deoxyhemo-
globin (5.0 g-dl
-1
), and, in addition, sulfhemoglo-
bin moves the P
50
of the oxygen dissociation curve
to the right, enhancing oxygen delivery to the tis-
sues. The arterial blood is described as having a
mauve or lavender hue.
10. Involvement of Mechanisms of
Hypoxemia in Specific Diseases
In any given pulmonary disease, many of the
mechanisms of hypoxemia just listed may be in-
volved in producing hypoxemia. Pulmonary em-
bolism (air, fat, thrombi) (Fig. 3-49) and the evo-
lution of the adult respiratory distress syndrome
(Fig. 3-50) are used here to illustrate this point. A
Pulmonary Embolus: Mechanisms of Hypoxemia
Figure 3-49 Mechanisms of hypoxemia during pulmonary embolism. See text for explanation of the pathophysiologic flow
diagram. (CAP PERM = capillary permeability; AV = arteriovenous; HPV = hypoxic pulmonary vasoconstriction; PA =
pulmonary artery; FRC = functional residual capacity; CC = closing capacity.)
Shock and ARDS: Mechanisms of Hypoxemia
Figure 3-50 Mechanisms of hypoxemia during the adult respiratory distress syndrome (ARDS). See text for explanation of
pathophysiologic flow diagram. (PA - pulmonary artery; CAP PERM = capillary permeability; HPV = hypoxic pulmonary
vasoconstriction; FRC = functional residual capacity; CC = closing capacity.)
significant pulmonary embolus
332
can cause severe
increases in pulmonary artery pressure, and these
increases can cause right-to-left transpulmonary
shunting through opened arteriovenous anastomo-
ses and the foramen ovale (this is possible in 20
per cent of patients), pulmonary edema in nonem-
bolized regions of the lung, and inhibition of HPV.
The embolus may cause hypoventilation via in-
creased dead-space ventilation. If the embolus
contains platelets, serotonin may be released, and
this release can cause hypoventilation via broncho-
constriction and pulmonary edema via increased
pulmonary capillary permeability. Finally, the pul-
monary embolus increases pulmonary vascular re-
sistance (in proportion with the degree of vascular
obstruction; it may be also increased by active
vasoconstriction) and decreases the cardiac output.
Following major hypotension, shock, or blood
loss, respiratory failure often ensues, and this syn-
drome has been called ARDS. This syndrome can
evolve during and after anesthesia and has the
hallmark characteristics of decreased FRC and
compliance and hypoxemia. Following shock and
trauma, increased plasma levels of serotonin, his-
tamine, plasmakinins, lysozymes, superoxides, fi-
brin degradation products, products of comple-
ment metabolism, and fatty acids occur. Sepsis and
endotoxemia may be present. Increased levels of
activated complement activate neutrophils into
chemotaxis in patients with trauma and pancreati-
tis; activated neutrophils can damage endothelial
cells. These factors, along with pulmonary contu-
sion (if it occurs), may individually or collectively
increase pulmonary capillary permeability. Fol-
lowing shock, it has been shown that acidosis,
increased circulating catecholamines and sympa-
thetic nervous system activity, prostaglandin re-
lease, histamine release, microembolism (with
serotonin release), increased intracranial pressure
(with head injury), and alveolar hypoxia may oc-
cur and may individually or collectively, particu-
larly postresuscitation, cause a moderate increase
in pulmonary artery pressure. Following shock, the
normal compensatory response to hypovolemia is
movement of a protein-free fluid from the intersti-
tial space into the vascular space in order to restore
vascular volume. The dilution of vascular proteins
by protein-free interstitial fluid can cause a de-
creased capillary colloid osmotic pressure. In-
creased pulmonary capillary permeability and pul-
monary artery pressure along with decreased
capillary colloid osmotic pressure will cause fluid
transudation and pulmonary edema. Additionally,
a decreased cardiac output, inhibition of HPV, im-
mobility, supine position, excessive fluid adminis-
tration, and an excessively high F,0
2
can contrib-
ute to the development of ARDS.
G. Mechanisms of Hypercapnia and
Hypocapnia During Anesthesia
1. Hypercapnia (Fig. 3-51)
The following factors can all cause hypercapnia.
a. HYPOVENTILATION
Patients spontaneously hypoventilate during an-
esthesia because it is more difficult to breathe (ab-
normal surgical position, increased airway resis-
tance, decreased compliance), and they are less
willing to breathe (decreased respiratory drive
caused by anesthetics). Hypoventilation will result
in hypercapnia (see Fig. 3-30).
b. INCREASED DEAD-SPACE VENTILATION
253
A decrease in pulmonary artery pressure, as dur-
ing deliberate hypotension,
333
may cause an in-
crease in zone 1 and alveolar dead-space ventila-
tion. An increase in airway pressure (as with
PEEP) may cause an increase in zone 1 and alveo-
lar dead-space ventilation. Pulmonary embolus,
thrombosis, and vascular obliteration (kinking,
clamping, blocking of pulmonary artery during
surgery) may increase the amount of lung that is
ventilated but unperfused. Vascular obliteration
may be responsible for the increase in dead-space
ventilation with age (, /
= 33 + age/3).
Rapid, short inspirations may be distributed pref-
erentially to noncompliant (short time constant for
inflation) and badly perfused alveoli, while a slow
inspiration allows time for distribution to more
compliant (long time constant for inflation) and
better perfused alveoli. Thus, rapid, short inspira-
tions may have a dead-space ventilation effect.
The anesthesia apparatus increases total dead
space (Vc/V-r) for two reasons. First, the apparatus
simply increases the anatomic dead space. Inclu-
sion of normal apparatus dead space increases the
total Vj/V
T
ratio from 33 per cent to about 46 per
cent in intubated patients and to about 64 per cent
in patients breathing via a mask.
334
Second, anes-
thesia circuits cause rebreathing of expired gases,
which is the equivalent to dead-space ventilation.
The rebreathing classification by Mapleson is
widely accepted.
335
The order of increasing re-
breathing (decreasing clinical merit) with sponta-
neous ventilation with Mapleson circuits is A (Ma-
gill), D, C, and B. The order of increasing
rebreathing (decreasing clinical merit) with con-
trolled ventilation is D, B, C, and A. There will be
no rebreathing in system (Ayre T-piece) if the
patient's respiratory diastole is long enough to per-
mit washout with a given fresh gas flow (common
event) or if the fresh gas flow is greater than the
peak inspiratory flow rate (uncommon event).
Mechanisms of Hypercapnia during Anesthesia
The effects of an increase in dead space can
usually be counteracted by a corresponding in-
crease in the respiratory minute volume. If, for
example, the minute volume is 10 L/min and the
Vp/ ratio is 30 per cent, the alveolar ventilation
will be 7 L/min. If a pulmonary embolism oc-
curred resulting in an increase of the V/VT ratio
to 50 per cent, the minute volume would need to
be increased to 14 L/min to maintain an alveolar
ventilation of 7 L/min (14 L/min X 0.5).
c. INCREASED CARBON DIOXIDE PRODUCTION
All of the causes of increased 0
2
consumption
will also increase C0
2
production, namely, hyper-
thermia, shivering, catecholamine release (light
anesthesia), hypertension, and thyroid storm. If
minute ventilation, total dead space, and ventila-
tion-perfusion relationships are constant, an in-
crease in C0
2
production will result in hypercap-
nia.
d. INADVERTENT SWITCHING OFF OF A
CARBON DIOXIDE ABSORBER
Many factors, such as patient ventilatory re-
sponsiveness to C0
2
accumulation, fresh gas flow,
circle system design, and C0
2
production, deter-
mine whether hypercapnia will result from inad-
vertent switching off or using up of a circle CO
:
absorber. However, high fresh-gas flows (>5
L/min) minimize the problem with almost all sys-
tems for almost all patients.
2. Hypocapnia
The mechanisms of hypocapnia are the reverse
of those that produce hypercapnia. Thus, with all
other factors being equal, hyperventilation (spon-
taneous or controlled ventilation), decreased dead-
space ventilation (change from mask airway to
endotracheal tube airway, decreased PEEP, in-
creased pulmonary artery pressure, or decreased
110
pends on the severity of the hypoxia; the severity
of hypoxia determines the magnitude of and bal-
ance between the inhibitory and excitatory com-
ponents; the balance may vary according to the
type and depth of anesthesia and the degree of pre-
existing cardiovascular disease.
Mild arterial hypoxemia (arterial saturation less
than normal but still 80 per cent or higher) causes
a general activation of the sympathetic nervous
system and release of catecholamines. Conse-
quently, heart rate, mean circulatory pressure (and
index of venous tone and cardiac preload), sys-
temic blood pressure, stroke volume, cardiac out-
put, and myocardial contractility (as measured by
a shortened pre-ejection period [PEP], left ventric-
ular ejection time [LVET], and a decreased
PEP/LVET ratio) are increased (Fig. 3-52).
339
In
healthy, awake humans, most of the increase in
cardiac output with mild isocapnic hypoxemia is
due to the increase in heart rate.
340
Changes in
systemic vascular resistance are usually
slight.
336
340
However, in patients under anesthesia
with beta blockers, hypoxia (and hypercapnia
when present) may cause circulating catechola-
mines to have only an alpha-receptor effect, and
the heart may be unstimulated (even depressed by
a local hypoxic effect), and systemic vascular re-
sistance may be increased. Consequently, cardiac
output may be decreased in these patients. With
moderate hypoxemia (arterial oxygen saturation 60
to 80 per cent), local vasodilatation begins to pre-
dominate, and systemic vascular resistance and
blood pressure decrease, but heart rate may con-
tinue to be increased because of a systemic hyper-
tension-induced stimulation of baroreceptors.
However, even in halothane-anesthetized sheep,
moderate hypoxemia does not cause left ventricu-
lar contractility dysfunction.
341
Finally, with severe
hypoxemia (arterial saturation less than 60 per
cent), local depressant effects dominate, and blood
pressure falls rapidly, the pulse slows, shock de-
velops, and the heart either fibrillates or becomes
asystolic. It should be remembered that significant
pre-existing hypotension will convert a mild hy-
rebreathing), and decreased C0
2
production (hy-
pothermia, deep anesthesia, hypotension) will lead
to hypocapnia. By far, the most common mecha-
nism of hypocapnia is passive hyperventilation by
mechanical means.
H. Physiologic Effects of Abnormalities
in the Respiratory Gases
1. Hypoxia
The end-products of aerobic metabolism (oxi-
dative phosphorylation) are carbon dioxide and
water, both of which are easily diffusible and lost
from the body. The essential feature of hypoxia is
the cessation of oxidative phosphorylation when
mitochondrial Po
2
falls below a critical level. An-
aerobic pathways, which produce energy (ATP)
inefficiently, are then utilized. The main anaerobic
metabolites are hydrogen and lactate ions, which
are not easily excreted. They accumulate in the
circulation where they may be quantified in terms
of the base deficit and the lactate-pyruvate ratio.
Since the various organs have different blood
flow and oxygen consumption rates, the presenta-
tion and clinical diagnosis of hypoxia are usually
related to symptoms arising from the most vulner-
able organ. This is usually the brain in an awake
patient and the heart in the anesthetized patient
(see the following), but in special circumstances it
may be the spinal cord (aortic surgery), kidney
(acute tubular necrosis), liver (hepatitis), or limb
(claudication, gangrene).
The cardiovascular response to acute hypoxemia
is a product of both reflex (neural and humoral)
and direct effects (Table 3-9).
336
-
338
The reflex ef-
fects occur first and are excitatory and vasocon-
strictive. The neuroreflex effects result from aortic
and carotid chemoreceptor, baroreceptor, and cen-
tral cerebral stimulation, and the humoral reflex
effects result from catecholamine and renin-angio-
tensin release. The direct local vascular effects of
hypoxia are inhibitory and vasodilatory and occur
late. The net response to hypoxia in a subject de-
Figure 3-52 Changes in the minute ventilation and circula-
tion of normal awake humans during progressive isocapnic
hypoxia and hyperoxic hypercapnia. V
F
= expired minute ven-
tilation; Q = minute cardiac output; ,.,, = end-tidal Po
2
;
P
ET
CO
:
= end-tidal Pco
2
; S, = slope during the first phase of
slowly increasing ventilation and/or circulation; S
:
= slope
during the second phase of sharply increasing ventilation
and/or circulation. (From Serebrouskaya TV: Comparison of
respiratory and circulatory human responses to progressive hy-
poxia and hypercapnia. Respiration 59:35-41, 1992. Used with
permission.)
poxemia-hemodynamic profile into a moderate hy-
poxemia-hemodynamic profile, and a moderate
hypoxemia-hemodynamic profile will convert into
a severe hypoxemia-hemodynamic profile. Simi-
larly, in well-anesthetized and/or well-sedated pa-
tients, early sympathetic nervous system reactivity
to hypoxemia may be reduced,
342
and the effects
of hypoxemia may be expressed only as bradycar-
dia with severe hypotension and, ultimately, cir-
culatory collapse.
Hypoxemia may also cause cardiac arrhythmias,
and the cardiac arrhythmias may in turn potentiate
the already mentioned deleterious cardiovascular
effects. Hypoxemia-induced arrhythmias may be
caused by multiple mechanisms; the mechanisms
are inter-related by virtue of the fact that they all
cause a decrease in the myocardial oxygen supply-
demand ratio, which in turn increases myocardial
irritability. First, arterial hypoxemia may directly
decrease myocardial oxygen supply. Second, early
tachycardia may cause an increased myocardial
oxygen consumption, and a decreased diastolic
filling time may cause a decreased myocardial ox-
ygen supply. Third, early increased systemic blood
pressure may cause an increased afterload on the
left ventricle, which increases left ventricular oxy-
gen demand. Fourth, late systemic hypotension
may decrease myocardial oxygen supply owing to
decreased diastolic perfusion pressure. Fifth, coro-
nary blood flow reserve may be exhausted by a
late maximally increased coronary blood flow (as
a result of maximal coronary vasodilation).
341
The
level of hypoxemia that will cause cardiac arrhyth-
mias cannot be predicted with certainty because
the myocardial oxygen supply and demand rela-
tionship in a given patient is not known (i.e., the
degree of coronary artery atherosclerosis may not
be known). However, if a myocardial area (or
areas) becomes hypoxic and/or ischemic, unifocal
or multifocal premature ventricular contractions,
ventricular tachycardia, and ventricular fibrillation
may occur.
The cardiovascular response to hypoxia includes
a number of other important effects. Cerebral
blood flow increases (even if hypocapnic hyper-
ventilation is present). Ventilation will be stimu-
lated no matter why hypoxia exists. Not surpris-
ingly, acute progressive isocapnic hypoxemia
causes bronchodilation in both normal subjects
and in patients with chronic lung disease (includ-
ing asthma).
343
The pulmonary distribution of
blood flow is more homogeneous owing to an in-
creased pulmonary pressure.
The most common causes of chronic hypoxia in
humans (excluding existence at high attitude) are
pulmonary and/or cardiac problems that lead to
insidious reduction of arterial blood and tissue ox-
ygen tension, such as chronic obstructive pulmo-
nary disease and cardiac failure. The most impor-
tant effects of chronic hypoxemia are hematologic
(increased RBC mass) and increased minute ven-
tilation.
344
The higher hemoglobin concentration
and hematocrit found after prolonged exposure to
hypoxia are believed to be caused by a sustained
increase in renal release of erythropoietin, the gly-
coprotein hormone that stimulates the formation of
RBCs. Chronic hypoxia does not affect the stan-
dard oxygen consumption of large mammals. To
maintain a normal level of oxygen consumption, a
certain degree of hyperventilation must be main-
tained. As hypoxemia increases, the ventilation re-
quired to maintain the same amount of oxygen
consumed per unit of time must increase.
344
The
bulk of evidence shows that, at rest, cardiac output
in chronic hypoxia is normal. Systemic oxygen
transport (i.e., the product of 0
2
content times car-
diac output) is maintained within normal levels
because the absolute increase in hemoglobin sus-
tains an unchanged arterial oxygen content. Al-
though chronic hypoxia will cause a right-shifted
oxygen-hemoglobin dissociation curve (because of
either an increase in 2,3-DPG or acidosis), which
is an effect that increases tissue Po
2
, oxygen up-
take by tissue may still be impaired, possibly be-
cause of increased blood viscosity.
2. Hyperoxia (Oxygen Toxicity)
The dangers associated with the inhalation of
excessive oxygen are multiple. Exposure to a high
0
2
tension clearly causes pulmonary damage in
healthy individuals.
345
346
Because pulmonary tis-
sue Po
2
is directly determined by P
A
0
2
, arterial
hypoxemia does not delay the onset of oxygen
toxicity at 1 atmosphere of pressure.
347
A dose-
time toxicity curve for humans is available from a
number of studies.
345
Because the lungs of normal
human volunteers cannot be directly examined to
determine the rate of onset and the course of tox-
icity, indirect measures such as the onset of symp-
toms have been used to construct the dose-time
toxicity curves. Examination of the curve indicates
that 200 to 500 per cent oxygen (2 to 5 atmos-
pheres) should not be administered for more than
6 hours, 100 per cent oxygen should not be admin-
istered for more than 12 hours, 80 per cent oxygen
should not be administered for more than 24 hours,
and 60 per cent oxygen should not be administered
for more than 36 hours.
348
No measurable changes
in pulmonary function or blood-gas exchange oc-
cur in humans during exposures to less than 50 per
cent oxygen or less even for long periods.
348
Nev-
ertheless, it is important to note that in the clinical
setting, these dose-time toxicity relationships are
often generally obscured
347
because of the complex
multivariable nature of the clinical setting.
The dominant first symptom of oxygen toxicity
in human volunteers is substernal distress, which
begins as a mild irritation in the area of the carina
and may be accompanied by occasional cough-
ing.
349
As exposure continues, pain becomes more
intense, and the urge to cough and to deep breathe
also becomes more intense. These symptoms will
progress to severe dyspnea, paroxysmal coughing,
and decreased vital capacity when the F,0
2
has
been 1.0 for greater than 12 hours. At this point,
recovery of mechanical lung function usually oc-
curs within 12 to 24 hours but may require more
than 24 hours in some individuals.
348
No individual
measure of pulmonary function has been found to
be uniquely satisfactory for monitoring rates of
development or reversal of pulmonary oxygen poi-
soning. As toxicity progresses, pulmonary func-
tions studies such as compliance and arterial blood
gases deteriorate. Pathologically, in animals the
lesion progresses from a tracheobronchitis (expo-
sure for 12 hours to a few days) to involvement of
the alveolar septa with pulmonary interstitial
edema (exposure for a few days to a week) to
pulmonary fibrosis of the edema (exposure greater
than a week).
350
Ventilatory depression may occur in those pa-
tients who, by reason of drugs or disease, have
been ventilating in response to a hypoxic drive. By
definition, ventilatory depression resulting from
removal of a hypoxic drive by increasing the in-
spired oxygen concentration will cause hypercap-
nia but does necessarily produce hypoxia (because
of the increase in F,0
2
).
Absorption atelectasis has been previously dis-
cussed (see High Inspired Oxygen Concentration
and Absorption Atelectasis) and is not covered
here. Retrolental fibroplasia, an abnormal prolifer-
ation of the immature retinal vasculature of the
prematurely born infant, can occur following ex-
posure to hyperoxia. Very premature infants are
most susceptible to retrolental fibroplasia (i.e.,
those of less than 1.0 kg birth weight and 28 weeks
gestation). The risk of retrolental fibroplasia exists
whenever an F,0
2
causes a P
a
0
2
of > 80 mm Hg
for more than 3 hours in an infant whose gesta-
tional age plus life age is less than 44 weeks. If
the ductus arteriosus is patent, arterial blood sam-
ples should be drawn from the right radial artery
(umbilical or lower extremity P
a
0
2
is lower than
the P
a
0
2
to which the eyes are exposed owing to
ductal shunting of unoxygenated blood [see chap-
ter 18]).
The mode of action of toxicity of oxygen in
tissues is complex, but interference with metabo-
lism seems to be widespread. Most important,
there is oxygen free radical (partially reduced ox-
ygen) inactivation of many enzymes, particularly
those with sulfhydryl groups.
347
Neutrophil recruit-
ment and release of mediators of inflammation oc-
cur next and greatly accelerate the extent of endo-
thelial and epithelial damage and impairment of
the surfactant systems.
347
The most acute toxic en-
zyme effect of oxygen in humans is a convulsive
effect that occurs during exposure to pressures in
excess of 2 atmospheres absolute.
The free radical theory has provided the princi-
ples on which to evaluate and possibly augment
tolerance to elevated oxygen pressures.
351
First,
various regimens of intermittent or pre-exposure to
oxygen or adaptation to hypoxia have been shown,
at least in some models, to augment oxygen toler-
ance. The apparent basis for increased tolerance is
repair of toxic effects and/or augmentation of an-
tioxidant defenses. A second approach is the ma-
nipulation of nutritional status, because dietary de-
ficiency of antioxidant factors can decrease
tolerance. These dietary components include pro-
tein to provide essential amino acids for synthesis
of glutathione, selenium as a component of gluta-
thione peroxidase, and alpha-tocopherol (vitamin
E) as a radical scavenger. Vitamins C and A may
also have important antioxidant roles. A third strat-
egy for increasing oxygen tolerance is to augment
the cellular antioxidant defenses. This very active
area of investigation has seen several major devel-
opments. The administration of antioxidant en-
zymes has been evaluated. Liposomes and RBCs
have been tried as delivery systems to promote
entry of exogenous enzymes into cellular compart-
ments. The administration of sulfhydryl antioxi-
dants has also shown promise with development
of newer agents that are internalized by cells. For
the future, genetic engineering may offer a more
suitable method for altering antioxidant defenses.
A final strategy for extending tolerance is to treat
the toxic manifestations of oxygen to prevent
death or disability. Although this approach is
clearly less appealing than the methods discussed
previously, it may have the most immediate prac-
ticality for extending oxygen tolerance. Possible
therapies include administration of exogenous sur-
factant or use of anti-inflammatory and antifibrotic
drugs.
High inspired oxygen concentrations can be of
use therapeutically. Clearance of gas loculi in the
body may be greatly accelerated by inhalation of
100 per cent oxygen. Inhalation of 100 per cent
oxygen creates a large N
2
gradient from the gas
space to the perfusing blood. As a result, N
2
leaves
the gas space, and the space diminishes in size.
The technique of using oxygen to remove gas may
be used to ease intestinal gas pressure in patients
with intestinal obstruction, to hasten recovery from
pneumoencephalography, to decrease the size of
an air embolus, and to aid in absorption of pneu-
moperitoneum and pneumothorax.
3. Hypercapnia
The effects of carbon dioxide on the cardiovas-
cular system are as complex as they are for hy-
poxia. As with hypoxemia, hypercapnia appears to
cause direct depression of both the cardiac muscle
and the vascular smooth muscle. At the same time,
however, it causes reflex stimulation of the sym-
pathoadrenal system, which compensates to a
greater or lesser extent for the primary cardiovas-
cular depression.
336
341
With moderate hypercapnia,
a hyperkinetic circulation results, with increased
cardiac output and systemic blood pressure (Fig.
3-52).
339
352
Even in patients under halothane an-
esthesia, plasma catecholamine levels increase in
response to increased C0
2
levels in much the same
way as they do in conscious subjects. Thus, hyper-
capnia, like hypoxemia, may cause increased myo-
cardial oxygen demand (tachycardia, early hyper-
tension) and decreased myocardial oxygen supply
(tachycardia, late hypotension).
Table 3-10 summarizes the interaction of anes-
thesia with hypercapnia in humans: The increased
cardiac output is maintained during anesthesia, but
the systemic blood pressure and vascular resis-
tance are decreased.
352
353
The increase in cardiac
output is most marked during anesthesia with
drugs and enhances sympathetic activity and is
least marked with halothane and nitrous oxide.
The decrease in systemic vascular resistance is
most marked during anesthesia with enflurane and
accompanying hypercapnia.
Arrhythmias have been reported in unanesthe-
tized humans during acute hypercapnia but have
seldom been of serious importance. A high P
a
CO
:
level, however, is more dangerous during general
anesthesia, and with halothane anesthesia arrhyth-
mias will frequently occur above a P
a
C0
2
ar-
rhythmic threshold, which is often constant for a
particular patient.
The maximal stimulant respiratory effect is at-
tained by a P
a
C0
2
of about 100 mm Hg. With a
higher P
a
C0
2
, stimulation is reduced, and at very
high levels respiration is depressed and later ceases
altogether. The Pco
2
-ventilation response curve is
generally displaced to the right, and its slope is
reduced by anesthetics and other depressant
drugs.
354
With profound anesthesia the response
curve may be flat, or even sloping downward, and
carbon dioxide then acts as a respiratory depres-
sant. In patients with ventilatory failure, carbon
dioxide narcosis occurs when the P
a
C0
2
rises
above 90 to 120 mm Hg. Thirty per cent carbon
dioxide is sufficient for the production of anesthe-
sia, and this concentration causes total but revers-
ible flattening of the electroencephalogram.
355
As
expected, hypercapnia causes bronchodilation in
normal and lung disease patients.
341
Quite apart from the effect of carbon dioxide on
ventilation, hypercapnia exerts three other impor-
tant effects that influence the oxygenation of the
blood and tissue. First, if the concentration of ni-
trogen (or other "inert" gas) remains constant, the
concentration of C0
2
in the alveolar gas can only
increase at the expense of 0
2
, which must be dis-
placed. Thus, P
A
0
2
and P
a
0
2
may decrease.
Second, hypercapnia shifts the oxygen-hemoglo-
bin curve to the right, facilitating tissue oxygena-
tion. Third, hypercapnia may depress spontaneous
diaphragmatic function.
356
Chronic hypercapnia results in increased resorp-
tion of bicarbonate by the kidneys, further raising
the plasma bicarbonate level and constituting a
secondary or compensatory "metabolic alka-
losis." Chronic hypocapnia decreases renal bicar-
bonate resorption, resulting in further fall of
plasma bicarbonate and producing a secondary or
compensatory "metabolic acidosis." In each case,
arterial pH returns toward the normal value, but
the bicarbonate ion concentration departs even fur-
ther from normal.
Hypercapnia is accompanied by a leakage of
potassium from the cells into the plasma. A good
deal of the potassium comes from the liver, prob-
ably from glucose release and mobilization, which
occurs in response to the rise in plasma catechola-
mine levels.
357
Since the plasma potassium level
takes an appreciable time to return to normal, re-
peated bouts of hypercapnia at short intervals re-
sult in a stepwise rise in plasma potassium.
4. Hypocapnia
In this section, hypocapnia is considered to be
produced by passive hyperventilation (by the anes-
thesiologist or ventilator).
Hypocapnia may cause a decrease in the cardiac
output by three separate mechanisms. First, if
present, an increase in intrathoracic pressure will
decrease the cardiac output. Second, hypocapnia is
associated with a withdrawal of sympathetic ner-
vous system activity, and this can decrease the
ionotropic state of the heart. Third, hypocapnia can
increase pH, which can in turn decrease ionized
Ca
++
, which may in turn decrease the ionotropic
state of the heart. Hypocapnia with an alkalosis
will also shift the oxygen-hemoglobin curve to the
left, which increases the hemoglobin affinity for
0
2
, which will impair O, unloading at the tissue
level. The decrease in peripheral flow and im-
paired ability to unload oxygen to the tissues is
compounded by an increase in whole body oxygen
consumption caused by an increased pH-mediated
uncoupling of oxidation from phosphorylation
357
;
P
a
C0
2
of 20 mm Hg will increase tissue 0
2
con-
sumption by 30 per cent. Consequently, hypocap-
nia may simultaneously increase tissue 0
2
demand
and decrease tissue 0
2
supply. Thus, in order to
have the same amount of 0
2
delivery to the tissues,
cardiac output or tissue perfusion has to increase
at a time when it may not be possible to do so. It
has been suggested that the cerebral effects of hy-
pocapnia may be related to a state of cerebral
acidosis and hypoxia, since hypocapnia may cause
a selective reduction in the cerebral blood flow
and also shifts the oxygen-hemoglobin curve to
the left.
Hypocapnia may cause V
A
/Q abnormalities by
inhibiting HPV or by causing bronchoconstriction
and a decreased lung compliance. Finally, passive
hypocapnia will produce apnea.
REFERENCES
1. West JB, Dollery CT, Naimark A: Distribution of blood
flow in isolated lung: Relation to vascular and alveolar
pressures. J Appl Physiol 19:713, 1964.
2. Puri GD, Venkataranan RK, Singh H, Jindal SK: Physio-
logical dead space and arterial to end-tidal CO\ difference
under controlled normocapnic ventilation in young an-
aesthetized subjects. Indian J Med Res 94:41-46, 1991.
3. Permutt S, Bramberger-Barnea B, Bane HN: Alveolar
pressure, pulmonary venous pressure and the vascular
waterfall. Med Thorac 19:239, 1962.
4. West JB, Dollery CT, Heard BE: Increased pulmonary
vascular resistance in the dependent zone of the isolated
dog lung caused by perivascular edema. Circ Res 17:191-
206, 1965.
5. West JB (ed): Regional Differences in the Lung. New
York, Academic Press, 1977.
6. Hughes JMB, Glazier JB, Maloney JE, et al: Effect of
lung volume on the distribution of pulmonary blood flow
in man. Respir Physiol 4:58-72, 1968.
7. Hughes JM, Glazier JB, Maloney JE, et al: Effect of
extra-alveolar vessels on the distribution of pulmonary
blood flow in the dog. J Appl Physiol 25:701-709, 1968.
8. Permutt S, Caldini P, Maseri A, et al: Recruitment versus
distensibility in the pulmonary vascular bed. In Fishman
AP, Hecht H (eds): The Pulmonary Circulation and Inter-
stitial Space. Chicago, University of Chicago Press, 1969,
pp 375-387.
9. Maseri A, Caldini P, Harward P, et al: Determinants of
pulmonary vascular volume. Recruitment versus distensi-
bility. Circ Res 31:218-228, 1972.
10. Hakim TS, Lisbona R, Dean GW: Gravity-independent
inequality in pulmonary blood flow in humans. J Appl
Physiol 63:1114-1121, 1987.
11. Hoppin FG Jr, Green ID, Mead J: Distribution of pleural
surface pressure. J Appl Physiol 27:863, 1969.
12. Milic-Emili J, Henderson JAM, Dolovich MB, et al: Re-
gional distribution of inspired gas in the lung. J Appl
Physiol 21:749, 1966.

CHAPTER 4
Special Respiratory Physiology
of the Lateral Decubitus
Position, the Open Chest, and
One-Lung Ventilation
I. Introduction
II. Physiology of Spontaneous
Ventilation With an Open Chest
A. Mediastinal Shift
B. Paradoxical Respiration
III. Physiology of the Lateral Decubitus
Position and the Open Chest During
Controlled Two-Lung Ventilation:
Distribution of Perfusion (Q) and
Ventilation (V)
A. Distribution of Q, V, V/QLateral
Decubitus Position, Awake, Closed
Chest
B. Distribution of Q, VLateral
Decubitus Position, Anesthetized,
Closed Chest
C. Distribution of Q, VLateral
Open Chest
D. Distribution of Q, VLateral
Open Chest, Paralyzed
E. Summary of Physiology of Lateral
Decubitus Position and the Open
Chest
IV. Physiology of One-Lung Ventilation
A. Comparison of Arterial Oxygenation
and C0
2
Elimination During Two-
Lung Versus One-Lung Ventilation
B. Blood Flow Distribution During One-
Lung Ventilation
Blood Flow to the Nondependent,
Nonventilated Lung
1. Distribution of Hypoxia
2. Low V/Q Versus Atelectasis
3. Vasodilator Drugs
4. Anesthetic Drugs
5. Cardiac Output and
Pulmonary Vascular
Pressure
6. PA
7. F,0
2
in the Normoxic
Compartment
8. Vasoconstrictor Drugs
9. Drugs That Increase HPV
(a) Almitrine
(b) Products of Arachidonic
Acid Metabolism
(Vasoconstrictor
Leukotrienes)
10. P
A
C0
2
and pH Changes
11. PEEP and Alveolar Pressure
12. Other Important Systemic
Factors (Hypertension, Age,
Sex, Infection)
Blood Flow to the Dependent,
Ventilated Lung
Miscellaneous Causes of
Hypoxemia During One-Lung
Ventilation
-. I
123
124 Special Respiratory Physiology of the Lateral Decubitus Position, the Open Chest, and One-Lung Ventilation
I. INTRODUCTION
It is not possible to have adequate gas exchange
during spontaneous ventilation with an open chest
because of the occurrence of mediastinal shift and
paradoxical respiration (see the "pneumothorax
problem" discussed in chapter 1). Consequently,
intrathoracic surgery cannot be performed with
spontaneous ventilation. The first part of this chap-
ter briefly reviews this physiology.
Patients undergoing thoracic surgery are usually
in the lateral decubitus position under general an-
esthesia, have an open chest wall (nondependent
hemithorax), are pharmacologically paralyzed,
and, of course, have ventilation controlled. Even if
both lungs are being ventilated, each of these an-
esthesia and surgical requirements can cause major
alterations in the distribution of perfusion (Q),
and/or ventilation (V), and ventilation-perfusion
relationships (V/Q) (compared with the awake
state in the upright position). The second part of
this chapter discusses the physiologic effects of
each one of these anesthetic and surgical events on
the distribution of Q, V, and V/Q during two-lung
ventilation.
In addition, a good deal of thoracic surgery must
be performed in the lateral decubitus position with
the nondependent lung nonventilated and the de-
pendent lung ventilated (one-lung ventilation).
One-lung ventilation imposes a new host of deter-
minants on the distribution of blood flow (and, of
course, ventilation). The third, and by far the larg-
est, part of this chapter considers the physiology
of one-lung ventilation.
II. PHYSIOLOGY OF
SPONTANEOUS VENTILATION
WITH AN OPEN CHEST
A. Mediastinal Shift
An examination of the physiology of the open
chest during spontaneous ventilation reveals why
controlled positive-pressure ventilation is the only
practical way to provide adequate gas exchange
during thoracotomy. (However, it should be noted
that spontaneous ventilation may be indicated in
patients with a large mediastinal mass.
1
) In the
spontaneously breathing, closed-chest patient in
the lateral decubitus position, gravity causes the
pleural pressure in the dependent hemithorax to be
less negative than that in the nondependent hemi-
thorax (see Fig. 3-5), but there is still negative
pressure in each hemithorax on each side of the
mediastinum. In addition, the weight of the me-
diastinum causes some compression of the lower
lung, contributing to the pleural pressure gradient.
With the nondependent hemithorax open, atmos-
pheric pressure in that cavity exceeds the negative
pleural pressure in the dependent hemithorax; this
imbalance of pressure on the two sides of the me-
diastinum causes a further downward displacement
of the mediastinum into the dependent thorax.
During inspiration, the caudad movement of the
dependent-lung diaphragm increases the negative
pressure in the dependent lung and causes a still
further displacement of the mediastinum into the
dependent hemithorax. During expiration, as the
dependent-lung diaphragm moves cephalad, the
pressure in the dependent hemithorax becomes rel-
atively positive, and the mediastinum is pushed
upward out of the dependent hemithorax (Fig. 4-
l).
2
Thus, the tidal volume in the dependent lung
is decreased by an amount equal to the inspiratory
displacement caused by mediastinal movement.
This phenomenon is called mediastinal shift and is
one mechanism that results in impaired ventilation
in the open-chest spontaneously breathing patient
in the lateral decubitus position. The mediastinal
shift can also cause circulatory changes (decreased
venous return) and reflexes (sympathetic activa-
tion) that result in a clinical picture similar to
shock: The patient is hypotensive, pale, and cold,
with dilated pupils. Local anesthetic infiltration of
the pulmonary plexus at the hilum and the vagus
nerve can diminish these reflexes. More practi-
cally, controlled positive-pressure ventilation abol-
ishes these ventilatory and circulatory changes as-
sociated with mediastinal shift.
B. Paradoxical Respiration
When a pleural cavity is exposed to atmospheric
pressure, the lung is no longer held open by nega-
tive intrapleural pressure, and it tends to collapse
because of unopposed elastic recoil. Thus, the lung
in an open chest is at least partially collapsed. It
has long been observed during spontaneous venti-
lation with an open hemithorax that lung collapse
is accentuated during inspiration, and, conversely,
the lung expands during expiration. This reversal
of lung movement with an open chest during res-
piration has been termed paradoxical respiration.
The mechanism of paradoxical respiration is simi-
lar to that of mediastinal shift. During inspiration,
the descent of the diaphragm on the side of the
open hemithorax causes air from the environment
to enter the pleural cavity on that side through the
thoracotomy opening and fill the space around the
exposed lung. The descent of the hemidiaphragm
on the closed-chest side causes gas to enter the
closed-chest lung in the normal manner. However,
Special Respiratory Physiology of the Lateral Decubitus Position, the Open Chest, and One-Lung Ventilation 125
Spontaneous Ventilation with Open Chest
Mediastinal Shift Paradoxical Respiration
Inspiration /
Figure 4-1 Schematic representation of mediastinal shift and paradoxical respiration in the spontaneously ventilating patient
with an open chest and placed in the lateral decubitus position. The open chest is always exposed to atmospheric pressure ().
During inspiration, negative pressure ( ) in the intact hemithorax causes the mediastinum to move vertically downward (medias-
tinal shift). In addition, during inspiration movement of gas from the nondependent lung in the open hemithorax into the dependent
lung in the closed hemithorax and movement of air from the environment into the open hemithorax causes the lung in the open
hemithorax to collapse (paradoxical respiration). During expiration, relative positive pressure () in the closed hemithorax causes
the mediastinum to move vertically upward (mediastinal shift). In addition, during expiration the gas moves from the dependent
lung to the nondependent lung and from the open hemithorax to the environment; consequently, the nondependent lung expands
during expiration (paradoxical respiration).
gas also enters the closed-chest lung (which has a
relatively negative pressure) from the open-chest
lung (which remains at atmospheric pressure); this
results in further reduction in the size of the open-
chest lung during inspiration. During expiration
the reverse occurs, with the collapsed, open-chest
lung filling from the intact lung and air moving
back out of the exposed hemithorax through the
thoracotomy. The phenomenon of paradoxical res-
piration is illustrated in Figure 4-1.
2
Paradoxical
breathing is increased by a large thoracotomy and
by increased airway resistance in the intact lung.
Paradoxical respiration may be prevented by either
manual collapse of the open-chest lung or, more
commonly, controlled positive-pressure ventila-
tion. In summary, in a spontaneously breathing
patient in the lateral decubitus position with an
open, nondependent chest, the mediastinum will
move away from the observer, the nondependent
lung will decrease in size during inhalation, the
mediastinum will move toward the observer, and
the nondependent lung will increase in size during
exhalation.
III. PHYSIOLOGY OF THE LATERAL
DECUBITUS POSITION AND
THE OPEN CHEST DURING
CONTROLLED TWO-LUNG
VENTILATION: DISTRIBUTION
OF PERFUSION (Q) AND
VENTILATION (V)
A. Distribution of Q, V, V/QLateral
Decubitus Position, Awake, Closed
Chest
Gravity causes a vertical gradient in the distri-
bution of pulmonary blood flow in the lateral de-
cubitus position for the same reason that it does in
the upright position (see Fig. 3-5). Since the ver-
tical hydrostatic gradient is less in the lateral de-
cubitus position than in the upright position, there
is ordinarily less zone 1 blood flow (in the nonde-
pendent lung) in the former position compared
with the latter position (Fig. 4-2). Nevertheless,
blood flow to the dependent lung is still signifi-
cantly greater than blood flow to the nondependent
Distribution of Blood Flow
Lateral Decubitus Position
Figure 4- 2 Schematic representation of the effects of gravity on the distribution of pulmonary blood flow in the lateral decubitus
position. The vertical gradient in the lateral decubitus position is less than that in the upright position (see Fig. 3-1). Consequently,
there is less zone 1 and more zone 2 and zone 3 blood flow in the lateral decubitus position compared with the upright position.
Nevertheless, pulmonary blood flow increases with lung dependency and is greater in the dependent lung compared with the
nondependent lung. (P
A
= alveolar pressure; P
pa
= pulmonary artery pressure; P
pv
= pulmonary venous pressure.) (Modified with
permission from Benumof JL: Physiology of the open chest and one-lung ventilation. In Kaplan JA (ed): Thoracic Anesthesia. New
York, Churchill Livingstone Inc., 1983, chapter 8.)
lung (Fig. 4-2). Thus, when the right lung is non-
dependent, it should receive approximately 45 per
cent of total blood flow as opposed to the 55 per
cent of the total blood flow that it receives in the
upright and supine positions.
3,4
When the left lung
is nondependent, it should receive approximately
35 per cent of total blood flow as opposed to the
45 per cent of the total blood flow that it receives
in the upright and supine positions.
3,4
Because gravity also causes a vertical gradient
in pleural pressure (P
pl
) in the lateral decubitus
position (as it does in the upright position, see
Figs. 3-4 and 3-5), ventilation is relatively in-
creased in the dependent lung compared with the
nondependent lung (Fig. 4-3). In addition, in the
lateral decubitus position, the dome of the lower
diaphragm is pushed higher into the chest than the
dome of the upper diaphragm, and therefore the
lower diaphragm is more sharply curved than the
upper diaphragm. As a result, the lower diaphragm
is able to contract more efficiently during sponta-
neous respiration. Thus, in the lateral decubitus
position in the awake patient, the lower lung is
normally better ventilated than the upper lung, re-
gardless of the side on which the patient is lying,
although there remains a tendency toward greater
ventilation of the larger right lung.
5
Since there is
greater perfusion to the lower lung, the preferential
ventilation to the lower lung is matched by its
increased perfusion, so that the distribution of the
ventilation-perfusion ratios of the two lungs is not
greatly altered when the awake subject assumes
the lateral decubitus position. Since the increase in
ventilation is less than the increase in perfusion
with lung dependency, the V/Q ratio decreases
from dependent to nondependent lung (just as it
does in upright and supine lungs).
B. Distribution of Q, VLateral
Closed Chest
In comparing the awake patient with the anes-
thetized patient in the lateral decubitus position,
there is no difference in the distribution of pulmo-
Awake, Closed Chest
Distribution of Ventilation
Upright
Position
Lateral Decubitus
Position
Figure 4-3 Pleural pressure (P
pl
) in the awake upright patient (A) is most positive in the dependent portion of the lung, and
alveoli in this region are therefore most compressed and have the least volume (see Fig. 3-5). Pleural pressure is least positive
(most negative) at the apex of the lung, and alveoli in this region are therefore least compressed and have the largest volume. When
these regional differences in alveolar volume are translated over to a regional transpulmonary pressure-alveolar volume curve, the
small, dependent alveoli are on a steep (large-slope) portion of the curve, and the large, nondependent alveoli are on a flat (small-
slope) portion of the curve (see also Fig. 3-6). In this diagram regional slope equals regional compliance. Thus, for a given and
equal change in transpulmonary pressure, the dependent part of the lung receives a much larger share of the tidal volume than the
nondependent part of the lung. In the awake patient in the lateral decubitus position (B) gravity also causes pleural pressure
gradients and therefore similarly affects the distribution of ventilation. The dependent lung lies on a relatively steep portion, and
the upper lung lies on a relatively flat portion of the pressure-volume curve. Thus, in the lateral decubitus position the dependent
lung receives the majority of the tidal ventilation. (V alveolar volume; = transpulmonary pressure.) (Modified with permission
from Benumof JL: Physiology of the open chest and one-lung ventilation. In Kaplan JA (ed): Thoracic Anesthesia. New York.
Churchill Livingstone Inc., 1983, chapter 8.)
nary blood flow between the dependent and non-
dependent lungs. Thus, in the anesthetized patient,
the dependent lung continues to receive relatively
more perfusion than the nondependent lung. The
induction of general anesthesia, however, does
cause significant changes in the distribution of
ventilation between the two lungs.
In the lateral decubitus position, the majority of
ventilation is switched from the dependent lung in
the awake subject to the nondependent lung in the
anesthetized patient (Fig. 4-4).
6
7
In fact, in the
lateral decubitus position, the nondependent lung
will receive approximately 55 per cent of each
tidal volume.
8
There are several interrelated rea-
sons for this change in the relative distribution of
ventilation between the nondependent and depen-
dent lung. First, the induction of general anesthesia
usually causes a decrease in functional residual
capacity (FRC), and both lungs share in the loss
of lung volume. Since each lung occupies a differ-
ent initial position on the pulmonary pressure-vol-
ume curve while the subject is awake, a general
anesthesia-induced reduction in the FRC of each
lung causes each lung to move to a lower but still
different portion of the pressure-volume curve
(Fig. 4-4). The dependent lung moves from an
initially steep part of the curve (with the subject
awake) to a lower and flatter part of the curve
(after anesthesia is induced), while the nondepen-
dent lung moves from an initially flat portion of
the pressure-volume curve (with the subject
awake) to a lower and steeper part of the curve
(after anesthesia is induced). In fact, in the lateral
decubitus position, the ratio of nondependent to
dependent lung FRC is approximately 1.5 (average
values are 1400 ml/900 ml) in the adult patient.*
Similar findings may be expected in children.
4
Thus, with the induction of general anesthesia, the
lower lung moves to a less favorable (flat, non-
compliant) portion and the upper lung to a more
favorable (steep, compliant) portion of the pres-
sure-volume curve. In fact, in the lateral decubitus
position, the compliance of the nondependent and
dependent lung is 30 and 23 cm H
2
0, respec-
tively.
8
Second, if the anesthetized patient in the
lateral decubitus position is also paralyzed and ar-
tificially ventilated, the high curved diaphragm of
the lower lung no longer confers any advantage in
ventilation (as it does in the awake state), since it
is no longer actively contracting.
10
Third, the me-
diastinum rests on the lower lung and physically
impedes lower lung expansion as well as selec-
tively decreases lower lung FRC. Fourth, the
weight of the abdominal contents pushing cepha-
128 Special Respiratory Physiology of the Uiteral Decubitus Position, the Open Chest, and One-Lung Ventilation
Closed Chest, Lateral Decubitus Position
Awake Anesthetized
Figure 4-4 This schematic diagram shows the distribution of ventilation in the awake patient in the lateral decubitus position
(A) and the distribution of ventilation in the anesthetized patient in the lateral decubitus position (B). The induction of anesthesia
has caused a loss in lung volume in both lungs, with the nondependent lung moving from a flat, noncompliant portion to a steep,
compliant portion of the pressure-volume curve and the dependent lung moving from a steep, compliant part to a flat, noncompliant
part of the pressure-volume curve. Thus, the anesthetized patient in a lateral decubitus position has the majority of the tidal
ventilation in the nondependent lung (where there is the least perfusion) and the minority of the tidal ventilation in the dependent
lung (where there is the most perfusion). (V = alveolar volume; = transpulmonary pressure.) (Modified with permission from
Benumof JL: Physiology of the open chest and one-lung ventilation. In Kaplan JA (ed): Thoracic Anesthesia. New York, Churchill
Livingstone Inc., 1983, chapter 8.)
lad against a passive, flaccid, paralyzed diaphragm
is greatest in the dependent lung, which physically
impedes lower lung expansion the most and
disproportionately decreases lower lung FRC. Fi-
nally, suboptimal positioning, which fails to pro-
vide room for lower lung expansion, may consid-
erably compress the dependent lung. Opening the
nondependent hemithorax further disproportion-
ately increases ventilation to the nondependent
lung (see the following discussion). Thus, in com-
paring the supine position to the lateral decubitus
position in both spontaneously breathing and me-
chanically ventilated children and adults, the lung
that is made the nondependent lung has an in-
crease in FRC, compliance, and relative ventila-
tion, and the lung that is made the dependent lung
has a decrease in FRC, compliance, and relative
ventilation.
8
9
" The change in each lung ventila-
tion is proportional to the change in each lung
volume.
8,9
"
In summary, the anesthetized patient, with or
without paralysis, in the lateral decubitus position
and with a closed chest has a nondependent lung
that is well ventilated but poorly perfused and has
a dependent lung that is well perfused but poorly
ventilated, which results in an increased degree of
mismatching of ventilation and perfusion. The ap-
plication of positive end-expiratory pressure
(PEEP) to both lungs restores the majority of ven-
tilation to the lower lung.
4 I2
Presumably, the
lower lung returns to a steeper, more favorable
part of the pressure-volume curve, and the upper
lung resumes its original position on a flat, unfa-
vorable portion of the curve.
C. Distribution of Q, VLateral
Open Chest
Compared with the condition of the anesthe-
tized, closed-chest patient in the lateral decubitus
position, opening the chest wall and pleural space
alone does not ordinarily cause any significant al-
teration in the partitioning of pulmonary blood
flow between the dependent and nondependent
lungs; thus, the dependent lung continues to re-
ceive relatively more perfusion than the nondepen-
dent lung. However, if the compliance of the non-
dependent lung increases so much (see discussion
immediately following) that nondependent lung
vascular resistance decreases significantly, then
nondependent-lung blood flow may increase rela-
tive to dependent-lung blood flow. In addition, the
vertical distance between the heart and the nonde-
pendent lung may be decreased, which in the face
of a constant pulmonary artery pressure might, in
theory, result in an increased perfusion of the non-
dependent lung.
13
Opening the chest wall causes
only very minor alterations in pulmonary and sys-
temic vascular pressures and cardiac output.
13
Opening the chest wall and pleural space, how-
ever, does have a significant impact on the distri-
bution of ventilation (which must now be deliv-
ered by positive pressure). The change in the
distribution of ventilation may result in a further
mismatching of ventilation with perfusion (Fig.
4-5).
14
If the upper lung is no longer restricted by a
chest wall and the total effective compliance of
that lung is equal to that of the lung parenchyma
alone, it will be relatively free to expand and will
consequently be overventilated (and remain under-
perfused). Conversely, the dependent lung may
continue to be relatively noncompliant and poorly
ventilated and overperfused.
3
From a practical
point of view, it is necessary to mention that sur-
gical retraction and compression of the exposed
upper lung can provide a partial, although non-
physiologic, solution to this problem in that if ex-
pansion of the exposed lung is mechanically or
externally restricted, ventilation will be diverted to
the dependent, better-perfused lung.
13
D. Distribution of Q, VLateral
Decubitus Position, Anesthetized Open
Chest, Paralyzed
In the open-chest anesthetized patient in the lat-
eral decubitus position, the induction of paralysis
alone does not cause any significant alteration in
the partitioning of pulmonary blood flow between
the dependent and nondependent lungs. Thus, the
dependent lung continues to receive relatively
more perfusion than the nondependent lung. There
are, however, strong theoretical and experimental
considerations that indicate that paralysis might
cause significant changes in the distribution of
ventilation between the two lungs under these con-
ditions.
In the supine and lateral decubitus positions, the
weight of the abdominal contents pressing against
the diaphragm is greatest on the dependent part of
the diaphragm (posterior lung and lower lung, re-
spectively) and least on the nondependent part of
the diaphragm (anterior lung and upper lung, re-
spectively) (Fig. 4-5). In the awake, sponta-
neously breathing patient, the normally present ac-
tive tension in the diaphragm overcomes the
weight of the abdominal contents, and the dia-
phragm moves the most (largest excursion) in the
dependent portion and least in the nondependent
portion. This is a healthy circumstance because
this is another factor that maintains the greatest
amount of ventilation where there is the most per-
fusion (dependent lung) and the least amount of
ventilation where there is the least perfusion (non-
dependent lung). During paralysis and positive-
pressure breathing, the passive and flaccid dia-
phragm is displaced preferentially in the nonde-
pendent area, where the resistance to passive dia-
phragmatic movement by the abdominal contents
is least; conversely, the diaphragm is displaced
minimally in the dependent portion where the re-
Anesthetized, Lateral Decubitus Position
Closed Chest Open Chest
Figure 4-5 This schematic of a patient in the lateral decubitus position compares the closed-chest anesthetized condition with
the open-chest anesthetized and paralyzed condition. Opening the chest increases nondependent lung compliance and reinforces or
maintains the larger part of the tidal ventilation going to the nondependent lung. Paralysis also reinforces or maintains the larger
part of tidal ventilation going to the nondependent lung because the pressure of the abdominal contents (P
AB
) pressing against the
upper diaphragm is minimal (smaller arrow), and it is therefore easier for positive-pressure ventilation to displace this lesser
resisting dome of the diaphragm. (V = alveolar volume; = transpulmonary pressure.) (Modified with permission from Benumof
JL: Physiology of the open chest and one-lung ventilation. In Kaplan JA (ed): Thoracic Anesthesia. New York, Churchill
Livingstone Inc., 1983, chapter 8.)
sistance to passive diaphragmatic movement by
the abdominal contents is greatest.
15
This is an
unhealthy circumstance because the greatest
amount of ventilation may occur where there is the
least perfusion (nondependent lung), and the least
amount of ventilation may occur where there is the
most perfusion (dependent lung).
15
E. Summary of Physiology of Lateral
Decubitus Position and the Open Chest
In summary (Fig. 4-6), the preceding section
has developed the concept that the anesthetized,
paralyzed patient in the lateral decubitus position
with an open chest may have a considerable ven-
tilation-perfusion mismatch, consisting of greater
ventilation but less perfusion to the nondependent
lung and less ventilation but more perfusion to the
dependent lung. The blood flow distribution is
mainly and simply determined by gravitational ef-
fects. The relatively good ventilation of the upper
lung is caused, in part, by the open chest and
paralysis. The relatively poor ventilation of the
dependent lung is caused, in part, by the loss of
dependent-lung volume with general anesthesia
and by compression of the dependent lung by the
Anesthetized
Open
Chest
/
Nondependent ^ ^
w
^ <
Lung
Mediastinum
1 1 i
Paralysis
(Flaccid
Diaphragm)
/
AB
v Suboptimal y
^Posi ti oni ng'
Effects
Figure 4-6 Schematic summary of ventilation-perfusion relationships in the anesthetized patient in the lateral decubitus position
who has an open chest and is paralyzed and suboptimally positioned. The nondependent lung is well ventilated (as indicated by the
large dashed lines) but poorly perfused (small perfusion vessel), and the dependent lung is poorly ventilated (small dashed lines)
but well perfused (large perfusion vessel). In addition, the dependent lung may also develop an atelectatic shunt compartment
(indicated on the left side of the lower lung) because of the circumferential compression of this lung. (P
AB
= pressure of the
abdominal contents.) (Modified with permission from Benumof JL: Physiology of the open chest and one-lung ventilation. In
Kaplan JA (ed): Thoracic Anesthesia. New York, Churchill Livingstone Inc., 1983, chapter 8.)
mediastinum, abdominal contents, and suboptimal
positioning effects. In addition, poor mucociliary
clearance and absorption atelectasis with an in-
creased F,0
2
may cause further dependent-lung
volume loss. Indeed, on rare occasion, the depen-
dent lung may be massively atelectatic and edem-
atous.
16
Consequently, two-lung ventilation under
these circumstances may result in an increased
alveolar-arterial oxygen tension difference
[P(A-a)0
2
] and less than optimal oxygenation.
A physiologic solution to the adverse effects of
anesthesia and surgery in the lateral decubitus po-
sition on the distribution of ventilation and perfu-
sion during two-lung ventilation would be the se-
lective application of PEEP to the dependent lung
(via a double-lumen endotracheal tube).
12
Selective
PEEP to the lower lung should increase the venti-
lation to this lung by moving it up to a steeper,
more favorable portion of the lung pressure-vol-
ume curve. Indeed, this has been done with rea-
sonably good success,
12
17
although all the results
have not been entirely uniform.
18
A series of 22
mechanically ventilated patients (both lungs)
undergoing thoracotomy in the lateral decubitus
position was divided into two groups.
17
Group 1
patients had 10 cm H
2
0 of PEEP applied to the
dependent lung while zero end-expiratory pressure
(ZEEP) was applied to the nondependent lung.
Group 2 (control) patients were intubated with a
standard endotracheal tube, and both lungs were
ventilated with ZEEP. Selective PEEP to the de-
pendent lung in group 1 patients resulted in an
adequate P
a
0
2
with a lower inspired 0
2
concentra-
tion during surgery and a smaller P(A-a)0
2
at the
end of surgery than when both lungs were venti-
lated with ZEEP. Thus, even if the selective PEEP
to the dependent lung increased dependent-lung
pulmonary vascular resistance and diverted some
blood flow to the nondependent lung, the diverted
blood flow could still participate in gas exchange
with the ZEEP-ventilated nondependent lung.
19
However, it should be noted that this technique
requires that the nondependent (and operative)
lung be ventilated, and this may impede the per-
formance of surgery.
IV. PHYSIOLOGY OF ONE-LUNG
VENTILATION
A. Comparison of Arterial Oxygenation
and C0
2
Elimination During Two-Lung
Versus One-Lung Ventilation
As discussed previously, the matching of venti-
lation and perfusion is impaired during two-lung
ventilation in an anesthetized, paralyzed, open-
chest patient in the lateral decubitus position. The
reason for the mismatching of ventilation and per-
fusion is relatively good ventilation but poor per-
fusion of the nondependent lung and poor ventila-
tion and good perfusion of the dependent lung
(Figs. 4-6 and 4-7A). The blood flow distribution
Two Lung
Ventilation
vs
One Lung
Ventilation
Figure 4-7 Schematic representation of two-lung ventilation versus one-lung ventilation. Typical values for fractional blood flow
to the nondependent and dependent lungs as well as P
a
0
2
and Q
s
/Q, for the two conditions are shown. The Q
s
/Q, during two-lung
ventilation is assumed to be distributed equally between the two lungs (5 per cent to each lung). The essential difference between
two-lung and one-lung ventilation is that during one-lung ventilation the nonventilated lung has some blood flow and, therefore, an
obligatory shunt, which is not present during two-lung ventilation. The 35 per cent of total flow perfusing the nondependent lung,
which was not shunt flow, was assumed to be able to reduce its blood flow by 50 per cent by hypoxic pulmonary vasconstriction.'
2
The increase in Q
s
/Q, from two-lung to one-lung ventilation is assumed to be solely due to the increase in blood flow through the
nonventilated, nondependent lung during one-lung ventilation.
was seen to be mainly and simply determined by
gravitational effects. The relatively good ventila-
tion of the nondependent lung was seen to be
caused, in part, by the open chest and paralysis.
The relatively poor ventilation of the dependent
lung was seen to be caused, in part, by the loss of
dependent-lung volume with general anesthesia
and by circumferential compression of the depen-
dent lung by the mediastinum, abdominal contents,
and suboptimal positioning effect. The compres-
sion of the dependent lung may cause the devel-
opment of a shunt compartment in the dependent
lung (Figs. 4-6 and 4-7). Consequently, two-
lung ventilation under these circumstances may
result in an increased P(A-a)0
2
and impaired oxy-
genation.
However, if the nondependent lung is nonventi-
lated, as during one-lung ventilation, then any
blood flow to the nonventilated lung becomes
shunt flow, in addition to whatever shunt flow
might exist in the dependent lung (Fig. 4-).
(Table 4-1 and discussion on page 135 show a
quantitative analysis of the two-lung ventilation to
the one-lung ventilation conversion process with
respect to arterial oxygenation.) Thus, one-lung
ventilation creates an obligatory right-to-left trans-
pulmonary shunt through the nonventilated, non-
dependent lung, which is not present during two-
lung ventilation. Consequently, it is not surprising
to find that, given the same inspired oxygen con-
centration (F,0
2
) and hemodynamic and metabolic
status, one-lung ventilation results in a much
larger alveolar-to-arterial oxygen tension differ-
ence P(A-a)0
2
and lower P
a
0
2
than does two-lung
ventilation. This contention is best supported by
studies that compare arterial oxygenation during
two-lung ventilation and one-lung ventilation,
wherein each patient serves as their own control.
20
One-lung ventilation has much less of a steady-
state effect on P
a
C0
2
in comparison with its effect
on P
a
0
2
. Blood passing through underventilated
alveoli retains more than a normal amount of C0
2
and does not take up a normal amount of 0
2
; blood
traversing overventilated alveoli gives off more
than a normal amount of C0
2
but cannot take up a
proportionately increased amount of 0
2
owing to
the flatness of the top end of the oxygen-hemoglo-
bin dissociation curve (see Fig. 3-32). Thus, dur-
ing one-lung ventilation (the one-lung minute ven-
tilation = the two-lung ventilation), the ventilated
lung can eliminate almost enough C0
2
to compen-
sate for the nonventilated lung, and P
A
C0
2
and
P
ET
C0
2
to P
a
C0
2
gradients are small; however, the
ventilated lung cannot take up enough 0
2
to com-
pensate for the nonventilated lung, and P
A
0
2
to
P
a
0
2
gradients are usually large. With a constant
minute ventilation (two-lung ventilation compared
with one-lung ventilation), the retention of C0
2
by
blood traversing the nonventilated lung usually
slightly exceeds the increased elimination of C0
2
from blood traversing the ventilated lung, and the
P
a
C0
2
will usually slowly increase (along with the
end-tidal C0
2
; see later discussion) over time un-
less the respiratory rate is increased (see chapter
11 and following paragraph).
The initiation of one-lung ventilation has much
more of an acute effect (first 5 min) on P
ET
C0
2
than it does on P
a
C0
2
. When one-lung ventilation
is begun (keeping total tidal volume and respira-
tory rate constant), the ventilated lung is immedi-
ately hyperventilated in relation to its perfusion
(i.e., has an increased V/Q ratio) and P
ET
C0
2
from
this lung decreases in the first minute (e.g., by 5
mm Hg).
21
Over the next 5 min, hypoxic pulmo-
nary vasoconstriction (HPV) in the nonventilated
lung shifts blood flow over to the ventilated lung,
increases ventilated lung perfusion, decreases ven-
tilated lung V/Q ratio and increases the P
ET
C0
2
back to the baseline two-lung ventilation value.
21
Thereafter, and as discussed previously, P
ET
C0
2
will slowly increase (along with P
a
C0
2
) because
the same total minute ventilation to one lung is not
as effective as when it is delivered to both lungs
(i.e., there is an increased alveolar dead space
within the one ventilated lung).
B. Blood Flow Distribution During
7. Blood Flow to the Nondependent,
Nonventilated Lung
Fortunately, there are both passive mechanical
and active vasoconstrictor mechanisms that are
usually operant during one-lung ventilation that
minimize the blood flow to the nondependent,
nonventilated lung and thereby prevent the P
a
0
2
from decreasing as much as might be expected on
the basis of the distribution of blood flow during
two-lung ventilation. The passive mechanical
mechanisms that decrease blood flow to the non-
dependent lung consist of gravity, surgical inter-
ference with blood flow, and perhaps the amount
of pre-existing disease in the nondependent lung
(Fig. 4-8). Gravity causes a vertical gradient in the
distribution of pulmonary blood flow in the lateral
decubitus position for the same reason that it does
in the upright position (see Figs. 3-4 and 4-2).
Consequently, blood flow to the nondependent
lung is less than blood flow to the dependent lung.
The gravity component of blood flow reduction to
the nondependent lung should be constant with
respect to both time and magnitude.
Severe surgical compression (directly compress-
ing lung vessels) and retraction (causing kinking
One Lung Ventilation: Determinants
Of Blood Flow Distribution
Surgica
Interference
Nondependent Lung
Gravity
Dependent Lung
Gravity
Hypoxic Pulmonary
Vasoconstriction
and/or
Lung Disease
Lung Disease
and/or
Hypoxic Pulmonary
Vasoconstriction
Figure 4-8 Schematic diagram of the determinants of blood flow distribution during one-lung ventilation. The major determinants
of blood flow to the nondependent lung are gravity, surgical interference with blood flow, the amount of nondependent lung disease,
and the magnitude of nondependent lung hypoxic pulmonary vasconstriction. The determinants of dependent lung blood flow are
gravity, amount of dependent lung disease, and dependent lung hypoxic pulmonary vasoconstriction. (RV right ventricle.)
and tortuosity of lung vessels) of the nondependent
lung may further passively reduce nondependent-
lung blood flow. In addition, ligation of pulmonary
vessels for pulmonary resection will greatly de-
crease nondependent-lung blood flow. The sur-
gical interference component of blood flow reduc-
tion to the nondependent lung should be variable
with respect to both time and magnitude.
However, it should be noted that there is evi-
dence indicating that some physical stimuli, such
as stroking of pulmonary tissue, may cause local
release of vasodilator prostaglandins,
22
and other
stimuli such as lung compression with a moist
gauze may cause release of a quick-acting and
terminating substance such as endothelium-derived
relaxing factor.
23
In view of these findings, it is
not surprising that one study has shown that the
shunt fraction increases significantly when the
nonventilated lung is exposed to mild-to-moderate
degrees of surgical manipulation.
24
Thus, it ap-
pears that the effect of surgical manipulation on
arterial oxygenation may depend on the exact na-
ture and force/strength of the physical stimulus.
The amount of disease in the nondependent lung
should also be a significant determinant of the
amount of blood flow to the nondependent lung. If
the nondependent lung is severely diseased, then
there may be a fixed reduction in blood flow to
this lung preoperatively, and collapse of such a
diseased lung may not cause much of an increase
in shunt (see Table 4-1). The notion that a dis-
eased pulmonary vasculature might be incapable
of HPV is supported by the observation that ad-
ministration of sodium nitroprusside and nitroglyc-
erin (which should abolish any pre-existing HPV)
to chronic obstructive pulmonary disease patients
(who have a fixed reduction in the cross-sectional
area of their pulmonary vascular bed) does not
cause an increase in shunt,
25
whereas these drugs
do increase shunt in patients with acute regional
lung disease who have an otherwise normal pul-
monary vascular bed.
26
If the nondependent lung
is normal and has a normal amount of blood flow,
then collapse of such a normal lung may be asso-
ciated with a higher nonventilated-, nondependent-
lung blood flow and shunt. A higher one-lung ven-
tilation shunt through the nondependent lung is
therefore theoretically more likely to occur in pa-
tients who require thoracotomy for nonpulmonary
disease.
27
There have been only two studies that
have systematically validated the inverse correla-
tion between the amount of nondependent-lung
disease and shunt during one-lung ventilation.
2
*
2i)
Figure 4-9 shows the correlation between the per-
centage of the cardiac output perfusing the opera-
tive lung, as measured by preoperative perfusion
1 3 4 Special Respiratory Physiology of the Lateral Decubitus Position, the Open Chest, and One-Lung Ventilation
Flow to Operative Lung
Figure 49 The relative preopera-
tive perfusion to the operative lung
correlated with the P
a
0
2
after 10 min
of one-lung anesthesia. Relative per-
fusion to the operative lungs was
measured by scintigraphy performed
after the intravenous injection of
technetium 99m macroaggregated
human albumin. (From Hurford WE,
Kolker AC, Strauss HW: The use of
ventilation/perfusion lung scans to
predict oxygenation during one-lung
anesthesia. Anesthesiology 67:841-
844, 1987. Used with permission.)
Figure 4-10 Effect of unilateral
hypoxia in humans. The pulmonary
vascular resistance of the hypoxic
lung is plotted against its P
A
0
2
.
(From Harris P, Heath D: The Hu-
man Pulmonary Circulation. 2nd ed.
Edinburgh, Churchill Livingstone,
1977, p. 456. Used with permission.)
scans, and the intraoperative one-lung ventilation
P,0
2
.
28
The most significant reduction in blood flow to
the nondependent lung is caused by an active
vasoconstrictor mechanism. The normal response
of the pulmonary vasculature to atelectasis is an
increase in pulmonary vascular resistance (in just
the atelectatic lung), and the increase in atelectatic
lung pulmonary vascular resistance is thought to
be due almost entirely to HPV.
303I
The effect of
unilateral hypoxia on the hypoxic lung pulmonary
vascular resistance in humans is shown in Figure
4-10.
32
The selective increase in atelectatic lung
pulmonary vascular resistance diverts blood flow
away from the atelectatic lung toward the remain-
ing normoxic or hyperoxic ventilated lung. The
diversion of blood flow minimizes the amount of
shunt flow that occurs through hypoxic lung. Fig-
ure 4-11 shows the theoretically expected effect
of HPV on arterial oxygen tension (P
a
0
2
) as the
amount of lung that becomes hypoxic increases.
33
When very little of the lung is hypoxic (near 0 per
cent), it does not matter, in terms of P
a
0
2
, whether
the small amount of lung has HPV or not because
in either case the shunt will be small. When most
of the lung is hypoxic (near 100 per cent), there is
no significant normoxic region to which the hy-
poxic region can divert flow, and, again, it does
not matter, in terms of P
a
0
2
, whether the hypoxic
region has HPV or not. When the percentage of
lung that is hypoxic is intermediate (between 30
and 70 per cent), which is the amount of lung that
is hypoxic during the one-lung ventilation/anes-
thesia condition, there is a large difference be-
tween the P
a
0
2
expected with a normal amount of
HPV (which is a 50 per cent blood flow reduction
for a single lung)
33
compared with when there is
no HPV. In fact, in this range of hypoxic lung,
HPV can increase P
a
0
2
from hypoxemic levels to
much higher and safer values. It is not surprising,
then, that numerous clinical studies on one-lung
ventilation
20
27
34
^
2
found that the shunt through
the nonventilated lung is usually 20 to 30 per cent
of the cardiac output as opposed to the 40 to 50
per cent shunt that might be expected if there was
no HPV in the nonventilated lung.
41
Thus, HPV is
an autoregulatory mechanism that protects the
P
a
0
2
by decreasing hypoxic lung.
It is possible to model the two-lung ventilation
conversion process for various initial two-lung
ventilation shunts; the model shown in Table 4-1
makes several assumptions. First, the initial two-
lung ventilation shunt flow is equally distributed
between the nondependent and dependent lungs.
Second, the remaining normal blood flow to the
nondependent lung can decrease its blood flow by
50 per cent owing to HPV,
33
and any initial shunt
flow (i.e., during two-lung ventilation) does not
Figure 4-11 The graph is a model of the effect of hypoxic
pulmonary vasoconstriction (HPV) on P
a
0
2
as a function of the
per cent of lung that is hypoxic. The model assumes an F,0
2
of
l.O, a normal hemoglobin, cardiac output, and oxygen con-
sumption. In the range of 30 to 70 per cent of the lung being
hypoxic, the normal expected amount of HPV can increase
P
a
0
2
significantly.
participate in an HPV response. Third, the total
one-lung ventilation shunt flow is a sum of half
the normal flow to the nondependent lung when it
was ventilated plus the original nondependent and
dependent lung shunt flows. Table 4-1 shows that
as the two-lung ventilation shunt increases, the
amount of nondependent lung flow that is able to
participate in an HPV response decreases; there-
fore, the amount of blood flow diversion due to
nondependent lung HPV decreases, and, thus, the
total one-lung ventilation shunt increases.
Figure 4-12 outlines the major determinants of
the amount of atelectatic lung HPV that might
occur during anesthesia. In the following discus-
sion, the HPV issues or considerations are num-
bered as they appear in Figure 4-12.
1. Distribution of Hypoxia
The distribution of the alveolar hypoxia is prob-
ably not a determinant of the amount of HPV; all
regions of the lung (either the basilar or dependent
parts of the lungs [supine or upright] or discrete
anatomic units such as a lobe or single lung) re-
spond to alveolar hypoxia with vasoconstriction.
43
Table 4-1 MODEL OF CONVERTING TWO-LUNG TO ONE-LUNG VENTILATION
Lung
Two-Lung Ventilation One-Lung Ventilation
Fractional
Normal
Flow
Total Shunt
Flow
Fractional
Shunt Flow
Fractional
Normal
Flow
Total
Shunt
Flow*
Fractional Shunt
Flow
ND :: 0.400 0 0 0.2001
(0.200 + 0)
0 0.200
D 0.600 0 0.800 0
ND 0.350 0.050 0 0.225t
(0.175 + 0.050)
0.100 0.275
D 0.550 0.050 0.725 0.050
ND 0.300 0.100 0 0.250t
(0.150 + 0.100)
0.200 0.350
D 0.500 0.100 0.650 0.100
ND 0.200 0.200 0 0.300t
(0.100 + 0.200)
0.400 0.500
D 0.400 0.200 0.500 0.200
ND 0.100 0.300 0 0.350t
(0.050 + 0.300)
0.600 0.650
D 0.300 0.300 0.350 0.300
*Sum of ND and D lung fractional shunt flows.
t Hal f of two-lung ventilation fractional normal flow (due to hypoxic pulmonary vasoconstriction) plus all of two-lung ventilation
fractional shunt flow.
Abbreviations: ND = nondependent lung; D = dependent lung.
Anesthetic Experience and Regional HPV
NORMOXIC COMPARTMENT HYPOXIC COMPARTMENT
Figure 4-12 This figure lists many of the components of the anesthetic experience that might determine the amount of regional
hypoxic pulmonary vasoconstriction (HPV). The clockwise numbering of considerations corresponds to the order in which these
considerations are discussed in the text. (PVP = pulmonary vascular pressure; PEEP = positive end-expiratory pressure; Q, =
pulmonary blood flow; V/Q = ventilation-perfusion ratio.)
However, recent evidence suggests that, on a sub-
lobar level, collateral ventilation may be the first
line and HPV the second line of defense against
the development of arterial hypoxemia (Fig. 4-
13).
44
In species with extensive collateral ventila-
tion, such as canines, the development of sublobar
atelectasis or low V/Q areas does not cause as
much sublobar HPV because collateral ventilation
prevents the sublobar area in question from be-
coming very hypoxic. This protective phenomenon
seems reasonable when one considers that air (0
2
)
is less dense than blood and therefore easier to
redistribute. Collateral ventilation is most efficient
when very small airways (1.6 mm) are occluded,
because many channels are available for by-pass-
ing the obstruction, whereas collateral ventilation
is least efficient when large airways (4.8 mm) are
occluded, because fewer by-pass channels are
available (see Figs. 2-19 and 2-20).
45
On the other
hand, in consolidated lesions and in species with
no collateral ventilation,
46
such as the coatimundi,
ferret, cattle, and swine, the development of sub-
lobar atelectasis and low V/Q areas does elicit a
great deal of sublobar HPV, which minimizes the
decrease in P
a
0
2
.
2. Low V/Q Versus Atelectasis
As with low V/Q and nitrogen-ventilated lungs,
it appears that the vast majority of blood flow
reduction in the acutely atelectatic lung is due to
HPV, and none of the blood flow reduction is due
to passive mechanical factors (such as vessel
tortuosity).
30
31
This conclusion is based on the
observation that re-expansion and ventilation of a
collapsed lung with nitrogen (removing any me-
chanical factor) does not increase the blood flow
to the lung, whereas ventilation with oxygen re-
stores all of the blood flow back to precollapse
values. This conclusion applies whether ventilation
is spontaneous or with positive pressure and
whether the chest is open or closed.
47
In canines, a
slight amount of further subacute (greater than 30
min) decrease in blood flow to the atelectatic lung
may have been due to some mechanical effect of
the atelectasis on lung blood vessels.
48
However,
in humans, a prolonged unilateral hypoxic chal-
lenge during anesthesia results in an immediate
vasoconstrictor response with no further potentia-
tion or diminution of the response.
49
3. Vasodilator Drugs
A major concern with administering a drug that
can cause pulmonary vasodilatation is that the
drug will inhibit pre-existing HPV and thereby
increase Q
s
/Q
t
and decrease P
a
0
2
. However, the in
vivo physiology that follows vasodilator therapy
(usually both systemic and pulmonary circulations
vasodilate) is complex, and there are three major
reasons why, in a given experiment, administration
of a vasodilator drug may not decrease HPV and.
if it does decrease HPV, does not result in a de-
crease in P
a
0
2
. First, the pre-existing hypoxia may
involve all of the lung (e.g., as a result of de-
creased F,0
2
); by definition, P
a
0
2
cannot change
Sublobar Ventilation-Perfusion Regulation
Collateral Ventilation
First Line of Defense
Hypoxic Normoxic
HPV (No Collateral Ventilation)
Second Line of Defense
Hypoxic Normoxic
(Air is less dense than blood and therefore easier to redistribute)
Figure 4- 13 Because gas is much less dense than blood and, therefore, easier to redistribute, collateral ventilation may be the
first line of defense and HPV the second line of defense against the development of arterial hypoxemia caused by sublobar
atelectasis (ATEL) or low V/Q (ventilation-perfusion ratio) areas. (HPV = hypoxic pulmonary vasoconstriction.)
significantly, yet hypoxic pulmonary vasoconstric-
tion may be greatly decreased. Second, the amount
of pre-existing HPV may be slight (e.g., healthy
awake or anesthetized patients, presence of high
concentrations of halogenated anesthetics or other
inhibiting factors); by definition, inhibition of a
slight amount of HPV has no effect on arterial
oxygenation. Third, most pulmonary vasodilators
are also systemic vasodilators and therefore cause
an increase in cardiac output. The increase in car-
diac output may cause an increase in P
9
0
2
(see
section 6) and the increase in P
0
2
may offset any
decrease in HPV and result in no change or even
increase in P
a
0
2
(see Table 3-5). However, and
adding to the complexity of the situation, changes
in cardiac output in either direction may increase
transpulmonary shunt (see section 5).
Virtually every drug known to cause pulmonary
vasodilatation has been studied with regard to ef-
fects on either HPV and/or arterial oxygenation.
Because most vasodilator drugs have been so ex-
tensively studied, Tables 4-2 to 4-6 list only those
studies performed after 1986 (since the first edi-
tion); these latest studies (and their references) and
the references in this text will allow the reader to
construct most of if not the entire literature tree for
each drug. Almost all of the studies with the very
potent systemic and pulmonary vasodilator drugs
(Table 4-2) have shown inhibition of HPV or have
a clinical effect (increased shunt, decreased P
a
0
2
)
that is consistent with decreased HPV. The potent
vasodilating drugs that have been shown to de-
crease HPV or have a clinical effect consistent
with decreased HPV are nitroprusside,
26
50-60
nitro-
glycerin,
26
6,_72
many
2
^88 (isoproterenol,
73-77
terbutaline,
78
), ritodrine,
79
orciprenaline,
80
salbuta-
mol,
81
adenosine,
82
and nitric oxide.
83-87
Interest-
ingly, nitroglycerin is metabolized intracellularly
Table 4-2 EFFECT OF THE VERY POTENT VASODILATORS ON HPV* AND/OR ARTERIAL
OXYGENATION
Table 4-3 EFFECT OF VASODILATOR PROSTAGLANDINS ON HPV* AND/OR ARTERIAL
OXYGENATION
into nitric oxide.
88
The vasodilator prostaglandins
(E and I
2
) (Table 4-3)
57
68
85
89
"
91
and the calcium
channel blockers (Table 4-4)
92
-'
03
inhibit HPV to
a much lesser extent and often have no effect on
arterial oxygenation.
57
90
'
l02
,03
Although most
animal studies with dobutamine show significant
inhibition of HPV,
73
74
77
, 04
-
,
7
human one-lung
ventilation studies
72
l<)8
demonstrate an improve-
ment in arterial oxygenation and HPV (Table 4-
5). On the basis of these latter human one-lung
ventilation studies,
72
l08
dobutamine should proba-
bly be considered the potent pulmonary vasodila-
tor of choice in clinical situations that demand
preservation of both gas exchange and hemody-
namic function. Minoxidil, diazoxide, isosorbide
dinitrate, phentolamine, and disodium cromogly-
cate may inhibit HPV.
92
I09
-"
3
Studies are conflict-
ing as to whether aminophylline"
4
"
5
and hydral-
azine"^"
8
decrease HPV and/or P
a
0
2
.
4. Anesthetic Drugs
The effect of anesthetic drugs on regional HPV
is covered extensively in chapter 8, Choice of An-
esthetic Drugs and Techniques.
5. Cardiac Output and Pulmonary
Vascular Pressure
Because changes in cardiac output (Q
t
) usually
cause directly proportional changes in pulmonary
vascular pressure (PVP) (e.g., see Fig. 3-11) (al-
though in heart failure Q, and PVP are inversely
related), this section assumes that the effect of
changes in Q
t
on HPV are exerted or mediated
through concomitant changes in PVP. It is under-
stood that Q,-induced changes in P-0
2
must be
taken into account when analyzing the effect of
changes in Q, on HPV and arterial oxygenation.
The HPV response is maximal when PVP is
normal and is decreased by either high or low PVP
(Fig. 4-14). The mechanism for high PVP inhi-
bition of HPV (whether the cardiac output is
high"
9-
'
24
or low"
9
122
is simple; the pulmonary
circulation is poorly endowed with smooth muscle
and cannot constrict against an increased vascular
pressure."
9-124
Furthermore, in one-lung ventila-
tion in the lateral decubitus position, it is obvious,
with all other factors remaining constant, that the
fraction of the cardiac output perfusing the col-
lapsed, nondependent lung will increase with in-
Pulmonary Vascular Pressure (PVP) and HPV:
One Lung Ventilation Conditions
LOW NORMAL HIGH
PULMONARY VASCULAR PRESSURE
igure 4-14 A schematic diagram of the effect of changes in pulmonary vascular pressure (PVP) on regional hypoxic pulmonary
asoconstriction (HPV). Both high and low pulmonary vascular pressures inhibit the HPV response. A shows an average PVP of
0 mm Hg and an average ventilated lung alveolar pressure of 5 mm Hg. Because vascular resistance is greater in the nonventilated
mg, blood flow is diverted to the ventilated lung (arrow, the HPV response). When PVP decreases to 5 mm Hg, it is possible for
>ne l to develop in the ventilated lung (i.e., collapse of pulmonary vessels as a result of alveolar pressure). The collapse of
ulmonary vessels increases vascular resistance in the ventilated lung and diverts blood flow to the nonventilated lung (arrow,
ihibition of the HPV response [ [ HPV]). With increased pulmonary vascular pressure (B) the poorly muscled pulmonary arteries
innot constrict as effectively as with a normal PVP, and the HPV response is decreased.
creasing pulmonary arterial pressure (i.e., the ef-
fect of gravity will be overcome).
125
The
mechanism for low PVP inhibition of HPV is
more complex. In order for this to occur, the hy-
poxic compartment must be atelectatic. Under
these circumstances, when PVP is decreased by
hemorrhage (low cardiac output), it is possible for
part of the ventilated lung (but not the atelectatic
lung) to be in a zone 1 condition (alveolar pressure
increases relative to pulmonary arterial pressure)
and experience a disproportionate increase in pul-
monary vascular resistance, which would divert
blood flow back over the atelectatic lung, thereby
inhibiting atelectatic lung HPV.
126
When PVP is
decreased by drug infusion (see section 3), HPV is
decreased, but simultaneous changes in P
0
2
(car-
diac output is usually high) may also influence
HPV (see next subsection) as well as have an
opposite effect on P
a
0
2
(shunted blood will have
an altered 0
2
content). Figures 4-14 and 4-18
summarize the effect of low and high PVP on
HPV.
6. P,0
2
The HPV response is also maximal when the
mixed venous Po
2
(Pv0
2
) is normal and is de-
creased by either high or low P
0
2
) (Fig. 4-15).
The mechanism for high P
0
2
inhibition of HPV
is presumably due to reverse diffusion of oxygen,
causing the oxygen tension of either the vessels or
interstitial or alveolar spaces or all these to be
increased above the HPV threshold.
127
That is, if
enough oxygen can get to some receptor in the
small arteriole-capillary-alveolar area, then the
vessels will not vasoconstrict. The mechanism for
low P
0
2
inhibition of HPV is a result of the low
P
0
2
decreasing alveolar oxygen tension in the
normoxic compartment down to a level sufficient
to induce HPV in the supposedly "normoxic"
Pv0
2
and HPV: One Lung Ventilation Conditions
Figure 4-15 Schematic diagram of the effect of changes in mixed venous oxygen tension (P*0
2
) on regional hypoxic pulmonary-
vasoconstriction (HPV). Both high and low P,o, inhibit the HPV response. Low P^0
2
lowers the alveolar Po
2
(P
A
O0 in the
ventilated lung, causing offsetting and competing HPV in the ventilated lung. Increased
,0
2
causes the oxygen tension to increase
in the nonventilated lung (perhaps in the pulmonary arteries, interstitium, alveolar space, and/or veins), thereby inhibiting nonven-
tilated lung HPV.
l ung.
1 2 8
1 2 9
The HPV in the "normoxic" lung com-
petes against and offsets the HPV in the originally
hypoxic lung and results in no blood flow diver-
sion away from the more obviously hypoxic lung.
Figures 4-15 and 4-18 summarize these mecha-
nisms.
7. F,0
2
in the Normoxic Compartment
Selectively decreasing the F,0
2
in the normoxic
compartment (from 1.0 to 0.5 to 0.3) will cause an
increase in normoxic lung vascular tone, thereby
decreasing blood flow diversion from the hypoxic
to the normoxic lung.
121, 129
Indeed, with unilateral
lung injury, ventilation of both lungs with hypoxia
gas mixture (F,0
2
= 0.12) induces much more
vasoconstriction in the normal, previously noncon-
stricted lung than in the injured and already hy-
poxically constricted lung, which redirects blood
flow to, increases the shunt through, and increases
edema in, the injured lung.
130
In addition, the de-
velopment of systemic hypoxemia with either bi-
lateral hypoxic ventilation or when there is a very
large hypoxic compartment and a small normoxic
compartment, may indirectly inhibit regional HPV
by stimulation of arterial chemoreceptors.
131
At the
other extreme, prolonged exposure to hyperoxia
(F,0
2
= 1.0) for 68 hours blunts a subsequent
whole-lung HPV response.
132
8. Vasoconstrictor Drugs
Older studies suggested that the vasoconstrictor
drugs (dopamine, epinephrine, phenylephrine)
constricted normoxic lung vessels preferentially,
thereby disproportionately increasing normoxic
lung pulmonary vascular resistance.
73
74
75, 12
The
increase in normoxic lung pulmonary vascular re-
sistance would be expected to decrease normoxic
lung blood flow and increase atelectatic lung blood
flow. The HPV-inhibiting effect of vasoconstrictor
drugs is similar to decreasing normoxic lung F,0
2
(see previous discussion).
In recent years, dopamine has been extensively
studied. Although one reasonably straightforward
study
133
agrees well with previous studies,
73
74
75
12

most recent studies
72
77
105
-
107
showed no signifi-
cant effect of dopamine on HPV and/or arterial
oxygenation (Table 4-6). On the basis of these
latter, more recent studies, dopamine appears to be
a reasonable cardiovascular stimulant to use in pa-
tients with lung disease, provided arterial oxygen-
ation is monitored.
9. Drugs That Increase HPV
a. ALMITRINE BISMESYLATE
This is a peripheral chemoreceptor agonist and
thus increases the ventilatory response to hy-
poxia.
134
Several studies suggested that HPV is
enhanced by low doses
135
"
138
of almitrine (includ-
ing the 1LV situation; see Fig. 4-16)
136
and atten-
uated by high doses of almitrine.
138-140
A possible explanation for this biphasic dose
response is that in lung regions that are partially
vasoconstricted, a low dose of almitrine increases
the amount of vasoconstriction in these areas so
that HPV is enhanced. As the dose of almitrine is
increased, nonspecific vasoconstriction occurs in
the nonhypoxic lung regions so that no effect on
HPV is detected. When the almitrine dose is high,
vasoconstriction of the hyperoxic lung regions oc-
curs so that blood flow is diverted back into the
hypoxic lung and HPV is diminished. This sug-
gests that almitrine, when used clinically, may be
useful only in a very narrow dose range. These
findings and hypothesis are consistent with studies
that show increased P
a
0
2
in patients with acute
Table 4-6 EFFECTS OF DOPAMINE ON HPV* AND/OR ARTERIAL OXYGENATION
Preparation Dose Effect on HPV
Arterial
Oxygenation Study
Canine whole-
lung hypoxia
Canine whole-
lung hypoxia
Canine unilobar
atelectasis
Canine unilobar
pulmonary
edema
1LV sheep
1LV human
10-20 g/kg/min No change or slight No change or slight decrease
decrease
5-20 g/kg/min No change No change
5-10 g/kg/min No change
5 g/kg/min No change or slight No change
increase
3-15 g/kg/min Moderate decrease Moderate decrease
5 g/kg/min No change No change or slight increase
Lejeuneetal.
l (, 7
(1987)
Lejeuneetal.
l 06
(1987)
Gardazetal.
l 05
(1988)
Light et al.
77
(I988)
Loick et al.'-" (1990)
Nomoto & Kawamura
72
(1989)
*Either the effect on HPV was a direct conclusion of the authors or the results are consistent with the interpretation in the table.
Abbreviations: HPV = hypoxic pulmonary vasoconstriction.
Figure 416 Changes in P
a
0
2
(
3
0
2
) from control values during one-lung hypoxia in group 1 (open circles) and after the
infusion of almitrine under one-lung hypoxia in group 2 (solid circles). Values are mean standard deviation. Asterisk denotes
significant difference between the two groups (p < .05). (From Takasaki M, Oh-Oka T, Saito Y, Kosaka Y: Low-dose almitrine
bismesylate improves pulmonary gas exchange during canine one-lung hypoxia. Crit Care Med 17:661-665, 1989. Used with
permission.)
and chronic respiratory failure associated with
chronic obstructive pulmonary disease and sep-
sis
141-146
and that the P
a
0
2
effects are dose depen-
dent.
147
-
149
b. PRODUCTS OF ARACHIDONIC ACID
METABOLISM (VASOCONSTRICTOR
LEUKOTRIENES)
As described in detail in chapter 3 (section
I.C.2.), the prostaglandins can decrease HPV and
the leukotrienes can increase HPV. Thus, blockage
of the prostaglandin pathway with indometha-
c
j
n
150-156 meclofenamate,
157158
and ibuprofen,
159160
results in predominance of leukotrienes and in-
creased HPV.
10. P
A
C0
2
and pH Changes
Hypocapnia has been thought to inhibit directly
regional HPV"
9
l 6
' -'
6 6
and hypercapnia to enhance
regional HPV directly."
9
I6
'
l66
l 67
Furthermore,
and consistent with the pH changes caused by
changes in C0
2
, alkalosis inhibits HPV (whether
respiratory or metabolic) and acidosis enhances
HPV (whether respiratory or metabolic)
168
l 69
;
however, it should be noted that there are many
discrepancies in the literature, and these dis-
crepancies have been the basis of several re-
views.
170
"
173
In addition, during one-lung ventilation condi-
tions, hypocapnia can be produced only by hyper-
ventilation of the one lung. The hyperventilation
requires an increased ventilated lung airway pres-
sure, which may cause increased ventilated lung
pulmonary vascular resistance, which, in turn, may
divert blood flow back into the hypoxic lung. Hy-
percapnia during one-lung ventilation seems to act
as a vasoconstrictor drug by selectively increasing
ventilated lung pulmonary vascular resistance
(which would divert blood flow back to the non-
ventilated lung) (see previous discussion). In ad-
dition, hypercapnia is ordinarily caused by hypo-
ventilation of the ventilated lung, which greatly
increases the risk of developing low V/Q and ate-
lectatic regions in the dependent lung. However, it
should be noted as a theoretical possibility that if
hypoventilation of the dependent lung is associ-
ated with decreased ventilated lung airway pres-
sure, ventilated lung pulmonary vascular resistance
may be decreased, which, in turn, would promote
or enhance HPV in the nonventilated lung. Figures
4-17 and 4-18 summarize these mechanisms.
C0
2
Changes and HPV:
One Lung Ventilation Conditions
Low Normal High
Figure 417 Schematic diagram of the effect of changes in CO, and regional hypoxic pulmonary vasoconstriction (HPV). Both
hypocapnia and hypercapnia inhibit the HPV response. Hypocapnia can directly pharmacologically dilate the hypoxic lung. In
addition, hypocapnia must be achieved by increasing minute ventilation (V
F
) and airway pressure (P
airw
) in the ventilated lung. The
increased airway pressure in the ventilated lung may selectively increase ventilated lung pulmonary vascular resistance (PVR),
thereby inhibiting nonventilated lung HPV. Hypercapnia may directly vasoconstrict the ventilated lung, thereby increasing ventilated
lung pulmonary vascular resistance, which will inhibit nonventilated lung HPV. In addition, hypercapnia may possibly be achieved
by decreasing minute ventilation, which would decrease airway pressure in the ventilated lung, which would decrease ventilated
lung pulmonary vascular resistance, which would enhance nonventilated lung pulmonary vascular resistance. This latter offsetting
mechanism to inhibition of nonventilated lung HPV is in brackets on the right-hand side of the figure.
Figure 418 Summary of the mechanisms of change in the hypoxic pulmonary vasoconstriction (HPV) response in Figures 4-
14, 4-15, and 4-17 caused by the changes in the physiologic variables of pulmonary vascular pressure, P
0
:
and CO, level. (P
ain
= airway pressure.)
11. PEEP and Alveolar Pressure
The effects of changes in airway pressure result-
ing from end-expiratory pressure and tidal volume
changes are discussed in detail in chapter 11. In
brief, selective application of PEEP to just the
normoxic, ventilated lung will selectively increase
pulmonary vascular resistance in the ventilated
lung and shunt blood flow back into the hypoxic,
nonventilated lung (i.e., decrease nonventilated
lung HPV).
19 I74
On the other hand, high-fre-
quency ventilation of the gas-exchanging lung is
associated with a low airway pressure and an en-
hancement of HPV in the collapsed lung.
175
The
effect of collateral ventilation HPV has already
been mentioned (see section IV.B.l.); it is obvious
from this discussion that there are important spe-
cies differences.
46
12. Other Important Systemic Factors
(Hypertension, Age, Sex, Infection)
HPV may be increased in patients with systemic
hypertension compared with normotensive pa-
tients.
176
Because calcium channel blockade with
nifedipine can almost abolish the HPV response in
normotensive and hypertensive patients, the find-
ings are consistent with the interpretation that
HPV is mediated by calcium ions and the availa-
bility of calcium ions is facilitated by hyperten-
sion.
In addition, the amount of smooth muscle on
the vessels may determine the degree of pulmo-
nary hypertension that develops during chronic ex-
posure to hypoxia (which again may be species
dependent).
46
Age is one of the factors that deter-
mine the amount of musculature of the pulmonary
circulation (newborns are invested with the most
muscle, which then regresses). Age appears to be
inversely related to the magnitude of the HPV
response,
46
l77
~
179
and the reason for this inverse
relationship may be the amount of muscle invest-
ment of the vessels. Furthermore, patients with a
large patent ductus arteriosus or a large unre-
stricted ventricular septal defect will have failure
of the normal regression of perinatal musculature
and extension of the muscle peripherally as well
as medial hypertrophy. This may explain why
these patients may exhibit an abnormal increase of
pulmonary vascular resistance in response to ex-
ercise or stress
180
m
and an exaggerated HPV re-
sponse to alveolar hypoxia.
182
In a given species, females appear to have a
more vigorous HPV response than males.
46
There
is some evidence that certain types of infections
(which may cause atelectasis), particularly granu-
lomatous and pneumococcal infections, may in-
hibit HPV.
183
I84
2. Blood Flow to the Dependent,
Ventilated Lung
The dependent lung usually has an increased
amount of blood flow owing to both passive grav-
itational effects and active nondependent lung
vasoconstrictor effects (Fig. 4-8, lower panel).
However, the dependent lung may also have a
hypoxic compartment (areas of low ventilation-
perfusion ratio and atelectasis) that was present
preoperatively or that developed intraoperatively.
The dependent-lung hypoxic compartment may
develop intraoperatively for several reasons. First,
in the lateral decubitus position, the ventilated,
dependent lung usually has a reduced lung volume
owing to the combined factors of induction of gen-
eral anesthesia and circumferential (and perhaps
severe) compression by the mediastinum from
above, by the abdominal contents pressing against
the diaphragm from the caudad side, and by sub-
optimal positioning effects (rolls, packs, shoulder
supports) pushing in from the dependent side and
axilla (Fig. 4-6).
4
l 0
l 5
l85
Second, absorption ate-
lectasis can also occur in regions of the dependent
lung that have low ventilation-perfusion ratios
when they are exposed to high inspired oxygen
concentration.
186
Third, difficulty in secretion re-
moval may also cause the development of poorly
ventilated and atelectatic areas in the dependent
lung. Finally, maintaining the lateral decubitus po-
sition for prolonged periods of time may cause
fluid to transudate into the dependent lung (which
may be vertically below the left atrium) and cause
further decreased lung volume and increased air-
way closure in the dependent lung.
187
A decrease
in lung volume and an increase in airway closure
in the dependent lung will create areas that have a
low ventilation-perfusion ratio or atelectasis (see
chapter 3).
The development of low ventilation-perfusion
ratio and/or atelectatic areas in the dependent lung
will increase vascular resistance in the dependent
lung
185
188
(due to dependent lung HPV),
43
thereby
decreasing dependent-lung blood flow and increas-
ing nondependent-lung blood flow.
189
Stated dif-
ferently, the pulmonary vascular resistance in the
ventilated compartment of the lung determines the
ability of the ventilated, and supposedly normoxic,
lung to accept redistributed blood flow from the
hypoxic lung. Clinical conditions that are indepen-
dent of specific dependent-lung disease, but which
may still increase dependent-lung vascular resis-
tance in a dose-dependent manner, are a decreas-
ing inspired oxygen tension in the dependent lung
(from 1.0 to 0.5 to 0.3
121
129
,89
and a decreasing
temperature (from 40 to 30C).
,9

In view of the factors just listed, which can
affect dependent- and nondependent-lung vascular
resistance and blood flow, it is obvious that the
method used to ventilate the dependent lung is an
extremely important determinant of blood flow
distribution during one-lung ventilation (see chap-
ters 11 and 12). For example, if the dependent lung
is hyperventilated, then the resultant hypocapnia
may inhibit HPV. If the method of ventilation in-
volves an excessive amount of airway pressure,
due either to use of high PEEP levels or to very
large tidal volumes, the deleterious effects of in-
creased dependent-lung airway pressure, namely
increasing dependent-lung vascular resistance
(which would increase nondependent-lung blood
flow), may outweigh the beneficial effects of the
opening of atelectatic and low V/Q areas in the
dependent lung. Finally, a high inspired oxygen
concentration to the dependent lung may cause
vasodilatation in it, enhancing nondependent lung
HPV; however, absorption atelectasis is promoted
by a high inspired oxygen concentration to low
V/Q areas in the dependent lung.
186
189
3. Miscellaneous Causes of Hypoxemia
During One-Lung Ventilation
Still other factors may contribute to hypoxemia
during one-lung ventilation (see chapter 3). Hy-
poxemia due to mechanical failure of the 0
2
sup-
ply system or of the anesthesia machine is a rec-
ognized hazard of any kind of anesthesia.
191-194
Gross hypoventilation of the dependent lung can
be a major cause of hypoxemia. Malfunction of
the dependent-lung airway lumen (blockage by se-
cretions) and malposition of the double-lumen en-
dotracheal tube are other common causes of an
increased P(A-a)0
2
and hypoxemia. Resorption of
residual oxygen from the clamped, nonventilated
lung is time dependent and may account for a
gradual increase in shunt and decrease in P
a
0
2
after one-lung ventilation is initiated.
188
With all
other anesthetic and surgical factors constant, any-
thing that decreases the mixed venous partial pres-
sure of oxygen P^0
2
(decreased cardiac output,
increased oxygen consumption [excessive sympa-
thetic nervous system stimulation, hyperthermia,
shivering]) will cause an increased P(A-a)0
2
.
195196
Finally, transfusion of blood may cause pulmonary
dysfunction, and the dysfunction has been attrib-
uted to the action of isoantibodies against leuko-
cytes, which causes cellular aggregation, micro-
vascular occlusion, and capillary leakage. Indeed,
such a reaction has been described during pro-
longed one-lung ventilation.
197
Interestingly, the
noncollapsed lung was preferentially injured and
the collapsed lung showed only minimal radio-
logic signs of edema after re-expansion.
197
REFERENCES
1. Sibert KS, Biondi JW, Hirsch NP: Spontaneous respira-
tion during thoracotomy in a patient with a mediastinal
mass. Anesth Analg 66:904-907, 1987.
2. Tarhan S, Moffitt EA: Principles of thoracic anesthesia.
Surg Clin North Am 53:813-826, 1973.
3. Wulff KE, Aulin I: The regional lung function in the
lateral decubitus position during anesthesia and operation.
Acta Anesthesiol Scand 16:195-205, 1972.
CHAPTER 5
Preoperative Cardiopulmonary
Evaluation
I. Introduction 1. History, General Risk Factors,
II. Tumors and Other Masses of the Physical Examination
Lung and Bronchi 2. Pulmonary Function Tests
A. Major Causative Factors and History a. Whole-Lung Function Tests
1. Bronchopulmonary Symptoms (1) Spirometry
a. Cough (2) Flow-Volume Curves
b. Sputum (3) Carbon Monoxide
c. Hemoptysis Diffusing Capacity
d. Chest Pain (4) Maximum Breathing
e. Dyspnea Capacity
f. Wheeze (5) Exercise Testing
2. Extrapulmonary Intrathoracic (6) Lung Volume
Symptoms b. Regional Lung Function Tests
3. Extrathoracic Metastatic (1) Radioisotope Regional
Symptoms Perfusion, Ventilation-
4. Extrathoracic Nonmetastatic Perfusion Studies
Symptoms (Radiospirometry)
5. Nonspecific Symptoms (2) Lateral Position Test
B. Physical Examination (3) Regional Bronchial
C. Common Laboratory Tests Balloon Occlusion
D. Diagnosis of Lung Cancer (4) Regional Pulmonary
1. The Carcinomas of the Lung Artery Balloon Occlusion
2. Diagnosis of the Presence of c. Sequence of Tests for Lung
Lung Cancer (Is Lung Carcinoma Resection Surgery
Present? Cell Type?) 3. Pulmonary Vascular and Right
a. Chest Roentgenogram Ventricular Function and Testing
b. Bronchoscopy 4. Left Ventricular and Coronary
c. Sputum Cytology Artery Function and Testing
d. Needle Biopsy a. Exercise-EKG Testing
e. Summary of Diagnosis of b. Exercise-Thallium Testing
Lung Carcinoma c. Dipyridamole-Thallium Testing
E. Staging of Lung Cancer (Has the d. Radionuclide Angiography
Carcinoma Spread?) e. Echocardiography
1. Staging (Tumor Size and Direct f. Coronary Angiography
Extension) 5. Cardiovascular Sequelae of
2. Staging (Metastasis to Lymph Therapeutic Thoracic Radiation
Nodes) 6. Anesthetic Implications of Cancer
3. M Staging (Metastasis) Chemotherapy
4. Intraoperative and Postsurgical III. Mediastinal Masses
Staging A. Types of Masses
5. Staging in Small-Cell Lung B. Signs and Symptoms
Cancer C. Diagnostic Workup Logic for
6. Surgical Procedure as Dictated Mediastinal Masses
by Staging IV. Pleural Disease/Effusion
7. Prognosis and Survival as a A. Anatomy and Pathophysiology
Function of the Staging of Lung B. Symptoms, Signs, and Diagnostic
Cancer Workup Logic
8. Summary of Results of Diagnosis V. Pericardial Disease/Effusion
and Staging of Lung Cancer A. Anatomy and Pathophysiology
F. Physiologic Assessment of the B. Diagnosis and Treatment
Patient for Surgery
152
Preoperative Cardiopulmonary Evaluation I 53
I. INTRODUCTION
The vast majority of noncardiac, noncardiopul-
monary by-pass thoracic surgery consists of resec-
tional or repair procedures for cancer and other
masses of the lung and bronchi (including infec-
tious processes such as tuberculosis, fungal dis-
eases, bronchiectasis, lung abscess, and empyema
and anomalies such as arteriovenous malforma-
tions and pulmonary sequestration), mediastinal
masses (including thoracic aortic aneurysms), and
esophageal lesions. Because all aspects of anesthe-
sia for esophageal surgery are considered in chap-
ter 16, this chapter is divided into two basic pre-
operative evaluation categories: lung and bronchia}
masses and mediastinal masses. However, because
bronchogenic carcinoma of the lung is by far the
most common indication for thoracic surgery (see
reference 6 of chapter 1 ), this chapter places spe-
cial emphasis on the evaluation of this type of
lesion. The chapter concludes with considerations
of pleural and pericardial disease/effusions.
II. TUMORS AND OTHER MASSES
OF THE LUNG AND BRONCHI
Figure 5-1 shows the overall preoperative eval-
uation logic plan for patients with lung and/or
bronchial masses. The plan involves three basic
steps: First, is lung cancer present (diagnosis of
mass lesion) and if so what is the cell type? Se-
cond, has the lung cancer spread (is the lung can-
cer surgically resectable based on the TNM stag-
ing system)? Third, can the patient tolerate the
planned procedure (is the lung mass operable on
the basis of physiologic assessment of the patient)?
The following text discusses these three steps, and
each step has its own logic tree.
A. Major Causative Factors and History
The history often raises suspicion of the diag-
nosis of lung cancer.
1-
* The incidence of cancer of
the lung (carcinoma makes up approximately 90
per cent, adenomas 8 to 10 per cent, and benign
masses 1 per cent of lung cancer) increases with
age (peaks in the sixth to seventh decades of life)
and has a male to female ratio of 2:1 (as of 1991,
with females rapidly catching up to males), and,
more than 90 per cent of the time, the patient has
a history of heavy cigarette smoking and recent
weight loss and resides in an urban area (greater
degree of pollution). Because incidence and risk
of cancer and death from all lung cancer are di-
rectly proportional to the pack-years (Table 5-1
6
and Fig. 5-2
7
), the smoking history should be
quantified in terms of pack-years (e.g., 2 packs per
day [ppd] X 40 years = 80 pack-years).
Lung cancer is far more likely to be present
when chronic airflow obstruction is present
8 9
(which is usual in long-term, heavy smokers).
However, a small percentage of lung carcinomas
occur in apparent nonsmokers (< 10 per cent), but
many of these can be traced to a history of passive
or involuntary smoking.
,a
~
12
Although there is no
longer any doubt that smoking causes primary
lung cancer
13-15
(as well as cancer of the orophar-
ynx, larynx, esophagus, bladder, and uterus
16
and
extrapulmonary sman-cen carcinoma
17
), host fac-
tors (such as immune surveillance and repair
mechanisms) may modu)ate susceptibiihy ar>o
severity.
18 |y
For one example, the incidence of
lung cancer increases with age, but the lung cancer
in the third, fourth, and fifth decades of life is
much more aggressive (brief duration of symp-
toms, advanced stage, decreased time of survival)
than lung cancer in older patients.
20-22
Conversely,
heavy smokers do not necessarily develop lung
cancer. For another example, dietary factors (beta
carotene, vitamins A and E, selenium) may play a
role in the causation/prevention of lung cancer.
23-25
Lung carcinoma has a high incidence of occur-
rence in workers in some chemical industries
(asbestos.
26
27
arsenic, chromtes, coal gas, and
nickel) than in the general population. Uranium
miners have a greatly increased risk for the devel-
opment of lung carcinoma, especially if they
smoke. Radon gas (sixth daughter of uranium 238)
is ubiquitous and is now recognized as perhaps the
second most important cause of lung cancer.
28
Five per cent of the patients with lung carci-
noma are completely asymptomatic at the time of
discovery,
5
and in this group the tumor is discov-
ered only on routine roentgenographic examina-
Table 5-1 STANDARDIZED MORTALITY
RATIOS FOR LUNG CANCER IN
WOMEN IN ACS-CPS II BY NUMBER
OF CIGARETTES CURRENTLY
SMOKED AND DURATION OF
SMOKING't
Duration of
Smoking
(Years)
21-30
31-40
41-70
1-10
2.9
7.9
10.0
Cigarettes Per
11-19
6.7
19.2
17.0
20
13.6
19.2
25.1
Day
21-30
18.4
26.5
34.3
31 +
18.9
25.3
38.3
*Adapted with permission from Garfinkel L, Stellman SD:
Smoking and lung cancer in women: Findings in a prospective
study. Cancer Res 48:6951-6955, 1988.
tin comparison with nonsmoking women.
Abbreviation: ACS-CPS II = American Cancer Society
Cancer Prevention Study II.
154 Preoperative Cardiopulmonary Evaluation
Figure 5-1 The preoperative evalua-
tion of masses of the lung and bronchi
involves three basic steps. Step 1 con-
sists of determining whether lung carci-
noma is present and, if so, the cell type.
Step 2 consists of determining whether
the carcinoma has spread beyond its lo-
cal confines. Step 3 involves physiologic
assessment of the patient for the planned
surgical procedure. This preoperative
evaluation diagram displays the logic
necessary for a patient to arrive at tho-
racotomy .
Death due to lung cancer in relation to
cigarette smoking
Figure 5- 2 Number of lung cancer deaths/100,000 smokers
and nonsmokers. (From Spiro S: Lung cancer: Presentation and
treatment. Medicine International 3798-3805, 1991. Used with
permission.)
tion of the chest. The vast majority of patients,
however, have one or more symptoms related to
the presence of the tumor. The symptoms may be
designated as broncho-pulmonary, extrapulmonary
intrathoracic, extrathoracic metastatic, extratho-
racic nonmetastatic, and nonspecific (Table 5-2).'~
5
On the average, symptoms have been present for 6
to 7 months prior to the time the patient seeks
medical advice; since the first chest x-ray findings
frequently antedate the first symptoms by several
months, lung carcinoma will be at least a year old
(and perhaps 2 to 5 years old) by the time of
clinical presentation.
1. Bronchopulmonary Symptoms
Bronchopulmonary symptoms arising from in-
volvement of the lung are due to bronchial irrita-
tion, ulceration, obstruction, infection distal to the
obstruction, or a combination of these processes.
a. COUGH
In a large series of patients with carcinoma of
the lung, 75 per cent had cough as one of the
major symptoms, and this symptom was severe in
40 per cent of the patients. However, cough is
possibly the most common manifestation of respi-
ratory disease in general. It is so common among
cigarette smokers that many of them regard a
morning cough as "normal." The most common
stimulus to cough is the formation of sputum in
the respiratory tract (see the following), and the
cough process is an essential element in keeping
the tract clear.
b. SPUTUM
The normal adult produces about 100 ml of
mucus from the respiratory tract in a day. When
excess mucus is formed, it may accumulate, stim-
ulate the mucous membrane, and be coughed up
as sputum. Sputum in patients with bronchogenic
Table 5-2 FREQUENCY OF SYMPTOM OCCURRENCE AT PRESENTATION IN
BRONCHOGENIC CARCINOMA*
Symptom
Frequency of
Occurrence
(Per Cent)
*Based on data from Shields,' Spiro,
2
Le Roux,
3
Jones,
4
and Ferguson.
5
carcinoma may be formed in response to physical,
chemical, or infective insult to the mucous mem-
brane of the airways.
Mucoid sputum is clear or white. Black sputum
is due to the detritus of cigarette or atmospheric
smoke. Purulent sputum contains pus mixed with
mucus. Purulent sputum is usually yellow, but if it
has been stagnant it may be green, owing to the
action of verdoperoxidase, derived from neutro-
phils. Failure to clear a recent change in the quality
and quantity of sputum within a few days of initi-
ating antibiotic therapy should raise suspicion of a
neoplasm. Blood-stained sputum can vary from
small streaks to gross hemoptysis (see the follow-
ing) and always warrants investigation for carci-
noma. Hemoptysis, generally episodic blood
streaking of the sputum, is present in 57 per cent
of patients with bronchogenic carcinoma and is the
first symptom in many.
C. HEMOPTYSIS
Hemoptysis affects 50 to 70 per cent of patients
with lung cancer at some point in their clinical
course. It is usually not severe, rarely is life-threat-
ening, and may be occasional and/or a one-time
event.
d. CHEST PAIN
Chest pain is present in 40 per cent of patients
presenting with a new carcinoma. It is usually a
mild, constant dull ache on the side of the tumor
and is often due to erosion of a rib (disease within
the lung is usually painless). Another important
form of chest pain with lung carcinoma is pleuritic
pain. It is due to direct tumor extension to the
parietal pleura and is characteristically sharp, is
worse on breathing and coughing, and can usually
be accurately localized by the patient. Mediastinal
tumors can cause pain that is usually aching and
retrosternal but poorly localized.
. DYSPNEA
Dyspnea is a common complaint in patients
with chronic lung disease and lung carcinoma (39
per cent). In chronic diseases, it is common to find
that patients begin to complain of dyspnea only
after the respiratory reserve is quite severely im-
paired, whereas in patients with lung carcinoma
dyspnea occurs more abruptly and with less objec-
tive functional impairment. The degree of dyspnea
should be approximately quantified (i.e., how far
can the patient walk, how many steps can be
climbed, and so on). The time period over which
dyspnea has developed is important in the diagno-
sis; dyspnea due to lung carcinoma develops over
weeks to months.
f. WHEEZE
This is described by 10 per cent of patients with
lung cancer and is frequently localized to one side.
It is due to airway obstruction by the tumor and, if
located in the trachea, severe dyspnea and stridor
(i.e., inspiratory wheeze) may develop.
2. Extrapulmonary Intrathoracic
Symptoms
Other symptoms of chest disease occur as a
result of growth of the tumor beyond the confines
of the lung. These symptoms are due to involve-
ment of the pleura, chest wall, diaphragm, medias-
tinal structures, and contiguous nerves. Approxi-
mately 15 per cent of patients with carcinoma of
the lung have these kinds of extrapulmonary intra-
thoracic symptoms.
Pleural effusion is due to either metastatic in-
volvement of the pleura (blood stained) or obstruc-
tion of peripheral lymphatic drainage (clear color)
or central (thoracic duct) lymphatic drainage (chy-
lous effusion). Chest wall pain is due to direct
involvement of the chest wall by tumor. Dys-
phagia is due to partial obstruction of the esopha-
gus by the tumor in the paraesophageal lymph
nodes. The superior vena cava syndrome (dyspnea,
dysphagia, stridor, blackouts, and severe headache
on coughing) is due to obstruction of the superior
vena cava by right paratracheal lymphadenopathy.
Pain down the arm is due to involvement of the
branches of the brachial plexus from tumors lo-
cated in the superior sulcus. Horner's syndrome
(small pupil, partial ptosis, enophthalmos, lack of
thermal sweating on the ipsilateral half of the face)
is due to involvement of the cervical sympathetic
chain and will often be present in patients who
have pain down the arm. Hoarseness is due to
paralysis of the vocal cord as the result of involve-
ment of the left recurrent laryngeal nerve (at the
left hilum) or rarely, of the right recurrent laryn-
geal nerve. Paralysis of the recurrent laryngeal
nerve(s) may also cause difficulty with expectora-
tion and chronic aspiration (caused by failure of
the vocal cords to adduct). Brachial plexus neuritis
can occur with an apical tumor (usually T, distri-
bution), and pericarditis is caused by direct in-
volvement of the pericardium.
3. Extrathoracic Metastatic Symptoms
Symptoms resulting from metastatic spread of
the tumor outside the thorax account for a small
percentage of the presenting or major complaints
of patients with carcinoma of the lung (although
as many as 50 per cent of patients with lung car-
cinoma actually have asymptomatic extrathoracic
Preoperative Cardiopulmonary
1
Evaluation 157
metastases).
5
These extrathoracic metastatic symp-
toms can be referable, in order of general decreas-
ing frequency, to brain, skeleton, liver, adrenals,
gastrointestinal tract, kidneys, and pancreas. In
these cases, the history is extremely important be-
cause any positive history referable to these organs
requires specific organ workup for metastatic dis-
ease (see staging later) and, if found, precludes
surgery. Extrapulmonary small-cell carcinoma
(without lung involvement) is a distinct clinico-
pathologic entity, is associated with smoking, and
can occur in virtually every organ.
17
4. Extrathoracic Nonmetastatic
Symptoms
The extrathoracic nonmetastatic symptoms (may
occur in up to 10 per cent of patients with lung
carcinoma)
5
are usually due to paraneoplastic syn-
drome caused by secretion of endocrine or endo-
crine-like substances by the tumor (see chapter 3,
Table 3-7). The endocrine-like manifestations in-
clude Cushing's syndrome, excessive antidiuretic
hormone secretions (low serum sodium and
plasma osmolarity, high urine osmolarity), carci-
noid syndrome, hypercalcemia, ectopic gonadotro-
pin secretion, and hypoglycemia. The neuromus-
cular manifestations consist of carcinomatous
myopathies (Eaton-Lambert syndrome) and var-
ious myopathies related to brain dysfunction.
Other manifestations can be skeletal (clubbing,
pulmonary hypertrophic osteoarthropathy), derma-
tologie (scleroderma, acanthosis nigricans), vascu-
lar (thrombophlebitis), and hematologic.
5. Nonspecific Symptoms
Weight loss, anemia, protein-energy malnutri-
tion,
29
weakness, anorexia, lethargy, and malaise
occur in a large number of patients. Vague febrile
respiratory syndromes (cold-like) may be present
in 22 per cent of these patients. In 10 to 15 per
cent, these symptoms instigate the initial visit to
the physician.
B. Physical Examination
The basic tools of observation, inspection, pal-
pation, and percussion should allow the physician
to assess, in a gross way, the overall severity of
chronic lung disease, whether major consolidation,
atelectasis, or pleural effusion is present, and
whether there is any obvious extrathoracic compli-
cation of thoracic carcinoma (Table 5-3). How-
ever, it should be noted that the most common
manifestation of lung cancer on physical exami-
nation is the presence of palpable supraclavicular
lymph nodes because lung carcinoma is most often
in an advanced, inoperable stage at the time of
presentation. Clubbing of the fingers should al-
ways raise the possibility of intrathoracic neo-
plasm. Because it is mandatory subsequently to
use much more sensitive radiologic means of de-
termining resectability, further discussion of phys-
ical examination methods for making these deter-
minations is not continued here. Similarly, the
questions of overall severity of chronic lung dis-
ease and whether the patient can tolerate the
planned procedure (operability) can be much more
quantitatively answered by pulmonary function
testing than by the findings of the physical exami-
nation (see Physiologic Assessment of the Patient
for Surgery).
C. Common Laboratory Tests
Some of the routine laboratory tests that are
performed on all patients are especially relevant to
the preoperative evaluation of the patient with a
lung or bronchial mass. These tests can be divided
into those that help establish the diagnosis of lung
cancer (e.g., chest x-ray and sputum for cytology),
those that help establish the diagnosis of metastatic
lung cancer (e.g., liver and bone enzymes, blood
urea nitrogen and creatinine, and the urinalysis),
and those that help physiologic assessment (e.g.,
hemoglobin concentration). Each of these three di-
agnostic areas is fully discussed in the following
pages (along with the role of the common labora-
tory test in the workup logic).
D. Diagnosis of Lung Cancer
1. The Carcinomas of the Lung
Lung cancer
1
2
-
30
31
is the most common malig-
nant disease and cause of cancer death in both
sexes. On the basis of ordinary light microscopy
histopathologic findings, lung cancer was formerly
divided into four major types: squamous (epider-
mal), adeno, large cell, and small cell. However,
there is growing acceptance of the hypothesis that
all lung cancer starts by activation of an oncogene
(e.g., by smoke, radiation) (an activated oncogene
is called a proto-oncogene) or by loss of an anti-
oncogene in a cell capable of differentiating into
the various pathologic forms (Fig. 5-3).
32
~
36
The
cancer cell generation scheme in Figure 5-3 sug-
gests that there is a common stem cell for all types
of lung cancer, but that there is a basic genetic
difference between small-cell lung carcinoma
(SCLC) and non-small-cell lung carcinoma (non-
SCLC) (i.e., the adeno, large, squamous type).
Table 5-3 PHYSICAL FINDINGS THAT OCCUR WITH PULMONARY PATHOLOGY
Condition
Inspection
and Palpation Percussion Fremitus Breath Sounds
Adventitious
Sounds Other
Normal
Consolidation
Major atelectasis
Pleural effusion or empyema
Cavitation
Diffuse pulmonary fibrosis;
interstitial lung disease
Emphysema
Bronchitis
Bronchial
asthma
Pulmonary
edema
Pneumothorax
Fibrothorax
Equal rib and diaphragm Resonant Present Vesicular None
movement
Slight restriction of motion on Dull Increased Bronchial Rales
side affected
Slightly small and restricted on Dull Normal Diminished Rales after deep
side affected breath or cough
Reduced movement on side Dull or flat Absent Diminished or absent Friction rub early
affected
Normal Usually normal Usually present Amphoric Coarse rales
Symmetrically diminished Normal Normal Harsh vesicular with Coarse rales
prolonged expiration uninfluenced by
coughing
Enlarged and restricted Hyper-resonant Normal or reduced Diminished with Occasional rhonchi
bilaterally prolonged expiratory fine rales late in
phase inspiration
Normal Normal Normal Vesicular with Rhonchi with
prolonged expiration coarse rales
Normal or enlarged Hyper-resonant Reduced Diminished with Wheezes
prolonged expiratory
phase
Normal Normal Normal Bronchial if interstitial; Moist rales
vesicular if alveolar
Slightly enlarged and restricted Hyper-resonant Absent Diminished or absent None
movement
Small and very restricted Dull Present Reduced to None
movement absent
"Bronchophony" in normal
spoken voice
"Egophony" (E to A) and
whispered pectoriloquy
Tracheal and mediastinal shift
toward
Mediastinal shift away
Coin sign
Hoover's sign; high clavicle;
muscular wasting (pink
puffer)
Cyanotic (blue bloater)
Distress
Distress
Mediastinal shift away
Mediastinal shift toward
Figure 5- 3 Hypothesis for the development of histologic types of lung cancer.
12
^
This is supported by the occurrence of mixed
pathologic types within the non-small-cell group
such as adenosquamous carcinoma (by light mi-
croscopy). Indeed, lung carcinoma can be seen to
differentiate spontaneously into various histologies
in tissue culture, and the presence of adenosqua-
mous histologies may simply represent a phase in
this ongoing differentiation from a common stem
cell.
37
This hypothesis is even more dramatically sup-
ported by the finding that, when large-cell carci-
noma is examined by electron microscopy and im-
munoperoxidase studies, it is obvious that this
tumor may be further divided into five groups as
follows: squamous, adenomatous, adenosquamous,
neuroendocrine, and undifferentiated.
38
Finally,
lung cancers of both small-cell and non-small-cell
variety share a number of common antigens and
neuroendocrine markers
32
39
; this overlap in the
expression of biomarkers between small-cell and
non-small-cell lung cancer, and within the various
subtypes of non-small-cell lung cancer, also sug-
gests a common stem cell for all types of lung
cancer. The common ancestor of the epithelial
lung tumor is thought to be the simple cuboidal
endodermal cell of the primitive lung bud.
39
The most common carcinomas of the lung (epi-
dermoid, small cell, adeno, large cell), their rela-
tive incidence of occurrence, and most usual
growth rate characteristics are listed in Table 5-4.
However, it should be realized that all of these
carcinomas are capable of a wide range of growth
characteristics, and all cell type combinations are
possible (combined adenosquamous carcinoma is
the most common mixed tumor; this mixed carci-
noma is responsible for 2 per cent of all lung
carcinoma
37
40
). Other very rare, malignant pul-
monary neoplasms not listed in Table 5-4 are al-
veolar cell carcinoma, carcinoid, bronchial gland
tumors, papillary tumors, sarcomas, and melano-
mas.
Clinically, the most relevant differential diag-
nosis is that between SCLC on the one hand and
the other lung carcinomas (non-SCLC) on the
other. This is so because SCLC differs from non-
SCLC by showing a higher growth rate and a
propensity to earlier and more extensive metastas-
tic spread (see Table 5-4). Therefore, SCLC has a
much worse prognosis. In non-SCLC, surgery, if
possible, is the treatment of choice. The early oc-
currence of metastatic lesions in most SCLC cases
excludes surgical intervention as an effective treat-
ment for this type of lung cancer. However, SCLC
shows a remarkable initial sensitivity to chemo-
and radiotherapy.
42
In SCLC, a clinically complete
response is often observed after induction chemo-
therapy. Only few patients turn out to be cured,
however, and at relapse a tumor may emerge
whose cells are biologically different from those
present in the pretreatment tumor (i.e., they are
refractory to further treatment).
In order to understand fully the diagnosis and
the staging of lung carcinoma, it is necessary to
have some appreciation of the natural history of
each of the lung carcinoma cell types. Carcinoma
of the lung may spread by direct extension, by
lymphatic metastasis to lymph nodes, and by he-
matogenous metastasis to lymph nodes, and by
hematogenous metastasis to distant organs (Fig. 5-
4). Direct extension can be in any direction and
can involve any structure within the chest (pleura,
chest wall, diaphragm, all mediastinal structures).
Table 5-4 CARCINOMAS OF THE LUNG
*Taken in part from Bains
41
and Humphrey et al.
4
"
tThe incidence of squamous carcinomas is on the decrease, and the incidence of adenocarcinoma is on the increase.
^Central = Inner two thirds of lung or proximal to third to fourth bronchial generation and usually visualized with a fiberoptic
bronchoscope.
^Peripheral = Outer one third of lung or distal to third to fourth bronchial generation and not usually visualized with a fiberoptic
bronchoscope.
Spread of Carcinoma of the Lung
1. Direct
Extension
2. Lymphatic
Metastasis
3. Hematogenous Metastasis
Figure 5-4 Carcinoma of the lung spreads in three ways: first, by direct extension to the mediastinum, pleura and chest wall,
diaphragm, and bronchi; second, by lymphatic metastasis to hilar, paratracheal and supraclavicular, and cervical nodes (proceeding
from the distal to the proximal nodes, the incidence of involvement of the nodes decreases); third, by hematogenous metastasis to
brain, liver, bone, adrenals, and kidney.
Blockage of a bronchus can cause distal atelectasis
and infection. Cavitation may be due to either ne-
crosis within the tumor mass or abscess formation
distal to an obstructed bronchus. Lymphatic metas-
tasis follows the lymph sump pathway of hilar and
mediastinal nodal stations to the venous outlets
(see chapter 2). All mediastinal structures can be
potentially affected by lymph node enlargement
and erosion. Hematogenous spread is due to inva-
sion of the pulmonary veins by tumor cells. The
blood-transported tumor cells are most frequently
deposited and grow in brain, bone, liver, adrenal
glands, and kidney.
2. Diagnosis of the Presence of Lung
Cancer (Is Lung Carcinoma Present?
Cell Type?)
The diagnosis of the presence of lung cancer'"*
i 0
31
is made most often by use of the chest roent-
genogram, bronchoscopy, sputum cytology, and
percutaneous needle biopsy (Fig. 5-5). Although
clearly positive results from any of these four tests
establishes the diagnosis of lung carcinoma, the
diagnosis is most often established by a combina-
tion of chest roentgenogram consistent with the
diagnosis along with a positive result from one of
the other three tests. The latter three tests permit
diagnosis of cell type in 75 per cent of cases.
a. CHEST ROENTGENOGRAM
It is useful to think of the radiograph as consist-
ing broadly of four optical densities: black (air).
dark gray (fat), light gray (soft tissue/fluid), and
white or colorless (bone/calcification). To be seen
as a separate structure on a chest radiograph, an
object needs to have a radiograph density that con-
trasts with its environment and borders that are
tangential to the X-ray beam. Lesions of soft tissue
density in the air-density lung often fulfill these
conditions, making the high-voltage (penetrating)
130- to 140-kV chest radiograph the most impor-
tant preliminary investigative modality once lung
cancer is suspected.
It has been estimated that when a tumor of the
lung is first detected on a chest roentgenogram, it
has completed three fourths of its natural history,
4
^
and this first roentgenographic abnormality fre-
quently antedates the first symptoms or signs of
the disease by 7 or more months.
44
By the time
bronchial carcinoma becomes symptomatic, the
chest roentgenogram is abnormal in 98 per cent,
and the abnormality is most suggestive of tumor
in more than four fifths of all these patients.
The roentgenographic findings present due to
carcinoma of the lung (Fig. 5-6) may be the result
of the presence of the tumor itself within the lung
(70 per cent are centrally located), of changes in
the pulmonary parenchyma distal to a bronchus
Preoperative Evaluation of Masses of the Lung and Bronchi
epJ: Is Lung Carcinoma Present? Cell Type?
History, Physical Examination, Chest Roentgenogram
Diagnosis of Lung Carcinoma and Cell Type Confirmed
Figure 5-5 Preoperative evaluation logic of step 1 for determining the presence of lung carcinoma type. See text for full
explanation.
obstructed by the tumor (atelectasis, infection, and
cavitation), and of spread of the tumor to extrapul-
monary intrathoracic sites (hilar and mediastinal
lymph nodes, pleura, chest wall, and diaphragm).
Because the primary route of lymphatic spread of
lung carcinoma is to the hilum and mediastinum,
familiarity with the mediastinal silhouette is an
important diagnostic concern (Fig. 5-7).
The early roentgenographic features are, unfor-
tunately, subtle in nature and often are appreciated
only in retrospect.
44
The earliest signs visible in
the roentgenogram of the chest are locally pro-
duced by the tumor itself. These signs may include
any abnormal density within the lung parenchyma
(most common), lobulation and cavitation of a
mass (most specific for lung carcinoma),
45
seg-
mental atelectasis, a simple cavitation within the
lung, and a mediastinal mass (uncommon). A
newly appreciated pulmonary density must be
compared with old films to establish how long it
Chest Roentgenographic Findings Due to Lung Carcinoma
Hilar and
Mediastinal Masses
Parenchymal Mass
Distal Atelectasis
and Infection
rect Extension
\f_^y ( Pathology
Figure 5-6 In patients with lung carcinoma the chest roentgenographic findings result from the presence of the tumor itself
within the lung (parenchymal mass), changes in the pulmonary parenchyma distal to a bronchus obstructed by the tumor (atelectasis
and infection), and spread of the tumor to extrapulmonary intrathoracic sites (hilar and mediastinal masses and other direct extension
pathology).
Mediastinal Silhouette and Major Mediastinal Lines and Interfaces
Figure 5-7 The mediastinal sil-
houette is determined by the borders
of the great vessels (aorta, great
veins, pulmonary artery) and the
heart.
has been present. Malignant pulmonary lesions
usually have doubling times of less than 1 year. A
lesion that remains the same size for at least 2
years can be presumed to be benign (see Table 5-
5).
46
The usual roentgenographic manifestations of
lung carcinoma more frequently include the hilar
Table 5-5 RADIOLOGIC CRITERIA FOR
DIFFERENTIATING MALIGNANT
FROM BENIGN PULMONARY
OPACITIES*
I. More Likely Malignantf
1. Opacity larger than 3 cm in diameter
2. Spicular margins
3. Noncalcified
4. Increasing size (doubling time 30-490 days)
II. More Likely Benignf
1. Stable size for 2 yearst or doubling time less than 30
days (probably infectious) or more than 490 days
(probably benign)
2. Benign pattern of calcification
3. Well-circumscribed margins
4. Small (< 2 cm) size
5. Nearby satellite lesions
6. Cavitated with thin walls or with air-fluid level
III. Indeterminate or Noncontributory Factors
1. Age of lesion is unknown (no prior radiographs)
2. Noncalcified or eccentric calcification
3. 2-3 cm in size, with smooth margins
*From Batra P, Brown K, Aberle DR, et al: Imaging tech-
niques in the evaluation of pulmonary neoplasms. Chest
101:239-244, 1992. Used with permission.
tThe criteria in each section are "additive" (e.g., the pres-
ence of two or three criteria has greater impact than one alone).
Major criteria of benignancy.
and extrapulmonary intrathoracic manifestations in
addition to the pulmonary parenchymal manifes-
tations (see Table 5-5).
46
In a review of the chest
roentgenograms of 600 patients with carcinoma of
the lung,
35
the average lung cancer mass at radio-
logic presentation was 3 to 4 cm in diameter. A
larger parenchymal mass was present in 22 per
cent and a smaller mass in 20 per cent; multiple
masses were present in only 1 per cent. Obstruc-
tive pneumonitis, collapse, or consolidation was
present in 41 per cent. A hilar abnormality, either
alone or associated with other abnormalities, was
present in 41 per cent of the patients. The various
extrapulmonary intrathoracic manifestations, with
mediastinal widening, pleural effusion, and raised
hemidiaphragm being the most common of these,
were present in 11 per cent.
Certain roentgenographic patterns are character-
istic of the various cell types.
47
Calcification is the
most reliable sign of benign disease, especially if
the lesion is a peripheral nodule. Since squamous
cell carcinoma has a propensity to obstruct a bron-
chus, squamous cell carcinoma most often presents
the picture of obstructive pneumonitis, collapse,
consolidation, or cavitation. A hilar abnormality is
also usually present with a squamous cell carci-
noma. Adenocarcinomas are most often peripheral
masses, and two thirds of these are larger than 4
cm. Hilar abnormalities, obstructive parenchymal
lesions (atelectasis, abscess), and cavitation are in-
frequent to rare. Large-cell undifferentiated carci-
nomas are most likely to be peripheral lesions and
are larger than 4 cm. Cavitation is infrequent, and
Preoperative Cardiopulmonary Evaluation
hilar abnormalities and parenchymal changes are
present in approximately 30 per cent of cases.
Small-cell undifferentiated tumors appear primar-
ily as hilar abnormalities (80 per cent), and a par-
enchymal obstructive lesion occurs in approxi-
mately 40 per cent.
b. BRONCHOSCOPY
The examination of the tracheobronchial tree
with either flexible fiberoptic (by far most com-
mon) or rigid bronchoscope should be done in
almost all patients suspected of having a tumor of
the lung. Indeed, 44 per cent of lesions that com-
pletely obstruct a bronchus (segmental to lobar),
that are observed with a flexible fiberoptic bron-
choscope, will have no radiographic sign of the
obstruction.
36
An exception to performing bron-
choscopy may be made in patients with a very
small peripheral lesion with no evidence of hilar
or mediastinal lymph adenopathy. Direct visuali-
zation of the tumor, positive biopsy findings, pos-
itive bronchial brushing or trap suction specimens,
or some combination of these three bronchoscopic
findings is obtained in a high percentage of the
patients. The technique of transbronchial needle
aspiration has clearly made mediastinal and hilar
nodes accessible to biopsy, thereby obviating the
need for further invasive and costly surgical stag-
ing (e.g., mediastinoscopy) in some cases.
37
Trans-
bronchial needle aspiration of peripheral masses
and subtle endobronchial lesions has greatly in-
creased the diagnostic yield of bronchoscopy.
37
Cell type influences the rate of positive finding;
small-cell tumors are identified proportionately
more often than are squamous cell or large-cell,
undifferentiated tumors, and adenocarcinomas are
identified least frequently of all.
In addition to actual assessment of the tumor,
other valuable information may be obtained at
bronchoscopy. For example, the length of normal
bronchus proximal to the tumor and the status of
the carina (subcarinal nodes) may be determined,
both of which are determinants of the exact sur-
gical procedure to be performed (lobectomy, pneu-
monectomy, sleeve resection, or inoperable). In
addition, bronchoscopy may reveal the presence of
a second central tumor that was not visible on
chest X-ray.
C. SPUTUM CYTOLOGY
Cytologic examination of sputum has been
found to be positive in approximately half of pa-
tients suspected of having carcinoma of the lung.
With appropriate cytologic study of several spu-
tum specimens, tumor cells may be found in a
higher percentage of patients. In one study, one or
two sputum samples yielded a 59 per cent positive
result, three sputa a 69 per cent, and four sputa a
90 per cent positive result.
50
A false-positive inci-
dence of only 1 per cent was found, although in
most laboratories this incidence is reported to be
in the range of 2 to 3 per cent. Cell type as deter-
mined by cytologic study agrees with that of the
final histologic diagnosis in approximately 85 per
cent of the patients. Well-differentiated epider-
moid carcinomas, undifferentiated small-cell car-
cinomas, and adenocarcinomas can all be effec-
tively typed by cytology. The undifferentiated
large-cell carcinomas, the poorly differentiated ep-
idermoid carcinomas, and combined carcinomas
are more difficult to type correctly. Cytologic stud-
ies are most often positive in patients with large
central tumors that communicate with the main
bronchi (e.g., squamous cell). Peripheral parenchy-
mal lesions frequently do not communicate with a
large bronchus, and cytologic studies in patients
with such lesions are less rewarding (e.g., adeno
cell).
d. NEEDLE BIOPSY
Percutaneous, transthoracic needle biopsy (with
either fluoroscopic or computed tomographic [CT]
guidance) has been suggested as a routine proce-
dure for indeterminate (and noncommunicating)
solitary peripheral lesions.
48-50
With these lesions,
this procedure is very useful when sputum and
bronchoscopic methods fail to establish a defini-
tive histologic diagnosis.
51
The procedure can be
performed on central lung, mediastinal, and pleural
lesions and abscesses as well. Although percuta-
neous fine-needle aspiration is very accurate when
positive (97 per cent), the false-negative rate may
be as high as 30 per cent.
52
53
Usually fine needles (18-23 gauge) are used but
with CT scan guidance, larger needles (providing
more accurate diagnosis) can be used.
54
Percuta-
neous needle biopsy causes a significant incidence
(7-14 per cent) of a small pneumothorax, which
unfortunately requires chest tube drainage in 25 to
50 per cent of the patients with this complica-
tion.
51
55
56
The anesthesiologist must be aware of the his-
tory of a recent needle biopsy because of the pos-
sibility of the development of a tension pneumo-
thorax with the commencement of positive-
pressure ventilation. Finally, when all else fails to
establish a diagnosis, open-lung biopsy may rarely
be necessary.
57
e. SUMMARY OF DIAGNOSIS OF LUNG
CARCINOMA
The diagnosis of lung carcinoma is almost al-
ways certain, but in 20 to 25 per cent of patients a
histologic cell type diagnosis is not known preop-
erative^. For these patients, the findings at surgery
(which often begins with an open-lung biopsy)
complete the first step in the workup.
E. Staging of Lung Cancer (Has the
Carcinoma Spread?)
Simply making the diagnosis of lung cancer
alone is a grossly inadequate preoperative evalua-
tion. In order to devise a rational approach to treat-
ment, it is absolutely essential to determine the
nature and extent of the disease in terms of any
direct extension of the tumor to adjacent struc-
tures, metastasis of the tumor to the thoracic
lymph node system, and extension of the tumor to
extrathoracic structures. Any of these three forms
of tumor extension may render the tumor inopera-
ble, and the preoperative diagnosis of such exten-
sion will greatly decrease the incidence of unnec-
essary thoracotomy and surgical morbidity and
mortality. The need for precise classification of the
anatomic extent or stage of lung cancers led to the
application of the size of the tumor-nodal involve-
ment-metastasis staging system'
2,58-66
to the dis-
ease; the international staging system for lung can-
cer
64
provides a reference standard that has
consistent meaning worldwide (Table 5-6).
Staging is the quantitative assessment of malig-
nant disease and allows logical grouping of pa-
tients with a similar extent of disease of prog-
nostic, therapeutic, and analytic purposes. In
bronchogenic carcinoma, a stage is assigned based
on size, location, and the extent of invasion of the
primary tumor as well as the presence of any re-
gional or metastatic disease. Selecting the most
appropriate treatment for a patient with broncho-
genic carcinoma depends on precise staging.
In brief, the (for tumor) classification de-
Table 5-6 NEW INTERNATIONAL NON-SMALL-CELL LUNG CARCINOMA PRIMARY
SIZE OF TUMOR (T)-NODAL INVOLVEMENT (N)- DISTANT METASTASIS
(M) STAGING SYSTEM
Factor Classification Description
Primary tumor size (T) Tumors proven by the presence of malignant cells in bronchopulmonary
secretions but not visualized roentgenographically or bronchoscopically, or
any tumor that cannot be assessed, as in a retreatment staging
No evidence of primary tumor
Carcinoma in situ
A tumor that is 3.0 cm or less in greatest dimension, surrounded by lung or
visceral pleura, and without evidence of invasion proximal to a lobar bronchus
at bronchoscopy
A tumor more than 3.0 cm in greatest dimension, or a tumor of any size that
either invades the visceral pleura or has associated atelectasis or obstructive
pneumonitis extending to the hilar region; at bronchoscopy, the proximal
extent of demonstrable tumor must be within a lobar bronchus or at least 2.0
cm distal to the carina; any associated atelectasis or obstructive pneumonitis
must involve less than an entire lung
A tumor of any size with direct extension into the chest wall (including superior
sulcus tumors), diaphragm, or the mediastinal pleura or pericardium without
involving the heart, great vessels, trachea, esophagus, or vertebral body, or a
tumor in the main bronchus within 2 cm of the carina without involving the
carina
A tumor of any size with invasion of the mediastinum or involving heart, great
vessels, trachea, esophagus, vertebral body, or carina, or presence of
malignant pleural effusion.
Nodal involvement (N) No demonstrable metastasis to regional lymph nodes
Metastasis to lymph nodes in the peribronchial or the ipsilateral hilar region, or
both, including direct extension
Metastasis to ipsilateral mediastinal lymph nodes and subcarinal lymph nodes
Metastasis to contralateral mediastinal lymph nodes, contralateral hilar lymph
nodes, ipsilateral or contralateral scalene, or supraclavicular lymph nodes
Distant metastasis (M) No (known) distant metastasis
Distant metastasis presentspecify site(s)
Stage grouping
scribes the size of the tumor and any direct exten-
sion of the tumor into surrounding tissues. T
0
in-
dicates no evidence of a primary tumor; Tx
indicates malignant cytology, but tumor is not seen
roentgenographically or bronchoscopically. The
remaining descriptors for the primary tumor sub-
divide the category into four levels. T, tumors are
less than 3 cm in diameter and are completely
surrounded by lung tissue. In contrast, T
2
tumors
are greater than 3 cm or involve the visceral pleura
or cause lobar atelectasis. A tumor of any size that
extends into the chest wall, diaphragm, or medias-
tinal pleura, which grows to within 2 cm of the
carina or cause atelectasis of an entire lung, is
labeled T
3
. A new class, T
4
tumors, includes those
involving structures that are typically unresectable
(heart, great vessels, trachea, esophagus) or are
associated with malignant pleural effusions.
Regional lymph nodes (N) include N, nodes,
those totally enclosed by visceral pleura, or N
2
nodes, located in the ipsilateral mediastinum or
subcarinal region. A new category, N-,, is assigned
to nodes in the contralateral mediastinum, contra-
lateral hilum, or either supraclavicular or scalene
region.
The presence or absence of distant metastases
(M) is normally specified according to site.
The various T, N, and M categories are organ-
ized into various stage groupings (Table 5-6 and
Figures 5-8 through 5-11). Stage I (Fig. 5-)
64
includes patients with T, or T
2
tumors without
evidence for nodal (N
0
) or distant metastatic
spread (M
0
). Stage II cancers (Fig. 5-9)
M
are sim-
ilar to those in stage I except that N, nodes (peri-
bronchial, lobar, hilar nodes) are involved. Stage
III has been subdivided into stage Ilia (Fig. 5-
10)
64
patients (=s T
3
, ^ N
2
), who are usually can-
didates for definitive surgical treatment and stage
IHb (Fig. 5-11 )
M
patients (T
4
or N
3
), who are
normally not candidates for operative intervention.
Stage IV includes patients with distant metastatic
disease (M,).
With this new, revised TNM staging system,
both the designation of stage (I, II, Ilia, IHb, IV)
and the T, N, and M status are all significant indi-
cators of prognosis. Each successive T, N, and M
descriptor is associated with a worse prognosis,
and additionally each successive stage also carries
a poorer prognosis.
65
66
The breakdown of stages into three groups
serves three purposes. First, the stage of the dis-
ease closely correlates with survival rate (see Prog-
nosis and Survival as a Function of the Staging of
Lung Cancer) except for small-cell carcinoma (see
section II.E.5. Staging in Small-Cell Lung Cancer,
and section H.E.7., Prognosis and Survival as a
Function of the Staging of Lung Cancer). Small-
cell carcinoma is thought to have metastatic spread
by the time of diagnosis; therefore, its natural his-
tory and behavior are independent of (and much
more lethal than) the TNM system predictions.
The diagnosis of small-cell carcinoma should be
T4 N3 MO
Involvement of mediastinum,
(ipsilateral and) contralateral
mediastinal lymph nodes,
contralateral hilar nodes,
supraclavicular lymph
nodes
Figure 5-11 Stage Illb disease. (From Moutain CF: A new international staging system for lung cancer. Chest 89:225S, 1986.
Used with permission.)
made from the bronchoscopic procedures (brush-
ings, biopsy) or sputum cytology or needle biopsy
(see preceding discussion). Second, the stage of
the disease (I and II, approximately 30 to 35 per
cent of all patients) usually dictates the surgical
procedure of choice (see section II.E.6., Surgical
Procedures as Dictated by Staging). Third, the di-
agnosis of stage III (approximately 60 to 65 per
cent of all patients) may preclude surgery; there-
fore, staging tests are most vitally concerned with
determining T
3
, N
2
, or M, disease. However, it
should be noted that outstanding 5-year survival
results (30 per cent) have been obtained with some
stage III subsets (T
3
, N
0
-N
2
, M
0
).
67
In summary, the introduction of the TNM stag-
ing system has encouraged an orderly assessment
for selecting those cases most suitable for surgery
and the type of surgery that should be performed.
As a corollary to this appropriate staging should
reduce the incidence of unnecessary thoracotomy
to less than 20 per cent.
68
69
As a consequence of
these improvements, the staging should result in
an overall improvement in the present postsurgery
5- and 10-year survival rates of approximately 30
and 16 to 18 per cent, respectively. Although T,
N, and M staging are actually done in parallel, for
the sake of clarity they are discussed separately in
the following sections (Fig. 5-12).
1. Staging (Tumor Size and Direct
Extension)
staging is primarily done by CT scanning and
bronchoscopy. The advent of CT scans of the
thorax (lung parenchyma, pleura, and medias-
tinum) has had a profound and simplifying effect
on the staging of lung cancer (see Fig. 5-12).
CT scans of the thorax have been shown to pro-
vide a clear delineation of the tumor mass and can
suggest direct tumor extension (particularly direct
spread to the pleura with or without an accompa-
nying effusion and direct spread to the mediastinal
structures) in many patients in whom the more
conventional diagnostic radiologic methods fail to
do so. For example, CT scanning in non-small-cell
cancer when compared with conventional chest ra-
diology, tomography, and bronchoscopy increases
the stage in 40 per cent of cases
70
and in small-
cell cancer, it increases the stage from I or II to
III in anywhere from 30 to 84 per cent of cases.
71
72
These stage changes were due to direct tumor
extension into the mediastinum, pleura, or dia-
phragm. Visualization of the mediastinum and hi-
lar structures by CT scanning may be markedly
enhanced by intravenous contrast; this permits in-
tense opacification of the vessels and heart during
the scan, allowing improved visualization of non-
vessel structures.
72
Figure 5-12 Preoperative evaluation logic of step II to determine whether lung carcinoma has spread beyond its local confines.
Definition of T, N, and M staging and the various T, N, and M subsets are described in Table 5-6. The dashed lines indicate that
surgery has recently been performed with an encouraging degree of success.
76
78
79, M S5
Step III is physiologic assessment of the
patients for the planned surgical procedure. (CT = computed tomography.)
The results of and extra information gained
from staging with CT have to be interpreted
carefully with regard to pulmonary nodules. CT
detects 50 per cent more pulmonary nodules than
whole-lung tomograms.
73
Most of these nodules
are small, less than 6 mm, and serial follow-up
shows that the majority (60 per cent) are benign.
Thus, an additional intrapulmonary nodule (other
than the primary) seen only with CT does not
necessarily represent an intrapulmonary metastasis
and is not necessarily a contraindication to sur-
gery.
The proximal extent of a tumor in the airway is
determined by flexible or rigid bronchoscopy.
Bronchoscopy should be carried out to assess the
lesion fully, to exclude synchronous tumors, and
to plan the operative approach. Tumors within a
major lobar or main bronchus but greater than 2
cm from the carina are classified as T
2
. Tumors
less than 2 cm from the carina but not involving
the carina are T
3
lesions. Tumors that invade the
trachea or carina are T
4
tumors. The geographic
distribution of lung carcinoma is shown in Table
5-7.
40
Several other roentgenographic studies may
demonstrate direct extension involvement of tho-
racic structures, which may preclude resection.
Cardiac angiocardiography may demonstrate
main-stem pulmonary artery invasion (at least l to
2 cm of a main-stem pulmonary artery that is free
of tumor is required for pneumonectomy). Pulmo-
nary artery angiography may be necessary to dem-
onstrate pulmonary arteriovenous malformations
and aortography to demonstrate pulmonary se-
questrations. A barium swallow may demonstrate
fixation and/or distortion of the esophagus. Bron-
chography may be necessary for defining bron-
chopleural and bronchoesophageal fistulas. Azy-
gography may reveal mediastinal lymph node
involvement as well as involvement of the vena
cava by tumor.
In a patient with suspected or diagnosed lung
Table 5-7 LOCATION OF LUNG CARCINOMA*
Side of Lung Main Bronchus (%) Upper Lobe (%) Middle Lobe (%) Lower Lobe (%)
Right lung 4 33(19.4) 5(8.3) 14(25.2)
Left lung 4 27(22.4) 12(24.7)
Both lungs 2
*From Humphrey WS, Smart CR, Winchester DP, et al: National Survey of the pattern of care for carcinoma of the lung. J
Thorac Cardiovasc Surg 100:837-843, 1990. Used with permission.
The data are given as a percentage of the total number of patients. The numbers in parentheses are the tissue mass of the given
lobe as a percentage of the whole-lung mass.
cancer, a pleural effusion may be evidence for
extension of disease beyond the primary tumor.
Thoracentesis with cytologic analysis of the fluid
provides specific staging and in some cases diag-
nostic information. If cytologic analysis of the
fluid obtained by thoracentesis is not diagnostic,
thoracoscopy should be considered. This can facil-
itate collection of greater amounts of pleural fluid
for cytologic evaluation and allows a directed bi-
opsy of pleural based masses and peripherally lo-
cated primary tumors.
5
2. Staging (Metastasis to Lymph
Nodes)
The goal of preoperative evaluation of the me-
dastinum is to identify those patients with N
3
dis-
ease, exclude them from surgical consideration,
and select those patients with operable N
2
or less
disease. N, nodes are segmental, lobar, interlobar,
and hilar. N
2
nodes are inferior mediastinal, aortic,
and superior mediastinal. N-, nodes are any contra-
lateral nodes and ipsilateral scalene or supraclavic-
ular lymph nodes.
74
The final precise enumeration
of nodal involvement often requires staging at the
time of operation (see section II.E.4., Intraopera-
tive and Postsurgical Staging).
The preoperative assignment of classification
involves the use of CT of the thorax (lung paren-
chyma, hilum, and mediastinum) and mediastinos-
copy (Fig. 5-12). As with staging, use of CT
has also had a profound and simplifying effect on
tumor staging. Although paratracheal and hilar
lymph node adenopathy can usually be adequately
identified by conventional radiology, CT scans are
clearly superior in detecting subcarinal node en-
largement. Lymph nodes larger than 15 mm
should be regarded as CT positive (but do not
necessarily contraindicate surgery).
72
CT com-
pared with conventional radiology causes an up-
staging of status in 32 per cent of cases of lung
carcinoma.
70
The results of and extra information gained
from CT scanning for nodes also have to be inter-
preted carefully. The demonstration of enlarged
mediastinal glands should not automatically lead
to the conclusion that they are infiltrated by tumor.
Glands are frequently enlarged and fleshy and may
be subsequently found to be free of disease (e.g.,
centrally located squamous cell carcinoma can
cause postobstructive pneumonia with reactive re-
gional lymph node involvement), and the false-
positive rate of CT mediastinal lymphadenopathy
is about 25 per cent of cases.
75
Consequently, a
positive CT scan result still requires mediastinos-
copy. However, because CT tends to "overstage"
tumors, thoracic surgical units without CT scan-
ners will achieve equally good staging results
(and perhaps better)
76
with routine preoperative
mediastinal exploration. On the other hand, the
predictive value of a negative CT scan is on the
order of 90 to 95 per cent, and in such cases
mediastinoscopy (see the following) is omitted by
some surgeons before thoracotomy.
75
However, it
should be noted that, for the remaining 5 to 10 per
cent, a normal-sized mediastinal gland does not
mean that it is tumor free, and the results of studies
comparing CT findings in the mediastinum with
histologic evaluation show that there is a definite
small false-negative rate.
75
77
Indeed, two authors
found a significant rate of metastatic disease in
nodes that were small on CT scan (< 1 cm) at
surgery (i.e., the nodes were completely replaced
by tumor but yet of normal size), especially with
adenocarcinoma.
76, 78
79
It is clear from this discussion that staging
with CT involves a moderate false-positive rate
and a small false-negative rate. Many recent stud-
ies (N = 18; these studies are conveniently
summarized in Daly et al.'s
80
Table 3) carefully
quantitated the sensitivity, specificity, accuracy,
positive predictive index, and negative predictive
index (see legend of Table 5-8 for the definition
of these indices) of CT scanning. Table 5-8 shows
the accuracy of each staging modality from one
typical study and concludes (as do most studies)
that mediastinoscopy is the most accurate staging
investigation and recommends its continued rou-
tine use.
81
Table 5-9 shows one author's actual
guidelines for the use of mediastinoscopy (see
chapter 14).
82
In recent years, magnetic resonance imaging
Table 5-8 ACCURACY OF EACH STAGING MODALITY'
Statistical Parameter! MR CT CXR Mediastinoscopy
Sensitivity
Specificity
Positive predictive value
Megative predictive value
accuracy
71.0 71.0 80.7
90.6 88.7 43.4
81.5 78.6 45.5
84.2 83.9 79.3
83.3 82.1 57.1
87.1
100.0
100.0
93.0
95.2
*From Patterson GA, Ginsberg RJ. Poon PY, et al: A prospective evaluation of magnetic resonance imaging, computed
tomography, and mediastinoscopy in the preoperative assessment of mediastinal node status in bronchogenic carcinoma. J Thorac
Cardiovasc Surg 94:679-684, 1987. Used with permission.
tThe sensitivity, specificity, accuracy, positive predictive value, and negative predictive value were calculated by the following
formulas:
Sensitivity
Specificity =
Accuracy
Number of true positives
Number of true positives +
Number of false negatives
Number of true negatives
Number of true negatives +
Number of false positives
Total number of instances
X 100
X 100
X 100
Positive
predictive value =
Negative
predictive value
Number true positives
Number of false positives
Number of true negatives +
Number of false negatives
x 100
X 100
Abbreviations: CXR = chest roentgenography; MR = magnetic resonance; CT = computed tomography.
(MRI) has been compared with CT scanning in its
usefulness for staging lung cancer patients. MRI
has advantages in that it is able to distinguish
vascular from solid structures without the use of
contrast materials (which CT scanning requires).
MRI is still in its infancy from a technologic
standpoint, requiring a long scanning time that re-
sults in motion degradation of images and poor
spatial resolution. For these reasons, comparative
studies between CT and MRI show no difference
between the two or a slight to moderate advantage
for CT in the assessment of both primary cancer
and nodal involvement.
5
81
-
83
The present role of
MRI should be confined to resolving specific ques-
Table 5-9 ACTUAL GUIDELINES FOR THE
USE OF MEDIASTINOSCOPY*
So mediastinoscopy
Peripheral squamous cell carcinoma
Undiagnosed peripheral nodule < 3 cm
Cervical mediastinoscopy
Right-sided centrally located squamous cell carcinoma
Every undifferentiated carcinoma and adenocarcinoma
(except for left upper lobe and left central tumors)
\nterior mediastinoscopy
Left upper lobe tumors
Left-sided centrally located tumors (when anterior
mediastinoscopy is negative, a cervical approach should
be performed)
*From Van Schil PEY, Van Hee RH GG, Schoots ELG: The
ralue of mediastinoscopy in preoperative staging of broncho-
genic carcinoma. J Thorac Cardiovasc Surg 97:240-244, 1989.
Jsed with permission.
tions in patients with possible cord involvement by
the primary tumor or in whom the use of iodinated
contrast material is contraindicated.
5
Mediastinoscopy or mediastinotomy (anterior
resection of a costosternal cartilage) should be car-
ried out as the final procedure to assess staging
when CT results are positive or when CT is not
available.
84
85
Mediastinoscopy or mediastinotomy
should be performed even if the mediastinum
looks normal with conventional radiologic tech-
niques, since almost 50 per cent of patients with
mediastinal nodal involvement do not have me-
diastinal widening on chest X-ray.
86
Mediastinot-
omy is particularly important for left upper lobe
lesions because it can provide better access to the
left anterior mediastinal lymph nodes. If glands
resected at mediastinoscopy or mediastinotomy are
found to contain tumor (N
2
), most would classify
the case as inoperable.
3. M Staging (Metastasis)
M staging begins with a complete history and
physical examination that covers the function of
all organ systems, with particular emphasis on
brain, bone, and liver (Fig. 5-12). If the history or
physical examination is positive for any given or-
gan system, then that particular organ system
should be further investigated for metastases. The
investigation should first involve an organ scan of
some type (radioisotopic, CT). Organ scanning
should not be done in the absence of symptoms;
5. Staging in Small-Cell Lung Cancer
The very high incidence of mediastinal lymph
node disease in SCLC and coexisting evidence of
extrathoracic dissemination at the time of presen-
tation mean that the T, N, and M classification has
less bearing on SCLC prognosis. In fact, the situ-
ation at one time was thought to be so dismal that
patients were classified only into limited (involve-
ment of only one hemithorax) versus extensive
disease, surgical resection had no place in the
treatment in these patients, and chemotherapy was
considered the main treatment modality.
95
96
How-
ever, concern arose that these understandings and
dictums needed revision when it was noted that
some patients with ' ' limited" disease had an ap-
parently complete response to chemotherapy and/
or radiation (i.e., SCLC is often initially exqui-
sitely responsive to chemotherapy.
97
-
|0
These re-
ports suggested that surgery might be of value as
a component of combined-modality therapy, under
carefully defined conditions, and, indeed, subse-
quent studies have shown improved survival in
stage I and II disease (for both stages the reported
5-year range is as great (and as good) as 10 to 70
per cent; see also Table 5-11).
1()|
-" The dominant
opinion in 1992 appears to be that the treatment of
choice for stage I and II SCLC is curative surgical
resection combined with pre- and postoperative
chemotherapy and radiation treatment. Thus, TNM
staging of clinically localized small-cell carcinoma
is necessary to define the lesions for which sur-
gical resection might be of benefit. Nevertheless,
one must still keep in mind that, of all patients
seen with limited SCLC, only about 10 per cent
will have sufficiently early disease to warrant con-
sideration for surgical treatment.
1
"
6. Surgical Procedure as Dictated by
Staging
All agree that, in the absence of contraindica-
tions, surgical excision of lung carcinoma offers
the best likelihood of survival and quality of life
and that the appropriate surgical procedure is one
that excises all the carcinoma and yet, at the same
time, preserves as much normal lung as possible."
2
In this regard, the neodymium:yttrium-aluminum-
garnet laser has been shown to be an excellent tool
for removing parenchymal lesions while sparing
surrounding normal lung tissue."
3
It is possible
that socioeconomic factors may impact on the ex-
act choice of lung cancer treatment."
4,
"
5
Segmental resection is indicated for small (< 3
cm in diameter) stage I (T,, N
0
, M
0
) peripheral
lesions and for patients with severely compro-
mised cardiopulmonary status. Relative contrain-
dications to segmental resection are T
2
lesions or a
Lobectomy Pneumonectomy
or Bronchoplastic Procedure
Figure 5-13 Algorithm of operating-room decision for stage I and stage II disease based on the size of the primary tumor, its
location, and the involvement of the hilar nodes. (From DeMeester TR, Albertucci M: Stage I non-small cell lung cancer: Surgical
therapy. In Bitran JD, Golomb HM, Little AG. et al (eds): Lung Cancer: A Comprehensive Treatise. Orlando, Grune & Stratton.
1988, I4l. Used with permission.)
lesion that crosses an intersegmental plane. Mor-
tality for segmental resection is less than 0.5 per
cent."
6
At present, lobectomy is the operation of choice
for carcinoma of the lung. It is indicated when
technically feasible for all stage I and II disease
and some stage III lesions. Lobectomy is not indi-
cated when the tumor involves structures in the
pulmonary hilum, crosses an interlobar fissure, or
involves the main bronchus. Lobectomy is associ-
ated with a mortality rate of l to 3 per cent."
6
""*
Indications for bilobectomy are tumor extending
across the fissure, absent fissure, endobronchial tu-
mor, extrinsic tumor or nodal invasion of the bron-
chus intermedius, and vascular invasion."
9
The
mortality rate associated with bilobectomy is 4 per
cent."
9
Pneumonectomy is indicated with more exten-
sive disease, generally T
2
, N, M
()
or T
;
, N,, M
0
with hilar involvement. Figure 5-13 shows an ex-
ample of how the intraoperative decision/choice
between pneumonectomy versus lobectomy is
made based on intraoperative findings. The mor-
tality for pneumonectomy is about 5 per cent."
6
""
8
T
3
lesions (stage ) will require additional re-
section of involved structures (chest wall, dia-
phragm, pericardium, etc.). N
2
lesions (ipsilateral
mediastinal nodes, stage ) may benefit from
pneumonectomy combined with neoadjuvant ra-
diation and/or chemotherapy.
120-122
With few exceptions, patients with stage IIIB
disease are not candidates for definitive surgical
resection.
67
The exceptions may include certain pa-
tients with central lesions and only hilar node in-
volvement, Pancoast's tumors, or certain lesions
with chest wall involvement. Specially selected
patients with limited carinal involvement (T
4
) have
undergone resection for cure by sleeve pneumo-
nectomy or carinal resection (see Fig. 5-12).
84
'
85
In the rare case of resectable solitary brain metas-
tasis, M, disease is potentially surgically curable
(see Fig. 5_i2).
7 K
7 9
^
1 2 4
Patients who are not candidates for surgical re-
section may receive radiation therapy, chemother-
apy, and/or immunotherapy. Unfortunately, none
of these latter therapeutic modalities have signifi-
cantly or dramatically impacted on survival rates
for bronchogenic carcinoma. The use of chemo-
therapy for non-SCLC is toxic and at present very
controversial.
125-129
Surgery, when it can be per-
formed, is still a patient's best and only hope for
cure of bronchogenic carcinoma.
1 7 4 Preoperative Cardiopulmonc Evaluation
I g 135.136 Approximately 65 per cent of the patients
would be considered inoperable at presentation be-
cause of intrathoracic spread, the detection of ex-
trathoracic metastases, or extremely poor lung
function. The remaining 35 per cent would
undergo a thoracotomy, although the figure may
be nearer 20 to 30 per cent if mediastinoscopy was
routinely performed prior to definitive surgery. At
thoracotomy, 15 per cent of the original 100 cases
would be found to be inoperable because of pre-
viously unrecognized tumor spread. The remaining
20 per cent would undergo "curative" resection.
The 5- and 10-year survival rates for all persons
undergoing resection would be 40 per cent and 16
to 18 per cent, respectively (eight and three to four
patients, respectively, of the original 100 patients).
"Curative" resection may fail so often in the
lower stage because of the inability to detect small
asymptomatic metastases before operation.
o 61.4%
* - .
37.4%
25.8%
8.4%
48 60

Figure 5-14 Survival rates for
1479 patients in M0 group after re-
section of lung cancer, according to
classification (includes one patient
with TX and five with TO disease).
Differences between groups: Tl ver-
sus T2, < .01; T2 versus T3, <
.01; T3 versus T4, < .01. (From
Naruke T, Goya T, Tsuchiya R, et al:
Prognosis and survival in resected
lung carcinoma based on the new in-
ternational staging system. J Thorac
Cardiovasc Surg 96:440-447, 1988.
1479 patients in MO group after re-
section of lung cancer, according to
postoperative classification. Dif-
ferences between groups: NO versus
Nl, < .01; NI versus N2, < .01;
N2 versus N3, < .01. (From Na-
ruke T, Goya T, Tsuchiya R, et al:
1737 patients after resection of lung
cancer according to postoperative M
classification. Differences between
groups; M0 versus Ml. < .01.
(From Naruke T, Goya T, Tsuchiya
R, et al: Prognosis and survival in
resected lung carcinoma based on the
new international staging system. J
Thorac Cardiovasc Surg 96:440-447.
1988. Used with permission.)
1737 patients after resection of lung
cancer, classified by postoperative
stage (four patients in stage 0 in-
cluded). Differences between groups:
stage I versus stage , < .01 ; stage
II versus stage IIIA, < .01; stage
III A versus stage IIIB, < .01; stage
IIIB versus stage IV, NS. (From Na-
ruke T, Goya T, Tsuchiya R, et al:
tors being equal, e.g., cell type); possibly hor-
monal factor(s) may contribute to the prognosis.
137
Although most studies have found that age greater
than 80 years does not impose a higher risk of
complications after lung resection for carci-
noma,
138-140
two studies found that age greater than
60 years does.
141
142
Preoperative protein depletion
has been found to be as important as site of inci-
sion as a risk factor for postoperative pneumo-
nia.
143
The available data suggest that postopera-
tive atelectasis is much more common in obese
patients, but the risk of death is, in general, not
significantly increased.
144
Aside from smoking/stopping smoking, there
are three other aspects of poor lung function that
can be easily diagnosed by history and/or physical
Table 5-10 POSTOPERATIVE 5-YEAR
SURVIVAL IN PATIENTS WITH
NON-SMALL-CELL LUNG CANCER
ACCORDING TO THE NEW (1986)
INTERNATIONAL STAGING
SYSTEM
examination and can be treated or improved pre-
operatively; therefore, the patient should always
be questioned about these areas. If the patient re-
ports a significant departure from his or her usual
state of health with regard to any of these three
aspects, then a preoperative pulmonary preparation
regimen should be instituted (see chapter 6). First,
all patients should be questioned about broncho-
spasm. Most patients who have bronchospasm (or
wheezing) are very aware of varying degrees of
airway resistance and chest tightness and of which
medications relieve these symptoms. Auscultation
of the chest for bronchospasm should also always
be performed. If the amount of bronchospasm as
reported by the patient, or heard by auscultation,
is more than usual, then therapeutic levels of bron-
chodilator drugs should be administered (see chap-
ter 6). Second, all patients should be questioned
about the amount of their secretions. If secretions
are much more copious than usual, a few days of
secretion removal (see chapter 6) may be of great
benefit. Third, if the color of the sputum has re-
cently changed from mucoid to yellow or green, a
sputum culture and sensitivity test should be per-
formed and appropriate antibiotic therapy insti-
tuted.
2. Pulmonary Function Tests
If only studies that permit determination of pre-
operative and postoperative probabilities of mor-
bidity, mortality, sensitivity, and specificity are
Table 5-11 SURVIVAL RATES ACCORDING TO HISTOLOGIC TYPE AND STAGE
OF DISEASES
*From Naruke T, Goya T, Tsuchiya R, et al: Prognosis and survival in resected lung carcinoma based on the new international
staging system. J Thorac Cardiovasc Surg 96:440-447. 1988. Used with permission.
tSixty patients with adenosquamous carcinoma and 27 patients with unclassified carcinoma are excluded.
considered, then it can be concluded that preoper-
ative pulmonary function testing (spirometer) has
a measurable benefit in predicting outcome in lung
resection candidates.
145
In selected patients, split
perfusion lung scanning and pulmonary exercise
testing are also useful.
145
Therefore, it is thought
that pulmonary function testing is more sensitive
than history or physical examination for detecting
preoperative lung disease and the post-test proba-
bility of pulmonary complications and that patients
with abnormal studies would benefit from preop-
erative respiratory therapy designed to decrease
Fate of 100 "Typical" Patients with Non-Small
Cell Lung Carcinoma
Figure 5-18 The fate of 100 "typical" patients with non-small cell lung carcinoma.
their risk for complication. In addition, in recent
years, an effort has been made to use preoperative
pulmonary function testing to predict which pa-
tients will develop intraoperative hypoxemia dur-
ing one-lung ventilation (see chapter 11). How-
ever, it should be noted that it is only common
sense to realize that not all the risk of thoracotomy
for lung cancer is predictable and that some of it
is of a random nature.
140
In this section, the various
individual whole-lung and regional lung function
tests are described first, and then the sequence in
which these tests should be performed is detailed.
a. WHOLE-LUNG FUNCTION TESTS
All patients undergoing resectional thoracic sur-
gery should have a timed expiratory spirogram
(spirometry), a maximum breathing capacity, and
an arterial blood gas analysis. Pulse oximetry may
prove to be a useful preoperative screening test.
146
Increasing consideration is being given to the rou-
tine performance of a flow-volume loop and some
form of exercise tolerance test (see the following
appropriate sections). The minimal pulmonary
function criteria indicating increased risk to per-
forming various-sized pulmonary resectional sur-
gery are shown in Table 5-12.
30
l47
Values below
the ones listed in Table 5-12 are associated with a
greatly increased incidence of postoperative respi-
ratory complications.
30
I47
The preoperative values listed in Table 5-12
Table 5-12 MINIMAL PULMONARY FUNCTION TEST CRITERIA FOR VARIOUS-
SIZED PULMONARY RESECTIONS*
*Data are based on Miller,
30
Gass and Olsen,
147
Bechard and Wetstein,
14
* Wahi et al.,
144
Putman et al.,
1
-" Markos et al.,
151
Nakahara et al.,
153
Miller and Hatcher,
1
" Ferguson et al.,
154
Eugene et al.,
,5
Smith et al.,
156
and Batton et al.
1
"
Abbreviations: MBC = maximum breathing capacity; FVC = forced vital capacity; FEV, = forced expiratory volume in 1 sec;
FEV,
V
,__
7
= forced expiratory volume between 25% and 75% of FVC; Vo
:m
.
lx
= maximum oxygen consumption.
should be interpreted with three very important
cautions in mind. First, technical and/or iatrogenic
factors, such as bronchopleural fistula, overhydra-
tion, dehydration, bleeding, and transfusion, can
certainly modify the postoperative course. Second,
pulmonary function tests (spirometry) only test
one essential component of the gas exchange sys-
tem, namely, the ventilatory pump (lung mechan-
ics) and ignore the other two essential components,
namely, the transport system (the heart, vascula-
ture and blood), and the gas-exchange systems (the
V/Q matching in the lungs).
I4H
~
|S7
Third, the pre-
dicted postoperative pulmonary function values
(based on preoperative V/Q scanning) (see radio-
isotope regional lung functions later) are more pre-
dictive of postoperative function than preoperative
whole-lung function tests
14
*'
157
(see Fig. 5-23 to
5-25).
(1) SPIROMETRY. The simplest spirometry test
is measuring vital capacity. To measure vital ca-
pacity, the patient inspires maximally to total lung
capacity (TLC) and then exhales either slowly
(slow vital capacity) or forcefully (forced vital ca-
pacity, FVC) and completely, and the exhaled vol-
ume is recorded. With regard to the lung volumes
shown in Figure 3-24, inspiratory capacity, inspi-
ratory reserve volume, and expiratory reserve vol-
ume can all be measured directly from the vital
capacity spirogram.
The volume exhaled rapidly from one breath has
been quantified in many ways; the most commonly
used value is the volume of gas exhaled in the first
second and is usually expressed as a percentage of
the FVC (forced expired volume in 1 sec/FVC or
FEV,%) (see Fig. 5-19). Normally, an individual
can exhale 70 to 80 per cent of the VC in 1 sec;
the remainder may take an additional 2 sec (FEV
3
).
Patients with significant airway obstruction are
able to exhale much less volume in the first second
(the FEV,/FVC ratio is decreased), and they re-
quire a much longer time to deliver the entire
volume, whereas in patients with lung parenchy-
mal disease (i.e., restriction) the FEV,/FVC ratio
is normal (but the FVC is decreased).
The measurement of the maximum instanta-
neous "peak" expiratory flow rate (PEFR) is
measured with an automated device such as an
electronic flowmeter. As a patient forcefully ex-
hales, expiratory flow rate increases rapidly to the
PEFR (see Fig. 5-20, the flow-volume loop), and
then the flow rate gradually decreases in parallel
with the decrease in lung volume. The PEFR is
defined as the highest expiratory flow rate sus-
tained for at least 10 msec. Because flow rate at
any given lung volume or moment is equal to the
slope of the tangent line to the timed expiratory
spirogram curve (such as the one shown in Fig. 5-
19), and it may be difficult to accurately pick out
the steepest slope, the PEFR cannot be measured
easily or with precision by hand from a recorded
spirogram. However, a comparable, albeit slightly
lower value, the maximum expiratory flow rate
(MEFR), can be derived by taking the average rate
of flow (slope) over a 1-L segment of the early
portion of the spirogram trace (Fig. 5-19). Usu-
ally, the segment between 200 and 1200 ml of
expired volume is chosen; hence, the measurement
is often referred to as FEF
200
_,
200
the subscript des-
ignating the volume segment used.
FEV,, PEFR, and MEFR are easily measured
(electronically or with a dry-wedge spirometer) in-
dices of large airway obstruction (normally 80 per
cent of the airway resistance is in the large air-
ways) and correlate well with gross functional im-
pairment. Nevertheless, they are relatively insen-
sitive to increases in the frictional resistance in
small airways, which normally are responsible for
20 per cent of the airway resistance and therefore
could be very diseased, but the disease would not
be reflected in the PEFR, FEV,, or MEFR. The
small airways are increasingly important at low
lung volumes near the terminal portion of the FVC
maneuver. Rather, FEV,, PEFR, and MEFR rep-
resent measurements of flow occurring within the
first second of the FVC maneuver and at relatively
high lung volumes. These measurements are
highly dependent on patient effort (and are there-
fore variable) as well as on the resistance of the
0 1 2 3
TIME (seconds)
Figure 5-19 Spirogram of a patient taking two small tidal
volume breaths, then an inspiration to total lung capacity, and
then a forced exhaled breath to residual volume cycle. (MEFR
= maximum expiratory flow ratethe average rate of flow
[slope] over a l-L segment of the early portion of the expira-
tory spirogram trace. FEV,
0
= forced expired volume in l sec.
FVC = forced vital capacity, which equals 100 per cent of the
y-axis. MMEF = maximum midexpiratory flow rate, which is
measured on the timed expiratory spirogram between 25 and
75 per cent of the FVC curve.)
larger airways. The value of FEV,, PEFR, and
MEFR as screening tests is thus limited by the
extent to which disease involves the larger air-
ways. This helps to explain the relative lack of
sensitivity of these measurements in detecting
early mild (small airway) obstructive disease.
However, these effort-dependent large-airway pul-
monary function tests are the ones most responsive
to and capable of detecting bronchodilator drug
effects; they are, therefore, the most useful tests in
the assessment of the presence of bronchospasm
and the therapeutic benefit of these drugs. Indeed,
in view of recent demonstrations that cigarette
smoking causes an acute decrease in airway cali-
ber and that cigarette smokers have a much in-
creased airway responsiveness, pre- and postbron-
chodilator spirometry may be or should be more
frequently performed.
158
I59
The assessment of the
reversibility of increased airway resistance em-
ploys beta-adrenergic drugs and spirometry.
The maximum midexpiratory flow rate (MMEF)
is measured on the timed expiratory spirogram
during the middle half of the FVC curve between
25 and 75 per cent (see Fig. 5-19). However, it is
not truly the "maximum" flow the (PEFR) be-
cause the latter occurs at a point much closer to
the TLC. For this reason, use of "forced midexpir-
atory flow" has been advocated, with the more
precise symbol of FEV
25
_
75%
The MMEF, by con-
sidering only the middle segment of the FVC.
eliminates the initial highly effort-dependent and
more variable segment of the trace. The flow rate
during this segment has been shown to be largely
independent of patient effort and is slowed by ob-
struction of smaller airways. Because of this, the
MMEF is accepted as an indirect measure of
small-airway resistance and has been advocated as
a sensitive test for the early detection of small-
airway disease.
160
The fact that successful treat-
ment of left-sided cardiac failure results in marked
improvement of FEV
5(MK
FEV
73%
FEV
25%
_
75%
but
not in PEFR and FEV
25%
is strong support for the
effort-independent small-airway hypothesis.
161
(2) FLOW-VOLUME CURVES. The recording, dur-
ing spirometry, of the expiratory flow plotted
against volume, instead of time, produces a trian-
gle-shaped envelope (Fig. 5-20). The virtually in-
stantaneous peak is the peak expiratory flow
(PEF), and the gradually decreasing flow rates fol-
low the progressive airway narrowing down to
zero flow at residual volume (RV). An inspiratory
loop can also be obtained by asking the patient to
take a maximal breath in from RV back to TLC
(see Fig. 5-20). The semicircular shape of the nor-
mal inspiratory flow curve reflects the time taken
for the respiratory muscles to generate maximal
inspiratory force and flow; the maximal rate has
usually been achieved by the time 50 per cent of
the vital capacity (VC) has been inspired. See
chapter 15 (Fig. 15-1) for the effect of relief of
major obstruction of a main-stem bronchus on the
flow-volume curve.
(3) CARBON MONOXIDE DIFFUSING CAPACITY.
The magnitude of carbon monoxide gas transfer
(DL
co
) depends on the volume and flow of blood
through the pulmonary capillary bed (Q,) and the
matching between ventilation and perfusion (V/Q)
within the lungs. In the single-breath method, the
patient breaths out to RV and then inspires a gas
mixture of helium and 0.03 per cent carbon mon-
oxide in air, to TLC. This is held in the lungs for
10 sec, and then the patient slowly exhales. An
initial volume of 750 ml is discarded, and then a
similarly sized sample of expired gas is collected
and analyzed for the alveolar concentrations of
helium and carbon monoxide. The change in con-
centration of the gases from the original inspired
mixture enables two measurements to be obtained.
First, helium is inert and is not absorbed; the
change in helium concentration as a result of dilu-
tion within the lungs, therefore, provides a meas-
urement of the volume of alveolar gas into which
the gas mixture was inspired (V
A
). Second, the rate
of uptake of carbon monoxide, a highly soluble
gas that combines rapidly with hemoglobin in the
capillaries, gives the DL
C()
as the uptake of carbon
monoxide/minute/unit of partial pressure of the gas
Figure 5-20 Flow-volume loops
demonstrating normal pattern. Dia-
gram includes tidal volume (inner
loop) and vital capacity (outer loop)
loops. The vital capacity loop shows
the V 50% and V 25% points. (PEF
= peak expiratory flow; RV = re-
sidual volume.
(mmol/min/KpJ. DL
c o
is reduced when the alveo-
lar capillaries are reduced to number, there is in-
creased ventilation-perfusion mismatch, the hemo-
globin concentration is decreased, and the patient
is a heavy smoker with a very high level of car-
boxyhemoglobin. DL
c o
is difficult to assess in pa-
tients in -whom line VC is iess than \ L in
children younger than 7 years.
The uptake of carbon monoxide per unit of lung
can be determined by dividing the DL
c o
by the
alveolar volume, thereby determining the ^
0
The
Kco is useful in distinguishing a reduced transfer
coefficient caused simply by a reduction in lung
volume (reduced DL
c o
and normal Ko), as, for
example, after pneumonectomy, from situations
when the lung and its volume are both abnormal
(reduced DL
co
and reduced K^) , as in emphy-
sema.
In one study, of 38 preoperative and operative
risk factors, the DL
co
was the most important pre-
dictor of morality (p < .01) and was the sole
predictor of postoperative pulmonary complica-
tions (p < .005).
I54
The authors concluded that the
DL
co
can reveal structural and functional abnor-
malities of emphysema even when spirometric val-
ues are acceptable. A DL
co
less than 60 per cent
of that predicted
154
or a predicted postoperative
percentage of less than 40 per cent
151
is considered
to be indicative of very high risk (Table 5-12).
(4) MAXIMUM BREATHING CAPACITY. Thi s IS
the maximum amount of air that can be breathed
in 1 min. It is expiratory effort dependent and
reflects the total function of the entire cardiorespi-
ratory apparatus. However, the maximum breath-
ing capacity (MBC) may have unique value in that
it also depends on the intangible variables of co-
operation, motivation, and stamina in addition to
cardiorespiratory function. Thus, the predictive
value of MBC should be similar to an exercise
test, such as bicycle ergometry or treadmill walk-
ing (see next subsection).
162
(5) EXERCISE TESTING. Exercise increases utili-
zation of oxygen peripherally and requires the en-
tire interlocking lung/heart/vascular oxygen trans-
port system to react. Thus, the potential exists to
evaluate much of the cardiopulmonary system
with just one test.
163
Although exercise may be as simple as walking
or stair climbing,
157
exercise is usually undertaken
on a cycle ergometer or treadmill. Initially, the
patient cycles at rest with no resistance to the
pedals or walks at a slow pace on the treadmill. At
regular intervals of 1 to 3 min (according to pro-
tocol), the resistance to the cycle ergometer pedals
or the speed and/or slope of the treadmill is in-
creased by standard increments. At any increment,
a steady state may be achieved for a submaximal
end point, or the increments may be increased to a
maximal (symptom-limited) end point (onset of
chest pain, breathlessness, or other discomfort or
maximal allowable heart rate).
During exercise, the following measurements
should be made: minute ventilation, heart rate
from an electrocardiogram (EKG), and mixed ex-
pired carbon aioxie and mixed expired oxygen
(F
E
C0
2
and F
0
2
) concentrations. These measure-
ments are used to calculate minute ventilation, ox-
ygen uptake (Vo
2
), carbon dioxide production
(Vco
2
) and hence the respiratory exchange ratio
R(Vco
2
/Vo
2
). For each patient, heart rate and gas
exchange are documented minute by minute and
can be plotted against normal values (Fig. 5-21).
The normal response to exercise consists of a
linear increase in heart rate when plotted against
oxygen consumption (Vo
2
) until maximum heart
rate is reached. As workload is progressively in-
creased, Vo
2
increases until a plateau is reached
where further work produces no further increase in
Vo
2
. This point is called
a
Vo
2max
Typical values
for a healthy 70-kg adult man are a resting Vo
2
=
245 ml/min (see chapter 3, Fick Principle) and a
Vo
2
maximum of 2800 ml/min (cardiac output can
increase fourfold and the arteriovenous 0
2
content
can increase threefold).
164
Minute ventilation in-
creases progressively during exercise. The tidal
volume increases until it reaches 60 per cent of the
VC, and then ventilation increases solely by an
increased breathing rate. Maximal exercise venti-
lation in the normal subject is 60 to 70 per cent of
the MBC.
In patients with lung disease, there is an in-
crease in ventilation-perfusion mismatch and in-
creased dead space. The increase in dead space
requires a high minute ventilation to achieve an
alveolar ventilation sufficient to maintain normal
arterial blood gas tensions. Many patients with
lung disease can maintain normal arterial gas ten-
sions only at the expense of excessive minute ven-
tilation. When lung disease is severe, arterial Po
2
drops, and this is immediately detected by the ox-
imeter if severe. Physiologic abnormalities capable
of limiting Vo
2
at maximal exercise are anemia/
carboxyhemoglobinemia, heart disease, metabolic
disease, muscular disease, peripheral vascular dis-
ease, pulmonary disease, and reduced effort. At
some point, the utilization of oxygen by the mus-
cles may exceed the oxygen availability provided
by the transport axis. At this point, anaerobic
mechanisms are used by the muscles, and lactate
is produced. Information concerning this point
may be obtained by direct measurement of the
level of lactate in the blood.
165
Considering these physiologic responses to ex-
ercise in patients with lung disease and cardiac
disease, it is apparent there are similarities and
dissimilarities (Table 5-13).
166
Similar responses
Normal Response to Progressive Exercise Test
Diagnosis: normal control
Fi gure 5- 21 The V
E
and heart rate data for a 62-year-old man from a progressive exercise test plotted against the range of
values for normal controls. (Modified from Spiro SG, Roberts CM. Lung function tests. Medicine International 3661, I 99l . Used
with permission.)
Table 5-13 CARDIOPULMONARY EXERCISE TEST RESULTS
Type of Test Test Pulmonary Limitations Cardiac Limitations
Tests that do not differentiate
between pulmonary and
cardiac limitation (i.e.,
nonspecific)
Tests that do differentiate
between pulmonary and
cardiac limitations (i.e.,
specific)
Vo
2max
Reduced Reduced
Maximal heart rate Reduced Reduced
Ventilation/work Increased Increased
Pattern of breathing Rapid, shallow Rapid, shallow
Breathing reserve Reduced Normal
Blood gas Hypoxemia, hypercapnia,
respiratory acidosis
Anaerobic threshold Indeterminate or normal Indeterminate or reduced
Blood pressure response Normal Blunted
Electrocardiogram Normal Ischemia, arrhythmia
include reduced Vo
2max
and maximal heart rate,
increased ventilation relative to workload, and
rapid, shallow breathing. Dissimilar responses in-
clude reduced breathing reserve, hypoxemia, and
hypercapnia in lung disease and low anaerobic
threshold (expressed in ml 0
2
/kg/min), blunted
blood pressure response, and abnormal EKG in
cardiac disease.
Patients may have both cardiac and pulmonary
disease, and exercise testing may be helpful in
establishing the disease process that is contributing
the most to limitation. However, the precise per-
centages of contributions of cardiac and pulmo-
nary limitations are difficult to determine. In a
practical sense, it is most important to be aware
that both diseases are present. Specific therapeutic
trials can then be undertaken and the effect on
exercise response measured.
Several studies used exercise tolerance as a pre-
dictor of postpulmonary resection morbidity and
mortality.
148
,51
155, l56
The results to date have been
very encouraging; if the preoperative value is be-
low that shown in Table 5-12, then there is a 75
to 100 per cent chance of significant morbidity or
mortality.
(6) LUNG VOLUME. The functional residual ca-
pacity (FRC) is defined as the volume of gas in
the lung that exists at the end of a normal expira-
tion when there is no airflow and alveolar pressure
equals the ambient pressure. The FRC is approxi-
mately 45 per cent of TLC. The expiratory reserve
volume is additional gas below FRC that can be
consciously exhaled and results in the minimum
volume of lung possible, known as the RV. The
RV is approximately 30 per cent of the TLC. Thus,
the FRC equals the RV plus the expiratory reserve
volume (Fig. 3-24). Total lung volume, FRC, and
residual volume all contain a fraction (the residual
volume) that cannot be measured by simple spi-
rometry. However, since FRC is usually measured
(by N
2
washout, by He dilution, or plethysmo-
graphically; see chapter 3), TLC and RV can be
easily derived by using the other lung volumes that
are measured by simple spirometry.
The size of RV is determined by the force of
the expiratory muscles opposed by the tendency of
the thorax to recoil outward at low lung volumes.
Beyond the third decade of life, the RV increases
as a result of dynamic compression or closure of
airways; as a fraction of TLC, RV increases from
approximately 25 per cent at the age of 20 years
to 40 per cent at 70 years.
Airway disease accelerates the increase in RV.
Any airway narrowing or loss of elastic recoil,
allowing dynamic compression, facilitates the trap-
ping of gas within the lungs, and an increasing RV
is a characteristic feature of early obstructive dis-
ease. The RV/TLC per cent is a sensitive measure
of air trapping subsequent to airflow obstruction.
To improve the efficiency of breathing, the ob-
structed patient is obliged to breathe at a volume
above normal FRC, causing overinflation. These
patients' lungs are overinflated in terms of the lung
volume at which they pursue their tidal breathing.
Recently, abnormal permanent enlargement of air-
spaces (emphysema) has been noninvasively and
successfully assessed using CT scans taken at full
expiration.
167
b. REGIONAL LUNG FUNCTION TESTS
When a tumor completely occludes a mainstem
bronchus, then tests of whole-lung function evalu-
ate the function of the nonaffected side; in other
words, physiologically, the patient has already had
a pneumonectomy. However, in the large majority
of cases, tests of whole-lung function do not sepa-
rate out the relative contribution made by the lung
tissue to be resected from the contribution made
by the lung tissue that would remain after resection
to the total preoperative ventilation and gas ex-
change. In fact, ventilation-perfusion scans have
shown that, apparently, small tumors may greatly
distort the distribution of ventilation or perfusion
or both, and this may result in a misleading inter-
pretation of whole-lung function, especially if
there is coexistent underlying generalized disease
such as chronic bronchitis.
168
Thus, tests of whole-
lung function may fail to resolve whether the pa-
tient would survive a resection without being left
unduly dyspneic in cor pulmonale. The regional
lung function tests consist of radioisotope studies
(radiospirometry, which is the most important one)
and somewhat less important nonradioisotope
studies (the lateral position test, bronchial block-
ade, main-stem pulmonary artery blockade).
(1) RADIOISOTOPE REGIONAL PERFUSION, VEN-
TILATION-PERFUSION STUDIES (RADIOSPIROME-
TRY).
169
"
171
In the past, regional lung function was
measured by differential bronchospirometry. The
procedure involved considerable patient discom-
fort due to double-lumen endotracheal tube inser-
tion (usually into an awake patient) and technical
and physiologic uncertainties related to the re-
quired use of the Fick principle. Presently, right-
left split pulmonary functions are obtained with
easily performed noninvasive ' "Xe radiospirome-
try and ' "Xe and macroaggregate ("Te) perfusion
scanning.
Usually, three studies are performed to deter-
mine right-left split pulmonary function (regional
ventilation, regional perfusion, regional lung vol-
ume [and regional ventilation/perfusion, ventila-
tion/volume, perfusion/volume relationships by
appropriate division of the primary three num-
184 Preoperative Cardiopulmonary Evaluaii
Figure 5-22 Three studies are usually performed to deter-
mine regional lung function. Regional perfusion (A) is deter-
mined by intravenous injection of an insoluble radioisotope,
which distributes to the two lungs according to the blood flow
to each lung. Peak radioactivity over each lung (peak counts
with time) is proportional to each lung blood flow. Regional
ventilation (B) is determined by having the patient inhale an
insoluble gaseous radioisotope. Peak radioactivity over each
lung (peak counts with time) is proportional to each lung ven-
tilation. Regional volume (C) is determined by equilibrating
radioactive material within the lungs with a connected closed
space. The plateau radioactivity of each lung (plateau count
with time) is then proportional to the volume of each lung. See
text for more detailed explanation.
bers]; Fig. 5-22). First,
l33
Xe (or
w
Te or
l 3l
I
MAA)'
47
is infused as a bolus intravenously (Fig
5-22A). Because
l 33
Xe is a very insoluble gas
almost all of the
l 33
Xe (95 per cent) evolves out o:
the blood into the alveoli, and this is respired ou
of the body into an open system. Regional externa
chest scintillation counters record these events ai
a time (x-axis) versus activity (y-axis) curve. The
time versus activity curve shows a rapid vascular
to-alveolar wash-in peak and then an exponentia
alveolar-to-environment washout. Since
l33
Xe ap-
pears in the airspace only in proportion to the
perfusion of the region, the regional perfusion
wash-in peak count divided by the total counl
equals the regional fractional perfusion. Since "Te
and
l3l
I-MAA simply lodge in the pulmonary mi-
crocirculation, the regional activity count divided
by the total chest count with these isotopes more
directly yields the regional fractional perfusion.
In the second study, a vital capacity breath of a
bolus of
l 33
Xe results in a ventilation wash-in peak
(environment-to-alveolar) followed by an expo-
nential alveolar-to-environment washout of
l33
Xe
(Fig. 5-22Z?). Regional ventilation wash-in peak
counts divided by the total chest count yield the
regional fractional ventilation.
In the third study,
l 33
Xe is again administered as
an intravenous bolus, but after the
l 33
Xe evolves
into the alveoli from the vascular space, it is res-
pired into a closed system (Fig. 5-22C). After 10
to 15 min of equilibration, the
l33
Xe should have
reached all gas spaces within the lungs (slow and
fast time constants), and the concentration of
133
Xe
should be uniform throughout the lungs and the
attached closed system. The regional radioactive
counts/unit time then is proportional to the re-
gional volume of lung.
Radioactive-scanning radiospirometry therefore
yields regional perfusion, ventilation, and lung
volume. Dividing regional perfusion and ventila-
tion by regional lung volume results in regional
perfusion and ventilation per unit lung volume.
Finally, dividing regional ventilation by regional
perfusion results in the regional ventilation-perfu-
sion ratio.
Conventional whole-lung function tests (VC,
FEV,, maximal breathing capacity) cannot predict
postpneumonectomy lung function because the
amount of lung function to be removed is other-
wise unknown. However, multiplication of preop-
erative whole-lung pulmonary function tests by the
percentage of lung to be removed (or that percent-
age of lung that will remain), as determined by
radioactive-scanning radiospirometry, should the-
oretically predict postoperative pulmonary func-
tion testing.
147
Specifically,
Predicted postoperative FEV, =
Preoperative FEV, X [% perfusion to
"remaining" lung/100]
For example, if the preoperative VC and FEV, was
2.00 L and 1.40 L, respectively, and the perfusion
and/or ventilation of the lung to be removed was
40 per cent of the total ventilation, then the pre-
dicted postoperative VC and FEV, would be 1.20
L and 0.84 L, respectively. In the absence of ade-
quate resolution of the V/Q scans (especially at the
lobar level) prediction of postlobectomy function
and the loss in function can be calculated as:
172
Loss of function = preoperative FEV, X
[# of functional segments in lobe to be
resected/total # of segments in both lungs]
(where the total number of segments in both lungs
= 42).
Combining radiospirometry with conventional
pulmonary function tests has, in fact, resulted in a
fair to good correlation between predicted and
measured postpneumonectomy pulmonary func-
tion testing, including FVC, MBC, FEV,, FRC,
DL
co
and exercise capacity (Vo
2max
).
173-178
Figures
5-23, 5-24, and 5-25 show the results of one
study that compared per cent preoperative perfu-
sion to the lung to be resected to the change in
several different indices of pulmonary function
(FEV,, DL
co
%, K
co
%, V
Emax
, Vo
2 m
J; obviously,
the correlation for all indices was very good.
178
On the basis of this type of pulmonary function
analysis, the literature suggests that a predicted
postoperative FEV, of 0.8 or greater yields an "ac-
ceptable" incidence of postoperative morbidity
and mortality.
147,169
l73, l75
However, in view of the
fact that a short, elderly, thin women probably
does not need as large a postpneumonectomy
FEV, as a tall, young, muscular man, the predicted
postoperative percentage of predicted FEV, should
be used for determinations of operability rather
than the absolute value of FEV, (800 ml in Table
5-12).
Specifically,
Postpneumonectomy FEV, as % normal =
[predicted postpneumonectomy
FEV,/normalFEV,] X 100
The lowest proposed cutoff for FEV, is 30 per
cent (see Table 5-12); for other acceptable pre-
dicted postpneumonectomy values as a per cent of
normal, see Table 5-12.
Preoperative ventilation-perfusion scans appear
to identify those patients at greatest risk of hypox-
emia during one-lung anesthesia (as a result of
continued perfusion of the nonventilated, operative
/ preoperative perfusion to resected lung
10 20 30 40 50 60
Figure 5-23 Relationship between observed changes in FEV, (y-axis) and percentage of preoperative perfusion to resected lung
(x-axis) in 28 patients (r = - 0. 86, y = -0.788x + 0.14). The broken lines indicate 95% confidence limits. (From Corris PA.
Ellis DA, Hawkins T, Gibson GJ: Use of radionuclide scanning in the preoperative estimation of pulmonary function after
pneumonectomy. Thorax 42:285-291, 1987. Used with permission.)
lung). Figure 4-9 provides good evidence that hy-
poxemia during one-lung ventilation results, in
part, from persistent blood flow to the operative
lung.
179
Although the correlation between the per
cent of flow to the operative lung and the one-lung
ventilation P
a
0
2
is good, the preoperative scan
cannot take into account the influence of posture,
anesthesia, and surgery on pulmonary shunt and
ventilation-perfusion matching (see chapter 8).
Taken altogether, the studies discussed in the
section on regional lung scanning and their high
degree of accuracy in predicting intraoperative and
postoperative function, combined with the availa-
bility of lung scanning in most hospitals and the
ready acceptance of the test by patients, make the
quantitative lung scan a most useful preoperative
regional lung function test.
(2) LATERAL POSITION TEST. This has had
some success in approximating individual lung
ventilation.
180
When a patient with normal lungs
turns from the supine position to the lateral decu-
bitus position, there is an increase in total lung
FRC because the increase in nondependent lung
FRC (due to less exposure of the nondependent
lung to the weight of the mediastinum and pres-
sure of the abdominal contents) exceeds the de-
crease in dependent lung FRC (due to greater ex-
posure of the dependent lung to the weight of the
mediastinum and the pressure of the abdominal
contents).
181-183
Thus, the procedure is thought to
be a test of the expansile and ventilatory capabili-
ties of the nondependent lung.
The test is performed by having the patient
breathe continuously into a spirometer while se-
Figure 5-25 (a) Relationship be-
tween observed changes in maxi-
mum ventilation (V
F
max (y-axis)
and percentage of preoperative per-
fusion to resected lung (x-axis) in 14
patients (r = -0. 89, y = -0. 56x
0.03). The broken lines indicate 95%
confidence limits, (b) Relationship
between observed changes in Vo
2
max (y-axis) and percentage preop-
erative perfusion to resected lung (x)
in 14 patients (r = -0. 90, y =
-9.7x -0.09). The broken lines in-
dicate 95% confidence limits. (From
Corns PA, Ellis DA, Hawkins T,
Gibson GJ: Use of radionuclide scan-
ning in the preoperative estimation of
pulmonary function after pneumo-
nectomy. Thorax 42:285-291, 1987.
quentially turning from the supine position to one
of the lateral decubitus positions, back to the su-
pine position, then into the other lateral decubitus
position, and finally back into the supine position.
The increase in total lung FRC is easily and non-
invasive^ determined by noting the rise in the
end-expiratory level of the continuously recorded
spirogram; that is, the sine wave tidal ventilation
on the spirogram first shows a gradual slope up-
ward and then a plateau at a new stable level. The
results are expressed as the increase in FRC in one
lateral decubitus position divided by the increase
in FRC in both lateral decubitus positions.
The relative increase in FRC in each lung has
correlated well with oxygen consumption and min-
ute ventilation determined bronchospirometri-
cally
184
in each lung and with radionuclide venti-
lation-perfusion studies when the lung disease was
symmetrical,
185
and moderately well when the lung
disease was asymmetrical,
186
but predicts post-
pneumonectomy FEV, only fairly well.
187
The lat-
eral position test is most likely to give false infor-
mation whenever the disease process has little
effect on the compliance of the lung in question.
It should be stressed that since radionuclide venti-
lation-perfusion studies are well tolerated, safe,
and simple to perform and provide exquisitely de-
tailed information about both regional ventilation
and perfusion, they must be considered the tests of
choice of regional lung function. However, radio-
nuclide studies require the use of sophisticated
equipment, which may not be available in all hos-
pitals, and it would seem that the lateral position
test might be a useful alternative in this situation.
(3) REGIONAL BRONCHIAL BALLOON OCCLU-
SION. Postpneumonectomy (or after any resection)
ventilatory function can also be simulated preop-
eratively by passing, with the aid of a fiberoptic
bronchoscope, a balloon occlusion catheter, which
can occlude either lung (or any lobe), and then
measuring spirometry of the remaining lung tissue
(after careful withdrawal of the bronchoscope).
Supplemental oxygen must be administered during
bronchial blockade because the blocked segment
would still be perfused, and all of this perfusion
would be right-to-left shunt flow, which would
create a risk of hypoxemia.
The regional bronchial balloon occlusion
method of predicting postpneumonectomy ventila-
tory function has been studied during cycling ex-
ercise with a steady-state load equivalent to walk-
ing at a brisk pace for that patient.
188
The effects
on minute ventilation and oxygen uptake were ob-
served during occlusion of the bronchus to the
diseased lobe. If the patient was able to continue
cycling and maintain the same workload during
occlusion, this was regarded as evidence that they
would withstand resection of the occluded lung
tissue. All patients who could maintain the work-
load during preoperative bronchial occlusion were
able to do so postoperatively.
(4) REGIONAL PULMONARY ARTERY BALLOON
OCCLUSION. Postpneumonectomy pulmonary cir-
culatory and right ventricular function can be sim-
ulated preoperatively by passing into the main pul-
monary artery on the side to be resected a
pulmonary artery catheter (using fluoroscopy) that
has a 5-ml balloon at the tip of the catheter and a
port for measuring pressure that is proximal to the
balloon. Inflation of the distal balloon functionally
resects the vasculature distal to the balloon; the
pulmonary artery balloon inflation can be done
with and without exercise. Under these conditions,
all the pulmonary blood flow is diverted to the
lung that will remain after the pneumonectomy,
and the distensibility and compliance of the re-
maining pulmonary vascular bed are therefore
tested. However, the test is unphysiologic in that
the blocked lung would still be ventilated, and all
the ventilation to this lung would be dead-space
ventilation, which would not be present postpneu-
monectomy. If the mean pulmonary artery pres-
sure increases above 40 mm Hg, the P
u
C0
2
in-
creases above 60 mm Hg, or the P
a
0
2
decreases
below 45 mm Hg, it is likely that the patient will
not be able to tolerate resection of that amount of
pulmonary vascular bed without development of
right ventricular failure and cor pulmonale.
147
I74
189 l y3
Because this test is highly invasive, requires
special equipment, and is technically difficult, it is
rarely clinically used today.
Simultaneous balloon occlusion of both a main-
stem pulmonary artery and bronchus should com-
pletely simulate the physiologic effects of pneu-
monectomy and provide the most realistic assess-
ment of total postpneumonectomy lung function;
there would be no acute increase in shunt and/or
dead space. However, since this potentially very
accurate simulation of the postresection condition
is so invasive and complicated, it can only be
regarded as a research tool; the combined blockade
test has not yet been reported in humans.
C. SEQUENCE OF TESTS FOR LUNG
RESECTION SURGERY
With special reference to the performance of a
pneumonectomy, there is a consensus that pulmo-
nary function testing should proceed in three
phases (Fig. 5-26).
I47
l74
l89
m
~
1 9 6
The first phase
evaluates total lung (both, bilateral lung) function
and consists of arterial blood gas measurements,
simple spirometry, and perhaps carbon monoxide
diffusion capacity and exercise capability. In-
creased risk is present when hypercapnia is found
on a room air blood gas sample, the FEV, and/or
the MBC is less than 50 per cent of that predicted,
the RV/TLC is greater than 50 per cent, and/or
DL
c o
per cent or exercise tolerance is markedly
reduced. If any of these whole-lung pulmonary
function values are worse than these stated limits,
testing should proceed to the second phase, which
evaluates the function of each lung separately (sin-
gle, unilateral lung function) and consists of meas-
urement of the ventilation and perfusion of each
individual lung (as a fraction of the total) by ra-
dioisotopic ('"Xe, "Te) scanning. Combining
right-left fractional lung function tests with con-
ventional spirometry and/or DL
c o
or Vo
2max
should
yield the appropriate predicted postoperative pul-
monary function. If this second level criterion can-
not be met and surgery is still contemplated or
desired, the postoperative condition of the patient
can be simulated (the third phase of testing; see
also next section on pulmonary vascular function
testing) by functionally resecting the vascular bed
of the lung to be taken out by temporary balloon
occlusion of the major pulmonary artery on that
side with and without exercise. An increase in
mean pulmonary artery pressure above 40 mm Hg
(or an increase in P
a
C0
2
above 60 mm Hg, or a
decrease in P
a
0
2
below 45 mm Hg) indicates an
inability to tolerate the removal of this amount of
lung.
147
I74
89
-
| 9
This pulmonary function test cas-
cade is logical because it starts out with simple,
inexpensive, and noninvasive tests and only in-
creases the degree of difficulty, expense, and in-
vasiveness as necessary; thus, in practice, the third
phase of testing is rarely performed. In interpreting
the results of this preoperative pulmonary function
test cascade, physicians should always ask them-
selves what is an acceptable surgical mortality risk
in a disease that has close to 100 per cent natural
history 5-year mortality rate (even if the patient
receives radiotherapy).
197
Although less restrictive pulmonary function
test criteria for operability for pulmonary resec-
tions less than pneumonectomy have been pub-
lished (see Table 5-12),
198
there are several rea-
Preoperative Evaluation of Masses of the
Lung and Bronchi
Step III: Respiratory Function
Whole Lung
Function
Post-Operative
Function
Figure 5-26 Sequence of tests to determine pulmonary function for lung resection surgery (step III). The first group of tests
determines whole lung function, the second test determines regional lung function, and the third test mimics postoperative lung
function. See text for full explanation and definition of abbreviations.
sons why in some patients it may be prudent to
think of a lobectomy (and lesser procedures) as a
functional pneumonectomy.
199
First, in the imme-
diate postoperative period, the function of the lung
tissue remaining on the operative side may be sig-
nificantly impaired by atelectasis and perhaps in-
fection; consequently, these patients may experi-
ence significant transient postoperative functional
impairment.
200
Patients who are most likely to
have a stormy postoperative course with minor
resections are those in whom the surgeon had in-
traoperative exposure problems that required se-
vere and prolonged lung handling, retraction, com-
pression, and packing. Intraoperative exposure
problems are more likely to occur when the lung
under operation is large and moving (large tidal
volume positive-pressure ventilation). Second, at
the time of thoracotomy, more accurate staging of
the disease is possible, and it may become appar-
ent that it is necessary to do a pneumonectomy.
194
Third, the function of the lung on the nonoperated
side may be impaired preoperative^
200
and may
acutely deteriorate further intraoperatively (aspira-
tion and/or spillage of blood and/or pus from the
operated to the nonoperated lung, inability of the
nonoperated lung to tolerate dependency and com-
pression in the lateral decubitus position). Finally,
postoperative studies have shown that, although
the ventilation and perfusion of the lung remaining
on the operated side increase significantly during
the long-term interval (3 to 51 months) after lobec-
tomy, the volume of the remaining lung increases
even more, so that the ventilation and perfusion
per unit volume of the remaining lung decrease;
this is equivalent to hyperinflation.
176
The compen-
satory hyperinflation represents dilatation of the
pre-existing respiratory units (as opposed to hyper-
plasia [alveolar multiplication] or hypertrophy [in-
crease in existing mass by active growth])
201
with-
out disruption or fragmentation of the elastic tissue
as seen in pathologic emphysema; however, the
pulmonary hyperinflation decreases the compli-
ance and, therefore, the ventilation per unit volume
of the ipsilateral remaining pulmonary tissue. In
addition, the hyperinflated lung stretches and thins
out the capillaries in the alveolar walls, which de-
creases the perfusion per unit volume of the re-
maining pulmonary tissue on the ipsilateral side.
3. Pulmonary Vascular and Right
Ventricular Function and Testing
Assuming that there are no intracardiac shunts,
right ventricular output is equivalent to that of the
left ventricle, but because the pulmonary artery
pressure is much lower than systemic blood pres-
sure, the design and performance of the two ven-
tricles are quite different. The right ventricle has a
thin lateral free wall (it pumps against a low pres-
sure) and is a highly compliant chamber that better
accommodates increases in filling pressure. Sys-
tolic pressures in the right ventricle and pulmonary
artery are usually less than 30 mm Hg. At rest,
right ventricular end-diastolic pressure is less than
6 mm Hg and mean PAP ranges from 10 to 18
mm Hg. During exercise, recruitment of small ar-
terioles and capillaries of the pulmonary vascular
bed compensates for an increase in cardiac output,
and mean PAP rarely exceeds 20 to 30 mm Hg.
Thus, in the healthy individual, vigorous demands
are seldom made of the right ventricle (if the de-
mands are excessive, acute dilation may occur, but
hypertrophy will not).
202
Most patients with pulmonary tumors have had
a long history of smoking; consequently, they have
varying degrees of chronic obstructive pulmonary
disease (COPD). The cardiovascular response to
the chronic pathologic alveolar and airway
changes in COPD consists of the development of
pulmonary hypertension and increased pulmonary
vascular resistance (secondary to hypoxic pulmo-
nary vasodilation, the destruction of alveolar sep-
tae, increased blood viscosity resulting from poly-
cythemia and thromboembolism, and vascular
compression caused by air trapping)
202
followed by
right ventricular hypertrophy, dilation, and dys-
function (cor pulmonale).
203
Right ventricular dys-
function can lead to a shift of the interventricular
septum within the pericardium and impair left ven-
tricular filling and function. The prognosis of cor
pulmonale in patients with COPD is poor; 3-year
mortality is estimated at 60 per cent.
204
Increased PVR has very important implications
for patients undergoing pulmonary resection.
Whereas a normal pulmonary vasculature is dis-
tensible and capable of accommodating large in-
creases in pulmonary blood flow (to approximately
2 to 2.5 times greater than normal, as would occur
through the remaining lung following a pneumo-
nectomy) with only minor increases in PAP (Fig.
5-27), the relatively rigid and restricted pulmonary
vascular bed of patients with chronic lung disease
cannot accommodate even small increases in pul-
monary blood flow without concomitant increases
in pulmonary vascular pressure.
205
The inability to
tolerate increases in blood flow occurs over the
entire range of physiologic cardiac output and may
be an important contributing factor to the devel-
opment of postpneumonectomy pulmonary edema
when it occurs.
206
The preoperative pulmonary function testing
cascade as outlined in phases one and two in the
section Sequence of Tests for Lung Resection Sur-
gery (which, in reality, is the extent to which the
large majority of patients are studied preopera-
tively) does not allow for specific diagnosis of
Cardiac Output
(in multiples of resting output)
Figure 5-27 Mean pulmonary artery pressure (y-axis) does
not increase until cardiac output (x-axis) has been increased to
2 to 2.5 times when the pulmonary vascular bed is normal,
whereas mean pulmonary artery pressure increases linearly
when cardiac output is increased when the pulmonary vascular
bed is restricted.
204
increased PVR and right ventricular disease. In-
creased PVR may be noninvasively suspected pre-
operative^ by the presence of the auscultatory and
radiographic signs of pulmonary hypertension and
by electrocardiographic evidence of right atrial and
ventricular hypertrophy (Table 5-14). In COPD
patients without waking hypoxemia, cor pulmo-
nale can be detected twice as sensitively and fre-
quently by echocardiography (definition criteria
are pulmonary hypertension and right ventricular
enlargement and/or hypertrophy) as by the clinical
methods just discussed.
207
Indeed, invasive preop-
erative monitoring (pulmonary artery catheteriza-
tion) in a consecutive series of 148 patients older
than 65 years awaiting all types of major surgery
showed a 50 per cent incidence of elevated mean
PAP.
208
Clinically, the onset of cor pulmonale is
most often indicated by the development of a pos-
itive hepatojuglar reflex, ascites, and peripheral
edema. Not surprisingly, the incidence of death
and total cardiac complications is increased in pa-
tients with pulmonary hypertension in the periop-
erative period.
209
Measurements of PVR have been directly made
by determining mean PAP and pulmonary wedge
pressure at various levels of cardiac output pro-
duced by varying treadmill exercises (Fig. 5-28).
Thus, using the patient's own cardiac output, pul-
monary vascular compliance can be determined.
PVR measurements made in this way have been
good indicators of risk for pneumonectomy.
2I(V
~
:
"
Operative risk was considered to be increased if
PVR was greater than 190 dynes/sec/cm~
5
. How-
ever, if the risk, expense, and time have been taken
to insert a pulmonary artery catheter, then it is
logical to take one further step and measure pul-
monary vascular pressures during temporary uni-
lateral pulmonary artery balloon occlusion in states
of rest and exercise. This specifically tests the
Table 5-14 NONINVASIVE DIAGNOSIS OF PULMONARY HYPERTENSION ( PAP),
INCREASED PULMONARY VASCULAR RESISTANCE ( f PVR), RIGHT ATRIAL AND
VENTRICULAR HYPERTROPHY ( f RA AND RV), AND COR PULMONALE (CP)
Abbreviations: P-A = posterior-anterior.
Figure 5-28 Preoperative evaluation logic to determine cardiovascular function in patients with lung and bronchial carcinoma (step III). See text
for full explanation. (CABG = coronary artery bypass grafting; CAD = coronary artery disease; EKG = electrocardiogram.)
compliance of just the pulmonary vascular bed that
will remain after pneumonectomy.
Temporary unilateral pulmonary artery balloon
occlusion simulates the pulmonary vascular con-
ditions to be expected after pneumonectomy (see
Fig. 5-28). If significant pulmonary hypertension
(P- > 40 mm Hg) or arterial hypoxemia ensues,
pneumonectomy will likely not be tolerated be-
cause of the high risk of causing cor pulmonale,
pulmonary edema, and low ventilation-perfusion
relationships. Performing this procedure during
exercise is the most realistic preoperative approxi-
mation of pulmonary vascular and right ventricular
function that can be expected in the ambulatory
postpneumonectomy patient.
174189 nl93
4. Left Ventricular and Coronary Artery
Function and Testing
Ischemic heart disease is the number one cause
of morbidity and mortality in the United
States.
212
213
Cigarette smoking is thought to be
responsible for an estimated 80 per cent of all
heart attacks
214,215
and 30 per cent of the fatalities
caused by this disease.
216
The lethality of smoking
is understandable in view of the fact that patients
with coronary artery diseases who smoke have sig-
nificantly and substantially more active myocardial
ischemia during daily life than patients who do
not.
213
The 1979 report of the U.S. Surgeon Gen-
eral on smoking and health
217
leaves no doubt that
smoking is a major independent cause of coronary
artery disease and myocardial infarction. The ad-
verse cardiovascular effects of cigarette smoking
generally demonstrate a dose-response relationship
in the sense that the risk increases with increased
duration of smoking, increased quantity of ciga-
rettes smoked, and the increased depth of inhala-
tion of smoke.
218
For example, the risk of a fatal
or nonfatal coronary event in smokers of 25 or
more cigarettes a day was more than 500 per cent
the risk in nonsmokers; the risk in even the lightest
smokers (one to four cigarettes a day) was more
than doubled.
214
215
Thus, considering the usual
age, the long and heavy smoking history, and the
frequently sedentary lifestyle of patients undergo-
ing thoracic surgery, it is not surprising that most
of these patients have coronary artery disease and
ischemic heart disease to some extent and that
coronary artery disease is by far the most likely
independent cause of left ventricular dysfunction.
A number of mechanisms might explain the re-
lationship between smoking and coronary disease.
Carbon monoxide, one of the most poisonous by-
products of cigarette smoking, makes up approxi-
mately 2.7 to 6 per cent of the smoke
219
and may
cause damage by injuring vascular endothelium
and enhancing the deposition of cholesterol in the
development of atherosclerosis. Nicotine, a key
ingredient of cigarettes, acts to release catechola-
mines, which alter blood pressure and heart rate,
thereby increasing myocardial oxygen demand. In
combination with carbon monoxide, nicotine may
be the predisposing factor in the development of
coronary artery disease and myocardial infarction.
In addition to increasing blood pressure and heart
rate, nicotine also causes constriction of muscular
arteries, which may interfere with regional tissue
flow
220
and has been shown to lower thresholds of
ventricular fibrillation.
221
Additionally, cigarette
smoking increases the adhesiveness of plate-
lets
222
223
and lowers high-density lipoprotein cho-
lesterol levels.
224,225
Myocardial ischemia leading to infarction may
occur throughout the perioperative period, al-
though peaks of incidence occur during opera-
tion
226
and on the third day after operation.
227
The
first peak is caused by intraoperative changes in
hemodynamics, and the second peak is caused by
episodes of hypoxia, uneven administration of pain
medication and withdrawal, or alteration of drug
therapy.
228
There are only two proven preoperative clinical
predictors of perioperative cardiac morbidity (de-
fined as occurrence of myocardial infarction, un-
stable angina, congestive heart failure, serious dys-
rhythmia or cardiac death during the intraoperative
or in-hospital postoperative periods), and they
were recent (less than 6 months) myocardial in-
farction and current congestive heart failure.
229
The
value of other historic predictors, such as previous
(old) myocardial infarction, angina, previous
congestive heart failure, hypertension, diabetes
mellitus, and age is still unresolved. The classic
historic intraoperative predictors of emergency
surgery, prolonged (more than 3 hours) surgery,
and thoracic or upper abdominal surgery also ap-
pear to be independent predictors of perioperative
morbidity, whereas choice of anesthetic is not. The
dynamic intraoperative predictors of perioperative
cardiac morbidity are intraoperative hypotension
and tachycardia. Hypertension remains a contro-
versial predictor.
Angina pectoris is a constellation of symptoms
that reflect transient inadequacy of myocardial
blood flow and can be classified as follows
230
: ( 1 )
classic angina, a transient precordial or substernal
discomfort typically provoked by exertion and
promptly relieved by rest or nitrates; (2) atypical
angina, similar symptoms but with the absence of
one or more of the criteria for classic angina (e.g..
the pain may not be consistently related to exertion
or relieved by rest); (3) angina equivalent, usually
the sensation of dyspnea as the sole or major man-
ifestation (some physicians also label isolated pain
referral, e.g., localized left arm or shoulder pain,
as an angina equivalent); and (4) variant or Prinz-
metal's angina, angina pectoris that occurs at rest
and may manifest in stereotyped patterns such as
nocturnal symptoms or symptoms only after exer-
cise. Variant angina is thought to be caused by
coronary artery spasm with or without underlying
coronary artery disease. The daily activities of the
patient can be used to classify the frequency of
angina (Table 5-15). Limitations of these classifi-
cations include the inability to categorize patients
who are not accurate observers, overlap between
classes, and inapplicability to the patient who does
not have exertion-related angina.
The EKG and the chest X-ray may help to con-
firm the presence of coronary artery disease and
ischemic heart disease. The EKG may show Q
waves (previous infarction), left bundle branch
block, ST-segment elevation (transmural ische-
mia), ST-segment depression (subendocardial is-
chemia), T-wave inversions, and positive U wave
(left main coronary artery disease). On the chest
X-ray, cardiomegaly may be found in as many as
15 per cent of patients with coronary artery dis-
ease; the absence of cardiomegaly does not ex-
clude left ventricular dysfunction or even dila-
tion.
231
If the history, physical examination, and EKG
raise suspicion of coronary artery disease, then
further perioperative evaluation of coronary artery
function is necessary (see Fig. 5-29).
a. EXERCISE-EKG TESTING
The first step should be noninvasive exercise
testing.
232
Electrocardiography and thallium scans
in that order appear to be the best such exercise
tests at this time. Because exercise raises heart rate
and blood pressure and increases myocardial oxy-
gen consumption, it provides an excellent assess-
ment for the probability of developing ischemia
during the perioperative period. Exercise is most
commonly performed on a treadmill in the United
States. There are three basic treadmill protocols;
the treadmill gradient may be increased progres-
sively with a constant speed (e.g., the Balke and
Naughton protocols), the speed is increased at a
constant gradient (e.g., the Ellestead protocol), or
both are increased (e.g., the Bruce protocol). Ex-
ercise EKG testing is highly predictive of subse-
quent cardiac events when the ST-segment
changes are characteristic, large (> 2.5 mm), im-
mediate (first 1-3 min), sustained into the recovery
period, or associated with the decrease in blood
pressure.
229
There are two major causes of false-negative
results. First, intraoperative ischemia may not be
associated with any hemodynamic changes.
229
233. 234 s
e c o n (
j
false-negative results most fre-

quently are due to inadequate exercise stress (gen-
eral debility, orthopedic limitation, or claudica-
tion).
If an adequate exercise EKG stress testing
workload is achieved, the sensitivity of exercise of
EKG stress testing for the diagnosis of coronary
artery disease is 62 to 80 per cent (mean, about 65
per cent) and the specificity is 83 to 96 per cent
(mean, about 85 per cent).
229
The end points in-
clude workload and duration of exercise, ST-seg-
ment and T-wave changes, blood pressure and
heart rate responses, and symptoms (chest pain or
dyspnea). If the exercise EKG is normal, then sur-
gery should proceed: If the exercise EKG indicates
ischemia, then coronary angiography should be
considered. If the exercise EKG cannot be prop-
erly interpreted/performed (pre-existing EKG ab-
normalities, cannot exercise), then radionuclide
angiography or exercise-thallium testing should be
done.
b. EXERCISE-THALLIUM TESTING
Exercise-thallium imaging is another preopera-
tive test that can be used in patients who have
abnormal EKGs suggestive of ischemic heart dis-
ease (and which may preclude interpretation of the
EKG) (see Fig. 5-28).
2
-" In addition, thallium im-
aging can better identify the location and extent of
myocardium at risk compared with conventional
exercise electrocardiography.
232
After intravenous
injection of thallium-201, myocardial uptake of the
Table 5-15 NEW YORK HEART ASSOCIATION FUNCTIONAL CLASSIFICATION OF
ANGINA PECTORIS
Functional
Class Occurrence of Symptoms
Exercise Tolerance
(Met) Functional Impairment
I With unusual activity 7-8 (or more)
II With prolonged or slightly more 5-6
than usual activity
III With usual activity of daily living 3-4
IV At rest 1-2
Minimal or none
Mild (can do light and general industrial
work)
Moderate (may be able to have desk job)
Severe (incapacitated)
From Shub C: Angina pectoris and coronary heart disease. In (eds): Cardiology: Fundamentals and Practice. Vol. 2. Brandenburg
RO\ Fuster V, Giuliani ER, McGoon DC. Chicago, Year Book Medical Publishers, 1987, pp 1073-1104. Used with permission.
radioactive tracer is dependent on regional myo-
cardial blood flow. In normal persons, the myocar-
dial uptake is homogeneous, whereas in areas of
markedly reduced perfusion (> 90 per cent steno-
sis), the uptake after 30 to 60 min is markedly
reduced, and a perfusion defect appears (cold-spot
imaging) (as opposed to hot-spot imaging with
technetium, which indicates increased uptake in an
ischemia area). The appearance of thallium perfu-
sion defects in two or three separate left ventricu-
lar segments identifies approximately 75 per cent
of patients with multivessel coronary artery dis-
ease, and approximately 95 per cent of patients
with three-vessel coronary artery disease have ab-
normal scans.
231
It should be remembered that pos-
itive results indicate areas of reduced blood flow
(at a time when the rest of the coronary circulation
is fully dilated secondary to exercise), but that
does not necessarily mean that these areas are con-
currently ischemic. A period of thallium redistri-
bution occurs at 2 to 4 hours. A repeat image is
performed at 4 hours. Defects in the 1-hour image
and 4-hour image indicate infarcted tissue,
whereas a defect in the 1 -hour tests alone indicates
reversible or collateralized reduced perfusion
areas. If the thallium exercise test is negative, then
the planned pulmonary resection should proceed.
If the thallium exercise scan is positive for ische-
mia, the coronary angiography should be done (see
Fig. 5-28).
C. DIPYRIDAMOLE-THALLIUM TESTING
For patients who cannot exercise adequately
(which eliminates exercise EKG and thallium test-
ing) and for patients with already abnormal EKG
(eliminates exercise EKG testing), dipyridamole-
thallium testing is an attractive alternative (see Fig.
5-28). Dipyridamole, like strenuous exercise,
causes marked coronary vasodilation and a general
increase in coronary blood flow. Because highly
stenotic coronary vessels cannot dilate normally,
areas of myocardium supplied by them will take
up less thallium during scanning than areas sup-
plied by normal vessels. Later, after the vasodila-
tion has resolved, these underperfused areas will
"fill i n" as the thallium redistributes. Thus, a
"cold" area on the scan that fills in during the
later (redistribution) phase of the procedure im-
plies the presence of a highly stenotic coronary
artery.
236
The sensitivity and specificity of dipyridamole-
thallium imaging for detecting coronary artery
disease are both approximately 90 per cent, ap-
proaching the sensitivity and specificity of thal-
lium-exercise scintigraphy.
231
Like the exercise-
thallium scan, the dipyridamole-thallium scan
reflects characteristics of coronary flow rather than
ischemia, and therefore the test is limited to deter-
mining the quantity of myocardium at risk from
reduced perfusion distal to a coronary artery ste-
nosis and not the risk of developing myocardial
ischemia at a critical heart rate.
d. RADIONUCLIDE ANGIOGRAPHY
Radionuclide angiography provides useful in-
formation about global and regional left ventricu-
lar systolic function. The ejection fraction values
obtained are reproducible and compare favorably
with the results of contrast ventriculography. The
development of exercise-induced regional wall
motion abnormalities and an abnormal ejection
fraction response during exercise radionuclide an-
giography provide sensitivity of 85 to 90 per cent
for the diagnosis of coronary artery disease, which
is higher than that with treadmill testing alone.
231
. ECHOCARDIOGRAPHY
The major use of two-dimensional echocardi-
ography in patients with coronary artery disease is
to evaluate resting regional and global left ventric-
ular systolic function, although additional infor-
mation, such as assessment of cardiac valves, peri-
cardium, ventricular wall thickness, and the
presence of left ventricular aneurysm and throm-
bus, can be obtained concurrently. Exercise two-
dimensional echocardiography has been shown to
be useful in selected patients with coronary artery
disease and compares favorably with radionuclide
techniques.
f. CORONARY ANGIOGRAPHY
Patients with a positive preoperative nuclear
cardiology imaging test (see sections N.F.4. b., c,
d.) that demonstrates a large reversible defect
should undergo coronary catheterization if surgical
intervention is considered a possibility. If, for any
reason, there is a strong suspicion that the patient
is indeed having significant angina, even though
exercise testing is negative or equivocal, coronary
angiography is indicated. Consideration should al-
ways be given to coronary angiography in the pa-
tient with proven previous myocardial infarction,
especially if the patient currently has angina.
If significant coronary artery disease is present,
the patient needs coronary artery by-pass grafting
before or at the time of pulmonary resection. For
lesser degrees of coronary artery disease, pulmo-
nary resection for carcinoma of the lung should be
done after appropriate medical therapy for coro-
nary insufficiency has been initiated. If the patient
needs coronary artery by-pass grafting, and a lim-
ited resection can encompass the cancer, both pro-
cedures can be done under the same anesthetic, but
the coronary artery by-pass grafting should be
done before pulmonary resection.
237 238
After by-
pass, if the patient is stable with good myocardial
Preoperative Cardiopulmonary Evaluation
function and not bleeding, a pulmonary wedge re-
section can be done. For patients who require cor-
onary artery by-pass grafting and have pulmonary
lesions that require segmentectomy, lobectomy, or
pneumonectomy, there is a good possibility that
the prolonged nature of the pulmonary procedure
will increase the operative mortality (and therefore
should not be done), although a small number
of successful combined procedures have been
reported.
238
23y
In one series of 21 patients, the
pulmonary mass was discovered on the preopera-
tive chest X-ray (for cardiac surgery), and there-
fore before the occurrence of any symptoms, and
by definition constituted a fortuitously early diag-
nosis. Not surprisingly, resection at the time of
cardiac surgery (17 wedge resections, four lobec-
tomies) resulted in a 95 per cent 5-year survival.
240
In cases that require large resections in very com-
prised patients, coronary artery by-pass grafting
should be done first, and pulmonary resection
should be delayed until the patient has gained
weight and muscle mass (usually 4 to 6 weeks).
The risk of general anesthesia for a noncardiac
operation in the patient with previous coronary
artery by-pass grafting is similar to that in patients
without proven coronary artery disease.
241
242
Al-
though it is not possible to estimate the true effects
of delay in pulmonary resection, in terms of tumor
spread in a possibly immunocompromised patient
(especially after general anesthesia
243
), it seems
reasonable that in the latter group (those requiring
by-pass grafting and major pulmonary resection)
the operative risk of combined procedures proba-
bly exceeds the risk of tumor spread.
5. Cardiovascular Sequelae of
Therapeutic Thoracic Radiation
Heart disease resulting from therapeutic radia-
tion is now well recognized. Once thought to be
radioresistant, the heart, as well as the spinal cord,
appears to be the dose-limiting organ for thoracic
irradiation.
244
The pericardium, myocardium, en-
docardium, valves, conduction system, and coro-
nary arteries all may be affected by radiation. The
common pathogenetic factor involves microcircu-
latory injury. How frequently radiation damages
the heart depends significantly on the total dose,
how it is given, how long the patient is observed,
and the intensity with which cardiac abnormalities
are sought (Table 5-16).
Clinically significant radiation heart disease
most frequently takes the form of pericarditis (in-
cluding tamponade and constriction). Radiation-
related myocardial dysfunction is typically mild or
subclinical but can include cardiomyopathic heart
failure, particularly of the right ventricule. Valvu-
lar thickening and deformity are common findings
Table 5-16. FACTORS INFLUENCING
INCIDENCE OF RADIATION
HEART DISEASE
Total number rad delivered (must be > 3000 rads)
Number rad per treatment (inversely proportional)
Time span over which radiation is given (inversely
proportional)
Techniques used
Anterior port versus anterior posterior ports
Use of apical and subcarinal shields
(?) Presence and bulk of mediastinal tumor
(?) Age of patient (older than 40 years)
Length of follow-up
in irradiated hearts that come to postmortem ex-
amination. The aortic valve is involved most often,
with the mitral valve next. Multiple causes for
coronary artery disease in cancer patients exist,
including age-related atherosclerosis (most com-
mon), compression or infiltration of an artery by
tumor, embolization of tumor fragments into an
artery, a hypercoagulable state, and chemotherapy
(principally fluorouracil), as well as radiotherapy.
Therefore, it is understandable that coronary
events occur in young patients previously exposed
to mediastinal radiation who have no other risk
factor.
The most devasting noncardiac but rare sequela
or mediastinal radiation is myelitis. Characteristi-
cally beginning 6 months to 2 years after radiation,
the process is generally irreversible, and the end
result is functional transection of the cord. Medias-
tinal radiation may cause acute pneumonitis as
well as a more chronic form of pulmonary fibrosis.
Significant mediastinal fibrosis must also be con-
sidered in the patient with dyspnea after radiation.
6. Anesthetic Implications of Cancer
Chemotherapy
Cancer chemotherapeutic agents act through a
variety of mechanisms. In general, these agents are
more effective against small rather than large tu-
mor burdens; kill a fixed percentage of tumor cells,
not a fixed number; affect rapidly dividing normal
as well as neoplastic cells, especially those of gas-
trointestinal or surface epithelium and bone mar-
row; and are more effective against dividing rather
than testing cells. After repeated exposure to anti-
neoplastic agents, tumor cells may become resis-
tant, and combinations of drugs with different
mechanisms of action must be devised.
The antineoplastic drugs are classified into sev-
eral groups on the basis of the point in the cell
cycle at which they act, their chemical nature, and
their biologic source. Table 5-17 summarizes the
major agents included in each category (common
Table 5-17. COMMON CHEMOTHERAPEUTIC AGENTS
Type of Drug Generic Name Trade Name Chemical Name
Alkylating agents* Nitrogen mustard* Mustargen Mechlorethamine
Melphalan (L-PAM) Alkeran 1-phenylalanine mustard
Cytoxan* Cytoxan Cyclophosphamide
Chlorambucil Leukeran Chlorambucil
Myleran Myleran Busulfan
Thiotepa Triethylenethiophosphoramide
Amsacrine M-amsa
Antimetabolites Methotrexate* Methotrexate Amethopteran
5-FU 5-fluorouracil
5-FUDR 5-floxuridine
Ara-C Cytosar Cytosine arabinoside
6-TG 6-thioguanine
6-MP 6-mercaptopurine
Mitotic inhibitors (vinca Vincristine* Oncovin
alkyloids) Vinblastine Velban Vinca leukoblastine
Vindesine Eldisine
Antibiotics Adriamycin* Adriamycin Doxorubicin hydrochloride
Daunomycin Daunomycin
Bleomycin* Blenoxane Bleomycin
Mithramycin Mithracin Mithramycin
Mitomycin C Mutamycin Mitomycin C
Actinomycin D Cosmegen Dactinomycin
Streptozocin Streptozocin
Nitrosoureas and Carmustine BiCNU Bis-chloroethylnitrosurea
miscellaneous agents Lomustine* CeeNu Cyclohexylchloroethylnitrosourea
Methyl-CeeNU
Dacarbazine DTIC-Dome Dimethyltrianzeno-imidazole carboxdamide
Procarbazine* Matulane
L-asparaginase
C/i-platinum* Cisplatin Cis-dichlorodiamminoplatinum
Hydroxyurea Hydrea Hydroxyurea
Etoposide* VePesid
''Drugs used against cancers relevant to thoracic surgery.
name, trade name, chemical name). Table 5-18
shows the commonly used dosages of chemother-
apeutic agents against SCLC.
245
The alkylating agents covalently cross-link
strands of deoxyribonucleic acid (DNA), thus pre-
venting replication and subsequent transcription
into ribonucleic acid (RNA). In contrast, the nitro-
soureas with alkylating properties inhibit both
DNA and RNA synthesis; they also demonstrate
significant lipid solubility. The antibiotics, doxo-
rubicin and daunorubicin, bind to DNA by inter-
calation, inhibit both DNA and RNA synthesis,
and cause chromosome breaks. The antibiotics
cause some of the most diverse cell toxicities,
making them particularly notable for anesthesiolo-
gists. The mitotic inhibitors are plant alkaloids that
cause metaphase arrest by binding to microtubular
protein involved in the formation of the mitotic
spindle. The antimetabolites act by interfering with
the synthesis of new nucleic acids. Because this
synthesis occurs only in actively dividing cells,
antimetabolites are considered to be cycle specific.
The enzyme L-asparaginase is unique in that it
depletes tumor cells of the amino acid asparagine.
which is essential for their propagation. The clas-
sic chemotherapy complications are bleomycin
pulmonary toxicity, doxorubicin cardiac toxicity,
vincristine neurotoxicity, and the effects of the al-
kylating agents on neuromuscular function.
Many different antineoplastic drugs have been
associated with pulmonary toxicity. They are sum-
marized in Table 5-19. No matter the particular
agent, antineoplastic drug-induced pulmonary tox-
icity presents in a similar manner. In general, most
of the patients develop a pneumonitis that can ul-
timately lead to pulmonary fibrosis.
Bleomycin's action is similar to the effect of
radiation, and bleomycin and radiation may act
synergistically during simultaneous therapy.
246
Pulmonary toxicity is the most life-threatening,
drug-limiting effect, reported in 15 to 25 per cent
of patients.
247248
The predisposing factors for bleo-
mycin pulmonary toxicity appear to be patient age
greater than 70 years, a total dose greater than 450
Table 5-18. SMALL-CELL LUNG CANCER:
COMMONLY USED REGIMENTS
OF COMBINATION
CHEMOTHERAPY*
Drug Regimen
Cyclophosphamide
Lomustine
Vincristine

Etoposide
Cyclophosphamide
Doxorubicin
Vincristine "
Etoposide
'
Doxorubicin
Cyclophosphamide
1 g/m
2
intravenously every 4 weeks,
day 1
70 mg/m
2
by mouth every 4 weeks,
day 1
1.3 mg/m
2
intravenously (maximum =
2.0 mg) every 4 weeks, day 1, except
first 4 weeks, when administered
weekly
70 mg/m
2
by mouth, on days 3,4, 5 and
6, every 4 weeks
1500 mg/m
2
intravenously every 3
weeks
40 mg/m
2
intravenously every 3 weeks
2 mg/m
2
50 mg/m
2
intravenously 5 days every 3
weeks
45 mg/m
2
1 g/m
2
From Hanson HH, Kristijansen PEG: Changing concepts in
the management of patients with lung cancer. Med J Aus
149:77-84,1988. Used with permission.
patients, with death occurring within 3 weeks of
the onset of the symptoms.
252
The left ventricular
failure that occurs with doxorubicin is refractory
to inotropic drugs. EKGabnormalities are also part
of doxorubicin toxicity, but they resolve 1 to 2
months after cessation of therapy.
253
Once doxo-
rubicin administration is discontinued, left ventric-
ular function generally continues to show deterio-
ration in patients older than 40 years, whereas it
may improve in younger patients.
254
Table 5-21 lists the drugs that have been asso-
ciated with neurotoxicity. The mitotic inhibitors,
vinblastine and especially vincristine, pose the
greatest challenge to the anesthetic management of
patients treated with them. Essential nervous sys-
tem effects are rare (decreased consciousness),
whereas peripheral neuropathy, such as induced by
vincristine, is frequent.
Antineoplastic drugs associated with reduced
plasma cholinesterase activity include the alkylat-
ing agents, cyclophosphamide, nitrogen mustard,
and thiotepa, listed in decreasing order of effect on
cholinesterase. Cisplatin, methotrexate, 6-mercap-
topurine, mithramycin, and streptozocin have been
associated with major alterations of renal function.
The hypomagnesemia induced by cisplatin and
mithramycin is of particular interest. It may be
associated with tetany and increased susceptibility
to digoxin toxicity.
III. MEDIASTINAL MASSES

A. Types of Masses
units, radiation therapy to the chest, and pre-exist-
ing lung disease.
249
Signs and symptoms of pul-
monary toxicity are cough, dyspnea, and basal
rales. The disease may manifest itself with mini-
mal radiologic changes and normal resting P
a
0
2
,
or it may progress to severe hypoxemia at rest,
with radiologic changes similar to severe adult res-
piratory distress syndrome. A controversial factor
that may predispose patients to pulmonary toxicity
is the administration of oxygen in high concentra-
tions (see chapter ll).
250
Mitomycin C is also highly toxic and can cause
pulmonary fibrosis and nephrotoxicity.
251
Thus,
patients receiving mitomycin C should have their
pulmonary and renal status fully evaluated in the
preoperative period.
Drugs associated with cardiac toxicity are listed
in Table 5-20. The most severe cardiac toxicity
occurs as a cardiomyopathy, which can be lethal.
Toxicity secondary to doxorubicin includes severe
cardiomyopathy, seen in 1.8 per cent of patients
treated. When cardiomyopathy develops, it has
been shown to be irreversible in 60 per cent of the
..
Table 5-19. DRUGS ASSOCIATED WITH PULMONARY TOXICITY BY CATEGORY
The anatomy of the mediastinum is complex
and contains many structures (see chapter 2) and
cell types (see Fig. 2-23, 2-24, and 2-25). The
most common tumors by location are as follows
(Fig. 5-29): In the anterior mediastinum are thy-
moma, mesenchymal tumors, dermoid cysts, thy-
roid and parathyroid tumors, and lymphoma; in
the middle mediastinum are pericardial cysts,
bronchogenic cysts, and lymphomas; and in the
posterior mediastinum are neurogenic and enter-
ogenous tumors and cysts, aortic aneurysms, and
paravertebral abscess.
30
The most common tumors
by cell type are neurogenic (14 to 24 per cent),
1
cysts (pericardial, bronchogenic, enteric, and non-

Alkylating Agents Antimetabolites Antibiotics Nitrosoureas Miscellaneous
Busulfan
Chlorambucil
Cyclophosphamide
Melphalan
Azathioprine Bleomycin
Cytarabine Mitomycin C
6-mercaptopurine
Methotrexate
BiCNU
Methyl-CeeNU
Procarbazine
Table 5-20 DRUGS ASSOCIATED WITH
CARDIAC TOXICITY
Cisplatinum*
Cyclophosphamide
Daunorubicin
Doxorubicin*
*Drugs used against cancer relevant to thoracic surgery.
specific, totaling 19 to 25 per cent), and teratoder-
moids, thymomas, lymphomas, and others (each
10 to 20 per cent).
253 2 5 5
2 5 6
Approximately 30 to
40 per cent of the nonvascular masses are
malignant.
256
257
B. Signs and Symptoms
Signs and symptoms of mediastinal masses can
be referable to any of several of the many organs
within the mediastinum (heart, great vessels, the
airway, esophagus, vertebral column). The life-
threatening complications are superior vena cava
Table 5-21 DRUGS ASSOCIATED WITH
NEUROTOXICITY
L-Asparaginase
Cisplatinum
5-Fluorouracil
Methotrexate
Nitrogen mustard
Procarbazine
Vinblastine
Vincristine
obstruction (dyspnea, cough, orthopnea, facial
swelling), pulmonary artery obstruction (postural
hypotension), carinal obstruction (respiratory dis-
tress), and cardiac compression. Major, but non-
life-threatening, complications include esophageal
compression, which can cause dysphagia, pressure
on a nerve and erosion of bone, which can cause
severe pain, and compression of the spinal cord,
which may cause paralysis. More minor compli-
cations include hoarseness resulting from entrap-
ment of the recurrent laryngeal nerve, Horner's
syndrome caused by involvement of the high pos-
Mediastinal Masses
ANTERIOR
MEDIASTINUM
Thyroid Adenoma
Parathyroid Adenoma
Thymoma
Mesenchymal
Tumors
Terato
Dermoid Cyst
Lymphoma
MIDDLE
MEDIASTINUM
POSTERIOR
MEDIASTINUM
Aortic Aneurysm
Lymphoma
Esophageal Lesions
Enterogenous Cysts
Neurogenic Tumors:
Pheochromocytoma
Ganglioneuroma
Neurofibroma
Neuroblastoma
Hiatus Hernia
Meningocele
Paravertebral Abscess
Bronchial
Cysts
Figure 5-29 The location of commonly occurring mediastinal tumors. See Figure 2-23 for the division of the mediastinum into
conventional compartments.
terior mediastinal sympathetic nerves, lethargy,
and weight loss. In one large series, the symptoms
consisted of cough (40 per cent), pain (40 per
cent), dyspnea (20 per cent), dysphagia (20 per
cent), and hoarseness (3 per cent). The signs con-
sisted of weight loss (24 per cent), fever (24 per
cent), superior vena cava obstruction (16 per cent),
tracheal deviation (12 per cent), Horner's syn-
drome (7 per cent), spinal cord compression (5 per
cent), cyanosis (3 per cent), and mediastinal wid-
ening (3 per cent). Symptoms were absent in 22
per cent, signs were absent in 20 per cent, and
both signs and symptoms were absent in 20 per
cent.
257
Usually, mediastinal masses are suggested
by a combination of clinical signs and symptoms
and standard and high-voltage anteroposterior and
lateral chest roentgenograms. Contour distortions
of the normal interfaces between the lung and the
mediastinum may occur even with small medias-
tinal masses, which, with enlargement, may pro-
ject into the adjacent hemithorax. Obvious points
of diagnostic differentiation include vascular ver-
sus nonvascular mass, mass that can be surgically
cured (thymomas, teratomas, neurogenic tumors)
versus those that cannot be surgically cured (e.g.,
lymphoma), metastatic versus nonmetastatic, and
those that require specific tissue typing for correct
choice of chemotherapy and irradiation alone
(lymphoma) versus those that require chemother-
apy, irradiation, plus resection (germ cell tu-
mor).
2583
C. Diagnostic Workup Logic for
Mediastinal Masses
Once a mediastinal mass is suggested, CT is the
single best test to perform in a stable patient to
identify the nature and the location of the mass
(Fig. 5-30). In fact, CT has been shown to be
more useful in evaluating the mediastinum than
any other region in the thorax. Mediastinal disease
may easily be detected by CT even in the presence
of a normal chest radiograph. CT can determine
more accurately than other diagnostic procedures
whether the mass is primarily vascular, fatty, cys-
tic, or soft tissue in nature. For example, fluoros-
copy, conventional tomography, barium esopha-
gogram, radionuclide angiography and thyroid
imaging, angiography, bronchoscopy, scalene
node biopsy, and mediastinoscopy will all be nor-
mal in patients with mediastinal lipomatosis.
259,260
Echocardiography may be helpful in the recogni-
tion of the anatomic and functional aspects of me-
diastinal masses.
261
Once a soft tissue mass is identified in the me-
diastinum, the next logical question is whether the
process is benign or malignant. Approximately 30
to 40 per cent of mediastinal masses are malig-
nant.
257
Unfortunately, CT cannot distinguish be-
nign from malignant masses based on any intrinsic
appearance of the mass itself. However, CT may
demonstrate invasion of the pulmonary arteries,
airway, pericardium, or myocardium or may dem-
Preoperative Evaluation of Mediastinal Masses

History, Physical Examination, Chest X-rays
I
Diagnosis Suspected
I
Figure 5-30 Preoperative evalua-
tion logic of mediastinal masses. (CT
= computed tomography.) See text
for full explanation.
onstrate pleural or parenchymal metastases. In
each of these instances, the malignant nature of the
mediastinal mass is more apparent. More com-
monly, the diagnosis of benign versus malignant
will depend on a diagnosis of cell type, and this
must come from more invasive procedures such as
mediastinoscopy, mediastinotomy, needle biopsy,
bronchoscopy, or esophagoscopy. Biopsy may be
guided by fluoroscope, CT, or ultrasonography.
262
Lymphomas or metastases require only diagnos-
tic biopsy because irradiation or chemotherapy (or
both) is the treatment of choice. The accurate di-
agnosis of thoracic lymphoma or new thoracic le-
sions in patients with lymphomas usually requires
that enough tissue be taken for immunophenotyp-
ing. Providing adequate tissue samples and treat-
ing new lesions that are not lymphomas often
requires major thoracomediastinotomies for im-
munophenotyping and more than one operation.
263
Many germ cell tumors, even though they display
elevated levels of -fetoprotein or beta-human
chorionic gonadotropin, require a confirming bi-
opsy because they are treated initially with multi-
modality therapy followed by resection.
25H
If the
mass is not suspected of being a metastasis, lym-
phoma, or germ cell tumor, then plans to resect
the lesion primarily can proceed without preoper-
ative biopsy if the patient is otherwise in good
health. Thymomas, benign teratomas, cysts, and
neurogenic tumors are types of lesions that can be
resected without a preoperative biopsy. Resection
not only resolves any symptoms caused by a
space-occupying lesion but also determines the ex-
act histologic diagnosis.
258
If the lesion may potentially involve the aorta
or any of its branches, this is best demonstrated by
intra-arterial digital or conventional angiographic
examinations.
264
The evaluation of acute traumatic
injuries of the thorax is still best accomplished
using conventional arteriography and angiography
(especially if the diagnosis of dissection of the
aorta or its major branches is suspected). The as-
sessment of physiologic function of the patient
Table 5-22 CAUSES OF PLEURAL EFFUSION
Transudates
A. Increased hydrostatic pressure
1. Cardiac failure
2. Constrictive pericarditis
3. Obstruction of superior vena cava
B. Decreased oncotic pressure
I. Hypoalbuminemia (liver and renal
disease)
C. Miscellaneous
1. Peritoneal dialysis
2. Acute atelectasis
3. Subclavian catheter misplacement
with a mediastinal mass is similar to that for a
patient with lung cancer.
IV. PLEURAL DISEASE/EFFUSION
A. Anatomy and Pathophysiology
The pleura is the double serous membrane that
separates the lung from the chest wall and medias-
tinal structures. The surface of each pleural mem-
brane consists of a uniform layer of mesothelial
cells supported on a connective tissue framework,
which is well supplied with capillaries and lymph
vessels. The parietal pleura also contains pain-sen-
sitive nerve fibers supplied by the intercostal and
phrenic nerves.
In health, there is a continuous movement of
water, electrolytes, and a little protein into the
pleural space from the parietal surface; the hypo-
tonic pleural fluid thus formed is reabsorbed across
the visceral membrane, under the influence of hy-
drostatic and osmotic gradients (Starling's law).
The parietal pleura is perfused by systemic arte-
ries, whereas the visceral layer receives blood
from the low-pressure pulmonary circulation. In a
normal individual, the net volume of pleural fluid
may be as little as 1.0 ml. The volume of pleural
fluid will increase when there is a rise in intravas-
cular hydrostatic pressure (e.g., in cardiac failure)
or a fall in plasma oncotic pressure (e.g., in hypo-
proteinemic states). An increase in the permeabil-
ity of the pleural capillaries, caused by inflamma-
tory or neoplastic disease, leads to accumulation
of fluid in the pleural space. Lymphatic obstruc-
tion can also contribute to the accumulation of
pleural fluid. The causes of pleural effusion are
listed in Table 5-22.
B. Symptoms, Signs and Diagnostic
Workup Logic
The normally negative intrapleural pressure is
due to the elastic recoil of the lung pulling inward
Exudates
A. Infections
B. Neoplasms
C. Collagen
D. Intra-abdominal disease
E. Miscellaneous
1. Drug-induced pleural disease
2. Pulmonary infarction
3. Hemothorax
4. Chylothorax or pseudochylothorax

in opposition to the thoracic wall. At the relaxed,
end-expiratory position of the breathing cycle, the
intrapleural pressure is about 5 cm H
2
0 below
atmospheric pressure. Accumulation of liquid or
leakage of air into the pleural space is accompa-
nied by an increase in pleural pressure, a reduction
in lung volume, and a slight expansion of the ip-
silateral hemithorax. This results in a restrictive
ventilatory defect, which is in proportion to the
volume of liquid or gas in the pleural space. VC,
FRC, and TLC are all reduced, and the transfer
factor is also slightly reduced.
The most common symptoms of pleurisy are
localized inspiratory pain, dry cough, slowly in-
creasing breathlessness in larger effusions, and fe-
ver in infectious disease. On physical examination,
there may be a pleural friction rub at the onset,
restriction of chest wall movement, diminished
tactile vocal fremitus, dullness on percussion, and
decreased breath sounds with pleural fluid accu-
mulations of at least 300 ml. A scheme for diag-
nosis and management is outlined in Figure 5-31.
V. PERI CARDI AL DISEASE/
EFFUSI ON
A. Anatomy and Pathophysiology
The pericardium has three functions: mechani-
cal, membranous, and ligamentous. Mechanically,
it protects against acute dilation of the ventricles
and mediates diastolic coupling between ventri-
cles, and it is involved with the pressure-volume
relationship between the ventricles. Membranous
functions include lubrication and protection
against infection; ligamentous function consists of
limiting heart displacement.
265
There are numerous causes of cardiac tampon-
ade (see Table 5-23). In medical patients, neo-
plasms, primarily of the breast and lung, are most
Table 5-23 CAUSES OF CARDIAC
TAMPONADE
1. Malignant disease (primarily breast and lung)
2. Trauma
3. Iatrogenic causes, including radiation therapy,
central venous pressure placement, pacemaker
placement, and cardiac surgery
4. Acute cardiac infarct with anticoagulation
5. Infections
6. Dissecting aortic aneurysm
7. Postpericardiotomy syndrome
8. Chylous effusions
9. Connective tissue diseases, primarily systemic
lupus erythematosus and rheumatoid arthritis
10. Amyloidosis
11. Myxedema
12. Uremia
Table 5-24 HEMODYNAMIC ALTERATIONS IN
CARDIAC TAMPONADE
1. Limited atrial and ventricular diastolic filling
2. Increase and equalization of diastolic pressures
3. Absent or positive y descent and prominent descent
4. Decreased stroke volume
5. Increased heart rate
6. Decreased pulse pressure
7. Increased systemic vascular resistance
8. Decreased cardiac output
commonly responsible for cardiac tamponade.
266
Traumatic causes include direct, penetrating, or
blunt insults to the heart. Hemopericardium also
may be seen after dissecting ascending aortic
aneurysms, myocardial infarction, or anticoagula-
tion.
267
Table 5-24 summarizes the hemodynamic alter-
ations during cardiac tamponade. Limited diastolic
ventricular filling resulting from increased intra-
pericardial pressure causes a fall in cardiac output
and stroke volume. The fall in stroke volume and
subsequently hypotension elicit a generalized sym-
pathetic response, which, in turn, causes tachy-
cardia, increased systemic vascular resistance,
decreased pulse pressure, increased preload, de-
creased diastolic filling time, and increased con-
tractility with subsequent decrease in cardiac per-
fusion.
268
B. Diagnosis and Treatment
Symptoms seen with cardiac tamponade are
usually nonspecific. The patient may complain of
dyspnea, fullness of the head and neck, abdominal
pain, nausea, or a vague oppressive feeling in the
chest. The dyspnea is believed to be secondary to
decreased cardiac output and to restriction of lung
volumes by pericardial and pleural effusions.
Signs of cardiac tamponade are summarized in
Table 5-25.
Acute tamponade is well described by Beck's
triad: a small, quiet heart; increased venous pres-
sures, and hypotension. The triad is caused by
pericardial fluid softening heart sounds; resistance
Table 5-25 SIGNS OF CARDIAC TAMPONADE
1. Beck's triad: small, quiet heart; increased venous
pressures; and hypotension
2. Pulsus paradoxus
3. Equalization of diastolic pressures
4. Large and globular cardiac shadow radiographically
5. Nonspecific electrocardiogram changes
6. Right-heart compression during diastole by
echocardiography
2 U4 Preoperative Cardiopulmonary Evaluation
to right-heart filling, resulting in increased jugular
venous distention and neck vein dilation; and a
reduction in cardiac output with subsequent hypo-
tension, decreased pulse pressure, pulsus para-
doxus, and tachycardia. Increased central venous
pressure (CVP) may not be seen in the patient with
extreme volume depletion.
Cardiac catheterization shows equalization of
CVP, diastolic right ventricular pressure, pulmo-
nary capillary wedge pressure, left atrial pressure,
and LVEDP. Radiographically, one sees an en-
larged and globular cardiac shadow with a convex
or straight left-heart border.
Definitive treatment of cardiac tamponade is
drainage. It may be accomplished through pericar-
diocentesis or surgical decompression. The former
is the treatment of choice for nontraumatic cardiac
tamponade. Surgical drainage is usually chosen for
traumatic tamponade, after cardiac surgery, or if
fluid reaccumulates after pericardiocentesis. The
surgical approaches include the subxiphoid ap-
proach, which is easier to perform and more easily
tolerated with local anesthesia but affords a limited
exposure. A left thoracotomy incision provides ex-
cellent exposure and is indicated if a larger field is
required, such as after penetrating chest injury.
Initial noninvasive management of cardiac tam-
ponade includes fluid administration and pharma-
cologic support. It is important to maintain ade-
quate preload to optimize stroke volume so that
administration of a fluid load is necessary; ideally
colloid, plasma, or blood is used. The volume
should be titrated to the response of the initial
load. CVP should be maintained above 15 to 20
cm H
2
0. The increases in filling pressures will
oppose pericardial pressure.
269
Positive inotropes may be considered. Isoproter-
enol may be ideal because it sustains a tachycardia
(because stroke volume is usually limited, heart
rate must remain high to preserve cardiac output),
increases stroke volume, and decreases systemic
vascular resistance.
It is important to select anesthetic agents that do
not depress the myocardium, severely decrease af-
terload, or cause bradycardia. With these limita-
tions in mind, low-dose ketamine is an attractive
choice. Controlled ventilation significantly de-
creases preload and decreases cardiac output from
the right ventricle. Positive end-expiratory pres-
sure further decreases preload and cardiac output.
If it is not possible to perform pericardiotomy
without general anesthesia, it is suggested that the
patient be allowed to breathe spontaneously until
the chest is opened and drainage of the pericardial
space is imminent. If spontaneous ventilation is
not possible, ventilation with high rates and low
tidal volumes should be considered.
Acute pulmonary edema has been reported after
relief of pericardial tamponade.
270
The acute in-
crease in venous return may overload the left ven-
tricle in the face of increased systemic vascular
resistance secondary to elevated catecholamine
levels. This mismatch of preload and afterload is
exacerbated in patients with decreased left ventric-
ular compliance. Slow drainage of the effusion is
recommended. Right ventricular filling pressures
should be monitored.
REFERENCES
1. Shields TW: Carcinoma of the lung. In General Thoracic
Surgery. Philadelphia, Lea & Febiger, 1983, chapter 54,
pp 729-769.
2. Spiro SG: The diagnosis and staging of lung cancer. In
Smyth JF (ed): The Management of Lung Cancer. Balti-
more, Edward Arnold, Ltd., 1984, chapter 3, pp 36-52.
3. Le Roux BT: Bronchial Carcinoma. London, E. & F.
Livingstone, Ltd., 1968.
4. Jones DP: Diagnostic work-up of chest disease. Sympo-
sium on noncardiac thoracic surgery. Surg Clin North Am
60:743-755, 1980.
5. Ferguson MK: Diagnosing and staging of non-small cell
lung cancer. Hemalol/Oncol Clin North Am 4(6): 1053-
1068. 1990.
6. Garfinkel L, Stellman SD: Smoking and lung cancer in
women: Findings in a prospective study. Cancer Res
48:6951-6955, 1988.
7. Spiro S: Lung cancer: Presentation and treatment. Med
Int 3798-3805, 1991.
8. Petty TL: Pulmonary medicine. JAMA 263:2677-2678.
1990.
9. Tockman MS. Antonisen NR. Wright EC. Donathan MG:
Airways obstruction and the risk of lung cancer. Ann
Intern Med 106:512-518, 1987.
10. Fielding JE, Phenow KJ: Health effects of involuntary
smoking. Engl J Med 319:1450-1460, 1988.
1 I. Capewell S, Sankaran R, Lamb D, Mclntyre M, Sudlow
MF: Lung cancer in lifelong non-smokers. Thorax
46:565-568, 1991.
12. Janerich DT, Thompson WD, Varela LR, Greenwald P,
Chorost S, Tucci C, Zaman MB, Melamed MR, Kiely M,
McKneally MF: Lung cancer and exposure to tobacco
smoke in the household. Engl J Med 323:632-636.
1990.
13. U.S. Office on Smoking and Health: Smoking and health:
A report to the Surgeon General of the Public Health
Service. Washington, DC, U.S. Department of Health and
Human Services, 1979.
14. U.S. Department of Health and Human Services: The
Health Consequences o\' Smoking for Women: A Report
of the United States Surgeon General. Washington DC.
U.S. Department of Health and Human Services, 1980.
15. Peto R, Doll R: The control of lung cancer. Scientist
24:26-30, 1985.
16. Mathe G, Reizenstein P: Extra-pulmonary tumors caused
by smoking. Biomed Pharmacother 42:87-88. 1988.
17. Remick SC, Hafez GR, Carbone PP: Extrapulmonary
small-cell carcinoma: A review of the literature with em-
phasis on therapy and outcome. Medicine 66:457-471,
1987.
18. Hill GB: Smoking and lung cancer. Arch Intern Med
148:2538-2539, 1988.
19. Petruzzelli S. Hietanen E, Bartsch H, Camus AM, Mussi
A, Angeletti CA, Sracei R, Giuntini C: Pulmonary lipid
peroxidation in cigarette smokers and lung cancer pa-
tients. Chest 98:93-935. 1990.
CHAPTER
6
Preoperative Respiratory
Preparation
I. Introduction
II. Correlation of Respiratory
Complications with Degree of Pre-
Existing Lung Disease
III. Correlation of Respiratory
Complications With Site of
Operation
IV. Proof That Preoperative Pulmonary
Preparation Decreases Incidence of
Postoperative Respiratory
Complications
V. Preoperative Respiratory
Preparation Maneuvers
A. Discontinue Smoking
B. Dilating the Airways
1. Overall Bronchodilating Plan
2.
2
-Agonists
3. Anticholinergic Drugs
4. Inhaled Steroids
5. Inhaled Sodium Cromoglycate
6. Methylxanthines
7. Oral and Parenteral Steroids
C. Loosening the Secretions
D. Removing the Secretions
E. Ancillary Measures/Issues
1. General Measures (Treatment of
Systemic Disease)
2. Treatment of Cor Pulmonale
3. Preoperative Digitalization
F. Measures to Increase Motivation
and Education and to Facilitate
Postoperative Respiratory Care
VI. Mechanism of Preoperative
Respiratory Preparation Benefit
VII. Premedication
212 Preoperative Respiratory Preparation
I. INTRODUCTION
Thoracic surgical patients are at high risk for
(he development of postoperative pulmonary com-
plications. In most of the literature, "postoperative
complications" refers to the development of ate-
lectasis and/or pneumonia.' The incidence of
pneumonia usually parallels the incidence of ate-
lectasis, and the onset of pneumonia lags behind
the onset of atelectasis because atelectasis provides
the ventilatory and mucociliary stasis condition
necessary for the development and culture of
pneumonia.
2,3
There are three major reasons why thoracic sur-
gery promotes postoperative pulmonary complica-
tions; these reasons originate in the preoperative,
intraoperative, and postoperative period (Fig. 6-
l ). First, the incidence of postoperative respiratory
complications following any surgery is positively
correlated with the degree of preoperative respira-
tory dysfunction, and most thoracic surgical pa-
tients come to surgery with some degree of
preoperative lung dysfunction. Preoperative pul-
monary function testing will identify the patients
at high risk due to poor preoperative lung function.
Second, the performance of thoracic surgery can
impair lung function in any patient. During sur-
gery, nondependent-lung function may be im-
paired by resection of functional lung and/or by
trauma to the remaining nondependent lung (as a
result of various nondependent-lung manipula-
tions) and/or by overexpansion of the remaining
nondependent lung, and dependent-lung function
may be impaired as a result of the development of
atelectasis and edema formation. Third, thoracot-
omy incisions are painful and cause patients to
resist deep breathing and coughing in the postop-
erative period, leading to retained secretions, ate-
lectasis, and pneumonia. The second factor (im-
paired lung function due to the performance of
thoracic surgery) can be minimized by appropriate
intraoperative management (such as one-lung ven-
tilation, positive end-expiratory pressure, continu-
ous positive airway pressure) (see chapter 8). The
third factor can be minimized by appropriate post-
operative pain management (e.g., epidural narcot-
ics) (see chapter 20). The impact of the first factor
(presence of preoperative respiratory dysfunction)
can be significantly reduced by preoperative pro-
phylactic respiratory preparation measures.
This chapter first documents the correlation be-
tween degree of preexisting respiratory disease and
incidence of postoperative respiratory complica-
tions. Next, the correlation between thoracic sur-
gery and the increased incidence of postoperative
respiratory complications is discussed. Proof is
then offered that preoperative respiratory prepara-
tion decreases the incidence of postoperative res-
piratory complications, and the main body of the
chapter details a full preoperative respiratory prep-
Thoracic Surgery Impairs Postoperative Lung Function
Patients Who Have
Thoracic Surgery Usually
Have Pre-existing Lung Disease
Nondependent Lung:
Functional Tissue
Resected or Traumatized
Dependent Lung:
Compressed, Edematous
Painful Incision
Fail to Deep
Breathe and Cough
Figure 6-1 There are preoperative, intraoperative, and postoperative reasons why thoracic surgery impairs postoperative lung
function.
aration regimen. The mechanism by which preop-
erative respiratory preparation decreases the inci-
dence of postoperative respiratory complications is
considered. Finally, premedication, as part of a
preoperative respiratory preparation plan, is dis-
cussed.
II. CORRELATION OF
RESPIRATORY COMPLICATIONS
WITH DEGREE OF PRE-EXISTING
LUNG DISEASE
The relationship between lack of preoperative
pulmonary reserve and postoperative respiratory
morbidity and mortality is very well recognized
and may be very dramatic. Compared with non-
smokers, smokers have decreased pulse oximetry
values, increased post anesthesia care unit (PACU)
stay, and a sixfold increase in the incidence of
postoperative pulmonary complications after major
operative procedures.
4-8
Three major mechanisms
appear to be responsible for the adverse smok-
ing/surgery interaction, and they are small airway
disease (spasm, collapse), hypersecretion of mu-
cus, and impairment of tracheobronchial tree clear-
ance. In patients with chronic lung disease, com-
pared with normal healthy patients, there is a
20-fold increase in the incidence of postoperative
pulmonary complications.
9
Thus, it is not surpris-
ing to find widespread agreement that the risk of
postoperative pulmonary complications progres-
sively increases as preoperative pulmonary func-
tion progressively decreases.
Preoperative pulmonary function is best quanti-
tated by preoperative pulmonary function tests,
and patients with a vital capacity, maximum
breathing capacity, FEV, or FEF
257f
_
75C/i
of less than
50 per cent of predicted capacity and/or grossly
hypercapnic patients are at very high risk (see
chapter 5 references and Handlin and Baker
7
for
extensive substantiation of this contention). How-
ever, even if pulmonary function tests are not
available, it should be remembered that gross ob-
servations such as the production of a great deal
of sputum by the patient (more than 2 ounces per
day),
10
minimal exercise tolerance capability, se-
vere cardiac disease, obesity, sepsis, and very ad-
vanced age also indicate great risk for developing
postoperative pulmonary complications."
12
III. CORRELATION OF
RESPIRATORY COMPLICATIONS
WITH SITE OF OPERATION
The correlation between postoperative pulmo-
nary complications in patients with and without
respiratory disease and the site and type of opera-
tion has long been known, with the highest inci-
dence of complications following major thoracic
and upper abdominal procedures.
10-14
In a series of
1500 surgical patients with a wide variety of res-
piratory diseases treated over a 30-year period, the
incidence of respiratory complications averaged 63
per cent following thoracic and gastric operations,
15 to 19 per cent following midabdominal opera-
tions, and 9 per cent following lower abdominal
procedures.
15
In a group of 464 patients with
chronic respiratory disease who did not have any
preoperative respiratory preparation, the highest
risk of pulmonary complications involved patients
with thoracotomy or abdominal operations (com-
pared with surgery on other parts of the body).
16
Even when the patient population was specifically
defined as those with chronic obstructive pulmo-
nary disease (COPD) who underwent preoperative
respiratory preparation, the incidence of respira-
tory complications was still highest for thoracic
and abdominal operations compared with surgery
on other parts of the body.
10
Similarly, other pa-
tients with severe chronic obstructive pulmonary
disease who had operations on the thorax or upper
abdomen were found to have twice the mortality
compared with those with a similar amount of
respiratory impairment and who were subjected to
operations on other body regions.
17
The mechanisms by which thoracic surgery (and
abdominal surgery) especially predisposes patients
to postoperative respiratory complications include
resection of, or trauma to, functional lung and the
degree to which the bellows function of the lung
is affected (Fig. 6-1). It is painful for patients who
have incisions in these areas to deep breathe and
to stretch either the chest or abdominal wall (i.e..
the incision); consequently, they fail to cough
(which requires a deep breath), and they retain
secretions (which promotes the development of
atelectasis and infection). When the lateral thora-
cotomy incision is limited (preservation of the la-
tissimus dorsi, splitting of the serratus anterior,
and cutting of only the intercostal muscles without
rib resection) and the bellows function is therefore
better preserved, postoperative pulmonary reserve
may be increased.
18
In addition, it has been
claimed (but not proven) that operative time, blood
loss and postoperative pain, intensive care unit
stay, and morbidity are decreased.
19
IV. PROOF THAT PREOPERATIVE
PULMONARY PREPARATION
DECREASES INCIDENCE OF
POSTOPERATIVE RESPIRATORY
COMPLICATIONS
Considerable evidence has accumulated over the
last three decades to demonstrate that vigorous
2 1 4 Preoperative Respiratory Preparation
preoperative pulmonary preparation can signifi-
cantly reduce postoperative pulmonary complica-
tions.
20
It is useful to monitor pulmonary function
during the preoperative "tune-up" because those
who improve their pulmonary function have a bet-
ter prognosis than those who do not.
10
Several
decades ago, it was demonstrated that simple pre-
operative physical therapy instruction, as opposed
to the same instruction given postoperatively or
not at all, decreased the incidence of atelectasis
from 42 per cent (no instruction), to 27 per cent
(postoperative instruction), to 12 per cent (preop-
erative instruction).
21
About the same time, it was
demonstrated that the use of nebulized isoproter-
enol three times daily before and after surgery, in
conjunction with postural drainage and chest
physiotherapy, reduced the incidence of postoper-
ative atelectasis from 43 to 9 per cent.
22
In a much more recent study, normal patients
were prospectively randomized into a supervised
preoperative breathing exercise treatment group
(diaphragmatic breathing, deep breathing, forced
expiration, coughing, and dead-space rebreathing
using a tube) and a nontreatment group. The inci-
dence of postoperative (upper abdominal surgery)
complications (criteria derived from chest roent-
genograms, arterial blood gas samples, and tem-
perature registration) in the treatment group and in
the control group were 19 per cent and 60 per
cent, respectively.
23
Patients with chronic obstructive pulmonary dis-
ease are the ones most likely to benefit from pre-
operative respiratory preparation. In 1959 it was
impressively shown in a group of 250 surgical
patients with obstructive pulmonary disease that
when an intensive pulmonary preparation was util-
ized, consisting of oral therapy with a bronchodi-
lator drug, intermittent positive-pressure ventila-
tion with nebulized bronchodilator therapy, use of
an expectorant, postural drainage, cough training,
and antibiotic treatment for purulent sputum, only
two patients (1 per cent) developed atelectasis.
24
In
1970 randomly selected "poor-risk" patients were
treated preoperatively and postoperatively with a
very comprehensive respiratory care regimen
(bronchodilators, antibiotics, inhalation of humidi-
fied gas, segmental postural drainage, and chest
physical therapy). When the treated patients were
compared with the nontreated patients, the treated
patients had a pulmonary complication rate of only
22 per cent (all mild), whereas the untreated pa-
tients had a pulmonary complication rate of 60 per
:ent (of which 60 per cent were severe).
25
Simi-
larly, in a retrospective study of patients with
;hronic obstructive pulmonary disease who under-
went various surgical procedures under inhalation
inesthesia, the incidence of postoperative respira-
ory complications was only 24 per cent in those
>iven preoperative pulmonary preparation com-
pared with 43 per cent in a control, nonprepared
group.
16
Since preoperative pulmonary preparation can
significantly reduce postoperative pulmonary com-
plications, especially in patients with lung disease,
it is not surprising to find that a high-risk group of
patients with moderate to severe respiratory dys-
function who required thoracic surgery also greatly
benefited from preoperative pulmonary prepara-
tion.
26
This high-risk group was managed with vig-
orous prophylactic measures and suffered less pul-
monary complications than did a more normal
respiratory function group that had not been ex-
posed to the same preoperative regimen. Similarly,
in another study, the differences in pulmonary
complication rate were insignificant between un-
treated young "good-risk" and treated "poor-
risk" patients.
25
Thus, the frequency and severity
of pulmonary complications may not increase in
going from nontreated young "good-risk" patients
to treated "poor-risk" patients (implying that the
preoperative treatments corrected, at least in part,
the preoperative respiratory dysfunction), but the
respiratory complication rate definitely increases
in going from treated "poor-risk" patients to un-
treated "poor-risk" patients.
16
24
"
26
More recently (1979), a very careful, meticulous
prospective report described the effect of a stan-
dardized preoperative pulmonary regimen on pre-
operative pulmonary function tests in a series of
157 patients with chronic obstructive pulmonary
disease.
10
The pulmonary regimen consisted of 48
to 72 hours of aerosolized isoproterenol in saline
given four times a day, oral therapy with theoph-
ylline, guaiacol glyceryl ether, hydration, and
chest physiotherapy. In addition, most of the pa-
tients discontinued smoking. The preoperative
respiratory preparation regimen improved preop-
erative pulmonary function tests but not in a pre-
dictable or consistent way. The rate of complica-
tions was highest in those patients who had upper
or long abdominal incisions or thoracotomy, al-
though the incidence of respiratory complications
was significantly reduced as a result of the preop-
erative preparation. The authors were unable to
determine which patients would develop signifi-
cant pulmonary complications not requiring me-
chanical ventilation, but those requiring respiratory
support were predictable on the basis of the sever-
ity of their pulmonary functional impairment and
minimal response to the pulmonary preparation
used (i.e., those who had irreversible disease).
V. PREOPERATIVE RESPIRATORY
PREPARATION MANEUVERS
The preceding data indicate that patients
undergoing thoracic surgery are particularly sus-
ceptible to postoperative respiratory complica-
tions
4
9-17
and that prophylactic measures do de-
crease postoperative respiratory complications.
10
I7,
21, 22. 24-26 Consequently, preoperative evaluation
should be followed by preoperative preparation ef-
forts directed toward optimally managing any pre-
existing pulmonary disease.
27
In general, a full pre-
operative respiratory preparation regimen involves
a five-pronged attack on airway disease. The five
elements of the preoperative regimen are stopping
smoking, dilating airways, loosening and remov-
ing secretions, and taking measures to increase
motivation and education and to facilitate postop-
erative care (Table 6-1 and Fig. 6-2).
The five treatment modalities are instituted and
proceed in parallel fashion. Before discussing each
element separately, it is important to point out that
the desirable results of these maneuvers should be,
for the purpose of understanding their interaction,
achieved in a sequential manner (Fig. 6-2). The
logic behind this concept is as follows. First, stop-
ping smoking eliminates the stimulation for the
production of airway secretions and bronchocon-
striction. Next, the airways should be dilated to
facilitate secretion removal. Similarly, thick, tena-
cious, and adherent secretions must be loosened in
order to be removed. Once the airways are dilated
and the secretions loosened, it makes sense to use
physical maneuvers to remove the secretions. Fi-
nally, the patients should assist, as much as possi-
ble, in their preoperative preparation and postop-
erative respiratory care. The studies cited later
indicate that using the maneuvers in this sequence
(dilating the airways, loosening the secretions, re-
moving the secretions) allows the maneuvers to
complement one another in improving mucociliary
Table 6-1 PREOPERATIVE RESPIRATORY
CARE REGIMEN
1. Stop smoking.
2. Dilate the airways.
a. 3
2
-agonists
b. Theophylline
c. Steroids
d. Cromolyn sodium
3. Loosen the secretions.
a. Airway hydration (humidifier/nebulizer)
b. Systemic hydration
c. ? Mucolytic and expectorant drugs
d. Antibiotics
4. Remove the secretions.
a. Postural drainage
b. Coughing
c. Chest physiotherapy (percussion and vibration)
5. Perform ancillary measures.
a. General measures (treatment of systemic disease)
b. Treatment of cor pulmonale
c. Preoperative digitalization
6. Provide increased education, motivation, and facilitation of
postoperative care.
a. Psychologic preparation
b. Incentive spirometry
c. Exposure to secretion removal maneuvers
d. Exercise
e. Weight loss/gain
f. Stabilization of other medical problems
transport function. The following section discusses
each of these five preoperative preparation maneu-
vers in the order and context of this conceptual
approach. Of course, it is recognized that patients
who do not have bronchospasm will not benefit
from bronchodilator treatment and that patients
who do not have secretions will not benefit from
measures to enhance secretion removal.
28
Preoperative Respiratory Preparation Regimen
1. Terminate Stimulus for Bronchoconstriction and Secretions
Stop Smoking
Figure 6-2 A full, aggressive, preoperative respiratory preparation regimen consists of a five-pronged attack: (1) Require the
patient to stop smoking, (2) dilate the airways, (3) loosen secretions, (4) remove secretions, and (5) increase patient participation.
Using these five maneuvers in the numbered sequence allows them to complement one another in improving secretion removal.
2 1 6 Preoperative Respiratory Preparation
A. Discontinue Smoking
An improvement in mucociliary transport and
small-airway function and a decrease in airway
secretions, reactivity, and markers of injury occur
over several weeks after cessation of smoking.
29-32
Indeed, in addition to the usual clinical syndrome
of chronic bronchitis (chronic cough and phlegm),
smoking may cause wheezing by increasing the
level of nonspecific airways responsiveness
32
"
and/or by narrowing of airways secondary to in-
flammation and structural changes.
34
Numerous
past studies have clearly shown that a period of at
least 6 to 8 weeks of abstinence is necessary to
effect substantial improvement in smoking-in-
duced small-airways disease,
3536
hypersecretion of
mucus," reduced closing volume,
38
and impaired
tracheobronchial clearance.
39
Consequently, it is
not surprising to find that preoperative cessation of
smoking for more than 4 to 8 weeks is associated
with a decrease in the incidence of postoperative
respiratory complications.
29
30
Thus, despite the
fact that as few as 20 per cent of smokers will quit
when advised to do so by their physician well in
advance of an elective operation,
40
it is a worth-
while effort.
Although stopping smoking for only 24 hours
will do nothing to actually decrease the amount of
secretions (at least l to 2 weeks are required),
30
41
airway irritability, and incidence of postoperative
respiratory complications, there are still a number
of other important, perhaps intraoperative, benefits
that accrue in the first 1 to 2 days of abstinence.
29

30 4I
The most important benefit results from a
decrease in carboxyhemoglobin levels. Carbon
monoxide avidly combines with hemoglobin, even
at low tensions of carbon monoxide, because of
the extraordinary chemical affinity of carbon mon-
oxide for the iron atoms in hemoglobin. The affin-
ity of hemoglobin for carbon monoxide is more
than 200 times that for oxygen. Carbon monoxide
readily displaces oxygen from hemoglobin and re-
duces the volume of oxygen carried by the hemo-
globin in the circulation. In addition, the presence
of carboxyhemoglobin in the blood causes a left-
ward shift in the oxygen-hemoglobin dissociation
curve. Thus, oxygen dissociates from hemoglobin
at a lower tension, which, in turn, results in a
lower driving pressure and reduced availability of
oxygen in the tissues and metabolizing cells.
42
43
Cessation of smoking for as short a time as 12
42
to 48
44
hours has been shown to decrease carboxy-
hemoglobin levels (e.g., after 9 and 12 hours of
abstinence from smoking one to two packs per
day, carboxyhemoglobin concentrations decrease
from 6 and 7 per cent to 1 and 4 per cent,
respectively),
42
45
(Fig. 6-3). The decrease in car-
boxyhemoglobin increases oxyhemoglobin and
Figure 6-3 Mean ( standard error of the mean) blood
nicotine and carboxyhemoglobin (COHb) concentrations in
cigarette smokers. Subjects smoked cigarettes every half hour
from 8:30 A.M. to 11:00 p.M for a total of 30 cigarettes per
day. (Adapted from Benowitz NL: Pharmacologic aspects of
cigarette smoking and nicotine addiction. Engl J Med
319:1318-1330, 1980. Used with the permission of the pub-
lisher.)
right shifts the oxygen hemoglobin curve (in-
creases availability of oxygen tissues).
42 4 4 4 5
Numerous studies show that the smoking-in-
duced increase in carboxyhemoglobin may signif-
icantly impair cardiovascular performance in pa-
tients with coronary artery disease. In a double-
blind crossover study involving 30 patients with
ischemic heart disease who had carboxyhemoglo-
bin levels of 6 per cent, there was an early onset
of angina during exercise and abnormal left ven-
tricular function as detected by radionuclide ven-
triculography.
46
Similarly, and perhaps of even
more concern, in patients with coronary artery dis-
ease, cigarette smoking can produce an absolute
decrease in regional myocardial blood flow, usu-
ally without angina pain, comparable to exercise-
induced ischemia.
47
Indeed, there is a 12-fold Hoi-
ter monitor documented increase in the duration of
ischemic episodes in smokers (nearly all of which
were silent), even though the clinical and angio-
graphic characteristics of the smokers and non-
smokers were similar.
48
With respect to exposure to environmental car-
bon monoxide, it has been observed that there is a
reduction in exercise time to angina along with a
decrease in the product of systolic blood pressure
and heart rate at the time of the onset of angina in
10 patients with coronary artery disease who were
driven on a freeway in Los Angeles, in whom
carboxyhemoglobin levels reached 4.0 per cent.
49
Even low levels of carboxyhemoglobin (2 to 3 per
cent) exacerbate myocardial ischemia (as reflected
by shortening of the length of time to angina pain
and ST-segment depression) during graded exer-
cise in subjects with coronary artery disease.
50
The
levels of carboxyhemoglobin in this study (2 to 3
per cent) occur in smokers of one half to one pack
of cigarettes per day, passive smokers in an unven-
tilated room, firefighters actively fighting a fire,
and vehicular tunnel works.
51
At low doses of carboxyhemoglobin, similar to
those seen during cigarette smoking, the cardio-
vascular effects appear to be mediated by the cen-
tral nervous system, either through the activation
of chemoreceptor afferent pathways or by direct
effects on the brain stem. The net result is sympa-
thetic neural discharge, with an increase in blood
pressure and heart rate.
45
Smoking may also have
an arrhythmogenic effect, perhaps related to the
increase in sympathoadrenal activation.
52
In a
manner consistent with a half-life of 2 hours, nic-
otine accumulates over 6 to 8 hours of regular
smoking and persists overnight, even as the
smoker sleeps (see Fig. 6-3).
45
Nevertheless, absti-
nence from smoking for 9 hours results in mark-
edly lower level of nicotine in the blood (see Fig.
6-3) and can be expected to decrease nicotine-
induced tachycardia and hypertension
30
4I
and
lower the risk of arrhythmias. Thus, the acute ef-
fects of decreasing both carboxyhemoglobin and
nicotine levels from stopping smoking for several
hours may confer a critical benefit to the marginal
patient.
There a few select patients in whom the risks of
stopping smoking for a day or two may be thought
to outweigh the benefits. First, although stopping
smoking will cause some anxiety in many patients,
cessation of smoking may induce a great deal of
anxiety in some patients (due in part to acute nic-
otine withdrawal in nicotine-addicted patients).
45
53
In patients with significant coronary artery disease
the large increase in anxiety preoperatively may
lead to a critical ischemic event. Second, sudden
cessation in smoking can occasionally induce a
hypersecretory and bronchospastic state. In pa-
tients who have difficulty with secretions, a pre-
operative increase in secretions may lead to air-
ways that are more obstructed preoperatively.
29
Third, there are some experimental data to indicate
that continued smoking may result in a decreased
incidence of postoperative deep vein thrombo-
sis.
30, 4I
Nevertheless, a good argument can be
made that these potential benefits of continuing
smoking during the day or two prior to surgery
can just as easily be accomplished with anxioly-
tics, bronchodilators, and anticoagulants. Thus, the
potential benefits do not provide sufficient reason
to continue smoking.
B. Dilating The Airways
The next step should be to dilate the airways.
Patients who have increased airway responsive-
ness and therefore are candidates for preoperative
bronchodilation are smokers,
54
atopic individuals,
54
patients with airway symptoms of allergies,
54
pa-
tients with COPD
5 5 5 8
and asthmatics.
59 60
Perhaps
the least obvious candidates for preoperative bron-
chodilation among these groups of patients are
those with COPD. When FEV, is measured before
and after the one-time administration of an inhaled
bronchodilator (three puffs of metaproterenol me-
tered-dose inhaler [MDI]), in order to assess the
reversibility of chronic airway obstruction in non-
asthmatic COPD patients, 38 per cent had a signif-
icant response to the bronchodilator (15 per cent
increase in FEV,).
55
In another group of stable
COPD patients taking oral prednisolone (30 mg
every day), the one-time administration of salbu-
tamol, 200 g, from an MDI or 5.0 mg from a
nebulizer caused at least a 15 per cent increase in
FEV, in 56 per cent of patients.
56
The most important bronchodilator drugs (
:
-
agonist, steroids, and anticholinergics) are admin-
istered by inhalation of an aerosol. Therapeutic
aerosol can be generated by several means: MDIs
with fluorocarbon propellant, jet nebulizers pow-
ered by compressed air oxygen or an electric
pump, jet nebulizers coupled with an intermittent
positive-pressure breathing (IPPB) device, aerosol
generated by an ultrasonic nebulizer, or inhaled
fine powder. When correctly administered, all
these forms of delivery are probably equally effec-
tive.
Because MDIs are small, compact, and rela-
tively cheap and they administer a small dose of
medication with each puff, they are the ideal form
of therapy. Therefore, -agonists are most widely
administered from an MDI. Optimal operation of
the inhalation aids requires actuation of the MDI
followed by a slow, deep inhalation followed by
breath holding for several seconds.
MDIs have two major disadvantages. First, the
MDI requires the patient to learn the inhalation
technique. Numerous publications attest to the fact
that many patients with asthma have incorrect in-
halation technique. Even after repeated instruc-
tions, 10 to 20 per cent, particularly the very
young and the elderly, cannot learn the technique.
The most frequent problem is inability to coordi-
nate the actuation of the canister and inhalation by
mouth. The more fundamental problem is that the
patient is unable to control inhalation: when to
start inhalation and how fast to inhale through the
open mouth. Second, a sufficient amount of medi-
cation will not reach the airways when breathing
is rapid and shallow. Efficiency of penetration into
the lung can be improved by interposing spacers
and expansion chambers between the device and
the mouth (Fig. 6-4). Large-volume spacer cham-
bers, which contain the aerosol cloud delivered by
the MDI, are also useful in patients who cannot
coordinate the ordinary MDI, because they can
inhale the drug via the one-way valve after trigger-
ing the aerosol. Spacers ensure some increase in
delivery to peripheral lung and reduce oropharyn-
geal deposition. This is very important in high-
dose inhaled steroids (when spacers should always
be used to reduce oropharyngeal complications
[candidiasis]).
Nebulized bronchodilator solutions are valuable
for treating acute episodes of severe obstruction
when breathing is rapid and shallow and for pa-
tients who cannot master the coordination needed
to use MDIs.
1. Overall Bronchodilating Plan
Before presenting the individual types of bron-
chodilator drugs, it is important to consider the
order in which the drugs should be administered
(i.e., what is the pharmacologic bronchodilating
plan?). At present, there are two basic different
approaches to the use of bronchodilators in asthma
and COPD patients, and the essential difference
between the two approaches is whether the inhaled
-,-agonists and/or anticholinergics are considered
the front-line drugs, with inhaled steroids consid-
ered the second-line drugs or vice versa (Table 6-
2). In the first approach, espoused by Bone,
61
phase I consists of inhaled bronchodilators (
2
-
agonists for asthmatics, anticholinergics for COPD
patients [step 1, Table 6-2] and as adjunctive ther-
apy in both conditions [step 2, Table 6-2]).
61
Phase II for asthmatics consists of prophylaxis
and/or treatment of inflammation with agents that
have few side effects (inhaled steroids, cromolyn)
(step 3, Table 6-2). Theophylline can be used if
asthma is not adequately controlled with inhaled
,-agonists and either cromolyn and/or inhaled
steroids (step 4, Table 6-2). Phase III for asthmat-
ics is treatment of inflammation with agents that
have greater toxicity (oral or intravenous steroids)
(step 5, Table 6-2). For COPD patients, theophyl-
line and steroids follow the acute bronchodilators
(steps 1-5, Table 6-2). In the second approach,
espoused by Lam and Newhouse
62
and New-
house,
63
there is an increased appreciation of the
importance of inflammation in the pathogenesis of
Figure 6-4 Child using metered aerosol inhaler with
spacer.
Table 6-2 TWO BASIC DIFFERENT APPROACHES TO THE USE OF BRONCHODILATORS IN
ASTHMA AND COPD
Approach to Bronchodilation in Asthma and COPD
Bone' Lam, Newhouse
6263
asthma, and inhaled bronchodilators have been rel-
egated to second-line therapy for controlling
asthma by the increasingly potent inhaled steroids
and cromolyn, both of which have an excellent
therapeutic ratio. Thus, in this approach, first-line
therapy for moderate chronic asthma should be
dose-optimized inhaled steroids and possibly in-
haled cromolyn sodium (cromoglycate sodium),
especially in children (step 1, Table 6-2). Second-
line therapy is P
2
-agonists (step 2, Table 6-2).
Theophylline is considered rarely of additional
value except, perhaps, as a steroid-sparing strategy
in the maintenance therapy of patients with asthma
requiring systemic steroids for control despite
maximum tolerated doses of inhaled steroid, ad-
renoceptor agonists, and (rarely) anticholinergic
agents (step 4, Table 6-2).
For COPD patients, after the reversal of priority
for the first- and second-line drugs (steps 1 and 2,
Table 6-2), the approach is similar (steps 3-5,
Table 6-2). All the drugs (and their dosages) that
are contained within all approaches to the preven-
tion and treatment of bronchospasm in patients
with asthma and COPD are listed in Table 6-3.
2. fi
2
-Agonists
-Receptors are divided into , and
2
sub-
types; ,-receptor agonists subserve chronotropic
and inotropic cardiac effects, and
2
-08 re-
lax the airway and vascular smooth muscle. At the
molecular level, sympathomimetics (so-called first
messengers) cause -receptor activation, which
leads to the activation of adenylate cyclase, which
converts adenosine triphosphate to cyclic adeno-
sine monophosphate (cAMP), which acts as a se-
cond messenger for the production of protein ki-
nases, leading to smooth-muscle relaxation (Fig.
6-5).
64
When used by inhalation,
2
^ 8 8 cause a
peak bronchodilator effect within 15 min, which is
maintained for approximately 4 to 6 hours, de-
pending on the dose inhaled.
2
^8 8 may be
absorbed from the bronchial mucosa and distrib-
uted to more peripheral airways. Only about 10
per cent of inhaled drug reaches the lung; the rest
is swallowed. Substantial bronchodilation can be
achieved without detectable circulating levels of
2
^ 5 5. Therefore, side effects common to in-
travenous administration and overdoses, such as
muscle tremor, tachycardia, and hypokalemia, are
rarely observed with inhaled therapy. In addition,
adrenergic compounds (epinephrine, ephedrine,
isopropyl, norepinephrine) can also increase
ciliary activity, which may help in removing
secretions.
65
66
2
^^ drugs, such as albuterol,
terbutaline, and metaproterenol, are administered
to patients who have a demonstrable reversible
bronchospastic airway component to their respira-
tory disease.
61
-
67
68
Their use in asthma is obvious.
Studies in stable outpatients with COPD demon-
strate that, with adequate dosing and delivery, an
inhaled bronchodilator (either a
2
^ 5 or an
antimuscarinic) can result in complete
57 5
* or
partial
55
56
bronchodilation (as measured by FEV,).
In these studies, addition of a second agent did not
result in further spirometric improvement. In ad-
dition, if a therapeutic blood level of a p
2
-agonist
is achieved, then the ventricular performance of
COPD patients may also be improved because of
a decrease in afterload (particularly on the right
ventricle, which may be pumping against chronic
pulmonary hypertension) and by a positive ino-
tropic effect.
69
ro
)

Figure 6-5 Occupation of the B-receptor by agonist activates adenylate cyclase (AC) via coupling protein, guanine nucleotide
regulatory protein (N). Cyclic adenosine monophosphate (AMP) is broken down by phosphodiesterase (PDE). (ATP = adenosine
triphosphate.) From Chung KF, Barnes PJ: Drugs for respiratory disease. Medicine International 2469-2473, 1988. Used with
permission.)
3. Anticholinergic Drugs
The main anticholinergic agent of clinical use
as a bronchodilator is ipratropium bromide, a qua-
ternary ammonium derivative of atropine (N-iso-
propylnoratropine). Ipratropium bromide can be
prescribed only as an aerosol, either from an MDI
inhaler or from a nebulizer. It is also available
combined with a p
2
-agonist. Ipratropium bromide
is a competitive antagonist of acetylcholine at the
parasympathetic muscarinic receptor. It causes
bronchodilatation by relieving the increase in
bronchomotor tone as a result of tonic vagal nerve
impulses releasing acetylcholine at cholinergic
nerve terminals in airways. Muscarinic receptors
are more densely distributed in the larger, proxi-
mal airways, consistent with physiologic studies
that demonstrate a greater effect on larger airways
than on small airways. Compared to
2
^8 5,
ipratropium bromide is slower in achieving a max-
imal bronchodilator effect, with approximately 75
per cent of the effect occurring at 15 min and the
complete effect by 1 hour. The duration of signif-
icant bronchodilation is about 6 to 8 hours. Maxi-
mal responses are usually achieved with 80 to 120
g of ipratropium bromide. At the recommended
doses, ipratropium bromide aerosol causes rela-
tively few adverse reactions. Anticholinergic side
effects, such as blurred vision, dry mouth, hesi-
tancy of micturition, are uncommon (< 1 per
cent), because of its poor absorption.
Most asthmatics respond better to a p
2
-agonist
than to ipratropium bromide, in contrast to patients
with chronic bronchitis and COPD in whom ipra-
tropium may be/should be used as an alternative
to a -agonist as first-line treatment.
61
70
Ipratro-
pium bromide works as well in COPD patients
because they have an increased amount of cholin-
ergic tone compared with normal patients.
70
How-
ever, in asthma, ipratropium may be tried after
failure to respond to a -agonist,
60
71
and ipratro-
pium may also be useful as additional therapy in
patients with chronic bronchitis or asthma who are
already on therapeutic doses of
2
^5.
6 0 7:
Ipratropium (10 puffs or 200 g nebulized) may
also be useful in patients with nonallergic bron-
chial asthma
73
and in COPD patients being me-
chanically ventilated after cardiac operations.
74
4. Inhaled Steroids
Inflammation of the airways is the main patho-
physiologic process in asthma, and it can be found
even in patients with newly diagnosed, mild, or
asymptomatic asthma.
75
76
Thus, the disease is now
often described as chronic desquamating eosino-
philic bronchitis.
77
Airway hyperresponsiveness in
asthma is now viewed as a secondary consequence
of the inflammation that narrows the lumen ana-
tomically and increases the responsiveness of neu-
romuscular control of the airways through the ac-
tion of the mediators of inflammation. Inhaled
corticosteroids therapy (beclomethasone dipro-
pionate, flunisolide acetate, budesonide, and tri-
amcinolone acetate) is the best anti-inflammatory
treatment available for asthma.
62
It decreases
airway inflammation and nonspecific bronchial hy-
perresponsiveness, potentiates response to sympa-
thomimetics and aminophylline, improves pulmo-
nary function, and decreases the frequency and
severity of asthma symptoms. An expert panel,
convened by the National Institutes of Health and
a number of interested professional societies, has
recommended that the care of patients with daily
(or nightly) asthma include one of the anti-inflam-
matory aerosolscromolyn or a glucocorticoid
along with bronchodilators for symptomatic re-
lief.
7
"
A large fraction of the dose delivered by an
MDI held in the mouthup to 90 per centis
deposited in the pharynx and larynx, where ste-
roids can cause dysphonia and predispose the pa-
tient to oropharyngeal candidiasis. After it is swal-
lowed, the fraction of drug deposited in the
pharynx also contributes to systemic steroidal ac-
tion. Inhaling slowly through a spacer device re-
duces deposition in the mouth and pharynx and
reduces both local and systemic adverse effects.
Training the patient in aerosol-inhalation tech-
nique with a spacer adds to the complexity of
prescribing steroid aerosols, but it is essential for
success.
Recent studies showed that the effectiveness of
these medications increases if they are used with
an aerochamber (spacer). This device also de-
creases the incidence of oropharyngeal candidi-
asis.
7V
Inhaled corticosteroids have been underused
because patients who are accustomed to rapid re-
lief from an inhaled bronchodilator find that they
experienced no immediate relief from inhaled cor-
ticosteroids. In addition, the suppressant effect of
inhaled steroids (budesonide) on airway inflam-
mation, edema, and reactivity is dissipated within
2 weeks of dosing if it is not repeated.
80
The lack
of dramatic onset and dissipation of effect may
constitute an apparent lack of response to the pa-
tient and lead to noncompliance. Thus, patient ed-
ucation regarding therapeutic expectations for
these drugs is of paramount importance.
5. Inhaled Sodium Cromoglycate
Sodium cromoglycate is a synthetic derivative
of the herbal plant, khellin, and is used for the
prophylaxis and prevention of asthma attacks. An-
other closely related compound, nedocromil so-
dium, has recently become available. Sodium
cromoglycate is highly ionized and water soluble
and is therefore poorly absorbed when taken
orally. For the treatment of asthma, it is inhaled
either as a fine powder from a spinhaler or as an
aerosol from an MDI. A nebulizer solution is also
available.
Prevention of mast cell degranulation by aller-
gens was thought to be the main mechanism of
action of sodium cromoglycate in the prophylaxis
of asthma. However, other compounds with
greater potency as mast cell stabilizers have little
value in the prophylaxis of asthma. Other possible
modes of action of cromoglycate include an inhib-
itory effect of eosinophil activation to release me-
diators and inhibition of neurogenic reflex bron-
choconstriction perhaps by an action on sensory
nerves. It is not a bronchodilator.
Only a small proportion of inhaled sodium
cromoglycate can be detected in the circulation,
the amount depending on inspiratory flow rate.
Peak plasma levels are achieved within 15 to 20
min of inhalation, and sodium cromoglycate is ex-
creted unchanged in bile and urine. The time
course of action of sodium cromoglycate is diffi-
cult to assess, but some studies suggest that its
prophylactic effect may take a few weeks to be-
come optimum. Sodium cromoglycate is remarka-
bly devoid of side effects. Patients may complain
of irritation of the throat and of cough when in-
haled as a dry powder, probably because of the
irritant effect of the powder. Sodium cromoglycate
improves asthmatic symptoms and leads to a re-
duction in the use of bronchodilators in asthmatic
patients. However, in general, sodium cromogly-
cate is less effective as a prophylactic treatment
than inhaled steroids. Clinically, it is only effective
in a proportion of asthmatics, particularly in the
younger, atopic group. Sodium cromoglycate
should be considered in asthmatics whose symp-
toms are not well controlled by a -agonist inhaler
alone, particularly in children. If cromoglycate is
not effective, then inhaled steroids should be used.
However, in adults, there is a preference for in-
haled steroids as first-line prophylaxis. Sodium
cromoglycate should be used regularly over a pe-
riod of 2 to 3 weeks before assessing the response.
Improvement may be dose dependent. Sodium
cromoglycate should not be used for treatment of
acute asthma and in patients with COPD.
6. Methylxanthines
Theophylline is a methylxanthine found in var-
ious plants, including tea, and is closely related to
caffeine and theobromine. Theophylline can be
taken orally or by suppository. Aminophylline is a
theophylline salt, is more soluble than theophyl-
line, and is therefore used intravenously. Both
methylxanthines are difficult drugs to use because
of the large variation in metabolism between indi-
viduals and because of the narrow therapeutic ra-
tio.
c'AMP is broken down by a cytoplasmic en-
zyme, phosphodiesterase, whose activity can be
inhibited by methylxanthines, such as theophylline
and aminophylline. Thus, the methylxanthines also
increase c'AMP but by a mechanism different
from that of the p
2
-agonists (see Fig. 6-5). Be-
cause the methylxanthines and
2
^ 8 8 act by
different mechanisms, theophylline is often added
to the regimen of patients with bronchospasm al-
ready receiving beta-adrenergics, and thus they
work in synergy to increase intracellular concen-
trations of c' AMP.
81, 82
In addition, aminophylline
improves diaphragmatic contractility and renders
it less susceptible to fatigue.
83
In fact, in COPD
patients with severe "fixed" obstruction (FEV, <
30 per cent and unresponsive to
2
^ 8, the-
ophylline increases respiratory function and de-
creases dyspnea as a result of increased strength of
the respiratory muscles.
84
85
Thus, it is no surprise
that the methylxanthines cause subjective im-
provement in patients with chronic airflow ob-
struction.
86
Recently, however, there has been considerable
concern regarding the short-term toxicity of in-
haled beta-adrenergic agents when used in the
presence of methylxanthines; specifically, myocar-
dial ischemia might occur as a result of the drugs'
combined effect on the heart with resultant (and
possibly fatal) ventricular arrhythmias.
87 88
In one
study of patients in status asthmaticus, with a
mean age of 39 6 years, 17 per cent of the
patients exhibited severe ventricular and atrial ar-
rhythmias during combined therapy with amino-
phylline and the p
2
-agonist terbutaline.
89
Some
studies have not supported exercising great con-
cern with respect to arrhythmias with concurrent
use of aminophylline and
2
^8,
9 0
9|
but it
should be remembered that the dose required to
produce therapeutic/toxic levels can be quite vari-
able because of variation in hepatic metabolism.
Thus, appropriate caution should be exercised
when combining inhaled
2
^ 8 8 with meth-
ylxanthines.
88 92
Optimal therapeutic serum levels
of theophylline (10 to 20 mg/ml) can be safely
approached and toxicity avoided if an intravenous
loading dose (5 to 7 mg/kg) is given, followed by
continuous infusion (0.2 to 0.8 mg/kg/hr: dose de-
creases with increasing age).
93
Although the pa-
tient's subjective feeling of relief is an important
end point (see chapter 5), the effect of bronchodi-
lator drug treatment should be quantitated by pul-
monary function tests.
7. Oral and Parenteral Steroids
Oral or intravenous corticosteroids should be
used as a last-line measure. Because corticoste-
roids are rapidly and almost completely absorbed
from the gastrointestinal tract, and both routes of
administration have somewhat of a delayed effect
(but still less than a day),
94
it is not surprising that
oral corticosteroids are nearly as effective as intra-
venous corticosteroids for status asthmaticus.
61
Corticosteroids should be used in patients
whose attacks are not controlled with the combi-
nation of bronchodilators and inhaled anti-inflam-
matory medications. Oral dosage (for outpatients)
should be as high as 0.75 to 1.0 mg/kg/day, and
intravenous dosages (for inpatients) of corticoste-
roids should be in the range of 0.5 mg/kg every 6
hours, for 1 to 2 weeks, with daily monitoring of
peak flow rates (if possible), and the oral or intra-
venous dosage should be tapered over 1 to 2
weeks. If the patient is not taking inhaled steroids
or cromolyn, one of these should be added as the
systemic corticosteroid dosage is tapered. If oral
corticosteroids cannot be replaced by the inhaled
anti-inflammatory agents, alternate-day steroids in
conjunction with the anti-inflammatory agents
should be considered. The anti-inflammatory
agents constitute the ideal regimen in the patient
with severe asthma, with oral corticosteroids re-
served for severe exacerbation. The oral cortico-
steroid dosage is then tapered as soon as control is
achieved.
C. Loosening the Secretions
The next step should be to thin and to loosen
thick adherent secretions. The most efficacious
method is hydration. When tracheal mucus trans-
port velocity is quantitatively measured by radio-
active tracer methods, it can clearly be shown that
dehydration decreases and rehydration increases,
respectively, tracheal transport velocity.
95
The
most common method of hydrating secretions is
by use of a jet humidifier or ultrasonic nebulizer
to produce a heated, sterile water aerosol that is
delivered by a close-fitting mask for 20 min to a
deeply spontaneously breathing patient. Concur-
rently, continuous systemic hydration must be en-
sured orally or intravenously.
Very occasionally, administration of mucolytic
agents (such as acetylcysteine [Mucomyst]) by a
nebulizer and/or oral expectorants (such as guai-
fenesin, potassium iodide, iodinated glycerol) may
be of limited benefit in patients with very viscous
secretions, but both of these treatments (mucolytic
agents, oral expectorants) have side effects that
224 Preoperative Respirator} Preparation
would obviate their use in most patients. The ben-
efit of mucolytic agents is that they decrease the
viscosity of secretions (by depolymerizing muco-
polysaccharides),
96
but at the same time they may
induce irritability of the airways and broncho-
spasm. The benefit of oral expectorants is that they
increase the amount of secretions removed, but at
the same time they increase the absolute amount
of secretions (and may cause gastrointestinal upset
and skin lesions).
96
Pulmonary infection, if present, is treated ac-
cording to the results of culture and sensitivity
tests; broad-spectrum antibiotics such as ampicillin
or a cephalosporin frequently have the required
specificity and potency. If the antibiotic treatment
clears the infection to any extent, it may also de-
crease the tenacity, viscosity, and volume of secre-
tions.
D. Removing the Secretions
The next preoperative respiratory step is the ac-
tual removal of secretions, and this is accom-
plished by a combination of postural drainage
(several different positions may be required),
coughing or forced expiration (see later discus-
sion), and perhaps chest percussion and vibration
(common methods include tapping with cupped
hands and electric vibrators) for a period 15 to 20
min several times day.
97
98
Consequently, therapy
to remove secretions uses gravity, patient-gener-
ated expiratory airflow, and mechanical measures
to dislodge and to propel the secretions proxi-
mally. The additional use of inhaled
2
-agonists
may increase secretion removal.
99
In a patient who
has been in bed for a long period of time, getting
the patient out of the bed is one of the most im-
portant lung-expansion maneuvers that can be per-
formed, often generating a 10 per cent to 20 per
cent increase in functional residual capacity.
20
When removal of tracheobronchial secretions is
quantitatively measured with radioactive tracer
methods in patients with chronic obstructive lung
disease, chest physiotherapy (chest percussion and
vibration along with postural drainage) with cough
is effective in increasing both central and periph-
eral airway clearance and sputum yield, whereas
cough alone is effective in increasing only central
airway clearance and sputum yield.
100
Further evi-
dence that physiotherapy with coughing is most
effective at clearing secretions from the main
bronchi has come from a study of patients with
acute lobar collapse.
101
The success rate in regain-
ing lung volume was strongly related to whether
there was an air bronchogram: The resolution rate
was only 25 per cent when a bronchogram was
present compared with almost 90 per cent when
no bronchogram was present. The explanation wa
that, with a bronchogram, there was distal alveola
collapse but no bronchial sputum plug, whereas n<
bronchogram meant that sputum had blocked ;
bronchus and caused secondary collapse.
In patients with chronic bronchitis, coughing ex
ercises have been investigated with radioaeroso
techniques and have produced as much centra
lung clearance as physiotherapy
102
and greater tota
clearance than no treatment.
103
Thus, chest physio
therapy moves peripheral bronchial secretions t(
more central airways for expectoration by cough
ing. Cough alone cannot clear peripheral airway:
because an effective cough must attain a higr
enough airflow rate so as to shear secretions awa)
from the airway wall. In patients with chronic lun
disease, flow rates are low (especially peripher-
ally), and the shearing of secretions by cough ma>
well be limited to the trachea and perhaps the firsi
two airway generations.
104
Obviously, with eithei
chest physiotherapy, forced expiration technique,
or cough, it will be much easier to expel the secre-
tions if the airways have already been dilated and
the secretions loosened.
Chest physical therapy is relatively contraindi-
cated in patients with lung abscesses, bone metas-
tases, and a history of significant hemoptysis and
inability to tolerate the postural drainage positions
One review pointed out the possibility of short-
term hypoxemia lasting up to 30 min after physio-
therapy.
10
^ Most patients tolerate this hypoxia, but
postural drainage, percussion, and vibration must
be used with particular care in critically ill patients
and in those with low blood concentrations of ox-
ygen (e.g., monitor with pulse oximetry). Thus,
because of the risk of causing either bronchospasm
or short-term hypoxemia, postural drainage, per-
cussion, and vibration should be used only in pa-
tients in whom its value has been proved (e.g., a
patient with copious secretions and a weak cough).
The forced expiration technique combined with
postural drainage should be strongly considered as
an alternative (see later discussion). It is generally
agreed that IPPB regimens are not sufficiently ef-
ficacious to warrant the excessive cost ($60 per
treatment) of routine use.
106-109
The forced expiration technique (FET) is in-
creasingly regarded as more effective in removing
secretions than a cough.
110
The FET comprises a
forced expiration starting from midlung volume
(50 per cent of inspiratory reserve lung volume) to
a low lung volume, usually the residual volume
(Fig. 6-6), followed by a period of relaxation and
i
diaphragmatic breathing. This forceful expiration
maneuver differs from a cough because it is per-
formed without closure of the glottis and the ac-
companying compressive phase that characterizes
cough. The transpulmonary pressure is less during
Preoperative Respiratory Preparation
225
THE FORCED EXPIRATION TECHNIQUE
Figure 6-6 The forced expiration
technique begins at 50% of the inspira-
tory reserve volume and ends at residual
volume.
the FET than during cough, resulting in less air-
way compression and permitting improved proxi-
mal and distal clearance of bronchial secretions
compared with a conventional cough (as docu-
mented by radioaerosol techniques)."
0
The FET is
now being increasingly used as an alternative to
cough during chest physiotherapy."
0
E. Ancillary Measures/Issues
There are a number of ancillary preoperative
therapeutic measures that, in selected patients, may
be very important and contribute to the function of
the respiratory system. These therapies consist of
general measures to improve systemic and major
organ well-being and fitness for surgery and spe-
cific measures to treat cor pulmonale, including
the controversial issue of preoperative digitaliza-
tion.
1. General Measures (Treatment of
Systemic Disease)
Several general measures can significantly re-
duce respiratory morbidity and mortality. Graded
activity programs may provide considerable pre-
operative subjective improvement. If obese pa-
tients will cooperate, preoperative weight loss
should be strived for. Malnutrition, sometimes pre-
sent in patients with carcinoma or advanced pul-
monary disease, may require preoperative treat-
ment (by nasogastric or intravenous feeding).'"
Any other concurrent medical problems (e.g., dia-
betes, angina [see later discussion]) should be sta-
bilized. Oxygen supplementation by nasal cannula
at flows of l to 2 L/min may be helpful for patients
with severe hypoxemia and therefore have a risk
of arrhythmia.
Most people with angina require drug therapy.
Coronary thrombosis is prevented by aspirin. The
remaining pharmacologic goals are reduction of
myocardial oxygen demand and coronary vasodi-
lation. Nitrates produce venous dilatation and
therefore a reduction in venous return and cardiac
output. They also cause dilatation of the coronary
vasculature. The beta blockers and some calcium
antagonists (e.g., verapamil) reduce myocardial
oxygen demand by inhibiting exercise-induced
tachycardia. Other calcium antagonists (e.g., nifed-
ipine) primarily have a vasodilator effect, reducing
peripheral arteriolar resistance and thereby reduc-
ing left ventricular work. They may also have a
direct vasodilator effect on the coronary arteries.
Patients with hypertension are frequently treated
with diuretics. If the patient has a hypochloremic
alkalosis secondary to diuretic therapy, this should
be corrected to reduce any component of hypoven-
tilation in response to the metabolic abnormality.
The treatment of choice is KG supplementation.
Hypercalcemia caused by lytic bone metastases
should be treated with saline hydration, furosem-
ide, possibly drugs to prevent bone resorption (the
bisphosphonates etidronate or pamidronate), pli-
camycin (Mithracin), calcitonin, gallium nitrate,
and glucocorticoid."
2
Antibiotics are an important component of ther-
apy in COPD patients with acute pulmonary infec-
tions that may increase pulmonary vascular resis-
tance. The most common infecting organisms are
Haemophilus influenzae and pneumococci, both of
which usually can be treated effectively with am-
picillin or a cephalosporin."
3
Certainly, most cli-
nicians would treat a purulent bronchitis (without
evidence of a pneumonia) with these antibiotics.
Many surgeons use preoperative (1 week before
thoracotomy) antibiotic prophylaxis (deoxytetra-
cycline or cefuroxine) against postoperative infec-
tion."
4
2. Treatment of Cor Pulmonale
The prevalence of cor pulmonale increases with
the severity of airway obstruction, occurring in 40
per cent of patients with an FEV, of less than 1.0
L and in 70 per cent of patients with an FEV, of
less than 0.6 L."
5
It is also higher in patients with
hypoxemia, C0
2
retention, and polycythemia,
which likely explains why cor pulmonale is seen
more frequently in patients with chronic bronchitis
than in those with emphysema. The mainstays of
therapy for cor pulmonale are a vigorous pulmo-
nary toilet regimen, oxygen supplementation to re-
lieve hypoxic pulmonary vasoconstriction, diuret-
ics, and possibly digitalis (see the following
discussion). Diuretics have been shown to offer
significant benefit to patients with right ventricular
failure from cor pulmonale."
6
Excess lung water
that interferes with gas exchange and increases
pulmonary vascular resistance can be reduced by
systemic diuresis; careful monitoring of serum
electrolytes must be performed. Phlebotomy
should be considered whenever the hematocrit
reaches 55 per cent or greater. Secondary erytho-
cytosis increases blood viscosity as well as intra-
vascular volume and thereby increases both right
ventricular preload and afterload. Hypoxic patients
with COPD and polycythemia show improvements
in right ventricular ejection fraction, cerebral
blood flow, exercise tolerance, and neuropsycho-
logic function after phlebotomy that cannot be at-
tributed to volume reduction alone."
5
The hemo-
dynamic effects of many vasodilators have been
studied in patients with COPD but, despite their
theoretic potential benefit, vasodilator agents have
failed to consistently sustain favorable effects in
patients with COPD beyond those achieved by ox-
ygen and bronchodilators. Furthermore, some del-
eterious effects (e.g., hypoxemia) have been ob-
served with the use of these agents in patients with
COPD and secondary pulmonary hypertension
(see chapter 4). At present, vasodilator agents are
not recommended for routine use in patients with
COPD and cor pulmonale."
5
3. Preoperative Digitalization
The preoperative use of digitalis in the thoracic
surgical patient deserves special comment. Resec-
tion of pulmonary tissue reduces the available pul-
monary vascular bed for perfusion and can cause
postoperative right ventricular, and right atrial en-
largement. Thus, it is not surprising that the inci-
dence of postoperative arrhythmias (due to atrial
stretching) increases progressively with age and
amount of lung resected. In addition, there is a
higher incidence of atrial arrhythmias following
left pneumonectomy than right pneumonectomy
because a greater degree of manipulation of the
atrium occurs during the former operation. Al-
though the postoperative incidence of arrhythmias
has provided the basis for the prophylactic use of
digitalis in thoracic surgical patients without evi-
dence of congestive heart failure, this practice is
still controversial."
7-122
Widely accepted indica-
tions for preoperative digitalization in patients
without cor pulmonale undergoing thoracic sur-
gery include congestive (left-sided) heart failure
and supraventricular arrhythmias with a rapid ven-
tricular response.
123 I24
The indications for preoperative digitalization
are more straightforward in patients with cor pul-
monale. However, it is important to note that these
patients have a propensity to develop hypoxemia,
hypercarbia, and acidosis, and they are therefore
at an increased risk of developing digitalis toxic-
ity; the drug should therefore be used with
caution.
125
I26
If these patients do receive digitalis
preoperatively, it is important to normalize the
serum potassium in order to decrease the risk of
arrhythmias. Digitalis should probably be withheld
on the day of surgery to help avoid confusion with
digitalis intoxication if arrhythmias occur postop-
eratively.
127
F. Measures to Increase Motivation and
Education and to Facilitate
Postoperative Respiratory Care
The last step consists of general measures de-
signed to increase motivation and education and to
facilitate postoperative respiratory care. Preopera-
tive psychologic preparation (i.e., positive sugges-
tion and encouragement), including orientation to
the intensive care environment, can reduce fear
and improve patient outlook and cooperation. Still,
patients should be given realistic expectations
about postoperative pain and how it will be han-
dled. With this type of positive but realistic ap-
proach, patients will have a diminished postopera-
tive narcotic analgesic requirement and a shorter
period of hospitalization.
128
Preoperative patient education in the procedures
that will be used for respiratory care postopera-
tively and explaining why these procedures are
going to be used will greatly help to ensure opti-
mal postoperative compliance with and perform-
ance of the respiratory care maneuvers. There is
no question that a critical factor in treating patients
is to preserve and to enhance an individual's abil-
ity to do things for him- or herself.
129
In this re-
gard, preoperative practice with an incentive spi-
rometer prepares the patient to participate in
restoring lung volume and coughing postopera-
tively. Controlled studies have shown that incen-
tive spirometry is superior to IPPB in terms of
both effectiveness and cost.
130, ,31
Several incentive
spirometers are available, with the more popular
ones being Triflo II (which is a relatively inexpen-
sive, small, and easily carried flow-sensitive
device), the original volume-sensitive Bartlett-
Edwards Incentive Spirometer, and Spirocare, a
relatively expensive, nondisposable, volume-sen-
sitive device that must be plugged into an electric
outlet.
132
There does not appear to be any signifi-
cant difference in effectiveness among these three
different devices.
12
However, it has become emi-
nently clear that the most effective use of these
devices depends very heavily upon an effective
bedside coach (nurse, respiratory therapist, family
member, or friend) who provides frequent verbal
encouragement designed to maximize patient com-
pliance with the incentive of the device
12
(the de-
vices use either visual or auditory cues to signal
achievement of either the volume or flow end
point).
Other deep-breathing maneuvers, such as blow
bottles and blow gloves, actually depend on the
inspiratory maneuver that must precede the em-
phasized expiratory maneuver and, therefore, must
be regarded as mild, short-duration forms of incen-
tive spirometry. Carbon dioxide breathing devices
had a short period of popularity until it was shown
that the resulting increase in minute ventilation
was due to an increase in respiratory rate rather
than depth of breathing.
Preoperative exposure to chest physiotherapy
(percussion and vibration), postural drainage, and
deep-breathing exercises facilitates use of these
maneuvers postoperatively and thereby reduces the
incidence of atelectasis.
21
VI. MECHANISM OF
PREOPERATIVE RESPIRATORY
PREPARATION BENEFIT
There are several possible mechanisms to ex-
plain why preoperative respiratory care prepara-
tion maneuvers benefit the patient and result in a
decreased incidence of postoperative respiratory
complications. Preoperative removal of secretions
from the airways is probably the main beneficial
result of such intensive preparations for surgery.
The removal of secretions is accomplished by the
therapeutic cascade of dilating the airways, loos-
ening the secretions, and then removing the secre-
tions. Second, the patients may respond nonspecif-
ically to the attention given them, and because of
an improved psychologic and motivational status,
they may have improved compliance with postop-
erative respiratory care maneuvers and may am-
bulate earlier. Third, preoperative instructions
in deep-breathing exercises, incentive spirometry
and coughing maneuvers, and exposure to chest
physiotherapy probably improve the efficacious-
ness of these maneuvers in the postoperative pe-
riod. Fourth, there is evidence that many of the
techniques used (aminophylline, the various exer-
cises) improve respiratory muscle strength and en-
durance; however, it is not clear whether the usual
1- to 2-day preoperative regimen is capable of
producing significant muscular conditioning as
compared with the 5-week period allowed for in
more carefully controlled situations.
10
Finally, the
preoperative respiratory care regimen may actually
improve pulmonary function and pulmonary func-
tion testing, although this does not occur in every
prepared patient even though the incidence of res-
piratory complications decreases in prepared pa-
tients versus nonprepared patients.
10
It is likely that
some or all of these mechanisms contribute in part
to the ability of preoperative respiratory care prep-
aratory maneuvers to decrease the incidence of
postoperative respiratory complications.
VII. PREMEDICATION
Premedication is individualized according to the
psychologic needs of the patient, the severity of
pre-existing pulmonary disease, and the antici-
pated operation. An explanation of the need for
various vascular catheters, specific monitoring de-
vices, and use of the face mask for oxygenation
(and possibly for inhalation induction of anesthe-
sia) helps alleviate anxiety and promotes patient
cooperation in the operating room. For most pa-
tients with reasonably good preoperative pulmo-
nary function, a combination of a narcotic analge-
sic with a benzodiazepine (diazepam, lorazepam)
in moderate dosage provides sedation, periopera-
tive analgesia, reduction of anesthetic require-
ments, and amnesia without concern of predis-
posing the patient to preoperative respiratory
depression. Depression of spontaneous ventilation
during anesthesia caused by this type of premedi-
cation is rarely a concern because the vast majority
of these patients will have their ventilation con-
trolled intraoperatively. Excessively long-acting
drugs or very heavy sedation is to be avoided if
the operative procedure is short and early postop-
erative mobilization is desired.
The use of anticholinergic drugs in most normal
patients and in patients with mild to moderate
COPD often causes a fairly uncomfortable feeling
as a result of excessive drying; the drying may
theoretically cause difficulty in removing secre-
tions. In addition, modern inhalation anesthetics
are much less sialorrheic than their predecessors.
Consequently, anticholinergic drugs are not ordi-
narily used preoperatively. However, because
there is good evidence that atropine does not in-
crease the viscosity of secretions but merely de-
creases the volume of secretions,
133
these drugs
should be considered for patients with copious and
troublesome amounts of secretions. Although anti-
cholinergic drugs do increase the dead space to
tidal volume ratio, the increase is small.
For patients in whom histamine release might
be a problem (patients with chronic obstructive
lung disease and reactive airways, asthmatics),
premedication with the ,-blocker diphenhydra-
mine hydrochloride (Benadryl) would seem to be
a logical choice; Benadryl provides both sedation
and ,-receptor blockade of histamine-induced
bronchospasm. The use of H
2
-receptor blockers
(cimetidine, rantidine) poses the risk of provoking
bronchospasm. Histamine mediates bronchocon-
striction via ,-receptors, whereas bronchodilation
is mediated by H
2
-receptors; hence, selective
blockade of H
2
-receptors could unmask unopposed
bronchoconstriction. For patients who are thought
to have, or may develop, predominance of vagal
tone that could induce bronchospasm, atropine is
relatively indicated. For patients with a predomi-
nance of alpha-adrenergic activity that might result
in bronchospasm and hypertension, use of droper-
idol for premedication is relatively indicated. Both
hydroxyzine and droperidol decrease airway resis-
tance. The histamine-releasing opioids, such as
morphine and meperidine, also should theoreti-
cally be avoided in patients with bronchospastic
disease. Should opioid premedication be required,
fentanyl is suggested. Patients who are hypoxemic
on room air (P
a
0
2
< 60 mm Hg) or hypercarbic
(P
a
C0
2
> 45 mm Hg) are given little or no pre-
medicants that might further depress gas exchange.
Patients who are already on supplemental oxygen
preoperatively must be transported to the operating
room with the same oxygen flow as previously
administered, along with appropriate personnel in
constant attendance. Patients who experience or-
thopnea need to be transported in a semiupright
position. Manual or mechanical ventilation during
intrahospital transport of critically ill, mechani-
cally ventilated patients is safe provided the person
performing manual ventilation knows the inspired
oxygen fraction and minute ventilation required
before transport and is trained to approximate
them during transport.
134
REFERENCES
1. Ford GT, Guenta CA: Toward prevention of postopera-
tive pulmonary complications. Am Rev Respir Dis
130:4-5, 1984.
2. Lansing AM, Jamilson WG: Mechanisms of fever in ate-
lectasis. Arch Surg 87:168, 1963.
3. Harris JD, Johanson WG Jr, Pierce AK: Bacterial lung
clearance in hypoxic mice. Am Rev Respir Dis 1 11:910,
1975.
4. Latimer G, Dickman M, Clinton DW, Gunn MI, Du-
Wayne SC: Ventilatory patterns and pulmonary compli-
cations after upper abdominal surgery determined by pre-
operative and postoperative computerized spirometry and
blood gas analysis. Am J Surg 122:622, 1971.
5. Morton HJV, Camb DA: Tobacco smoking and pulmo-
nary complications after operation. Lancet 1:368, 1944.
6. Tait AR, Kyff JV, Crider B, Santibhavank V, Learned D:
Post-operative arterial oxygen saturationUp in a puff of
smoke? Anesth Analg 68:S284, 1989.
7. Handlin DS, Baker T: The effect of heavy and light
smoking on patient duration in the post anesthesia care
unit. Anesth Analg 72:S99, 1991.
8. Pearce AC, Jones RM: Smoking and anesthesia: Preop-
erative abstinence and perioperative morbidity. Anesthe-
siology 61:576-584, 1984.
9. Stein M, Koota GM, Simon M, et al: Pulmonary evalua-
tion of surgical patients. JAMA 181:765-770, 1962.
10. Gracey DR, Divertie MB, Didier EP: Preoperative pul-
monary preparation of patients with chronic obstructive
pulmonary disease. Chest 76:123-129, 1979.
11. Tisi GM: Preoperative evaluation of pulmonary function.
Am Rev Respir Dis 119:293-310, 1979.
12. Van De Water JM: Preoperative and postoperative tech-
niques in the prevention of pulmonary complications.
Surg Clin North Am 60:1339-1348, 1980.
13. Johnson WC: Postoperative ventilatory performance. De-
pendence upon surgical incision. Am J Surg 41:615-619.
1967.
14. Ali J, Weisel RD, Layng AB, Kripke BJ, Hechtman HB:
Consequences of postoperative alterations and respiratory
mechanics. Am J Surg 128:376-382, 1974.
15. Anderson WH, Dosett BE Jr, Hamilton GE: Prevention
of postoperative pulmonary complications. JAMA
186:763-766, 1963.
16. Tarhan S, Moffitt EA, Sessler AD, et al: Risk of anesthe-
sia and surgery in patients with chronic bronchitis and
chronic obstructive pulmonary disease. Surgery 74:720-
726, 1973.
17. Harmon E, Lillington G: Pulmonary risk factors in sur-
gery. Med Clin North Am 63:1289-1298, 1979.
18. Lemmer JH, Gomez MN, Symreng T, Ross AF, Rossi
NP: Limited lateral thoracotomy: Improved postoperative
pulmonary function. Arch Surg 125:873-877, 1990.
19. Mitchell RL: The lateral limited thoracotomy incision:
Standard for pulmonary operations. J Thorac Cardiovasc
Surg 99:590-596, 1990.
20. Jackson CV: Preoperative pulmonary function. Arch In-
tern Med 148:2120-2127, 1988.
21. Thoren L: Postoperative pulmonary complications: Ob-
servations on their prevention by means of physiotherapy.
Acta Chir Scand 107:193-205, 1954.
22. Palmer K.N, Sellick BA: The prevention of postoperative
pulmonary atelectasis. Lancet 1:164-168, 1953.
23. Roukema JA, Carol EJ, Prins JG: The prevention of pul-
monary complications after upper abdominal surgery in
patients with noncompromised pulmonary status. Arch
Surg 123:30-34, 1988.
24. Veith FJ, Rocco AG: Evaluation of respiratory function
CHAPTER 7
Monitoring
232
I. Introduction
A. Special Intraoperative Conditions
B. Pre-Existing Lung Disease
C. Tiered Monitoring System
II. Tier I: Essential Monitoring System
A. Checking the Anesthesia
Machine
B. Continuous Monitoring of the
Oxygen-Delivery System
C. Continuous Monitoring of Apnea
D. Minute Ventilation
E. Gas Exchange
1. Cyanosis
2. Pulse Oximetry (S
p
0
2
)
a. Basic Principles and Clinical
Use
b. Limitations of Pulse Oximetry
c. Use of S
p
0
2
From the Digit of
the Dependent Hand in the
During One-Lung Ventilation
3. Capnometry and Capnography
a. Basic Principles
b. Clinical Use
(1) Is Exhaled C0
2
(Waveform) Present?
(2) Monitoring Cardiac Output
and Cardiopulmonary
Resuscitation by End-
Tidal Carbon Dioxide
Concentration
(3) Phase I: Inspiratory
Baseline
(4) Phase II: Expiratory
Upstroke
(5) Phase III: Expiratory
(Alveolar) Plateau
(6) Phase IV: Inspiratory
Downstroke
4. Approximation of Arterial Blood-
Gas Tensions Using Various
Other Body and Equipment
Compartments
F. Airway Mechanics
G. Cardiovascular Parameters
H. Muscle Relaxation
I. Temperature
III. Tier II: Special Intermittent and/or
Continuous Monitoring
E. Gas Exchange
1. Arterial Blood-Gas Analysis
2. Venous Blood-Gas Analysis
3. Transcutaneous 0
2
and C0
2
Tensions
a. Basic Principles
b. Clinical Use
F. Airway Mechanics
IV. Tier III: Advanced Monitoring
Techniques
E. Gas Exchange
F. Airway Mechanics
1. Left Ventricular Preload and Left
Ventricular Function
2. Various Indices of Left Ventricular
End-Diastolic Pressure
a. Left Atrial Pressure
b. Pulmonary Artery Occluded
Pressure (P
pao
, Large
Pulmonary Vein Pressure)
(1) Definition of P
pao
(2) P
paoi
PEEP, and Zone III
Location
(3) P
pao)
PEEP, and Zone I
Location
(4) P
pao
and Determination of
Zone III Location
(5) Special P
pao
Considerations Relevant
to Preoperative and
Postoperative Thoracic
Surgery Patients
(Excluding the One-
Lung Ventilation
Situation)
c. Pulmonary Arterial (True)
Wedge Pressure (P
pdw
,
Pulmonary Distal Wedge
Pressure, Small Pulmonary
Vein Pressure)
d. Pulmonary Capillary Pressure
V cap/
e. Pulmonary Artery Diastolic
Pressure (P
pad
)
3. Various Indices of Right
Ventricular End-Diastolic
Pressure
a. Right Atrial Pressure
b. Central Venous Pressure
(1) Normal Heart
(2) Abnormal Heart
4. Clinical Value of the Pulmonary
Artery Catheter
0) Ppao
a n c l t n e
Morphology
of P
pao
(Giant a and cv
Waves)
(2) Cardiac Output
(3) Systemic Vascular
Resistance
Monitoring 233
(4) Monitoring Mixed Venous
Oxygen Saturation
(5) Total Benefit
5. Special Pulmonary Vascular
Monitoring Considerations
Related to Thoracotomy in the
6. Risks and Complications of
Pulmonary Vascular Pressure
Monitoring
a. Complications of Gaining
Central Venous Access
I. INTRODUCTION
Although thoracic surgery may affect both pul-
monary and cardiovascular function, it more com-
monly threatens respiratory function much more
so than cardiovascular function. Consequently, this
chapter emphasizes the monitoring of respiratory
function more than cardiovascular function for the
routine thoracic surgery patient. In addition, for
the very ill patient or for very physiologically in-
trusive and demanding thoracic surgery cases, ad-
vanced cardiovascular function monitoring tech-
niques are equally and fully discussed.
Patients undergoing thoracic surgery are most
prone to impaired gas exchange because with
either one- or two-lung ventilation both nondepen-
dent- and dependent-lung function (with reference
to lateral decubitus position) will be impaired (see
chapter 4). The consequences of both inadequate
oxygenation and carbon dioxide elimination are
serious. Even slight decreases in the content of
oxygen in arterial blood from a marginally normal
level may alter a delicate balance of supply and
demand to certain tissues and also have wide-
spread effects due to activation of the sympathetic
nervous system. Similarly, changes in blood car-
bon dioxide content can alter sympathetic nervous
system activity, and, in addition, concomitant
changes in pH may impair the function of several
vital organs. The heart, especially in patients with
coronary artery disease, may be the first organ to
dramatically show the occurrence of hypoxemia
and hypercapnia with the development of ischemia
and arrhythmias (see chapter 3). Thus, the need to
monitor respiratory function during thoracic anes-
thesia is a critical and continuous responsibility.
The monitoring responsibility can be fulfilled in
various ways with various degrees of sophistica-
tion depending on a particular patient's preopera-
tive condition and intraoperative requirements.
The philosophy espoused by this approach is based
upon the concept that there are two considerations
that dictate the type of monitoring used. First, pa-
tients undergoing thoracic operations have varying
b. Complications of Actually
Passing and Floating the
Pulmonary Artery Catheter
c. Complications of Maintaining
Central Venous Access
7. Transesophageal
Echocardiography
8. Lung Water Measurements
9. Computed Tomographic
Pulmonary Scan
degrees of pre-existing cardiorespiratory disease
(see chapter 5). Second, the very nature of thoracic
procedures causes further derangements in cardio-
respiratory function during the perioperative pe-
riod (see chapter 4). Thus, on the basis of these
two considerations and their interactions, individ-
ual patients can and should be categorized into a
progressively sophisticated and complex tier sys-
tem with regard to what monitoring is necessary
to make possible accurate and rapid diagnosis and
therapy during anesthesia (Table 7-1).
A. Special Intraoperative Conditions
Patients undergoing thoracic surgery may expe-
rience several hazardous intraoperative conditions.
First, most thoracic surgical procedures are per-
formed in the lateral decubitus position, which
may compromise gas exchange (with either one-
or two-lung ventilation, but especially with one-
lung ventilation; see chapter 4). Second, the sur-
gical procedure may adversely affect the function
of mediastinal organs (e.g., irritate the heart, ob-
struct the vena cava). Third, some thoracic proce-
dures may require massive transfusion (such as
excision of some arteriovenous malformations);
hypotension and massive transfusion are associ-
ated with the development of the adult respiratory
distress syndrome. Fourth, operations necessitating
deliberate hypotension increase monitoring re-
quirements because of changes in pulmonary vas-
cular autoregulation (hypoxic pulmonary vasocon-
striction), in VD/V, and perhaps in oxygen
transport. Fifth, operations of excessive duration
will promote transudation of fluid into dependent
regions of the lung and thereby cause progressive
oxygenation difficulties. Last, operations involving
the airway, such as laryngoscopy, which fre-
quently necessitates apnea, or bronchoscopy,
which frequently imposes restrictions on positive-
pressure ventilation, require increased respiratory
function monitoring.
234 Monitoring
Table 7-1 DERIVATION OF A TIERED MONITORING SYSTEM FOR ANESTHESIA
FOR THORACIC SURGERY
Risk of
Pre-Existing Special Respiratory
Lung Intraoperative Morbidity and Tiered Monitoring
Disease And/Or Conditions Mortality System
None None Very low I. Essential monitoring
None Moderate Moderate II. Special intermittent
Moderate None Moderate and/or continuous
Moderate Moderate High moderate monitoring
None Severe High III. Advanced monitoring
Severe None High
Severe Severe Very high
B. Pre-Existing Lung Disease
Patients undergoing thoracic surgery often have
significant pre-existing cardiopulmonary disease
(see chapter 5). Most of these patients have a long
history of heavy smoking and therefore have
chronic lung disease with reactive airways and ex-
cessive secretions. Due to the lifestyle associated
with smoking, many may have coronary artery
disease and some may be very obese (with a pro-
pensity for decreases in functional residual capac-
ity [FRC] during anesthesia; see chapter 3). Some
patients presenting for emergency thoracic surgery
will have acute chest disease (pulmonary infection,
hemorrhage, contusion, infarction) or acute sys-
temic diseases (sepsis; renal, cardiac, or liver fail-
ure; or multiple trauma). Some patients requiring
thoracic surgery may be very old.
C. Tiered Monitoring System
On the bases of the presence of special intra-
operative conditions and the degree of pre-existing
lung disease, and the interaction between these two
factors, a progressively sophisticated three-tiered
monitoring system should be used (see Table 7-
1 ). The first tier (tier I) includes healthy, young
patients without special intraoperative conditions,
such as a young patient undergoing pleurodesis.
This tier contains the minimal, yet essential, mon-
itoring that is required for any patient undergoing
a thoracic procedure. The second tier (tier II) rep-
resents an increase in risk, caused by either the
presence of special unfavorable intraoperative con-
ditions for relatively healthy patients or the pres-
ence of significant pre-existing cardiopulmonary
disease in patients who will not experience special
unfavorable intraoperative conditions. An example
of the former circumstance is a patient with mild
lung disease having a lobectomy. An example of
the latter circumstance is a patient with moderate
interstitial lung disease who requires an open-lung
biopsy. Anyone with some lung disease undergo-
ing one-lung ventilation for a major thoracic pro-
cedure must be considered to be in a high tier II
category, which actually represents a transition
category between the tier II and tier III continuum.
Finally, a third tier (tier III) of monitoring require-
ments is constructed for patients with significant
pre-existing cardiopulmonary disease and/or for
those who will experience major compromising
intraoperative conditions. An example of such a
patient is one with cor pulmonale undergoing lo-
bectomy or pneumonectomy. Thus, it should be
apparent from this monitoring approach that an
individual with severe pulmonary disease who is
to undergo a minor surgical procedure may well
require as extensive a monitoring system as a pa-
tient with normal lungs who is to have extensive
thoracic surgery.
II. TIER I: ESSENTIAL MONITORING
SYSTEM
An essential monitoring system is used for
healthy patients undergoing simple, physiologi-
cally nonintrusive thoracic procedures (Table 7-
2). Patients belonging exclusively to this monitor-
ing tier make up relatively few of the patients
undergoing thoracic surgery. However, since the
system represents the minimum amount of moni-
toring, it is a component of all levels of monitoring
for all patients and should allow one to anticipate
incipient ventilatory failure as well as to recognize
ventilatory failure when it does occur. Although
superficially this system may seem unsophisti-
cated, an alert anesthesiologist uses the senses of
sight, sound, and touch to gather automatically and
reflexly a great deal of information about the well-
being of the patient. In view of the many possible
mechanical mishaps described in chapter 3 that
can impair respiratory function, monitoring respi-
Table 7-2 TIERED MONITORING SYSTEM BASED ON AMOUNT OF PRE-EXISTING LUNG DISEASE AND PRESENCE OF
SPECIAL INTRAOPERATIVE CONDITIONS
Abbreviations: PIP = peak inspiratory pressure; EKG = electrocardiogram.
236 Monitoring
ratory function during anesthesia begins prior to
the induction of anesthesia with checking the an-
esthesia machine.
A. Checking the Anesthesia Machine
Failure to check equipment properly before the
induction of anesthesia is responsible for 22 per
cent of critical incidents that occur during anesthe-
sia.' Accordingly, preanesthesia check lists have
been written to help fulfill this responsibility com-
pletely.
2
In August, 1986, the U.S. Food and Drug Ad-
ministration (FDA) published its "Anesthesia Ap-
paratus Checkout Recommendations," shown in
Table 7-3. This check applies as a general guide-
line only, and the FDA encourages users to modify
this checkout "to accommodate differences in
equipment design and variations in local clinical
practice. . . . Users should refer to the operator's
manual for special procedures or precautions." In
this respect, the user must understand the basic
arrangement and functions of the anesthesia ma-
chine components so as to apply the correct check-
out. (Check lists based on the 1986 FDA Anesthe-
sia Apparatus Checkout Recommendations are
available from Organon Inc., West Orange, NJ
07052.)
The following mentions only some important
anesthesia machine checkout procedures. Inspect
the machine to ensure that component parts are
connected in the proper order. Vaporizers should
be filled and the caps and drains closed. The cen-
tral gas supply lines should be properly seated in
the appropriate pin-indexed wall or ceiling outlets.
The tanks should have a wrench, be color coded,
be on the proper yokes, and be pin indexed. The
tanks should be opened serially and adequately
pressurized and the flow of gas observed through
the appropriate flowmeter. The carbon dioxide ab-
sorber cannister should be full, functional as indi-
cated by color, and fastened securely. The breath-
ing circuit should be tested for leaks under
pressure, but the circuit should only be partially
filled, and the pressure should be released slowly
to avoid depositing soda lime in the inspiratory
limb.
3
Competence of unidirectional valves should
be tested by breathing into the circuit and by not-
ing any undue resistance, presence of irritating
gas, and motion (sticking) of the directional
valves. The ventilator, scavenger and suction sys-
tems, the monitors, and their alarm settings must
be checked and tested. A head strap should be
available.
For the vast majority of thoracic surgery cases,
the function of the anesthesia machine ventilator
needs to be assessed preoperatively. This can be
done easily for both pressure- and volume-limited,
ventilators by serially attaching a rebreathing bag
to the patient connection, closing the anesthesia
machine overflow valve, turning on the ventilator
to either a high pressure or a volume limit, and;
observing the movements of the bag. The bag,
serving as a test lung, should expand smoothly and
easily and deflate in a similar way through the
ventilator overflow valve.
B. Continuous Monitoring of the
Oxygen-Delivery System
Since some of the causes of failure of the oxy-
gen delivery system listed in chapter 3 can occur
even when the anesthesia machine appeared to
function properly preoperatively (e.g., wrong gas
in central storage tanks, crossed pipe lines, failure
of fail-safe mechanism), the oxygen concentration
in the inspired gas should be continuously moni-
tored.
4
Numerous paramagnetic, polarographic,
and fuel-cell analyzers are available. Ideally, the
oxygen analyzer should have a fast response time
and high and low alarm setting capabilities, which,
when exceeded, trigger audio and visual alarm sig-
nals. A model for the study of commercially avail-
able oxygen monitors has recently been pub-
lished.
5
It should be noted that a mass spectrometer
can function as a most elegant in-line oxygen ana-
lyzer.
If the oxygen sensor is placed on the expiratory
side of the anesthesia circle system (between the
corrugated tubing and the expiratory unidirectional
valve), then the oxygen sensor will detect the min-
imum oxygen concentration in the circuit (usually
the fraction of oxygen in expired gas is 0.05 less
than the fraction in the inspired gas). However, in
this position, the oxygen monitor can also double
as a circuit disconnection alarm (in addition to
continuous chest auscultation, chest observation,
and low-pressure alarms) if a falling bellows ven-
tilator (which is the most common type) is in use.
The falling bellows of the ventilator draws room
air past the oxygen probe, and within one or two
breaths following circuit disconnection the alarm
will sound if the low limit is set above 20 per cent
oxygen/'
(Fig. 7-1)
Precordial and esophageal stethoscopes enable
essential and continuous monitoring of breath
sounds. Bilateral breath sounds may be easily
monitored by joining two taped-on precordial
stethoscopes at a stopcock, which is attached to an
Monitoring 237
Table 7-3 ANESTHESIA APPARATUS CHECKOUT RECOMMENDATIONS'
This checkout, or a reasonable equivalent, should be conducted before administering anesthesia. This is a guideline which users are encouraged to
modify to accommodate differences in equipment design and variations in local clinical practice. Such local modifications should have appropriate
peer review. Users should refer to the operators manual for special procedures or precautions.
*1. INSPECT ANESTHESIA MACHINE FOR:
machine identification number
valid inspection sticker
undamaged flowmeters, vaporizers, gauges, supply hoses
complete, undamaged breathing system with adequate CO,
absorbent
correct mounting of cylinders in yokes
presence of cylinder wrench
2. INSPECT AND TURN ON:
electrical equipment warm-up (ECG/pressure monitor, oxygen
monitor, etc.)
*3. CONNECT WASTE GAS SCAVENGING SYSTEM:
adjust vacuum as required
4. CHECK THAT:
flow-control valves are off
vaporizers are off
vaporizers are filled (not overfilled)
filler caps are sealed tightly
CO, absorber by-pass (if any) is off
*5. CHECK OXYGEN (0
2
) CYLINDER SUPPLIES:
a. Disconnect pipeline supply (if connected) and return cylinder
and pipeline pressure gauges to zero with O, flush valve.
b. Open O, cylinder; check pressure; close cylinder and observe
gauge for evidence of high pressure leak.
c. With the O, flush valve, flush to empty piping.
d. Repeat as in b. and c. above for second O, cylinder, if present.
e. Replace any cylinder less than about 600 psig. At least one
should be nearly full.
f. Open less full cylinder.
*6. TURN ON MASTER SWITCH (if present)
7. CHECK NITROUS OXIDE (N
2
0) AND OTHER GAS
CYLINDER SUPPLIES:
Use same procedure as described in 5a. & b. above, but open and
CLOSE flow-control valve to empty piping.
Note: N,0 pressure below 745 psig. indicates that the cylinder is
less than VA full.
*8. TEST FLOWMETERS:
a. Check that float is at bottom of tube with flow-control valves
closed (or at min. O, flow if so equipped).
b. Adjust flow of all gases through their full range and check for
erratic movements of floats.
9. TEST RATIO PROTECTION/WARNING SYSTEM (if present):
Attempt to create hypoxic 0,/N,0 mixture, and verify correct
change in gas flows and/or alarm.
*10. TEST O, PRESSURE FAILURE SYSTEM:
a. Set O, and other gas flows to mid-range.
b. Close O, cylinder and flush to release O, pressure.
c. Verify that all flows fall to zero. Open O, cylinder.
d. Close all other cylinders and bleed piping pressures.
e. Close O, cylinder and bleed piping pressure.
f. CLOSE FLOW-CONTROL VALVES.
II. TEST CENTRAL PIPELINE GAS SUPPLIES:
a. Inspect supply hoses (should not be cracked or worn).
b. Connect supply hoses, verifying correct color coding.
c. Adjust all flows to at least mid-range.
d. Verify that supply pressures hold (45-55 psig.).
e. Shut off flow-control valves.
*If an anesthetist uses the same machine in successive cases, these
steps need not be repeated or may be abbreviated after the initial
checkout.
* 12. ADD ANY ACCESSORY EQUIPMENT TO THE
BREATHING SYSTEM:
Add PEEP valve, humidifier, etc., if they might be used (if
necessary remove after step 18 until needed).
13. CALIBRATE O, MONITOR:
*a. Calibrate O, monitor to read 21% in room air:
*b. Test low alarm.
c. Occlude breathing system at patient end; fill and empty
system several times with 100% O,.
d. Check that monitor reading is nearly 100%.
14. SNIFF INSPIRATORY GAS:
There should be no odor.
15. CHECK UNIDIRECTIONAL VALVES:
a. Inhale and exhale through a surgical mask into the breathing
system (each limb individually, if possible).
b. Verify unidirectional flow in each limb.
c. Reconnect tubing firmly.
*I6. TEST FOR LEAKS IN MACHINE AND BREATHING
SYSTEM:
a. Close APL (pop-off) valve and occlude system at patient
end.
b. Fill system via O, flush until bag just full, but negligible
pressure in system. Set O, flow to 5 L/min.
c. Slowly decrease 0
2
flow until pressure no longer rises
above about 20 cm ,. This approximates total leak rate,
which should be no greater than a few hundred ml/min. (less
for closed circuit techniques).
CAUTION: Check valves in some machines make it
imperative to measure flow in step c. above when pressure
just stops rising.
d. Squeeze bag to pressure of about 50 cm H,0 and verify that
system is tight.
17. EXHAUST VALVE AND SCAVENGER SYSTEM:
a. Open APL valve and observe release pressure.
b. Occlude breathing system at patient end and verify that
negligible positive or negative pressure appears with either
zero or 5 L/min. flow and exhaust relief valve (if present)
opens with flush flow.
18. TEST VENTILATOR:
a. If switching valve is present, test function in both bag and
ventilator mode.
b. Close APL valve if necessary and occlude system at patient
end.
c. Test for leak and pressure relief by appropriate cycling
(exact procedure will vary with type of ventilator).
d. Attach reservoir bag at mask fitting, fill system and cycle
ventilator. Assure filling/empyting of bag.
19. CHECK FOR APPROPRIATE LEVEL OF PATIENT
SUCTION.
20. CHECK. CONNECT, AND CALIBRATE OTHER
ELECTRONIC MONITORS.
21 CHECK FINAL POSITION OF ALL CONTROLS.
22. TURN ON AND SET OTHER APPROPRIATE ALARMS FOR
EQUIPMENT TO BE USED.
(Perform next two steps as soon as is practical)
23. SET O, MONITOR ALARM LIMITS.
24. SET AIRWAY PRESSURE AND/OR VOLUME MONITOR
ALARM LIMITS (if adjustable).
**A vaporizer leak can only be detected if the vaporizer is turned on
during this test. Even then, a relatively small but clinically significant
leak may still be obscured.
*As developed by the (U.S) Food and Drug Administration, August 1986.
Abbreviations: ECG = electrocardiogram; PEEP = positive end-expiratory pressure; APL = airway pressure line.
238 Monitoring
Monitoring of Only Apnea and Minute Ventilation
Figure 7-1 The number of things that an anesthesiologist does to monitor minute ventilation and to detect continuously the onset
of apnea is considerable. Most of the monitoring utilizes the basic senses of sight, sound, and touch and is done reflexively and
automatically by an experienced anesthesiologist. These automatic monitoring efforts, for the most part, make up the tier I
monitoring level (solid arrows). Additional tier I monitoring modalities include use of an in-line F, 0
:
monitor and low and high
airway pressure alarm systems. Tier II monitoring efforts include movement of in-line spirometer and observation of respiratory
variation in vascular pressures (dashed arrows). Both tier I and tier II monitoring efforts should be continuously conducted in a
systematic and frequently recurring circular pattern (cockpit analogy).
ear piece. This simple device allows one to listen
to both hemithoraces by simply turning the stop-
cock and is useful in a variety of surgical cases
with difficult access to the chest and when lung
isolation is required.
7
s
The system is particularly
useful in neonates and small infants because of
their short trachea and consequent high risk for
endobronchial intubation. Additionally, the anes-
thesiologist should frequently (every minute) scan
the chest, breathing bag, ventilator bellows, and
pressure manometers on the anesthesia machine
for appropriate movement. The anesthesia ma-
chine-mechanical ventilator system should have a
low and high positive-pressure audiovisual alarm
system. This latter monitoring aid must now be
viewed as essential because some accidental anes-
thesia-caused or -related deaths (brain and/or
whole body) have been, and presently are, due to
ventilator disconnections in otherwise healthy pa-
tients. As just discussed, oxygen sensors can serve
as disconnection alarms if placed on the expiratory
limb of an anesthesia circle system and if a falling
bellows ventilator is in use.
For the healthy patient having a physiologically
nonintrusive thoracic procedure, only a crude as-
sessment of minute ventilation (see Fig. 7-1) is
necessary. Minute ventilation is the product of res-
piratory rate per minute and tidal volume. Count-
ing respiratory rate with either spontaneous, as-
sisted, or mechanically controlled respiration
presents no difficulties. However, the limitation in
measuring minute ventilation precisely without in-
strumentation is in accurately measuring the tidal
volume. When a tier I patient is mechanically ven-
tilated, the setting of ventilator bellows (10 to 15
ml/kg) and patient chest movement usually suf-
fices to gauge the tidal volume. However, it should
be realized that the movement of ventilator bel-
lows may be in error by 50 per cent, depending on
respiratory rate, inspiratory-to-expiratory ratio, gas
flow rates, and patient compliance and resistance.
With most anesthesia machine ventilators, the
higher the flowmeter settings (0
2
, N
2
0, air) during
inspiration, the greater will be the tidal volume (by
Monitoring 239
an amount = flow rate X inspiratory time), and
the longer the inspiratory time (lower I:E) for any
given flow rate, the greater will be the tidal volume
(by an amount = flow rate X the inspiratory
time).
9
Thus, a high gas flow rate with a low I:E
ratio can increase minute ventilation by 50 per cent
(with correspondingly lower P
a
C0
2
) above what is
set by the ventilator bellows and respiratory rate.
9
In addition, it should also be realized that, with
spontaneous ventilation, visual or "educated"
hand assessments of chest or anesthesia reservoir
bag movements in questionable or marginal cases
may be misleading,
10
especially with inexperi-
enced observers." Conversely, with respect to con-
trolled ventilation, experienced anesthetists can
achieve a higher minute ventilation with lower
peak and end-expiratory alveolar pressures with
manual compared with mechanical ventilation
with two commonly used anesthesia ventilators
(Ohmeda 7000 and Drger AV-E).
12
In any pa-
tient, an unusually strong respiratory drive in the
presence of high doses of inhaled or intravenously
administered anesthetics may be the consequence
of hypercapnia and hypoxia.
E. Gas Exchange
1. Cyanosis
The simplest and perhaps most common signs
of severe hypoxemia are purple or dark shed
blood'
3 u
and/or the appearance of cyanosis.'
5
'
6
Most of the blood in skin and mucous membranes
is venous and is related to the C
a
0
2
as follows
(rearranged Fick equation; see equation ll in
chapter 3):
tissue Vo-,
C
v
0
2
= C
a
0
2
- C(a-v)0
2
= C
a
0
2
- -r
1
tissue Q
The oxygen consumption by skin (Vo
2
) is very
low in relation to its circulation (Q) so that the
quantity in the far right term is generally small.
Therefore, skin C
v
0
2
is close to C
a
0
2
. Cyanosis
can be detected by experienced observers in 95 per
cent of awake patients with an arterial saturation
(S
;1
0
2
) of 89 per cent with the aid of fluorescent
lighting
17
(corresponds to a P
a
0
2
= 55 mm Hg),
whereas cyanosis does not always become evident
in awake patients in ambient lighting until S
a
0
2
is
72 per cent (corresponds to P
a
0
2
= 36 mm Hg).'
H
It has been observed that cyanosis is not at all
consistently detected in anesthetized patients even
at S
p
0
2
values as low as 72 per cent.
18
This sug-
gests that, during inhalation anesthesia, cyanosis is
a less reliable sign than in awake patients.
15, ,6
l8
The greater difficulty in detecting cyanosis during
halogenated hydrocarbon anesthesia may be re-
lated to relative hypoperfusion of the skin and
mucous membranes.
14
It has also been noted that, during clinical epi-
sodes of hypoxemia (S
p
0
2
< 85 per cent), the
heart rate electrocardiogram (EKG), blood pres-
sure, and respiration hardly change.
18
This obser-
vation is explained by the results of studies of
volunteers that showed that both cardiovascular
and ventilatory responses to induced moderate hy-
poxemia (i.e., PET0
2
40-45 mm Hg) are severely
impaired or abolished by commonly used halogen-
ated anesthetics.
19
~
21
Clearly, cardiorespiratory
monitoring, including the EKG, has little or no
value as an indicator of moderate hypoxemia dur-
ing anesthesia with these agents.
14
In addition, in
cases of anemia in which it is not possible to
obtain the level of 5 g/100 ml of reduced hemoglo-
bin, which is generally accepted to be required for
the appearance of cyanosis, cyanosis may not oc-
cur in the presence of abnormal C
a
0
2
("central
cyanosis").
22
Clearly, cyanosis could never occur
if the hemoglobin concentration was only 5 g/100
ml. Alternatively, when skin circulation is reduced
in relation to skin oxygen consumption, as may
occur in hypovolemia (vasoconstriction) and in the
Trendelenburg position (stagnant flow), cyanosis
may occur in the presence of normal C
a
0
2
(periph-
eral cyanosis). Thus, cyanosis should always be
looked for, but since it can be falsely positive, it
should be regarded as a warning sign requiring
further investigation. Since the absence of cy-
anosis can be falsely negative, the anesthesiologist
should not be lulled into a sense of complacency.
In addition, recognition of cyanosis depends upon
the quality of lighting, reflection from drapes, and
the observers themselves.
2. Pulse Oximetry (SpOJ
Pulse oximetry is the technology whereby the
pulse oximeter provides a noninvasive estimate of
arterial hemoglobin saturation with oxygen. The
American Society of Anesthesiologists Standards
for Basic Intraoperative Monitoring (last amended
October 23, 1990, to become effective January 1,
1991) state (standard II, oxygenation, methods 2)
that "during all anesthesia a quantitative method
of assessing oxygenation, such as pulse oximetry,
shall be employed."
a. BASIC PRINCIPLES AND CLINICAL USE
Pulse oximetry is a combination of two technol-
ogies, namely, spectrophotometry (which meas-
ures hemoglobin saturation by the amount of light
absorbed by the blood) and optical or photople-
thysmography (which compares light absorbance
in the absence of a pulse [low finger blood vol-
240 M on it or H;
urne], its zero, and in the presence of the arterial
pulsation [high finger blood volume]). Pulse ox-
imeters provide instantaneous, in vivo measure-
ments of arterial oxygenation by determining the
color of the blood between a light source and a
photodetector. To be able to distinguish between
two species of hemoglobin (i.e., oxyhemoglobin
[HbO
:
] and deoxyhemoglobin [Hb]), it is neces-
sary to measure absorption at two different wave-
lengths. This is accomplished by using a light
source that consists of two different light-emitting
diodes, one emitting red (^ nm) and the other
infrared (^940 nm) light. Oxyhemoglobin absorbs
less red light than deoxyhemoglobin, accounting
for its red color; at infrared wavelengths, the op-
posite is true.
During each cardiac cycle, light absorption by
tissue beds varies cyclically (Fig. 7-2): During
"diastole," absorption is caused by venous blood,
tissue, bone, and pigments (melanin, nail polish,
etc.); during "systole," there is an increase in light
absorption that pulse oximeters assume to be cre-
ated by the influx of arterialized blood into the
tissue bed. The oximeter then determines the dif-
ference between background absorption during dias-
tole and peak absorption during systole at both red
and infrared wavelengths; these changes correspond
to the absorption caused by arterialized blood.
23
Ox-
ygen saturation determines the red:infrared absorp-
tion ratio; thus, the red:infrared ratio of these pul-
satile differences can be used to compute the pulse
oximeter reading (S
p
0
2
), which is an estimate of
arterial oxygen saturation (S
;1
0
2
).
23
The pulse oximeter will function anywhere it
can reliably detect a pulse (e.g., the digits, ear,
tongue,
24
2S
nose,
26
buccal mucosa
27
). The further
the sensing site is from the lung, the longer it will
take for the pulse oximeter to express the change
in oxygenation (i.e., acute changes in S
a
0
2
may
not be reflected by finger vs. ear S
p
0
2
for as long
as 30 sec vs. 5-10 sec, respectively
28
29
). The re-
sponse time at low S
p
0
2
is not nearly so dependent
on sampling site as are changes at high S
p
0
2
.
3()
Response time in change from S
p
0
2
= 100 pi
cent will also be a function of F,0
2
, whereas
sponse time for change from a low S
p
0
2
will n<
be.
31
A pulse waveform display enables the open
tor to determine whether the oximeter is detectir
valid pulses or interfering signals. By analogy, a
arterial catheter without a waveform display migl
give accurate pressure readings under most cond
tions, but erroneous readings resulting from art
fact would be difficult to recognize. S
p
0
2
valut
may be expressed as a compressed spectral arra
(i.e., as a function time and the amount of tim
spent at a particular S
p
0
2
) (Fig. 7-3).
2
Most mar
ufacturers specify that their oximeter readings ca
be expected to have a 2 to 3 per cent standar
deviation in the 70 to 100 per cent saturatio
range. At these relatively high saturations, the 9
per cent prediction limits of a single pulse oxime
ter reading are 6 per cent (i.e., there is a 95 pe
cent probability that an oximeter reading of 90 pe
cent corresponds to an arterial saturation betwee
84 and 96 per cent). In the range of saturatio
equal to 40 to 70 per cent, the variation may be a
great as 10 per cent.
29
The lack of reliability at lo\
S
p
0
2
is confirmed by studies in infants with S
a
C
less than 60 per cent.
13
The definition of saturation is the cause of muc
confusion. Laboratory co-oximeters separatei
measure deoxygenated hemoglobin (RHb), ox>
genated hemoglobin (Hb0
2
), and the dyshemoglc
bins methemoglobin (MetHb) and carboxyhemc
globin (COHb). From these can be calculated the
Figure 7-2 Components of lighi
absorption by a tissue bed as a func-
tion of time. (From Alexander CM
Teller LE, Gross JB: Principles ol
pulse oximetry: Theoretical anc
practical considerations. Anesth An
alg 68:368-376, 1989. Used with
permission.)
Monitoring 241
Figure 7-3 S
p
0
2
CSA of a 58-year-
old male after laparotomy/adhesiolysis.
Mild hypoxemia was noted in the
first hour, but then patient was in-
structed in the use of 0
2
nasal can-
nula (2 L/min), which he wore all
night. O-, was discontinued by physi-
cian or nurse (pt unsure) after 0800,
after which patient developed in-
creasing hypoxemia with S
p
0
2
in the
low 80s by midafternoon. (From
Cohen MS, Eichhorn JH, Darvish
AH: Postanesthetic hypoxemia due
to removal of supplemental oxygen
therapy. Anesth Analg 74:545,
I992.)
2 per cent, the S
a
0
2
would be 100 per cent (96/0
+ 96 = 96/96), whereas the Hb0
2
per cent would
be 96 per cent (96/0 + 96 + 2 + 2 = 96/100).
However, because pulse oximeters respond to the
dyshemoglobins in a very complex way (see later
discussion), the reading from a pulse oximeter is a
unique value and must be given its own designa-
tion S
p
0
2
. When the dyshemoglobinemia is
mild/moderate, the S
p
0
2
is a close approximation
of both Hb0
2
per cent and S
a
0
2
(see later discus-
sion).
The waveform that a pulse oximeter displays is
essentially a photoplethysmogram. Acute hypovo-
lemia may cause cyclic beat-to-beat variation in
both systolic blood pressure and the pulse oxime-
ter waveform that is synchronous with respiration.
Even though the display has an automatic gain and
hence cannot be considered quantitative, variation
in the pulse wave seems to correlate with systolic
pressure variation and was found to be a useful
guide to fluid therapy.
34
When the pulse oximeter
fails to detect the usual pulse during positive-pres-
sure ventilation (Fig. 7-4),
35
it may mean that the
patient is hypovolemic and the extremity blood
flow is further reduced by a positive-pressure ven-
tilation-induced decrease in venous return. In this
situation, instead of readjusting the pulse oximeter
or changing the pulse oximeter to a new site, pos-
itive-pressure ventilation should be briefly inter-
rupted to obtain an accurate reading (see Fig. 7-1 )
and fluid boluses administered to restore perfusion
and pulse oximeter function during positive-pres-
sure ventilation. Certainly, in oximeters that dis-
play the pulsatile waveform and that can fix gain
(signal strength), the amplitude of the waveform
may be helpful as an absolute monitor of the cir-
culatory volume and as an evaluation of the he-
modynamic significance of cardiac arrhythmias.
b. LI MI TATI ONS OF PULSE OXI METRY
A pulse oximeter will not function well under a
variety of conditions (Table 7-4). First, to measure
pulse-added absorbance, a pulse oximeter requires
a pulsatile arterial bed to be present. Readings may
Figure 7- 4 Top panel, Waveform
during positive-pressure ventilation.
Displayed rate is equal to respiratory
rate: Displayed saturation is low.
Bottom panel, Waveform during ex-
piratory pause in ventilation. Dis-
played rate is equal to heart rate; dis-
played saturation is normal. (From
Scheller J, Loeb P: Respiratory arti-
fact during pulse oximetry in criti-
cally ill patients. Anesthesiology
69:602-603, 1988. Used with per-
mission.)
242 Monitoring
Table 7-4 LIMITATIONS OF PULSE OXIMETRY
1. Will not function without a pulse
2. Motion artifact
3. Interference from large venous pulsation
4. Interference from dyes, nail polish, dark skin, jaundice
5. Interference by dyshemoglobins (MetHb and COHb)
6. Interference by electrocautery, ambient light, and infrared
radiation
7. Insensitive with high P^CX
not be obtainable when pulsatile flow is lost, such
as during intense vasoconstriction, hypothermia,
severe peripheral vascular disease, hypovolemia,
or cardiopulmonary bypass. Hypothermia-induced
loss of pulse oximeter function, presumably result-
ing from vasoconstriction, may be reversed by vo-
lar digital nerve blocks.
36
However, from a rheo-
metric standpoint, it should always be remembered
that the presence of a functioning pulse oximeter
should not be construed as evidence of adequate
tissue oxygenation or oxygen delivery to vital or-
gans."
Second, patient or probe movement may inter-
fere with oximeter function because these motions
cause changes in optical path length. The oximeter
may then fail to recognize actual pulsatile flow.
Flexible, adhesive probes, which are taped in
place, tend to be less susceptible to motion arti-
facts than clip-on probes (i.e., the orientation of
the probe to the finger is more stable).
Third, pulse oximetry requires a nonpulsatile
venous bed. Venous pulsations (e.g., caused by
tricuspid regurgitation) are not distinguishable
from arterial. Venous pulsations will contribute to
the pulse-added absorbance signal, causing a spu-
rious desaturation reading.
Fourth, pulse oximetry requires hemoglobin as
the dominant color species present. Intravascular
dyes, if present, may interfere with function. Meth-
ylene blue and indocyanine green cause spurious
desaturation, whereas indigo carmine has little or
no effect on the S
p
0
2
reading. Similarly, certain
colors of nail polish, skin pigmentation, and jaun-
dice may cause spurious readings.
38
Fifth, the saturation reading is based on absence
of significant amounts of dyshemoglobins (e.g.,
MetHb, COHb). Their presence may result in spu-
rious readings. Oximeters respond to COHb as if
it were HbO,, and the pulse oximeter reading
(S
p
O
;
) is the sum of COHb and Hb0
2
(i.e., if a
patient had 90 per cent Hb0
2
, 7 per cent COHb,
and 3 per cent Hb, a pulse oximeter would read
S 0
2
= 97 per cent). MetHb corresponds to an
S
p
0
2
reading of 85 per cent. Therefore, in the
presence of MetHb, the pulse oximeter reading
will be a weighted average of the 0
2
saturation of
the hemoglobin available for transport plus 85 per
cent. With fetal hemoglobin and the hemoglobin-
opathies, the heme moiety itself is unaffected, and
such changes do not produce clinically significant
alterations in S
p
0
2
readings.
Sixth, pulse oximeters are sensitive to interfer-
ence by electrocautery, intense ambient light, and
infrared radiation.
Seventh, pulse oximeters are insensitive to
changes in oxygenation at the high end of the
oxyhemoglobin dissociation curve. At arterial ox-
ygen tensions above about 120 mm Hg, saturation
is 100 per cent. The pulse oximeter is thus a poor
predictor of P
a
0
2
until the saturation falls below
100 per cent. A patient receiving an F,0
2
of 1.0
may have a large shunt and yet still show an S
p
0
2
of 100 per cent. For this reason, the pulse oximeter
is a poor indicator of endobronchial intubation or
a poor early indicator of esophageal intubation,
especially if the F,0
2
is greater than O.5.
31
c. USE OF S
p
0
2
FROM THE DIGIT OF THE
DEPENDENT HAND IN THE LATERAL DECUBITUS
POSITION DURING ONE-LUNG VENTILATION
It is clear that pulse oximetry performs well,
with a predictable error range of 2 per cent in
normal patients who do not have any of the limi-
tations just discussed. However, in a patient in the
lateral decubitus position undergoing one-lung
ventilation, there are many potential (and perhaps
unavoidable) sources of error. These potential
sources of error are rapidly changing oxygenation
levels with lung surgery (e.g., clamping, compres-
sion of vessels, and lung parenchyma), which are
reflected relatively slowly at a distal sensing site,
venous congestion in a dependent extremity, hy-
povolemia, decreased cardiac output with impaired
venous return, arrhythmias, hypothermia with the
chest open, large tidal volume positive-pressure
ventilation of one lung, and electrocautery.
One study very clearly highlights this problem
by showing that there was excellent agreement
between S
p
0
2
and S
a
0
2
while a series of thoracic
surgery patients were awake, but once anesthesia
had been induced, the patients were in the lateral
decubitus position, and surgery had begun, the
agreement between S
p
0
2
and S
;
,0
2
was much
worse.
39
These data support the conclusion that the
accuracy of pulse oximeters is altered during an-
esthesia for thoracic surgery and should be relied
on only as a warning of impending hypoxemia
when the S
p
0
2
falls. This recommendation is sim-
ilar to that of Chung et al.,
40
who also found that
transcutaneous monitors of oxygen were not accu-
rate during thoracic surgery. It was appropriately
concluded that, although the information derived
from pulse oximetry is useful in identifying im-
pending hemoglobin desaturation caused by hy-
poxemia, S
p
0
2
alone should not be used as a sub-
Monitoring 243
stitute for frequent arterial blood gas determinations
during thoracic surgery.
3. Capnometry and Capnography
a. BASIC PRINCIPLES
Airway C0
2
can be monitored continuously by
either small, freestanding operating room infrared
spectrophotometers or large, centrally housed res-
piratory mass spectrometers. Freestanding infrared
units are by far the most commonly used because
there is no delay in read-out and there are two
types: in-line and side-stream monitors. The in-
line C0
2
sensor is connected directly to the endo-
tracheal tube and receives the total alveolar gas
flow, whereas the side-stream type receives a por-
tion of the alveolar flow via a side-port connector.
The in-line type has the intrinsic advantage of
maximal frequency response and minimal delay
and is therefore most suitable for infants. Disad-
vantages of the in-line type include calibration in-
accuracy while in use and the encumbrance of a
sensing unit at the patient interface. The weight of
the transducer and its associated connections may
restrict patient movement and may necessitate se-
dation of restless patients. The side-stream monitor
adds essentially no encumbrance to the patient in-
terface and has the additional advantage of in-use
calibration capability. However, to achieve ade-
quate frequency response, side-stream flows be-
tween 125 and 500 ml/min are required. These
rates are too high for accurate end-tidal measure-
ments in neonates and infants.
The best measure of PETC0
2
is obtained when
(1) tidal volumes are large enough to displace dead
space; (2) fresh gas flow rates are low enough to
prevent dilution or washing out of C0
2
; (3) sample
aspiration rates are low enough that they do not
interfere with patient ventilation or entrain air that
may dilute the C0
2
; (4) the sampling site is close
to the patient, minimizing the dead space; and (5)
the waveform is displayed for end-tidal alveolar
plateau analysis.
41
Despite the use of elegant mu-
cus traps and filters, occlusion of sampling lines
and sensor by aspirated mucus and debris is a
common clinical problem.
In infants and small children (weighing 12 kg
or less) ventilated with a partial rebreathing circuit,
the sampling site becomes a significant dependent
variable. Capnographic waveforms from distal en-
dotracheal tube sampling sites must and do show
constant plateau phases during expiration, whereas
those from the proximal endotracheal tube sites
fail to achieve a plateau and underestimate the
P
a
C0
2
.
42
This is consistent with the hypothesis that
gas sampled at the proximal site is a mixture of
expired alveolar gas and fresh gas removed from
the circuit by sampling. In any situation in which
PETC0
2
does not correlate with what the clinician
expects (e.g., PETC0
2
monitoring in infants),
P
a
C0
2
should be measured as a guide for PETCO
:
tracking and interpretation. A change in the
P
a
C0
2
-PETC0
2
gradient in itself may indicate an
important pathophysiologic change (e.g., change
in dead space; see later discussion).
b. CLINICAL USE
The anesthesiologist will obtain the most infor-
mation out of a capnograph if it is examined sys-
tematically. First, the anesthesiologist must deter-
mine whether exhaled C0
2
(i.e., a waveform) is
present. Second, the shape of the waveform must
be analyzed systematically by looking at, and in
sequence, phase I (inspiratory baseline), phase II
(expiratory upstroke), phase III (expiratory pla-
teau), and phase IV (inspiratory downstroke) (Fig.
7-5). This discussion will follow this systematic
approach but will emphasize diagnoses that can be
obtained from the phase III alveolar plateau. In the
future, capnography may display carbon dioxide
concentration as a function of the volume of ex-
haled gas from each breath (the single-breath test,
or SBT) because the SBT allows on-line determi-
nation of Vco
2
(C0
2
concentration X tidal vol-
ume) and is a more sensitive test of, and reveals
more information about, gas elimination in the late
phase of each breath.
43
( 1) . IS EXHALED CO
z
(WAVEFORM) PRESENT? If
the presence or persistence of C0
2
is not detected
by the capnometer, or capnogram, failure to venti-
late the patient's lungs must be assumed. The dif-
ferential diagnosis of absent C0
2
includes esopha-
geal intubation, accidental tracheal extubation,
disconnection of the breathing circuit, complete
obstruction of the endotracheal tube, conducting
system (kink, inspissated blood or secretions, ex-
tremely severe bronchospasm), or breathing cir-
cuit, apnea, and cardiac arrest (Table 7-5).
44
With respect to esophageal intubation, capnom-
etry/capnography can make the diagnosis in one
breath and is therefore far superior to pulse oxime-
try, which usually requires some time for desatu-
ration to occur.
31
If the stomach contains exhaled
gas from previous mask ventilation attempts or
carbonated beverages, a few tidal ventilations
through an esophageal tube may contain minimal
and progressively diminishing concentrations of
carbon dioxide. It should be noted that very severe
bronchospasm (enough to cause complete airway
obstruction) can prevent carbon dioxide from
being registered by the capnogram even though
the trachea has been properly intubated.
45
Only
after all of these life-threatening possibilities have
been quickly ruled out and ventilation of the pa-
tient's lungs confirmed by clinical examination
244 M ont tarins
EXHALED
C02
CONCENTRATION
Figure 7-5 The four phases of the capnogram.
should failure of the capnometer or capnogram be
considered in the differential diagnosis. A rapid
qualitative check of the capnogram consists of
simply removing the C0
2
sensing or sampling port
and exhaling into it.
Of all the entities in the differential diagnosis
listed previously, monitoring cardiac output and
cardiopulmonary resuscitation during cardiac ar-
rest is a new application of capnography and there-
fore is the only entity further discussed.
(2). MONITORING CARDIAC OUTPUT AND CAR-
DIOPULMONARY RESUSCITATION BY END-TIDAL
CARBON DIOXIDE CONCENTRATION. During
steady-state gas exchange equilibrium, the alveolar
Pco
:
(
0
2
), tissue C0
2
production (Vco
2
), and
ABSENT EXHALED C0
2
I. Exhaled CO
:
usually present, then absent
1. Accidental tracheal extubation
2. Disconnection of breathing circuit
3. Monitor failure
4. Complete obstruction of endotracheal tube
5. Cardiac arrest
6. Patient becomes apneic
II. Exhaled C0
3
usually absent/minimal initially, remains ab-
sent
1. Esophageal intubation
alveolar ventilation (V
A
) are uniquely related as
given by P
A
C0
2
= (K)Vco/V
A
. During constant
minute ventilation and Vco
2
, an abrupt reduction
in cardiac output (Q,) reduces PETCO, by two
mechanisms.
46 47
First, a reduction in venous re-
turn causes a decrease in C0
2
delivered to the
alveolar compartment, resulting in decreased
P
A
C0
2
. Second, the increase in alveolar dead
space, which results from the reduced pulmonary
vascular pressures, will dilute the C0
2
from nor-
mally perfused alveolar spaces to decrease PETCO
:
below P
A
C0
2
(see later discussion). During a sus-
tained reduction in Q,, increasing C0
2
accumula-
tion in the peripheral tissues and in venous blood
will begin, after 10 to 20 min, to restore C0
2
delivery to the lung and PETC0
2
toward baseline
levels. Reciprocal changes in PETCO
:
will occur
during acute increases in Q,. These Q, versus
PETC0
2
observations have been made quantita-
tively and with very good correlation in both con-
trolled experiments in animals
46
and in patients
undergoing major vascular and cardiac surgery.
47
With cardiac arrest, there is no pulmonary blood
flow and therefore no delivery of carbon dioxide
to the lungs. Consequently, PETC0
2
exponentially
decreases over a dozen breaths, and there is no
steady-state exhaled C0
2
(waveform).
48
However,
with external cardiac compression, pulmonary
blood flow will begin again and the amount of
Monitoring 245
C0
2
excreted by the lungs (i.e., PETC0
2
) will be equipment being used. The inspiratory baseline be-
proportional to the amount of pulmonary blood
flow (see previous discussion). Indeed, the efficacy
of external cardiac compression can be continu-
ously and quantitatively followed by the amount
of C0
2
excreted (Fig. 7-6, top panel).
49
50
Exhaled C0
2
during cardiopulmonary resuscita-
tion can also be used prognostically.
5153
In one
study the initial PETC0
2
was 19 mm Hg in those
who eventually regained spontaneous pulses but
only 5 mm Hg in those who did not (p < .0001 ).
54
Furthermore, a sharp increase in ETco
2
often her-
alded, and was the first indicator of, the resump-
tion of spontaneous circulation.
52 54
Almost identi-
cal data were found in yet another study (Fig. 7-
6, bottom panel).
49
Figure 7-7 shows all of this, in
addition to the effects of sodium bicarbonate infu-
sion, in a patient undergoing ventricular fibrillation,
cardiopulmonary resuscitation, and resumption of
a spontaneous circulation.
51
Thus, capnography
provides an instantaneous and continuous guide to
the efficacy of external chest compression and the
resumption of spontaneous pulmonary perfusion.
(3). PHASE i: INSPIRATORY BASELINE. The in-
spiratory baseline is traced as fresh gas moves over
the C0
2
sensing or sampling site. The C0
2
level
during this phase should be zero; if it is not, C0
2
is being rebreathed. This may be intentional and/or
a characteristic (desirable or undesirable) of the
comes elevated if C0
2
is added to the fresh in-
spired gas, if a C0
2
rebreathing or by-pass valve
(which is present on older anesthesia machines) is
open, if the C0
2
absorbent is partially exhausted
or gas is channeling through the absorbent, if the
expiratory valve is missing or incompetent (ex-
haled gas in the exhalation limb goes back into the
patient's lungs during inhalation, thereby pulling
C0
2
containing gas by the sampling site during
inhalation), or if a Bain circuit is being used.
( 4) . PHASE II: EXPIRATORY UPSTROKE. Soon
after exhalation begins, C0
2
containing gas arrives
at the C0
2
sampling site, and it quickly washes
away the fresh gas from the previous inspiration.
Thus, the expiratory upstroke is steep. When the
expiratory upstroke phase of the capnogram be-
comes prolonged (i.e., the upstroke becomes less
steep), delivery of C0
2
from the lungs to the CO
:
sampling site is delayed. Possible causes include
mechanical obstruction in the equipment, such as
a kinked endotracheal tube, or slow emptying of
the lungs, such as with chronic obstructive pul-
monary disease (COPD) or bronchospasm. The ex-
piratory upstroke also becomes prolonged when a
side-stream capnogram samples gas too slowly or
when the capnogram has a slow response time and
the respiratory rate is fast.
( 5) . PHASE III: EXPIRATORY (ALVEOLAR) PLA-
Figure 7-6 Top panel, Infrared analyzer trace showing end-tidal carbon dioxide concentration with two different physicians
giving external chest compression. Bottom panel, infrared analyzer trace showing changes in end-tidal carbon dioxide concentration
with successful defibrillation and recurrence of ventricular fibrillation. (From Nielson MS, Fitchet A, Saunders DA: Monitoring
cardiopulmonary resuscitation by end-tidal carbon dioxide concentration. Br Med J 300:1012-1013, 1990. Used with permission.)
246 Monitoring
Figure 7-7 Serial changes in the end-tidal carbon dioxide concentration (ETco
:
) and arterial (A) and mixed venous (PA) bio
gases in a representative patient before and immediately after cardiac arrest, during precordial compression, and after detibrillati
(DF) and resuscitation. The transient increase in the ETco
:
after the administration of sodium bicarbonate (NaHCO,) is a
demonstrated. The original tracing has been modihed because of space limitations. (From Falk JL, Rackow EC, Weil :
tidal carbon dioxide concentration during cardiopulmonary resuscitation. Engl J Med 318:607-61 1. 1988. Used with permissio
TEAU. The expiratory plateau should detect mixed
alveolar gas at the CO-, sampling site. During ini-
tial exhalation, elimination of gas from the ana-
tomic dead space (infinite V/Q, zero CO
;
concen-
tration) is followed by elimination of gas from the
well-ventilated, low-resistance regions of the lung
(relatively high V/Q, low CO, concentration).
Later, gas from poorly ventilated, high-resistance
regions of the lung (relatively low V/Q, high CO-,
concentration) is eliminated. The continuum of
V/Q ratios between high and low V/Q areas cre-
ates a positive slope (upward to the right) to the
alveolar plateau part of the C0
2
elimination wave-
form, with the result that the end-tidal CO, con-
centration is the last and highest (the peak) con-
centration on the alveolar or the expiratory
plateau.
55,56
In addition, continued production and
evolution of C0
2
into the alveolar space during
exhalation contributes to the rise in C0
:
concentra-
tion during exhalation. Theoretically, the slope is
such that the end-tidal value should be about 2 per
cent higher than the time-weighted mean in normal
resting subjects and about 4 per cent higher in
exercising subjects, assuming tidal volumes are
large enough to displace dead space.
41
Because the alveolar plateau expresses the V/Q
continuum in the lung, analysis of the alveolar
plateau may result in a wealth of diagnostic infc
mation. First, the steepness of the alveolar plate
is directly related to the degree of airway res:
tance. Second, biphasic waveforms may reveal t
presence of a two-compartment lung. Third, lea
in the sampling system may alter the alveolar pi
teau in a characteristic and, at first glance, peculi
way. Fourth, the P. , CO
:
- PETC0
2
gradient is c
rectly related to alveolar dead space.
Steepness of the Slope of the Alveolar Plates
Is a Function of Expiratory Resistance. Whi
the lungs of a patient without lung disease a
being mechanically ventilated, the V/Q units in ti
lung are relatively uniform and homogeneo
(have the same C0
2
concentration), and the ex
ratory plateau is smooth and nearly horizont;
However, when there is significant lung disea
and a large spread in V/Q ratios within the lun
very well-ventilated, high-V/Q, low-CO
:
conce
tration areas empty first, causing the alveolar pi
teau to be relatively low. Following this, ve
poorly ventilated, low-V/Q, high-CO
:
concentr
tion areas empty, causing the alveolar plateau
be relatively high. Thus, with a large spread in tl
V/Q ratios, the upward positive slope of the alve
lar plateau will be very steep to the right (it
possible, but unusual, for there to be simultaneoi
Monitoring 247
Figure 7-8 The phase III alveolar plateau
slopes upward to the right because of a
spread in ventilation-perfusion (V/Q) ratios
from high (early phase III) to low (late phase
III). Bronchospasm spreads the distribution
of V/Q ratios and increases the slope of
phase III. (R
airvi
= airway resistance.)
emptying of alveoli with very different V/Q ratios
and therefore a minimal slope to the expiratory
plateau). Thus, it is not surprising that the increase
in airway resistance that is associated with bron-
chospasm, which causes a large spread in the V/Q
ratios and emptying times, to be tightly correlated
with an increase in the slope of the alveolar pla-
teau (Fig. 7-8).
57
Biphasic Waveform in Patients Who Have
Markedly Different Individual Lungs. Theoreti-
cally, if the continuum of the V/Q ratios is broken
into two distinctly different lung regions (a low-
resistance, high-V/Q, low-CO
:
concentration
region and a high-resistance, low-V/Q, high-C0
2
concentration region), a biphasic C0
2
excretion
waveform might be expected. Such a biphasic C0
2
waveform has been described in the patient who is
in the lateral decubitus position, in which the non-
dependent lung has relatively low airway resis-
tance, high V/Q ratio, and low C0
2
concentration
compared with the dependent lung; in a patient
with severe rotary kyphoscoliosis causing severe
compression of one lung (Fig. 7-9)
58
; and in a
major main-stem bronchial intubation (Fig. 7-
10).
39
Some patients with COPD may also display a
slight biphasic expiratory plateau if they have,
throughout both lungs, two distinct populations of
alveoli with very different time constants. In this
situation, rapidly exchanging alveolar spaces are
overinflated during inspiration (their compliance is
high) so that their C0
2
concentration is low.
whereas slower exchanging alveoli empty only
during the latter part of exhalation, releasing a
higher C0
2
content.
60
In patients with active expiratory efforts, a sim-
ilar pattern may also be precipitated by airway
closure because of increased intrapleural pressure
during expiration.
60
Finally, spontaneous breathing
effort during a mechanical positive-pressure breath
will create a cleft in the expiratory plateau and
therefore a biphasic appearance.
Sampling Line Leak Equals Very Unusual
Waveform. When there is a leak in the sampling
Figure 7-9 Biphasic CO, excre-
tion waveform during manual inter-
mittent positive-pressure breathing.
(From Nichols K, Benumof JL, Clau-
sen J, Ozaki G: Expiratory CO
:
pla-
teau slope predicts airway resistance.
Anesthesiology 71:A1072, 1989.
248 Monitoring
Figure 7-10 Left panel (read right to left), Biphasic capnogram during right main-stem bronchial intubation. Each horizonta
line indicates 10 mm Hg C0
2
concentration. The first C0
2
concentration peak ranged from 21 to 23 mm Hg, and the second CO)
concentration peak ranged from 26 to 29 mm Hg. No spontaneous ventilation was apparent during this period of controlled
ventilation. Right panel (read right to left), Normal appearing capnogram after the tip of the endotracheal tube was pulled baclj
above the tracheal carina. Each horizontal line indicates 10 mm Hg C0
2
concentration: The end-tidal C0
2
concentration at the en<j
of this breath was 28 mm Hg. (From Gilbert D, Benumof JL: Biphasic carbon dioxide elimination waveform with right mainsten
bronchial intubation. Anesth Analg 69:829-832, 1989. Used with permission.)
Figure 7-11 Top panel, Photograph of a C0
2
excretion waveform when the C0
2
sampling line was loosely connected to the
C0
2
analyzer sample port. The C0
2
excretion waveform consists of a long, low plateau followed by a brief peak. This patient was
being ventilated with an inspired 0
2
concentration of 30% and an inspired N
2
0 concentration of 70% (as indicated by flowmeter
settings). Bottom panel,-Photograph of the C0
2
excretion waveform from the same patient as in the top panel, but when the
connection between the C0
2
sample line and C0
2
analyzer sample port was made tight. The C0
2
excretion waveform is now almost
a square or rectangular wave. Note that the mean exhaled 0
2
and N
2
0 concentrations are now near the inspiratory settings. (From
Zupan J, Martin M, Benumof JL: End-tidal C0
2
excretion waveform and error with gas sampling line leak. Anesth Analg 67:579-
Monitoring 249
line and the sampling line can entrain room air,
the alveolar plateau will be artificially low. In ad-
dition, when the next positive pressure inspiration
forces gas through the sampling line at a faster
rate, so that room air is no longer entrained, a peak
will follow the low plateau and the peak will be
equal to the true ETco
2
(undiluted end-tidal gas
being pushed through the sampling line) (Fig. 7-
ll).
61
The P
8
C0
2
- P E T C0
2
Gradient Equals Alveo-
lar Dead Space. The P
a
C0
2
-PETC0
2
gradient is a
function of the temporal sequence of alveolar emp-
tying (i.e., the slope of the phase III of the SBT
and the level to which the C0
2
concentration rises)
and the total dead space in the lung.
56
62
The alveo-
lar concentration of C0
2
(P
A
C0
2
) is ordinarily just
slightly less than the P
v
C0
2
but slightly higher
than P
a
C0
2
(Fig. 7-12). However, the alveolar
concentration of C0
2
is ordinarily diluted by al-
veolar and anatomic dead-space gas, which has no
C0
2
in it so that the end-tidal C0
2
ordinarily is
less than the P
a
C0
2
(see Fig. 7-12).
If the lungs are small (reduced FRC) and ho-
mogeneous (normal V/Q relationship and no al-
veolar dead space), and the anatomic dead space
is small (these conditions occur in healthy, supine,
term, pregnant women about to undergo cesarean
section, patients having postpartum tubal ligations,
and women in early pregnancy)
63-
*
5
and in exer-
cise,
66
then P
A
C0
2
is only diluted to a small extent
by C0
2
free gas and PETC0
2
may be greater than
P
a
C0
2
. The existence of negative arterial to end-
tidal Pco
2
gradients is best understood if it is re-
membered that P
a
C0
2
represents the temporal and
spatial mean alveolar Pco
2
(i.e., physiologic inte-
grator), whereas the end-tidal C0
2
is the high-
est (peak) alveolar Pco
2
coming from slow and
low V/Q areas (especially when the tidal volume
is large and the respiratory rate is low).
62-66
How-
ever, normally the amount of anatomic dead space
is large enough so that the P
a
C0
2
-PETC0
2
is
slightly positive (by 2-4 mm Hg).
55
62
As alveolar dead space progressively increases,
the P
a
C0
2
-PETC0
2
gradient progressively in-
creases.
62
67
In patients with pulmonary disease,
the P
a
C0
2
-PETC0
2
gradient increases unpredicta-
bly by 10 to 20 mm Hg or more, with the result
that PETC0
2
is no longer a reliable reflection of
the effectiveness of ventilation and P
a
C0
2
.
68
Although trends in PETC0
2
may still be useful.
55
one author found that the trends were not useful in
patients with severe lung disease.
69
Trends in
P
a
C0
2
-PETC0
2
may not follow P
a
C0
2
in patients
with severe lung disease because a fall in PETCO
:
may be associated with an equal fall in P
a
CO
:
(no change in distribution of V/Q), with a greater
fall in P
a
C0
2
(as a result of better ventilation and
improvement in V/Q matching), or with a constant
or increased P
a
C0
2
(because of increased dead
space and worsening of V/Q matching).
All authors agree that, in patients with severe
respiratory failure, the P
3
C0
2
- P ETC0
2
gradient is
usually a good index of efficiency of ventila-
tion and VP/V. Causes of increased alveolar dead
space include decreased pulmonary blood flow
(decreased cardiac output and/or pulmonary artery
pressure), pulmonary embolization, kinking of pul-
monary vessels, thrombosis in the pulmonary cir-
culation, application of positive end-expiratory
pressure (PEEP), and significant dilation of the
tracheal bronchial tree (see chapter 3, Fig. 3-50,
and Fig. 7-12). A large shunt across the lungs will
increase the P
a
C0
2
-PETC0
2
gradient to a small
extent (because of an increase in P
a
C0
2
resulting
from admixture of P
c
C0
2
through the shunt).
67
During thoracotomy, the P
a
C0
2
-PETC0
2
from
the dependent and nondependent lungs alone (in-
dividually) is mainly a function of the pulmonary
artery pressure. In one study,
70
in the lateral posi-
tion, P
a
C0
2
-PETC0
2
was zero for the lower lung
Figure 7-12 The P
a
C0
2
-P
I T
C0
2
gradient is directly proportional to
the amount of alveolar dead space
because C0
2
-free gas from alveolar
dead space dilutes C0
2
-containing
gas from gas-exchanging alveolar
ventilation (P
A
C0
2
) (the origin of
P
A
CO
:
is C0
2
in the mixed venous
blood [PvC0
2
]) and decreases the
concentration, usually below the ar-
terial level (P
a
C0
2
).
250 Monitoring
and 11 mm Hg for the upper lung. It may be
surmised that if PETC0
2
had been measured in the
combined expirate, the P
a
C0
2
-combined PETC0
2
difference would be large, especially because the
upper lung has more volume to empty (i.e., it
makes the greater contribution to late expiration).
Incision of the chest wall produced an increase in
mean pulmonary artery pressure,
71
which was as-
sociated with an increase in C0
2
elimination by
the upper lung and a decrease in dead space. Con-
sequently, a surgical stimulation-induced increase
in mean pulmonary artery pressure causes a de-
crease in upper lung P
a
C0
2
-PETC0
2
.
(6). PHASE IV: INSPIRATORY DOWNSTROKE.
Soon after inspiration begins, fresh gas from the
breathing circuit washes C0
2
from the previous
exhalation away from the C0
2
sampling site. Be-
cause the volume of gas at the sampling site is
small, the inspiratory downstroke (like the expira-
tory upstroke) is steep. A missing inspiratory valve
or an inspiratory valve that is incompetent during
exhalation allows C0
2
containing gas to go up the
inspiratory limb. In this case, during inspiration,
the C0
2
previously exhaled into the inspiratory
hose is pushed back into the patient's lung and the
inspiratory downstroke becomes prolonged and
slanted.
4. Approximation of Arterial Blood-Gas
Tensions Using Various Other Body
and Equipment Compartments
Frequently, it may be desirable to obtain a rough
approximation of arterial blood gas tensions in tier
I patients by measuring gas tensions and compart-
ments that are in some sort of equilibrium with
arterial blood. Venous blood from the back of a
warmed hand or large neck vein obtained with no
tourniquet or a short-time low-pressure tourniquet
has a Pco
2
near enough to that of arterial blood
(usually 4 to 8 mm Hg) to be useful for most
clinical purposes.
72,73
In addition, under these sam-
pling conditions, a sufficiently high oxygen ten-
sion and saturation of venous blood (greater than
40 mm Hg and 75 per cent, respectively) provide
a good indication that arterial hypoxemia is most
probably absent.
As an alternative sampling site, capillary ear
lobe blood (obtained by needle prick or cut) is
usually acceptable for clinical management be-
cause this blood has a P
a
C0
2
2.0 per cent of
simultaneous arterial samples.
74
The ear lobe vas-
cular bed is more accurate than finger-tip or heel
vascular beds with respect to Pco
2
determina-
tions.
75
P
a
C0
2
can be very roughly approximated
by sampling gas from an endotracheal tube cuff
after 1 hour of cuff inflation at patient diastolic
blood pressure.
76
Last, the C0
2
absorber cannister should be
for warmth and the color observed, particula
during induction of anesthesia. The temperai
and color of the cannister are proportional to t
patient's production of C0
2
. When the dread
complication of malignant hyperpyrexia occurs,
early indication of what is happening is an increi
in cannister warmth and perhaps changes in coll
This situation demands special monitoring and \
gent treatment.
F. Airway Mechanics
Maintenance of a clear, compliant, low-resl
tance tracheobronchial tree is of paramount imp<
tance to the anesthesiologist, and in monitori
this aspect of respiratory function, the stethosco
and the feel of the breathing bag are again esse
tial monitoring tools. The stethoscope allows c
tection of adventitious breath sounds (rales, rhc
chi, and especially bronchospasm). The breathi
bag indicates bronchospasm by classically empl
ing and refilling slowly, and changes in the feel
the reservoir bag when squeezed indicate chang
in compliance. During the induction of anesthes
the anesthesiologist should continuously listen 1:
air movement with a stethoscope over the chest
trachea and, during the maintenance of anesthes
with an esophageal stethoscope (preferable)
with a stethoscope taped over the lateral aspect
the dependent lung.
The peak inspiratory pressure (PIP) should
noted. Assuming the tidal volume remains ccj
stant (as delivered by a volume-controlled
chanical ventilator), increases in PIP should refle
increases in airway resistance and decreases
lung compliance (see tier II: Airway Mechanic!
The slope of the alveolar or expiratory plate}
should be noted; a steep or increasing slope inc
cates an increase in airway resistance and vii
versa (see prior discussion).
57
Last, the volume of air used to inflate the end
tracheal tube cuff needs to be adjusted to a jus
seal volume. This maneuver is important prophj
lactically in preventing postoperative upper airwj
obstruction secondary to tracheal mucosal eden
(reactive hyperemia) caused by excessive ci
pressures. This maneuver will also avoid herni
tion of the balloon cuff into an obstructing positic
as well as invagination of compression of the
men of the tube. If nitrous oxide anesthesia
being used, the just-seal volume needs to be a
sessed periodically, since nitrous oxide will diffu
into the cuff and increase cuff volume and pre
sure.
77
Monitoring 251
Because cardiovascular function can affect res-
piratory function and vice versa, monitoring car-
diovascular function has important respiratory
function implications. Essential monitoring of the
cardiovascular system consists of heart rate, non-
invasive systemic blood pressure, and electrocar-
diogram (EKG). As previously mentioned, pulse
oximetry is also an independent heart rate monitor.
A multiple (five-wire) lead system should be
available for surgical patients with cardiac disease
and consists of leads I, II, III, aVR, aVL, aVF, and
V
5
; 90 per cent of ischemic events are first de-
tected by the true unipolar V
5
lead. However, most
noncardiac surgical operating room EKG systems
have only three electrode wires. Therefore, the
most popular modified three-lead system for diag-
nosing ischemia is the CS
5
bipolar lead in which
the left arm lead is placed in the V
3
position, the
right arm lead is put over the right shoulder or
arm, and the remaining lead is put on the leg (Fig.
7-13). Selecting lead I then records the right arm
to V
5
bipolar CS
5
lead, which is almost as good as
true V
5
to detect ischemia. If a good wave is
also desired, the CB
?
lead can be used by placing
the right-arm electrode on the back and the left-
arm electrode in the V
5
position (see Fig. 7-13).
ST-segment changes of myocardial ischemia
may be subtle and difficult to recognize from an
oscilloscope. Therefore, it is recommended that a
calibrated EKG recorder be used for all surgical
cases. ST-segment changes are not an early sign
of myocardial ischemia. In fact, many ischemic
events occur in the left ventricle before the EKG
changes. These include regional myocardial dys-
function as seen by echocardiography, decreased
ventricular compliance (increased P
pilo
), and the ap-
pearance of V waves on the wedge pressure trac-
ing, myocardial lactate production measured by
coronary sinus catheters, and angina pectoris in
some awake patients.
A noninvasive blood pressure monitor that con-
tinuously displays the arterial waveform using the
Penaz methodology has been recently introduced
into clinical practice (model Finapres 2300.
Ohmeda, Boulder, CO). Several studies in patients
not undergoing thoracic surgery have found a
good correlation between the intra-arterial pressure
and the noninvasive finger arterial pressure. Al-
though one study
78
found that the device did not
function properly, or at all, during one-lung venti-
lation in 4 of 9 patients undergoing thoracic sur-
gical procedures, another study found that the Fi-
napres was able to follow accurately the pressure
changes during thoracic surgery (lobectomy and
pneumonectomy) even during the period of one-
lung ventilation and surgical manipulation.
79
These
latter authors concluded that the Finapres was an
acceptable alternative to invasive blood pressure
monitoring during thoracic surgery.
H. Muscle Relaxation
To determine the depth of neuromuscular block-
ade and the adequacy of reversal of neuromuscular
blockade, neuromuscular transmission must be
monitored quantitatively. There are numerous neu-
romuscular blockade monitors available, and many
provide all or some combination of simple twitch,
tetanus, and train of four capability. At the tier I
level, it is not important to make a priority distinc-
tion between these various indices of neuromus-
cular blockade. However, for the sake of brevity
and efficiency, these indices were listed previously
in order of increasing ability to detect neuromus-
cular blockade. More crude methods of assessment
of neuromuscular function, especially postopera-
tively, include presence of intercostal muscle ac-
THREE LEAD SYSTEMS FOR DETECTION
OF MYOCARDIAL ISCHEMIA
RA Lead is
Clavicular
Figure 7-13 Three lead systems
for detecting myocardial ischemia.
The left panel shows the CS
5
system,
and the right panel shows the CB,
system.
RA Lead is
Scapular
tidal volume can be assessed only visually or b]
touch. Fairly precise measurement of the tidal vol
ume can be made by placing on the expirator
limb an in-line respirometer, which works by
vane anemometer principle (e.g., the Wright oj
Boehringer spirometer). As indicated previously
there are a number of situations in which the calii
brated plastic cylinder around the bellows of th<
ventilator can be quite inaccurate (50 per cent erroi
possible depending on fresh gas flow rates, inspi
ratiomexpiration ratio, and patient mechanics)
Since the anemometer respirometer is much mon
accurate than the ventilator bellows, the formei
may be used to check the latter. Both can be used
continuously, but the anemometer respirometeri
become less accurate after exposure to moistur<
for a long period of time and are therefore bettel
preserved by intermittent use. Several modern anj
esthesia machines now incorporate a minute venl
tilation read-out (integration of an in-line pneui
motachygraph) or at least a tidal volume anq
respiratory rate read-out (from which minute veni
tilation is easily calculated).
Monitoring 253
radial artery, occasionally dorsalis pedis) is con-
veniently available, whereas for all other cases
(middle- and high-level tier II situations, including
almost all one-lung ventilation cases), arterial
blood gases should be sampled from an indwelling
peripheral arterial catheter. An indwelling periph-
eral arterial catheter guarantees immediate, fre-
quent, and certain arterial blood samples.
2. Venous Blood-Gas Analysis
Blood gases are traditionally measured in arte-
rial blood because venous blood is unsuitable for
the assessment of oxygen tension. However, for
the measurement of carbon dioxide tension and
bicarbonate concentration, venous blood is usually
acceptable; its carbon dioxide tension may be
higher by 5 to 8 mm Hg and its concentration of
bicarbonate by 1 or 2 mmol per liter, but these
differencesnearly within the limits of measuring
errorare seldom clinically important.
81
There are, however, circumstances in which car-
bon dioxide tension is much higher in venous than
in arterial blood. Severe circulatory failure, shock,
and cardiac arrest prolong capillary transit time.
When less blood is available to sweep away car-
bon dioxide, the blood that does flow carries a
bigger load. Ventilation keeps arterial carbon diox-
ide tension at its normal value, but venous carbon
dioxide tension rises. In addition, circulatory col-
lapse also causes cellular anoxia and lactic aci-
dosis. The lactic acid is neutralized, but at the cost
of a reduction in the plasma bicarbonate concen-
tration (a decrease in plasma bicarbonate defines a
metabolic acidosis).
One study reported that 16 patients who were
undergoing cardiopulmonary resuscitation had a
mean venous carbon dioxide tension of 74 mm Hg
compared with a mean arterial value of 32 mm
Hg. During cardiac arrest, venous carbon dioxide
J tension increased sharply, whereas arterial carbon
dioxide tension remained near its previous level.
82
Another study demonstrated similar discrepancies
both in patients with hypotension and severe cir-
culatory failure and in patients undergoing cardio-
pulmonary resuscitation.
83
The discrepancies they
found were particularly large when ventilation
(and therefore arterial carbon dioxide tension) was
maintained mechanically.
In both studies, the venous blood of patients
with severe compromise or collapse of the circu-
lation had a markedly higher carbon dioxide ten-
sion, and therefore a markedly lower pH, than
simultaneously drawn arterial blood. Thus, the im-
X)rtance of studying central venous blood gases
md determining central venous hypercapnia lies in
issessing the environment of cells at the venous
end of the capillaries and the need to improve the
circulation.
81
3. Transcutaneous 0
2
and C0
2
Tensions
Although blood-gas analysis defines the effi-
ciency and quantity of ventilation, the usefulness
of the information is diminished by the fact that
the results are static intermittent values that existed
several minutes ago. In addition, the method suf-
fers from being invasive. In contrast, the new non-
invasive techniques of measuring oxygen gas ex-
changetranscutaneous oxygen tension and pulse
oximetry (see tier I)and the new noninvasive
techniques of measuring carbon dioxide ex-
changetranscutaneous C0
2
tension and end-tidal
C0
2
analysis (see tier I)allow, with certain im-
portant limitations (see tier I, discussed previously,
and transcutaneous monitoring discussed later),
continuous monitoring of these gases. Arterial
blood gases are precise for all values but slow and
intermittent; S
p
0
2
responds very quickly, but im-
portant changes are limited to the low end of the
oxygen-hemoglobin dissociation curve; PETC0
2
responds very quickly but may be moderately in-
accurate; P
tc
0
2
is uniquely useful for continuously
monitoring high P0
2
values; and P
tc
C0
2
is slow
responding but trends are useful for all values (Ta-
ble 7-6). Thus, intermittent invasive and continu-
ous noninvasive techniques should be viewed as
complementary rather than competitive.
a. BASIC PRINCIPLES
Transcutaneous blood-gas tension measure-
ments are probably the most inaccurate and blood
flow-dependent of the noninvasive continuous
gas-exchange monitoring methods. When the skin
is made sufficiently hyperemic (to 44C by the
transcutaneous probe), the skin 0
2
consumption
and C0
2
production become insignificant relative
to 0
2
supply and C0
2
washout, the right-hand term
in the equation on page 239 becomes insignificant,
and the tissue and transcutaneous Po
2
and Pco
2
(P
tc
0
2
and P
tc
C0
2
, respectively) approach the P
a
0
2
and P
a
C0
2
, respectively.
84-87
Heat also increases
the Po
2
of the capillary bed by decreasing the
aqueous 0
2
solubility, providing an apparent offset
to the drop in tissue Po
2
as a result of metabo-
lism.
41
b. CLINICAL USE
In general, in normal patients, P
tc
0
2
should be
expected to be 60 to 80 per cent of P
a
0
2
, and
P
tc
C0
2
should be expected to be 25 per cent
greater than (1.3 X) P^Oj.
84
"
87
However, when
cardiac index decreases below 2.2 L/min/m
2
, P
tc
0
2
254 Monitoring
Table 7-6 ADVANTAGES AND DISADVANTAGES OF VARIOUS RESPIRATORY
FUNCTION MONITORS
Monitor Advantage Disadvantage
Arterial blood gases Precise Intermittent
S
p
0
2
Fast response, continuous Limited to low end of oxygen-hemoglobin
dissociation curve
P
l c
0
2
Gives information about high end of Slow response, 20-40% lower than P
a
0
2
oxygen-hemoglobin dissociation
curve
PETC0
2
Fast response, continuous ?(Compared with P
l c
C0
2
) may be
significantly lower than P
a
C0
2
and
change unpredictably in very severe lung
disease
P
l c
C0
2
?(Compared with PETCO, ) trends Slow response, 25% higher than P
a
C0
2
may be useful
and fast-responding technique for monitoring ox-
ygenation is needed during transition from two-
lung ventilation to one-lung ventilation and during
one-lung ventilation because extreme hypoxemia
occasionally results unpredictably and rapidly. In
these instances, pulse oximetry is well suited and
P
t c
0
2
is poorly suited for detecting acute changes
in P
a
0
2
and hypoxemia.
97
The unique value of
P
t c
0
2
in anesthesia for thoracic surgery is in de-
tecting, on-line and continuously, changes in oxy-
gen tension when the changes are large and occur
at the high end of the oxygen-hemoglobin disso-
ciation curve (i.e., P
a
0
2
between 100-700 mm
Hg).
98
" Transcutaneous monitoring is also effec-
tive for following P
a
0
2
and P
a
C0
2
in neonates who
have a thin, immature epidermis (the increase in
effectiveness is primarily due to decreased metab-
olism, not a decrease in the barrier to diffusion of
oxygen), in whom an umbilical artery catheter is
not available and in whom avoidance of hyperoxia
is desired.
Because peripheral oximetry and transcutaneous
measurements are continuous, one would expect
that these continuous methods would be able to
diagnose more periods of hypoxemia in patients
who have medical problems or who are in situa-
tions in which P
a
0
2
may vary rapidly and signifi-
cantly compared with intermittent blood-gas
analysis.
84
85
Indeed, in neonates who require me-
chanical ventilation and elevated F,0
2
, continuous
oxygen monitoring by the transcutaneous method
detected significantly more hypoxia than intermit-
tent blood-gas samples (237 min/24 hours vs. 136
min/24 hours, < .Ol).
86
No instance in this study
had an arterial P
a
0
2
of < 30 mm Hg by blood-gas
analysis, whereas this low value was noted for 32
min/24 hours by the continuous transcutaneous ox-
ygen monitoring method. These extremely low ox-
ygenation values were most frequently observed in
association with airway suctioning or spontaneous
crying or during chest physiotherapy.
In view of the extensive experience with contin-
uous arterial oxygen monitoring in critically ill
jatients,
84
85
such as the neonatal experience just
iescribed,
86
it is not surprising that continuous ox-
rgen monitoring methods were quickly used for
jatients undergoing one-lung ventilation. Indeed,
)etween 1980 and 1990, nine different reports
[five using transcutaneous monitoring
87,90-93
and
four using pulse oximetry
39
40, 100
, 01
) documented
that continuous arterial oxygenation monitoring
during one-lung ventilation revealed fluctuations
in arterial oxygenation that might have been
missed by only sampling arterial blood gases inter-
mittently.
For low-level tier -type cases, gas exchange
may be adequately monitored by intermittent arte-
rial blood gases (drawn either from single punc-
tures or from an indwelling catheter). At present,
there is an insufficient amount of evidence to sup-
port the contention that gas exchange may be ade-
quately monitored at this low tier II level with only
one of the noninvasive continuous monitors of gas
exchange (see discussion of individual monitors
for their limitations). The absolute values from
continuous arterial oxygenation monitors should
be regarded as the minimum P
a
0
2
and S
a
0
2
possi-
ble, and a downward trend in the continuous arte-
rial oxygenation monitor value should be regarded
as an early warning sign of impending hypoxemia;
both a low absolute value and a downward trend
should be documented by arterial blood-gas anal-
ysis. For middle-level tier II patients, arterial blood
gases should be monitored from an indwelling pe-
ripheral arterial catheter. At this middle tier level,
a continuous oxygen monitor should be used as
complementary early indicators of gas-exchange
problems (usually the continuous monitor de-
creases the number of arterial blood gases re-
quired).
100
For high-level tier II cases, arterial
blood gases should be regularly monitored from
an indwelling arterial catheter (i.e., sampling
should not be based on the data from a continuous
monitor), although the continuous monitors of gas
exchange can certainly still be used to indicate any
interval problems.
The P
a
0
2
alone does not describe the efficiency
of oxygenation of the blood in terms of wasted
perfusion (Q/Qt) nor does the P
a
C0
2
alone de-
scribe the efficiency of ventilation in terms of
wasted ventilation (Vc/V
T
). More complex steps
(see the following), such as actually measuring
VD/V
T
and Qg/Qp cardiac output, and C
v
0
2
, are
necessary to do this with precision. However,
given a cardiac output and Oxygen consumption in
the normal range, a reasonably close approxima-
tion of Qg/Qt can be made if the P
a
0
2
is known in
relation to the inspired oxygen concentration
Monitoring 255
(Fi0
2
) (see Fig. 3-6). Likewise; a reasonably close
approximation of VE/VT can be made if the P
a
C0
2
is known in relation to the minute ventilation.
102
For example, a 70-kg patient whbse minute venti-
lation is twice normal (e.g., 10 L/min) and whose
P
a
C0
2
'.rs= 40 mm Hg will have twice normal
wasted ventilation (60 per cent). A simplified, but
relatively accurate, equation for Vf/V
T
requiring
only the patient's minute ventilation, expected
ventilation (Radford nomogram), and P
a
C0
2
has
been described.
102
i ta
In addition to detecting airway obstruction,
adventitious sounds, and bronchospasm by listen-
ing to breath sounds, airway mechanics can be
monitored in more quantitative terms. The anes-
thesia machine and ventilator have pressure
gauges/traces. Dividing the peak and/or plateau (if
present) pressure during inspiration into the tidal
volume (as measured by a spirometer) allows one
to calculate dynamic and static lung compliance,
respectively. At the mid- to high tier level,
whole-lung and individual-lung airway pressures
and tidal volumes should be measured separately
so that the compliance of the whole lung and each
individual lung can be calculated separately.
Knowledge of individual lung compliance allows
the anesthesiologist to know what compliance each
lung should have after any thoracic procedure <but
especially after unilateral lung lavage); this num-
ber should serve as the end point for postoperative
individual lung PEEP and sighing maneuvers.
:
In patients who have undergone long operations,
have experienced many physiologic interventions,
or have received large amounts of anesthetic
drugs, it is both reasonable and wise to question
whether they will be able to breathe adequately
spontaneously following such an anesthetic and
surgical experience. In an attempt to identify those
few variables that singly or in combination best
predict the outcome of the first trial of spontaneous
respiration following cardiac surgery, a stepwise
linear discriminant analysis of more than 50 phys-
iologic and clinical predictors of postoperative res-
pirator^ adequacy indicated that postoperative vital
capacity per kilogram and maximum inspiratory
force were the most useful variables. Theidiffer-
ence between extubation successes and failures is
represented by a vital capacity of 15 ml/kg and a
maximum inspiratory force of 28 cm H
2
0.
103
Measurements of passive pulmonary mechanics,
cardiac function, and arterial blood gases were sur-
prisingly poor predictors (however, see use of 0
2
and ST0
2
in section G of Tier ).
256 Monitoring
Because both hypovolemia and hypervolemia
have important implications for gas exchange, in-
creased monitoring efforts should be directed to-
ward monitoring the blood pressure and blood vol-
ume. Systemic arterial blood pressure and central
venous pressure can be accurately measured by
direct catheter monitoring and can be displayed on
an oscilloscope. Insertion of an indwelling periph-
eral arterial catheter is sufficiently justified for he-
modynamic monitoring in patients who have mod-
erate cardiovascular compromise or in patients who
will undergo moderately physiologically intrusive
operations. When used with a paper read-out, the
trend in arterial pressure is readily demonstrated
and frequently reveals changes in hemodynamic
status that might not otherwise be appreciated. An
electrical mean pressure can be calculated, allow-
ing for a more precise measurement of perfusion
pressure. Finally, a rough estimate of myocardial
contractility can be made by observing the steep-
ness of the upstroke of the arterial pressure wave-
form. The combined hemodynamic and blood-gas
measurement indications for arterial line insertion
at the tier II level virtually guarantee that almost
all patients at the tier II level should have an in-
dwelling arterial catheter placed. In patients with-
out significant cardiovascular disease, central ve-
nous pressure monitoring will track intravascular
volume satisfactorily (see Cardiovascular Parame-
ters under Tier III: Advanced Monitoring Tech-
niques).
Input and output balance can be followed more
closely by weighing all sponges and by measuring
the urine output from a Foley catheter every half
hour. Hematocrit values, osmolality, and serum
electrolytes can be measured intermittently. These
measurements, in addition to essential monitoring,
will reflect the contribution of the cardiovascular
system to respiratory function.
IV. TIER III: ADVANCED
MONITORING TECHNIQUES
Patients who have severe, pre-existing respira-
tory disease and/or especially severe special intra-
operative conditions are a challenge because the
potential for respiratory function to deteriorate is
great. For example, an elderly patient with severe
COPD undergoing a major pulmonary resection or
esophagogastrectomy is a likely candidate for hav-
ing some degree of postoperative respiratory fail-
ure. Similarly, some patients are so critically ill
preoperatively (e.g., thoracic trauma, bleeding,
sepsis) that anesthetic requirements are minimal
and the anesthesiologist's main concerns reside in
resuscitation and monitoring. Under these circun]
stances, maximum monitoring of cardiopulmonait
function is necessary to keep those physiologi
variables related to respiratory function as near t
normal as possible. The following sections discus
advanced monitoring techniques for respiratoi
functions to G in Table 7-2 that further increas
the degree of accuracy and rapidity of diagnosi
and treatment of cardiac and/or pulmonary failun
Monitoring techniques for respiratory functions
to D, H, and I in Table 7-2 are the same as thos
described in previous sections.
E. Gas Exchange
The amount of total venous admixture (Q
s
/Q
can be measured accurately in the clinical settinj
provided a pulmonary artery catheter is in situ. 1
determine Q
s
/Q
t
properly, the partial pressure <
oxygen in the pulmonary artery mixed venous an
systemic arterial blood (P;:0
2
and P
a
0
2
, respei
tively), the partial pressure of carbon dioxide
arterial blood (P
a
C0
2
), hemoglobin concentratioi
and the inspired oxygen concentration need to b
measured. From these data the content of oxygq
in mixed venous (Q0
2
), arterial (C
a
0
2
), and enj
pulmonary capillary blood (Q.0
2
) can be calq
lated (see chapter 3). Knowing these various ox]
gen contents enables calculation of Q
s
/Q
t
:
Q
s
_ QQ
2
- C
a
Q
2
Q
t
" Q0
2
- C
c
0
2
In addition, if the cardiac output is known (whic
should be the case if a pulmonary artery catheti
is in situ; see Cardiovascular Parameters, follov
ing), Vo
2
can be calculated by the Fick principle
Vo
2
= Q
t
X C(a - v)0
2
The normal arteriovenous oxygen content diffe
ence [C(a v)0
2
] is 5 vol per cent. It is importai
that the P^0
2
be measured directly. If the contei
of mixed venous blood is simply assumed to be
vol per cent less than arterial blood, rather tha
actually measured, large errors in the QJQ
t
dete:
mination may be encountered.
104
For example,
the C(a - v)0
2
is actually only half or twid
normal, then the QJQ
t
determined by assuming
vol per cent will be 50 per cent lower and 100
cent higher, respectively, than it actually is. Sim
larly, large errors in the shunt determination wi
be encountered if superior vena cava blood san
pies (S
cv
0
2
) are substituted for pulmonary artei
samples in the calculation of Q0
2
.
1 0 5 1 0 6
Althoug
absolute values of S
cv
0
2
are not sufficiently idei
tical to S
7
0
2
to calculate oxygen uptake or pulm(
nary shunt precisely, close tracking (correlation,
not accuracy) of changes in the two sites across a
wide range of hemodynamic conditions (r = 0.96)
warrant consideration of S
cv
0
2
for patient monitor-
ing of trends in oxygen supply and demand.
107
The
variability that can be expected in an individual
shunt determination, when shunt is repeatedly
measured, using mixed venous blood in hemody-
namically stable postoperative patients is 5 per
cent of the mean value, and increasing or decreas-
ing shunt values that exceed 2 per cent of the
cardiac output (approximately 2.0 standard devia-
tions) in response to therapeutic interventions
likely reflect actual physiologic change, whereas
lesser changes may represent technical prob-
lems.
108
In addition, direct measurement of P^0
2
can
serve as an indirect monitor of P
a
0
2
, oxygen con-
sumption, and cardiac output (see Cardiovascular
Parameters, following); a decrease in P-0
2
may
reflect either a decrease in P
a
0
2
or cardiac output
or an increase in oxygen consumption. Therefore,
the P-0
2
should not be regarded as having much
specificity as a monitor of arterial oxygenation.
Since the best use of ^0
2
and S^0
2
may be as a
monitor of global well-being, it is not surprising
that an S-0
2
of less than 60 per cent (Pv0
2
of
approximately 30 mm Hg) has been found to be
an extremely important criterion for the need for
mechanical ventilation following cardiac sur-
gery.
109
Measurement of the inspired and expired con-
centration of respiratory gases by mass spectrom-
etry is now extremely accurate, rapid, and com-
mercially available and, as such, has been in
progressively widespread clinical use over the past
20 years. A mass spectrometer causes the inspired
and expired respiratory gases that will be meas-
ured to pass through a magnetic field that differ-
entiates between and quantifies the different gases
by deflecting the gases in an amount that is in-
versely proportional to the molecular weight of the
gas.
End-tidal and inspired concentrations of eight
gases (0
2
, N
2
, C0
2
, N
2
0, argon, halothane, enflu-
rane, and isoflurane) can be measured from each
patient. Continuous monitoring of the inspired ox-
ygen concentration at the mouth or endotracheal
tube confirms that a gas mixture of the required
composition is being delivered and that rebreath-
ing is not occurring. An alarm system can be in-
cluded to detect any deviation of the gas mixture
outside preset, adjustable limits, which can be se-
lected appropriately for each patient. The most im-
portant breath-by-breath monitoring functions of a
mass spectrometer are end-tidal carbon dioxide
tension (to follow ventilation patterns), inspired
Monitoring 257
oxygen concentrations (which are now a standard
of anesthesia practice in all cases), and inspired
and end-tidal anesthetic drug concentrations (for
both clinical management and safety concerns).
The cost of a mass spectrometer alone (approx-
imately $35,000) inhibits its use for single-patient
monitoring. However, the monitoring cost per op-
erating room can be reduced significantly by using
one mass spectrometer that can rapidly sample
many operating rooms on a rotating basis. The
total equipment and installation cost of a single
mass spectrometer designed to serve an entire op-
erating room suite ($60,000) is competitive with
the summed cost of individual C0
2
, 0
2
, and halo-
genated drug analyzers for five to 10 operating
rooms (Puritan Corp., personal communication).
Finally, progress has been made toward meas-
uring S
a
0
2
and P
a
0
2
directly and continuously.
This is an important step forward because pres-
ently used indirect estimates of P
a
0
2
and P
a
C0
2
(i.e., P
t c
0
2
and P
tc
C0
2
) do not eliminate the need
for arterial blood-gas analysis, especially in unsta-
ble patients, and intermittent arterial blood sam-
ples document only what has occurred in the re-
cent past, which is obviously of less value than
present observation. Pulse oximetry can provide
rapid detection of hypoxia during one-lung venti-
lation in adults, but it does not provide early warn-
ing of a decreasing P
a
0
2
at high F,0
2
and can fail
when the pulse signal is weak. Transcutaneous
oxygen tension has also been recommended for
monitoring during one-lung ventilation because of
its sensitivity to both P
a
0
2
and cardiac output. Un-
fortunately, its response time is often too slow to
detect rapid changes in P
a
0
2
.
110
The intra-arterial
optode does not suffer from any of these limita-
tions, and it can provide continuous, real-time
monitoring of arterial pH, Po
2
, and Pco
2
.
The fiberoptic intra-arterial optode makes use of
the phenomenon of fluorescence quenching to de-
termine blood gases and pH. The three-component
optode (Po
2
, Pco
2
, pH) (Cardiovascular Devices,
Inc., Irvine, CA) is inserted through an 18-gauge
radial artery catheter and allows continuous arte-
rial pressure monitoring as well as access for blood
sampling. The triple-function (Po
2
, Pco
2
, pH) fi-
beroptic optode is inserted into the radial artery
catheter after in vitro calibration. This calibration
procedure consists of placing the sterile optode
within its storage container into an automated cal-
ibrator. The calibration device heats the optode
and solution to 37C and bubbles calibration gases
through the solution-filled container until stable
values are maintained, and the optode is then cali-
brated. This process is completed in an automated
manner and requires 20 min. Intra-arterial optode
oxygen tension (P
0
0
2
), carbon dioxide tension
258 Monitoring
(P
0
C0
2
), and pH are displayed continuously either
at blood temperature or temperature corrected to
37C.
In a study by Barker et al., intra-arterial P
0
0
2
showed the most rapid and significant changes
during endobronchial intubation when compared
with P,
c
0
2
and S
p
0
2
.
m
In addition, continuous
blood-gas monitoring with an intra-arterial optode
during one-lung anesthesia has also been found to
be very useful.
112
However, intra-arterial optode
monitoring is not without its own limitations. The
optode sensor is expensive (estimated cost =
$150) relative to the cost of using noninvasive
monitors of oxygenation and ventilation. The de-
vice also requires additional time for calibration
and insertion. In this case, the probe was placed in
an 18-gauge radial artery catheter, which is larger
than that used in routine clinical practice. These
catheters usually require recalibration once every
12 to 24 hours; therefore, they do not replace the
need for blood-gas analysis, although, when func-
tioning properly, they may greatly reduce the num-
ber of arterial blood-gas samples required.
F. Airway Mechanics
Determination of compliance is simple because
the measurements required are pressure and vol-
ume. Determination of airway resistance is more
difficult because the measurements required are
pressure and flow. Airway resistance is calculated
as follows:
Pressure, cm H
2
0
Airway resistance =
Flow, L/min
Airflow can be measured with a pneumotachy-
graph, which is a rapid-response measuring device
based on a form of constant-orifice flowmeter. It
consists of an airflow resistance, across which a
pressure drop occurs, a differential pressure trans-
ducer, and a means of recording the output (the
differential pressure translated into airflow rate).
The useful range of airflow resistances is limited
at low flow by a small pressure difference and at
high flows by nonlinearity as a result of turbu-
lence. If airway pressure is recorded as relative to
atmospheric, then total (lung and chest wall) resis-
tance is calculated. If airway pressure is relative to
pleural pressure, then lung resistance is calculated.
There does exist a clinically applicable technique
for direct measurement of intrapleural pressure.
113
The monitoring of resistance allows for detection
of subtle degrees of bronchospasm not possible by
other means. However, at present, airway resis-
tance measurements are largely a research tech-
nique.
The determination of FRC during mechanical
ventilation has become important in clinical inves-
tigation in anesthesiology and for diagnostic and
therapeutic evaluation of patients in acute respira-
tory failure. Most previous methods used a closed-
circuit helium dilution technique,
114-116
but a new
sulfahexafluoride method appears most promis-
ing.
117
Measurement of other lung volumes such as
closing volume and volume of trapped gas are
research tools.
Normally, and at the tier I and tier II levels, the
central venous pressure (CVP) is an adequate in-
dex of intravascular volume (see the following dis-
cussion of use of CVP in normal hearts). However,
in patients with abnormal hearts, and at the high
tier II and tier III levels, CVP may be an inade-
quate and misleading index of intravascular vol-
ume, and the more sensitive method of pulmonary
vascular pressure monitoring may be necessary to
follow intravascular volume status. In addition, the
cardiovascular status of the patient may be so pre-
carious that it is also necessary to know other
indices of cardiovascular function such as P^0
2
,
cardiac output, and systemic vascular resistance.
Consequently, this section considers the rationale
for pulmonary vascular pressure monitoring, when
it is inappropriate to use CVP monitoring instead
of pulmonary vascular pressure monitoring, the
clinical value of inserting pulmonary artery cathe-
ters, and, finally, the complications of pulmonary
artery catheter insertion.
1. Left Ventricular Preload and Left
The Frank-Starling myocardial function curve
states that, as initial ventricular fiber length is in-
creased, which is the ventricular preload, ventric-
ular performance increases (force of fiber contrac-
tion increases) (Fig. 7-14). In the intact heart,
initial fiber length is the end-diastolic fiber length,
which is determined by the left ventricular end-
diastolic volume (Fig. 7-15). Left ventricular end-
diastolic volume can be measured intraoperatively
by esophageal probe echocardiography, but it can
be difficult to obtain quantitatively accurate vol-
umes (although echocardiography can discern an
empty [grossly small left ventricular diastolic
chamber] vs. a full [grossly large diastolic left
ventricular chamber] heart), and the apparatus is
financially and generally technically demanding.
Consequently, other, more removed, indices of left
ventricular preload are used clinically in most non-
cardiac thoracic surgery cases. However, it should
be noted that esophageal echocardiography can
Monitoring 259
Left Ventricular Function Curve
Figure 7-14 The Frank-Starling principle states that ventricu-
lar performance (i.e., force of contraction) is directly proportional
to the preload (i.e., fiber length, ventricular end-diastolic volume
[VEDVJ).
Indices of Left Ventricular Preload
Figure 7-15 The figure dynamically describes the various indices of left ventricular (LV) preload. In the top portion of the
diagram, the various central vascular and cardiac chambers are schematically portrayed in the sequence in which blood flows
through them. The cardiac valves portray the end-diastolic moment; the pulmonary and aortic valves are closed, and the mitral
valve is open. In the graph below the cardiac chambers, the characteristic vascular pressure contours in each vascular compartment
are depicted directly below each compartment. Below this graph, the various indices of left ventricular preload are labelled; they
consist mainly of the pressures within the various vascular compartments as well as two non-pressure indices on the right-hand side
of the panel (volume and fiber length). At the bottom of the diagram, the requirements for each of the indices to reflect accurately
left ventricular preload are shown impacting between the index in question and the preceding closer approximation of left ventricular
preload. (CVP = central venous pressure; RA = right atrium; RV = right ventricle; PA = pulmonary artery; PV = pulmonary
vein; PC = pulmonary capillary; LA = left atrium; Ao = aortic valve; LVED = left ventricular end-diastolic.)
260 Monitoring
clearly demonstrate abnormalities of interventric-
ular septal and ventricular wall motion resulting
from ischemia (akinesia and dyskinesia) (which is
the most important monitoring function of the in-
strument) and valve function (see section 7).
The next few sections describe a series of pro-
gressively more removed and less accurate, but
usually easier to obtain, indices of left ventricular
preload (see Fig. 7-15). Any factor that decreases
the accuracy of a preceding index of left ventricu-
lar preload also decreases the accuracy of all of
the succeeding (further removed) indices of left
ventricular preload (see Fig. 7-15).
The next closest approximation of left ventricu-
lar preload to left ventricular end-diastolic volume
is left ventricular end-diastolic pressure. As left
ventricular end-diastolic volume increases, left
ventricular end-diastolic pressure increases (Fig.
716); the relationship between ventricular volume
and pressure (dv/dp, or the instantaneous slope to
the curve in Fig. 7-16) is called ventricular com-
pliance (compliance = volume/pressure) (ventric-
ular compliance is the inverse of chamber stiff-
ness). Thus, when left ventricular compliance is
unchanging, left ventricular end-diastolic pressure
should be directly proportional to left ventricular
end-diastolic volume, fiber length, and ventricular
output (see Fig. 7-16, middle curve).
LV COMPLIANCE
LV Volume
Figure 7-16 Left ventricular (LV) compliance is the slope
of the curve relating left ventricular volume to left ventricular
pressure. Three curves are shown: normal compliance, de-
creased compliance, and increased compliance. For any curve,
the ventricular compliance is greater at lower levels of ventric-
ular filling than at higher levels. A ventricle that has decreased
compliance is stiff and nondistensible and has a higher pressure
than normal for any given volume. A ventricle that has in-
creased compliance is soft and distensible and has a lower
pressure than normal for any given volume. LV pressure will
not accurately reflect LV volume when LV compliance is
changing.
There are two major reasons why ventricular
compliance may change at any time. First, changes
in left ventricular compliance may result from
changes in the level of left ventricular filling, in-
volving a shift along a single diastolic pressure-
volume curve."
8
Figure 7-16 demonstrates the ef-
fect of progressive left ventricular filling on left
ventricular compliance. Because the pressure-vol-
ume curve is nonlinear, an equivalent increase in
volume (AV) will produce a greater increase in
pressure when the ventricle is relatively distended
(
2
) than when it is relatively empty (,). This
change in left ventricular compliance does not im-
ply any alteration in the inherent stiffness proper-
ties of the ventricle because the modulus of cham-
ber stiffness remains unchanged.
Second, changes in left ventricular compliance
may also result from changes in the inherent stiff-
ness properties of the myocardium, represented by
a shift of the pressure-volume curve and defined
by a change in the modulus of chamber stiffness."
8
If left ventricular compliance is decreased (stiffer,
less distensible ventricle) so that the position of
the left ventricular end-diastolic pressure-volume
curve is shifted upward and to the left (see Fig. 7-
16, upper curve), then any given left ventricular
end-diastolic volume will be associated with a
higher left ventricular end-diastolic pressure. Con-
ditions that decrease left ventricular compliance
are left ventricular hypertrophy, myocardial ische-
mia, fibrosis, myocarditis, sepsis, right-to-left in-
terventricular septal shifts (as caused by right ven-
tricular pressure loading [as caused by lung
disease, PEEP, volume loading]), pericardial con-
striction, effusion and tamponade, systemic hyper-
tension, aortic stenosis, idiopathic hypertrophic
subaortic stenosis, and an increased juxtacardiac
pressure."
9, I2
With an increased juxtacardiac
pressure (positive pleural or pericardial pressure),
transmural ventricular pressure (preload) is very
decreased even though absolute intracavity pres-
sure (relative to atmospheric pressure) may be
very increased. Increased juxtacardiac pressure is
a mechanism common to PEEP and pericardial
diseases."
8
Although myocardial ischemia can be
detected by increases in left ventricular filling
pressure, the changes in filling pressure are not as
sensitive or specific as ST-segment depression on
the EKG or transesophageal echocardiographic
(TEE) wall motion abnormalities (rank order TEE
> EKG > left ventricular preload).
121
If left ventricular compliance is increased
(softer, more distensible ventricle) so that the posi-
tion of the left ventricular end-diastolic pressure-
volume curve is shifted downward and to the right
(see Fig. 7-16, lower curve), then any given left
ventricular end-diastolic volume will be associated
with a lower left ventricular end-diastolic pressure.
Monitoring
261
Conditions that increase left ventricular compli-
ance are administration of vasodilators such as
nitroglycerin and nitroprusside, left-to-right inter-
ventricular septal shifts (as caused by right ventric-
ular pressure unloading), congestive cardiomyop-
athies, and aortic and mitral regurgitation."
8-120
Thus, when left ventricular compliance changes,
changes in left ventricular end-diastolic pressure
(and the various other, more removed indices of
preload [P
pad
, P
paw
, and P
C
J) may not reflect left
ventricular end-diastolic volume and may be diffi-
cult to interpret.
Figure 7-17 shows examples of bidirectional
changes in compliance caused by interventricular
septal shifts. Because the two ventricles are physi-
cally coupled by the interventricular septum and
by the constraining pericardium, the end-diastolic
pressure-volume curve of either ventricle is depen-
dent on the diastolic volume of the other.
122
With
right-to-left shifts and decreased left ventricular
compliance (see Fig. 7-17, lower panel), de-
creased end-diastolic volumes are associated with
increased end-diastolic pressures; with left-to-right
shifts and increased left ventricular compliance
(see Fig. 7-17, middle panel), increased end-dia-
stolic volumes are associated with normal or
mildly decreased end-diastolic pressures.
The incidence of abnormalities in left ventricu-
lar compliance (and therefore lack of correlation
between left ventricular end-diastolic volume and
the various pressure indices of left ventricular end-
diastolic volume) may be particularly high in the
Figure 7-17 The effect of interventricular septal shifting (as might be caused by right ventricular volume and pressure loading
and unloading) on left ventricular (LV) compliance. A, The interventricular septum in a normal position and accompanied by a
normal (N) ventricular compliance curve. B, A left-to-right interventricular septal shift and an increase in left ventricular compliance
(increased volumes with decreased pressures). C, A right-to-left interventricular septal shift and a decrease in left ventricular
compliance (decreased volumes with increased pressures). (Data taken from Cariile.
120
)
262 Monitoring
elderly undergoing major surgery,
123
postoperative
coronary artery bypass surgery patients,
124
hemor-
rhagic hypotension,
125
and in the adult respiratory
distress syndrome
126
and pulmonary vascular in-
jury.
127
Aside from the issue of left ventricular compli-
ance (which to at least the same extent invalidates
all of the more remote pressure indices of left
ventricular end-diastolic pressure [see later discus-
sion] as a precise index of left ventricular end-
diastolic volume), left ventricular end-diastolic
pressure would frequently be a good index of left
ventricular preload. However, there is not a good,
safe (no puncture of ventricular wall), and easy
(avoid retrograde technique) way ol measuring left
ventricular end-diastolic pressure. Consequently,
other, more removed but easier to obtain, indices
of ventricular preload must be used clinically.
2. Various Indices of Left Ventricular
A schematic representation of the central circu-
lation is shown in Figure 7-15.
128
Above are the
various vascular compartments; below are charac-
teristic pressure contours recorded at each site. At
the end of diastole, the pulmonic and aortic valves
are closed, and the mitral valve is open. The entire
vascular bed between the closed pulmonic and aor-
tic valves behaves like a common 1 id-filled pres-
sure chamber. At the end of dia.Mole, pressure
should fall to the same level in all segments of the
common pressure chamber, provided there is a
sufficiently low resistance to blooc! flow and that
there is a sufficient amount of tim (slow enough
heart rate) for the diastolic pressui run-off. End-
diastolic pressure in the left ventr: ie determines
the pressure level to which there i a pulmonary
artery diastolic pressure run-off.
Therefore, at the end of diastole and under nor-
mal conditions, pressure at all pc i tits within the
continuous chamber (namely, pressure in the left
ventricle, left atrium, pulmonary veins, pulmonary
capillaries, and pulmonary artery is the same.
Therefore, the pulmonary artery diastolic pressure
(P
pad
) should normally equal pulmonary artery oc-
clusion (P
pao
), pulmonary capillary (P
cap
) pressure,
and left atrial (P
la
) pressure, which should equal
left ventricular end-diastolic pressure (P
LV
ED) AS
a
corollary, in the absence of end-diastolic pressure
gradients, there is no end-diastolic pulmonary
blood flow.
129
a. LEFT ATRIAL PRESSURE
Mitral and aortic valve disease will create a left
atrial pressure-to-left ventricular end-diastolic
pressure gradient. The effect of mitral stenosis is
straightforward, and, because it impairs left atrial
emptying, it causes left atrial pressure to be higher
than left ventricular end-diastolic pressure.
Mitral regurgitation increases the mean left
atrial pressure above the diastolic left atrial pres-
sure and hence above P
LV
ED
1 3 0
In
m e
normal left
atrium (see Fig. 7-34), the c wave reflects the
bulging of the mitral valve into the atrium during
early ventricular systole. The wave corresponds
to the flow of blood into the atrium against a
closed mitral valve at the end of systole. In cases
of mitral insufficiency, the regurgitant flow of
blood causes more prominent c and waves that
are usually fused together and defined as large (>
10 mm Hg) V waves. These can be seen on both
the wedge pressure trace and pulmonary artery
pressure trace (Fig. 7-18). It is important to rec-
ognize the tall V waves on the wedge trace (which
occur after the wave on the EKG, whereas the
pulmonary artery pressure pulse occurs just before
the wave on the EKG) because the tall V waves
can be misinterpreted as a pulmonary artery pres-
sure tracing, stimulating further attempts to wedge
the catheter, which may lead to a pulmonary artery
rupture.
Figure 7-18 Diagrammatic repre-
sentation of an electrocardiogram
tracing and synchronous pulmonary
artery and wedge pressure tracing in-
dicating the appearance of a large V
wave (combination of c and
waves). The crest of the pulse of the
phasic pulmonary artery pressure
trace is just before or at the begin-
ning of the wave, whereas the
large V wave is after the wave.
(Modified from Nadeau S, Noble
WH: Misinterpretation of pressure
measurements from the pulmonary
artery catheter. Can Anaesth Soc J
33:352-363, 1986. Used with per-
mission.)
Monitoring 263
Another rapid, simple beat-to-beat method for
differentiating pulmonary arterial from pulmonary
capillary wedge positions in the presence of giant
left atrial V waves is the superimposition of the
pulmonary arterial trace on the radial arterial trace;
the upstrokes and peaks occur nearly simultane-
ously.
131
When a wedge position is attained and V
waves are present, there is an immediate rightward
shift in the upstroke and peak of the pulmonary
arterial pressure trace, which is now a V wave
(Fig. 7-19). Thus, the relationship among the pul-
monary arterial trace, the V wave, and systemic
arterial trace is easily observed (see Fig. 7-19).
131
In the case of aortic regurgitation, left ventricu-
lar end-diastolic pressure can be higher left atrial
pressure or pulmonary artery end-diastolic pres-
sure. The regurgitant flow back into the left ventri-
cle causes premature closure of the mitral valve
(while the left ventricle continues to fill thereafter)
and prevents the left atrium and the pulmonary
vascular bed from estimating or equilibrating
with the elevated left ventricular end-diastolic
pressure."
8
I30
b. PULMONARY ARTERY OCCLUDED PRESSURE
(P
pao
, LARGE PULMONARY VEIN PRESSURE)
(1). DEFINITION OF p
pao
. Inflation of the flotation
balloon on a pulmonary artery catheter until pul-
monary artery blood flow is completely obstructed
creates a static column of fluid distal to the cathe-
ter tip, which extends across the pulmonary artery
and capillaries over to the large pulmonary veins
(Fig. 7-20). The static column of fluid serves as
an extension of the pulmonary artery catheter tip
and allows the catheter tip to "see" pressure on
the left side of the pulmonary circulation. The
pressure measured by the distal tip of the catheter
beyond the obstructing balloon is called the pul-
monary artery occluded pressure (P
pao
) and is pre-
cisely defined as the pressure in the large pulmo-
nary venous circulation where flow first begins
beyond the static column of fluid. With the excep-
tion of the rare condition of pulmonary venous
thrombosis, there are virtually no pressure gra-
dients from the large pulmonary venous circula-
tion to the left atrium; therefore, the P
pao
essentially
Figure 7-19 Composite of mean measured intervals with standard error of the mean taken from the electrocardiograms, radial
arterial traces, pulmonary arterial traces, and pulmonary capillary wedge traces for all patients in the study. (From Moore RA,
Neary MJ, Gallagher JD, Clark DL: Determination of the pulmonary capillary wedge position in patients with giant left atrial
waves. J Cardiothorac Anesthesiol 1:108 113, 1987. Used with permission.)
264 Monitoring
Pulmon iry Artery Occluded Pressure
Measurement
Zone 1
(PEEP)
Zone 2
Static Column of Fluid
Figure 7-20 The mechanism by whic
on the pulmonary artery catheter is inflai
of the catheter over to the venous side
will record the pressure that is present i
Consequently, the pulmonary artery occ
pressure gradient between the pulmonar
pressure. Because the pulmonary veins a
accurately reflect left atrial pressure in tl
expiratory pressure.)
Zone 3
of pulmonary parenchymal and chest wall compli-
ance, and the pressure changes occurring in the
pleural space during the ventilatory cycle.
(2). P
p au
, PEEP, AND ZONE III LOCATION. If the
fluid column beyond the tip of the catheter (and
the chambers of the heart) is simply compressed
but not collapsed by the surrounding and increased
alveolar and pleural pressures (i.e., the vessel in
which the pulmonary artery catheter tip is located
is still in a zone 3), then the pressure inside all
intrathoracic cavities (including all pulmonary ves-
sels, chambers of the heart, great veins, aortae)
relative to atmospheric pressure (which is the ref-
erence pressure for the transducer outside the
body) will be increased by an amount slightly less
than the increase in pleural pressure relative to
atmospheric pressure (see Fig. 19-ID, which
shows a representative example of these pressure
changes). Because the intrathoracic intracavitary
pressure (relative to atmosphere) increases slightly
but stil. less than the increase in pleural pressure
(or intrapericardial pressure) relative to atmos-
pheric pressure, transmural pressure (intracavitary
Monitoring 265
pressure minus the pressure immediately outside
the cavity [pleural or intrapericardial pressure],
which is the true filling pressure) decreases (see
Fig. 19-1D).'
35
Pleural pressure will increase as a function of
the compliance of the lung and chest wall (C
L
and
C
c w
, respectively) according to the equation:
P
PL
- PEEP (C
L
/C
L
+ C
c w
)
With normal C
L
and C
c w
, the P
L
at the end of
passive exhalation should be approximately a 0.4
to 0.5 fraction of the PEEP. This is necessary
because, in normal lungs with ventilation at nor-
mal tidal volumes, C
L
approximates C
c w
(C
L
/C
L
4-
C
c w
= 1/1 + 1 = 1/2). If C
L
increases, then
C
L
/C
L
+ C
c w
increases, the fraction of PEEP
transmitted to the pleural space increases, and in-
tracavitary pressure increases (compared with a
lower C
L
for the same PEEP).
(3). P
p a o
, PEEP, AND ZONE I LOCATION. If the
fluid column between the pulmonary artery cathe-
ter tip and left atrium is collapsed by alveolar
pressure (Figs. 7-20 and 7-21), the catheter tip
can only sense the interrupting alveolar pressure
and reflect PEEP rather than the downstream P,...
A study conducted with dogs with normal lungs
and normal P
la
during controlled positive-pressure
ventilation demonstrated these P
pao
, PEEP, and
zone of the lungs relationships very nicely; a pro-
gressive increase in PEEP > 5 mm Hg produced
a progressive positive discrepancy between P
pao
and P
la
(i.e., P
pao
> P
la
) when the pulmonary arte-
rial catheter was wedged above the left atrium. In
fact, when PEEP = 10 and 15 mm Hg, P
pao
equaled 10 and 15 mm Hg, respectively, whereas
the P
u
equaled 6 and 8 mm Hg, respectively.
136
When the pulmonary artery catheter was wedged
below the left atrium, it was not possible to create
a P
pao
-P|
a
gradient with even high levels of PEEP.
A zone I catheter position may be suspected by
observing the effects of sudden increases or de-
creases of PEEP on the measured P
pao
; if the P
pao
increases by greater than 50 per cent of the change
in the applied PEEP (or C
L
suddenly increases), a
non-zone III catheter tip is likely.
137
(4). P
pao
AND DETERMINATION OF ZONE III LO-
CATION. If the tip of the pulmonary artery catheter
is vertically below or at the level of the left atrium,
zone III conditions usually exist, unless extreme
hypovolemia or high levels of PEEP are applied.
137
Anteroposterior chest radiographs may not reliably
locate the position of the tip of the pulmonary
Without
PEEP
Static Col umn of Fluid
Catheter
With
PEEP
Factors Causing Preservation of Ppao
=
Pla During PEEP
1. Location of catheter in dependent lung (Zone 3)
2. Maintenance of some spontaneous ventilation (venous return)
3. Non-compliant lung (decreased transmission of PEEP)
Figure 7-21 In the clinical situation, zone l conditions are created by the application of positive end-expiratory pressure (PEEP).
PEEP may collapse the pulmonary capillaries, interrupting the static column of fluid normally created by inflation of the pulmonary
artery catheter balloon. Consequently, the pulmonary artery wedge pressure (P
p a
J will sense alveolar pressure (PEEP) rather than a
downstream of vascular pressure (left atrial pressure [PJ). Several commonly present clinical conditions tend to preserve the
ao
as a reflection of P
la
even when PEEP is being used; these consist of dependent-lung catheterization, presence of spontaneous
ventilation, and noncompliance of the lung.
266 Monitoring
Monitoring 267
Figure 7-22 Schematic representation of pulmonary artery pressure tracings in the absence of respiratory effort (muscle
paralysis), during spontaneous breathing and during controlled ventilation. The arrow indicates inflation of the balloon to obtain the
pulmonary artery wedge pressure, and the asterisks represent the points of end expiration. (Reproduced with permission from Tobin
MJ. Essentials of Critical Care Medicine. New York, Churchill Livingstone, 1989, 37.)
c. PULMONARY ARTERIAL (TRUE) WEDGE
PRESSURE (P
pdw
, PULMONARY DISTAL WEDGE
PRESSURE, SMALL PULMONARY VEIN
PRESSURE)
When the balloon on a pulmonary catheter is
inflated, the segment of the pulmonary circulation
that is occluded is the relatively large area sub-
tended by a large pulmonary artery, and flow will
first begin in a relatively large pulmonary vein
(Fig. 7-23, upper panel).
151
Therefore, the pressure
measured when the balloon on the pulmonary ar-
tery catheter is inflated is the pressure in a large
vein (P
pao
; see prior discussion). When the tip of a
pulmonary artery catheter is wedged in a small
pulmonary artery, the segment of the pulmonary
circulation that is occluded is subtended by a rela-
tively small pulmonary artery, and flow will first
begin again in a relatively small pulmonary vein
(see Fig. 7-23, bottom panel).
151
Therefore, the
pressure measured by the distal wedging of the
pulmonary artery catheter tip is the pressure in a
small pulmonary vein (denoted by P
pdw
). Ob-
viously, the pressure in a small pulmonary vein
(Ppdw) will t>
e
slightly higher than the pressure in a
large pulmonary vein (P
pao
), and the P
pdw
-P
p; 10
gra-
dient will increase as venous resistance increases.
The P
p d w
-P
p a o
gradient was 1.1 0.5 mm Hg in
nine patients with normal lungs and was signifi-
cantly higher in the 13 patients with chronic conges-
tive heart failure (3.8 0.8 mm Hg, < .01) and
22 patients with adult respiratory distress syn-
drome (3.8 0.8 mm Hg; < .01), but not in 20
patients with COPD (1.8 0.7 mm Hg).
151
The distribution of the pulmonary vascular re-
sistance was, therefore, clearly different among the
four groups.
151
The fraction of the total pulmonary
vascular resistance attributable to large and me-
dium pulmonary veins was significantly increased
(p < .01) in adult respiratory distress syndrome
(27.5 12 per cent) and cardiac patients (27.5
9 per cent) compared with patients with COPD ( 13
5 per cent) and normal lungs (13.5 6 per
cent). The P
pdw
has been used clinically to assess
268 Monitoring
Monitoring 269
ual decline in pressure until the classic occluded
pressure (P
pao
) is attained (see Fig. 7-24). The fast
decline is due to the diastolic pressure run-off in
the pulmonary artery (to the P
cap
), and the slow
component of the pressure decay curve represents
the discharging of the pressure/volume in the large
capillary bed to the pressure/volume in the large
pulmonary veins and left atrium. The inflection
point or transition from fast to slow decline is
thought to reflect the P
cap
, and the P
c a p
-P
pa o
gradient
is thought to reflect postcapillary resistance (Fig.
7-24). Large gradients between estimated P
cap
and
P
pa0
have been shown in patients with increased
pulmonary venous resistance after mitral valve re-
placement.
154
The P
cap
may be suspected of being
greater than the P
pao
by simply observing an in-
creased difference between the P
pad
and the P
pao
(e.g., greater than 3 mm Hg), whereas a difference
of 2 to 3 mm Hg means the P
pao
more closely
approximates the P
cap
.
Although numerous mathematical/computer
methods for determining the inflection point have
been described, the visual technique has been
thought to be adequate (see Fig. 7-24).' " The vis-
ual technique is best accomplished by placing a
straight edge directly on the tracing, adjusted for
the best fit, and marking the inflection point (i.e.,
when the curve clearly shows the slow component
as being separate from the rapid component). The
easiest and best determinations are made in me-
chanically ventilated, paralyzed patients (which
eliminates respiratory artifact) with a pulmonary
artery catheter in the proximal one third of the
lung. Interobserver variability in determining
using Electronic for Medicine recorder varied
0.2 mm Hg standard error of the mean.
1
" How-
ever, it should be noted that it still may be difficult
to identify an unequivocal break in the pressure-
decay curve.
158
e. PULMONARY ARTERY DIASTOLIC PRESSURE
(Ppad)
Normally, pulmonary artery diastolic pressure
equals left ventricular end-diastolic pressure. In
Figure 7-25, pulmonary arterial (heavy line) and
left ventricular (thin line) blood pressures are su-
perimposed on one another during three condi-
tions: (1) normal circumstance (A), (2) in the pres-
ence of left ventricular failure (B), and (3) in the
presence of increased pulmonary vascular resis-
tance (C).
m
In the first two instances, both dia-
stolic blood pressures fall to the same level at the
end of diastole (single arrow). The pulmonary hy-
pertension shown in Figure 1-25B is a passive
consequence of the increased left ventricular end-
diastolic pressure. The vessel walls remain thin
and distensible under these conditions and do not
impose an appreciable increase in resistance to
blood flow. In contrast, the panel on the right illus-
trates that when pulmonary hypertension stems
from abnormal structure or function of the pulmo-
nary vessels, pressure in the pulmonary artery will
exceed the left ventricular pressure at the end of
diastole, as indicated by the two arrows. Disease
has rendered the pulmonary vessels sufficiently
obstructed so that resistance to blood flow is
greatly increased. The vertical distance between
the two arrows, which represents the end-diastolic
pressure gradient between the pulmonary artery
and the left ventricle, is proportional to the in-
crease in pulmonary vascular resistance. In sum-
mary, when pulmonary vascular resistance is nor-
mal, pressure in the pulmonary artery during
diastole can run off and decrease to left ventricular
end-diastolic pressure (Fig. 7-26, left panel) (and
blood flow will momentarily cease), whereas when
pulmonary vascular resistance is increased, pres-
sure in the pulmonary artery during diastole cannot
run off and decrease to left ventricular end-dia-
stolic pressure (see Fig. 7-26, right panel).
Increased resistance to flow can be caused by
both passive mechanical and active vasoconstrictor
mechanisms. The passive mechanical mechanisms
consist of capillary and venous compression by
interstitial fluid and/or blood and/or fibrosis, en-
dothelial cell edema, capillary compression by
PEEP, arterial obstruction by thrombosis and/or
microembolism, and medial hypertrophy. The ac-
tive vasoconstriction mediators can be alveolar
hypoxia, systemically released vasoconstrictor
Figure 7-24 Estimation of P
cap
using the balloon pressure profile
after inflation of pulmonary artery
catheter (PAC) balloon (open arrow).
270 Monitoring
Monitoring 271
amines and peptides, decreased mixed venous ox-
ygen tension, and acidosis (Fig. 7-27).
159
These
changes result in an increase in pulmonary vascu-
lar resistance, which is "a universal feature of
acute respiratory failure."
160
Tachycardia (heart rate greater than 120/min)
can also independently cause an end-diastolic pres-
sure gradient (as much as 10 mm Hg) between the
pulmonary artery and left ventricle because of in-
sufficient diastolic run-off time.
161
In another
study, the left atrium was paced from 74/min to
124/min.
162
As heart rate increased, P
pad
increased
but P
L
VED decreased, thus creating a significant
pressure gradient (11 mm Hg). Therefore, this
study also demonstrates that, with less diastolic
filling time, another effect of tachycardia is that
less blood is transferred from the pulmonary vas-
culature to the left side of the heart. Finally, an-
other study showed that when the heart rate was
above 115/min, P
pad
was always greater than
Ppao
163
Thus, in patients with acute respiratory fail-
ure who have these pathologic changes and in-
creased pulmonary vascular resistance, and often
an associated tachycardia, it'is highly likely that
pulmonary artery diastolic pressure will be higher
than left ventricular end-diastolic pressure.
Several studies showed that an increased P
pad
-
Ppao gradient carries a very poor prognosis. In one
study of a large series of critically ill patients, the
Ppad
w a s
6.0 mm Hg or more higher than the P
pao
in 30 per cent of the patients.
140
The pulmonary
vascular resistance in these patients averaged (
standard deviation) 257 145 dyne/sec/cm
-5
(normal = 80 to 160 dyne/sec/cm~
5
). The mean
pulmonary vascular resistance in the other 74 pa-
tients was slightly lower (158 72 dyne/sec/cm
-5
).
The mortality rate in the patients with the in-
creased P
pad
-P
pao
gradients was 59 per cent. This
was significantly higher than the mortality rate of
34 per cent seen with lower P
pad
-P
pao
gradients. In
Causes of Increased Pulmonary Vascular Resistance
in Acute and Chronic Lung Disease
Figure 7-27 The causes of increased pulmonary vascular resistance during acute and chronic respiratory failure are multiple. In
an anatomic progression from the arterial to the venous side of the pulmonary circulation, the causes of increased pulmonary
vascular resistance consist of medial hypertrophy, endothelial cell edema, pulmonary thromboembolism, arteriolar constriction by
vasoactive amines, peptides, decreased mixed venous oxygen tension ( [ P0
2
), acidosis, and alveolar hypoxia ( [ P
A
0
2
). Positive
end-expiratory pressure (PEEP) can compress the pulmonary capillaries. Increased interstitial hydrostatic pressure resulting from
transudated fluid and blood, which can later fibrose, can compress the venous side of the pulmonary capillaries. All these causes of
increased pulmonary vascular resistance will create a pulmonary artery diastolic (P
pad
) to left ventricular end-diastolic pressure
(PLVED) gradient (see inset).
272 Monitoring
another study, 22 of 37 patients with sepsis had an
increased P
pa d
-P
pa o
gradient. Patients in whom the
P
pad
exceeded the P
pao
by more than 5.0 mm Hg
had an average pulmonary vascular resistance ex-
ceeding 300 dyne/sec/cm
-5
and had increased
mortality rate.
164
Finally, another study noted that
a persistent or increasing P
pa d
-P
pa o
gradient exceed-
ing 5 mm Hg was associated with a 91 per cent
mortality rate.
165
Thus, if the P
pad
exceeds the P
pao
by 6.0 mm Hg or more, the patient has probably
developed significant pulmonary hypertension and
has a much poorer prognosis.
A right bundle branch block (RBBB) also cre-
ates pressure gradients in the pulmonary vascula-
ture.
166
In the presence of RBBB (Fig. 7-28),
130
right ventricular systole is delayed, allowing the
pulmonary pressure to continue to fall during the
descent of the left atrial pressure trace.
130
The
P
pad
is then actually lower (may be up to 7 mm Hg
lower
166
) than the mean P
la
. This will occur only
when there is normal pulmonary vascular resis-
tance and thus rapid equilibrium between P
pad
and
Pu.
0
J
Figure 7-28 A normal left atrial (LA) and pulmonary artery
(PA) pressure tracing (A) is compared with tracings with right
bundle branch block (RBBB) (B). The a wave corresponds to
left atrial contraction, the c wave to the bulging of the mitral
valve in ventricular systole, and the wave to the passive
filling of the atrium. Mean LA pressure is read between a and
c waves and should correspond to left ventricular end-diastolic
pressure. The descent represents relaxation of the left atrium,
the y descent the opening of the mitral valve. Note that the
delayed right ventricular systole in RBBB is associated with a
low P
pad
(arrow). (Modified from Nadeau S, Noble WH: Mis-
interpretation of pressure measurements from the pulmonary
artery catheter. Can Anaesth Soc J 33:352, 1986. Used with
permission.)
A pulmonary artery catheter should be floate
to the most proximal position from which both tlj
wedge pressure and phasic pulmonary artery prej
sure can be obtained by simple sequential inflaie
and deflation of the flotation balloon. If the cathl
ter floats out too far in the peripheral lung parei
chyma, it may permanently wedge and greatly ii
crease the risk of pulmonary infarction. If tl
catheter is floated to a too proximal position in tl
main pulmonary artery, the catheter tip may inte
mittently whip or dip into and out of the rigl
ventricle. Under these circumstances, the "dii
stolic" pressure may be intermittently too low b
cause several of the recorded beats will have a
tually been right ventricular (diastolic pressure
the right ventricle is close to zero).

3. Various Indices of Right Ventricular
If the right and left ventricles have parallel fun
tion, then right ventricular end-diastolic pressu
should follow left ventricular end-diastolic pre
sure. The difference in absolute value between tl
two end-diastolic pressures would be caused on
by the difference in compliance between the tv
ventricles; the right ventricle, being more cor
pliant than the left ventricle, has a lower en
diastolic pressure (see Normal Heart, which fc
lows). With this reservation in mind, various inc
ces of right ventricular end-diastolic pressure cl
be used to assess preload clinically.
a. RIGHT ATRIAL PRESSURE
Tricuspid and pulmonic valve disease will ci
ate a right atrial to right ventricular end-diastoi
pressure gradient. Tricuspid stenosis causes rig
atrial pressure to be higher than right ventricul
end-diastolic pressure, and pulmonic regurgitatii
causes right atrial pressure to be lower than rig
ventricular end-diastolic pressure.
b. CENTRAL VENOUS PRESSURE
To clearly answer the question of whether
when central venous, as opposed to pulmona
vascular, pressure monitoring can be used (P
cv
j
P
pad
and P
pao
) to assess preload, it is necessary
consider the answer separately for a normal he
versus an abnormal heart.
(1). NORMAL HEART. The left ventricle is
thickly muscled (walled) cavity and therefore
relatively noncompliant (Fig. 7-29, right-ha
panel). When fluid is systemically infused (p
load increased), the increase in left ventricu
end-diastolic volume causes a large increase in 1
ventricular end-diastolic pressure (as measured
Ppad
or
Ppao) The right ventricle is a thinly muse
(walled) cavity and therefore is relatively co
Monitoring 273
Pressure
jure 7-29 The relationship between central venous pressure (P
cv
) and pulmonary artery occluded pressure (P
pao
) and pulmonary
ery diastolic pressure (P
pad
) is normally based on the differences in compliance between the right ventricle (RV) and the left
ntricle (LV). The right ventricle is relatively compliant because it is thinly muscled, and the left ventricle is relatively noncom-
ant because it is thickly muscled. Thus, when preload is changed in a normal heart (such as with fluid infusion), the central
nous pressure (as a reflection of right ventricular end-diastolic pressure) increases only a small amount for a given increase in
fdiac output, whereas the pulmonary artery occluded pressure and pulmonary artery diastolic pressure (as reflections of left
ntricular end-diastolic pressure) increase a large amount for the same increase in cardiac output. Thus, because of the differences
compliance between the two ventricles, the absolute central venous pressure is always less than the pulmonary artery occluded
d pulmonary artery diastolic pressures, and the change in the central venous pressure is always less than the change in pulmonary
ery occluded and pulmonary artery diastolic pressures.
iant (Fig. 7-29, left-hand panel). The infusion of
lids systemically causes the increase in right
intricular end-diastolic volume to be accompa-
ed by only a relatively small increase in right
intricular end-diastolic pressure (as measured by
mtral venous pressure [P
cv
]). Thus, when intra-
iscular volume is acutely increased, the P
cv
in-
eases only a small amount compared with a rel-
ively large increase in the wedge pressure.
167
deed, when preload is manipulated in normal
itients (fluid infusion or diuresis
168
or by position
langes
169
) and simultaneous measurements of the
L and P
pao
are made, the initial and final absolute
is always approximately 100 per cent greater
[in P
cv
(Figs. 7-29 and 7-30). Figure 7-31, upper
nel, shows the good predictability and high de-
ee of correlation (lowest individual patient r >
72) in patients with ejection fractions greater
an 0.5 undergoing position changes.
169
In other
words, initial and final P
pao
is twice as great as the
initial and final P
cv
, and the change in P
pao
is al-
ways approximately twice the change in the P
cv
(see Figs. 7-29 and 7-30). Thus, in a normal heart,
there is an orderly and predictable relationship be-
tween the filling pressures of the right and left
sides of the heart, and the P
cv
can be used to follow
left-heart function.
However, because the regression line relating
P
cv
(x-axis) to P
pad
and P
pao
(y-axis) in relatively
normal patients has an average slope of 2 (which
means P
pao
changes twice as much as the P
cv
) (see
Fig. 7-30),
167-170
the P
pao
may be easier, and per-
haps more reliable, to use than the P
cv
to follow
intravascular volume status. In other words, it may
be easier to distinguish physiologically meaningful
changes above and beyond the usual background
noise and variation in the P
pao
compared with the
P
cv
. In addition, it is not known how other changes
274 Monitoring
(e.g., afterload and changes in myocardial contrac-
tility) in normal patients will affect simultaneously
measured P
cv
and P
pad
and P
pao
.
171
In summary,
because the right and left sides of the heart have
an orderly and predictable functional relationship
to one another in the normal heart, the P
cv
may be
used as a substitute for pulmonary vascular pres-
sure monitoring in normal patients, but the
changes in P
cv
will be less marked than those that
might be observed in P
pao
.
(2). ABNORMAL HEART. In an abnormal heart,
left- and right-heart ventricular function curves
(filling pressure vs. output) may be markedly dif-
ferent from one another. For example, it is possible
to have a failing right ventricle and a normal left
ventricle, resulting in a high P
cv
and low P
pao
(which is an uncommon combination in the gen-
eral surgical population but very common in the
multiple-trauma patient,
172
whereas with an ische-
mic left ventricle and a normal right ventricle
(which is a common combination found in many
cardiac patients), P
pa0
will be high while P
cv
may
be low (Fig. 7-32). Consequently, the discrepancy
between the P
cv
and the P
pad
and P
pao
increases a
great deal when intravascular volume is aug-
mented in patients with heart disease,
169-171
and it
is usually misleading to use the P
cv
as a meaningful
guide to the filling pressure of the left side of the
heart. Figure 7-31, bottom panel, shows the unpre-
dictability and extremely poor correlation (highest
individual patient r < 0.53) between P
cv
and P
pao
in patients with ejection fraction less than 0.5
undergoing position changes.
169
Thus, candidates
for pulmonary vascular^ pressure monitoring in-
clude all patients who have any significant myo-
cardial compromise who are undergoing signifi-
cant perioperative stress.
Monitoring 275
Figure 731 Upper panel, Regression lines and correlation coefficients for the 15 patients in group I (ejection fraction > 0.5).
For each patient, the correlation between P
c v
and was high; little dispersion of data is found around each individual regression
line. Lower panel. Regression lines and correlation coefficients for the 15 patients in group II (ejection fraction < 0.5). The P
c v
and P
pao
were uncorrected or poorly correlated for each group II patient, and the dispersion of data around individual regression
lines was large. (From Mangano DT: Monitoring pulmonary arterial pressure in coronary artery disease. Anesthesiology 53:364-
276 Monitoring
Figure 7-32 Patients with abnormal hearts may have either a nom
(LV) function curve (most common) (solid line) or an abnormal RV fun
(dashed line).
ventricular (RV) and an abnormal left ventricular
ve and a normal LV function curve (less common)
Monitoring 277
Clinical Value of Pulmonary Artery Catheter
Figure 7-33 The clinical value of the pulmonary artery catheter is great. Measurement of pulmonary artery diastolic (P
pad
) and
pulmonary artery wedge (P
pil()
) pressures can be early indicators of left ventricular ischemia (increase in mean pressure and the
appearance of "giant" a and cv waves). These two pressures allow estimation of left ventricular end-diastolic pressure (P
LV
ED)'
which allows for assessment of intravascular volume, and based on this assessment decisions regarding preload changes (position,
fluid infusion, or diuresis) can be made objectively. Measurement of the cardiac output (CO.) along with P
pao
allows for estimation
of myocardiac contractility and for an objective decision to be made as to whether inotropic or suppressant drugs should be used.
Measurement of cardiac output along with systemic pressure allows for determination of systemic vascular resistance (SVR) and
for an objective decision to be made as to whether vasodilator or vasoconstrictor drugs should be used. Measurement of SvO, may
be used as an index of global well-being because it is a function of cardiac output, oxygen consumption (Vo
:
), Hb concentration,
and .,.
278 Monitoring
ECG
CVP
Monitoring 279
GIANT
cv WAVE
NORMAL
GIANT
a WAVE
Figure 7-35 Giant a waves are caused by atrial contraction into a noncompliant stiff ventricle (as may be caused by ventricular
ischemia). Giant cv waves are actually giant c waves and are due to atrioventricular valve regurgitation when the supporting
papillary muscles dysfunction (as may be caused by ventricular ischemia).
280 Monitoring
for assessment of myocardial contractility. Knowl-
edge of cardiac output along with systemic and
pulmonary vascular pressures allows for calcula-
tion of systemic and pulmonary vascular resis-
tances, respectively. Knowledge of cardiac output
along with heart rate allows for calculation of
stroke volume. Measurement of P-0
2
(pulmonary
artery blood) allows for calculation of Q
s
/Q and
P-0
2
alone can be used as an indirect gross meas-
ure of cardiac output and oxygen consumption.
174
(2). CARDIAC OUTPUT. The ability to obtain rap-
idly accurate measurements of cardiac output
(CO), and, in conjunction with P
pad
and P
pao
, assess
cardiac contractility in critically ill patients is one
of the principal advantages of the pulmonary ar-
tery catheter. The temperature of pulmonary artery
blood is instantly measured by a thermistor located
near the distal tip of the pulmonary artery catheter.
The injection of cold solution into the proximal
(right atrial) lumen of the catheter allows the CO
to be determined by the indicator-dilution tech-
nique. As right-sided CO increases, the indicator
(cold solution) becomes more and more diluted by
the warm venous blood, and less temperature drop
(from body temperature) is detected by the pul-
monary artery thermistor. CO is usually calculated
by a computer that measures the area under the
thermodilution curve (obtained by plotting the de-
cline in pulmonary artery temperature from body
temperature vs. time). Representative curves from
patients with low, normal, and high COs are illus-
trated in Figure 7-36, B, and C, respectively.
(3). SYSTEMIC VASCULAR RESISTANCE. The
physiologic parameter known as systemic vascular
resistance (SVR) is commonly calculated from the
following relationship:
where P
sa
equals mean arterial pressure and
equals right atrial pressure. The logical basis fc
this expression arises by analogy from Poiseuille
1
law for fluid flow within rigid pipes, which state
that flow is proportional to the difference betwee
the upstream and downstream pressures. Althoug
CO, stroke volume, and stroke work are usuall
indexed to body size, it is not yet common practic
to do so for vascular resistance. Because the pi
ripheral arterial tree branches in parallel and to a
extent that is proportional to body size, large indi
viduals have lower resistance (more vessels in paj
allel) than smaller individuals. For example, a 1
lb, 62-inch female with P
sa
of 90 mm Hg, P
cv
of
mm Hg, and CO of 4.2 L/min has a calculate
SVR of 1617 dynes-sec-cm
-5
, whereas a 210-lt
72-inch male with the same P
sa
and P
cv
and a C(
of 6.16 L/min (same cardiac index) has a calcu
lated SVR of 1103 dynes-sec-cm
-5
. Calculatin
SVR index using cardiac index instead of CO i
the denominator normalizes the parameter to bod
surface area and yields an SVR index of 242
dynes-sec-cm~
5
/m
2
for both individuals, suggestin
that their vascular trees were similarly matched t
their CO. It seems that a method of indexing SV1
should be universally adopted into clinical prac
tice."
9
(4). MONITORING MIXED VENOUS OXYGEN SAT
URATION. In specific capillary beds, factors detei
mined by local tissue metabolism induce releas
of oxygen from hemoglobin (Hb) and produc
unique, tissue-specific differences in arteriovenou
oxygen content or saturation (the Fick principle^
For example, the arteriovenous oxygen conter
difference in the normal working heart is quit
large, usually 11.4 cc of oxygen per 100 ml
blood; in the skin, it is significantly smaller: 1 c
of oxygen per 100 ml of blood. Because of wid
,
Variations in regional venous oxygen saturation,
)lood from the pulmonary artery, where total mix-
ng has occurred, represents the best average value
or SA.
For total body oxygen uptake, or (Vo
2
), the Fick
)rinciple can be expressed mathematically as fol-
lows:
Vo
2
= Q
t
(C
a
0
2
- CA)
The equation for the determination of S-O2
c a n De
derived from the Fick equation as follows (the
derivation involves substitution of S-O2
an
d S
a
0
2
X [Hb][1.36] for C-0
2
and C
a
0
2
, respectively,
factoring and transposing [Hb][1.36] and then
transposing S
a
0
2
):
SA = S
a
0
2
- [Vo
2
/(Hb)(1.36)(Q
t
)l X 10
From this equation, we see that a decrease in SA
can result from ( 1 ) a decrease in oxygen saturation
of arterial blood, (2) an increase in oxygen con-
sumption, (3) a decrease in cardiac output (Q
t
),
and (4) a decrease in hemoglobin.
When respiratory function, hemoglobin concen-
tration, and oxygen consumption are stable, it fol-
lows from the Fick equation that changes in SA2
may accurately reflect parallel changes in cardiac
output.
175
This situation, a most common occur-
rence in the operating room, provides the basis for
rapid correlation of SA2 with therapeutic pharma-
cologic interventions when cardiac output is low.
Indeed, in patients who undergo coronary artery
surgery, it has been shown that the probability of
a decrease in SA2 being due to a reduction in
cardiac output is 86 per cent.
175
Even if the clinical situation is slightly more
complex, the change in SA2
c a n
usually be cor-
rectly interpreted. The correct interpretation of
changes in SA2
mu s t
be correlated with informa-
tion about inspired oxygen concentration, changes
in respiratory function, hemoglobin concentration,
oxygen consumption, and cardiac index. In most
clinical circumstances, the cause of SA2 changes
can usually be inferred from other available labo-
ratory data or astute clinical observation. For ex-
ample, in hypovolemic shock, reductions in both
hemoglobin and cardiac output may be additive
factors in decreasing SA2 During myocardial is-
chemia, a reduction in cardiac index is most likely
the cause of reduced SA2 However, reductions in
arterial saturation secondary to pulmonary conges-
tion may also be contributory. An arterial blood-gas
sample or continuous monitor of arterial oxygen
saturation, together with measurement of cardiac
output, will usually establish the cause of reduced
SA Finally, the changes in SA2 during initiation
of one-lung ventilation usually cause a large de-
Monitoring 281
crease in P
a
0
2
(a factor that decreases SA2)
an
d
a
moderate increase in Q
t
(a factor that increases
SAX which results in a final small decrease in
SA
2
1 7 6
'
7 7
If three of the four primary variables that deter-
mine SA2 (Qt Vo
2
, Hb, P
a
0
2
) are not constant
and/or secondary compensatory mechanisms kick
in (e.g., interorgan and intraorgan redistribution of
blood flow, right-shifted oxygen-hemoglobin dis-
sociation curve), then changes in SA2
m a v
e
uninterpretable. In fact, large changes in oxygen
supply (Q Hb, P
a
0
2
) and demand (Vo
2
) may oc-
cur without any change in SA2
175
In
t ms
context,
a decrease in SA2 which is probably not a desir-
able change, can only serve as an early warning
device. For example, a large decrease in Q, may
be compensated for by increase in F,0
2
and a de-
crease in Vo
2
, which is a circumstance that com-
monly occurs during general anesthesia. On emer-
gence, the increased oxygen consumption and
decreased arterial oxygenation resulting from the
increased ventilation-perfusion mismatch associ-
ated with anesthesia and surgery may act in com-
bination to decrease SA2 despite a notable in-
crease in cardiac output. A relatively common
experience after cardiopulmonary by-pass is that
transfusion to a normal hemoglobin and hemato-
crit will increase SA2 considerably even as car-
diac output decreases, sometimes to worrisome
levels. Clinical situations in which changes in SA2
have been found to be uninterpretable involve crit-
ically ill patients in an ICU,
178
postoperative car-
diac surgical patients,
179
patients in circulatory
shock,
180
patients undergoing aortic surgery with
cross-clamping,
181
and patients with acute myocar-
dial infarction.
182
There are two other situations/factors that limit
the usefulness of SA2 monitoring. First, only
pooled oxygen uptake and saturation values can be
obtained, although the important information is
whether oxygen supply is adequate to meet the
needs of all organ systems, including those of the
heart and brain. Therefore, SA2 monitoring pro-
vides little information on specific organ perfu-
sion.
182
In that sense, and in the case of compli-
cated patients with many variables changing at the
same time, decreases in SA2 (which are probably
never a good sign) can serve at best only as a
general early warning indicator.
Second, diseases that uncouple total oxygen
transport from oxygen consumption either by cre-
ating anatomic or physiologic peripheral shunts
invalidate the role of SA2
m
following the net
oxygen supply and demand relationship. Renal pa-
tients with surgical arteriovenous fistulas com-
monly have elevated central and mixed venous
oxygen saturations despite anemia and margi-
nal arterial content. These patients clearly have
282 Monitoring
anatomic arteriovenous shunting. Patients with hy-
pophosphatemia, nitroprusside toxicity, carboxy-
hemoglobinemia, cyanide poisoning, carbon mon-
oxide poisoning, and methemoglobinemia have
lesions that limit hemoglobin unloading at the tis-
sue level. The combined anatomic and metabolic
lesions of hepatitis, sepsis, major burns, pancreati-
tis, some drug-intoxicated states, and perhaps adult
respiratory distress syndrome are poorly under-
stood but widely known to be associated with
elevated venous oxygen saturation despite clinical
evidence for anaerobic metabolism (lactic aci-
dosis).
175
(5). TOTAL BENEFIT. The use of advanced tech-
nology such as pulmonary artery catheterization is
based on the assumption that prompt recognition
and therapy of hemodynamic abnormalities should
improve outcome (i.e., the benefits outweigh the
complications [see later discussion], and there is
net benefit or the risk/benefit ratio is small); how-
ever, the correctness of this assumption has been
debated.
183-188
Nevertheless, pulmonary artery
catheterization has been shown, by careful analysis
of the literature, to certainly be more accurate than
clinical assessment alone in critically ill patients
for determining the cause of shock (hypovolemic,
cardiogenic, or septic) (Fig. 7-37, upper panel)
and for assessing the cause of severe pulmonary
edema (cardiogenic or noncardiogenic) (see Fig.
7-37; compare upper with lower panel).
183
184
189
~
191
In addition, it has been shown that the diagnosis
of cardiac failure in medical or surgical patients
with invasive hemodynamic monitoring provides
physiologic data that guide pharmacologic treat-
ment, which may favorably influence preload and
afterload in the failing or ischemic heart.
183
184
189_191
Figure 7-37 illustrates some of the obvious he-
modynamic insights that pulmonary artery catheter
monitoring can provide.
183
The upper panel of the
figure shows examples of typical and uncompli-
cated hemodynamic data for two common patho-
logic conditions (hypovolemia vs. cardiac failure
caused by myocardial infarction). In both hemo-
dynamic syndromes, or constellations, systemic
pressure (P
sa
), central venous pressure (P
cv
), heart
rate (HR), CO, and SVR are identical or at least
DECISION ANALYSIS HEMODYNAMIC DATA
DECISION ANALYSIS HEMODYNAMIC DATA
Figure 7-37 Data from a pulmonary artery catheter can be the differentiating factor among clinical syndromes that otherw
might appear hemodynamically similar. The shaded boxes provide the key hemodynamic data for diagnostic analysis and therapei
decision making. See text for more complete explanation and definition of abbreviations.
Monitoring 283
uite similar, and the differential diagnosis be-
ween the two conditions is not possible without
neasurement of pulmonary vascular pressure.
Vith simple hypovolemia, the heart is normal, and
he P
cv
and P
pao
correlate, whereas with myocardial
nfarction the heart is abnormal and the P
cv
and
5
pao
do not correlate. Consequently, the pulmonary
/ascular pressures are shaded in and are the key
iifferentiating factors. Note that the correct diag-
losis dictates therapy strategies that are quite dif-
ferent.
Similarly, the lower panel of Figure 7-37 shows
that measurement of P
sa
, P
cv
, and HR may not be
dramatically different during sepsis and cardiac
failure (because of very increased systemic after-
load). However, measurement of CO and SVR
clearly differentiates between the two conditions
and would result in very different treatment strat-
egies.
5. Special Pulmonary Vascular
Monitoring Considerations Related to
Thoracotomy in the Lateral Decubitus
_ Position
Pulmonary arterial catheters usually (greater
than 90 per cent) float to and locate in the right
__ lung
142
; it is possible to increase the incidence of
left-lung catheterization to 50 per cent using the
right lateral decubitus position (which implies that
balloon flotation vertically upward is as important
as the pathway and curvature of blood flow.
192
Consequently, during a right thoracotomy (left lat-
eral decubitus position), the pulmonary artery
catheter will usually be in the nondependent right
lung (if no special effort was made to place it in
the left lung) and, therefore, either in a collapsed
lung if one-lung ventilation is used or possibly in
a zone 1 or 2 region of the lung if large tidal
volume two-lung ventilation is used. Conversely,
_ when a left thoracotomy is performed (patient in
the right lateral decubitus position), the pulmonary
artery catheter will be in the dependent lung and
will probably be in a zone 3 region. Thus, it is
theoretically possible that the pulmonary artery
catheter might function differently or yield differ-
ent pulmonary vascular pressure and cardiac out-
put data during right versus left thoracotomies and
during two-lung versus one-lung ventilation.
Indeed, with the pulmonary artery catheter tip
~ located in the right lung, the cardiac output is
lower during right thoracotomy with one-lung ven-
tilation (right lung collapsed) than during left tho-
-7- racotomy with one-lung ventilation (left lung col-
lapsed) in patients who were otherwise similar
(Fig. 7-38).
193
Consequently, it is possible that
_ when the pulmonary artery catheter is located in
I
the collapsed lung, where blood flow patterns may
be distorted or the function of the thermistor inter-
fered with (not free in the lumen of the vessel),
the measured output is indeed lower. This hypoth-
esis is supported by the concurrent finding that
continuously measured mixed venous oxygen sat-
uration is also decreased during right thoracoto-
mies compared with left thoracotomies when the
pulmonary artery catheter is located in the col-
lapsed nondependent lung. The decrease in mixed
venous oxygen saturation may have been caused
by stagnant blood flow and, therefore, not truly
representative of whole patient mixed venous ox-
ygen saturation.
193
When the nondependent lung is ventilated with
varying levels of PEEP (in contrast to nondepen-
dent-lung collapse), there is no difference in the
cardiac output measured simultaneously from
thermistors located in the nondependent and de-
pendent lungs.
194
This finding implies that, when
the pulmonary artery catheter tip is in the nonde-
pendent lung and the nondependent lung is venti-
lated, blood flow to the nondependent lung is un-
distorted and/or there is no interference with the
function of the thermistor. Indeed, in a conven-
tional two-lung ventilation unilateral lung injury
model (hydrochloric acid aspiration) in sheep, all
hemodynamic and respiratory measurements ob-
tained with a pulmonary artery catheter are accu-
rate and identical whether the catheter is located
ipsilateral or contralateral to the injury.
195
When the pulmonary artery catheter is in the
nondependent lung and the nondependent lung is
ventilated with large tidal volume, PEEP, or con-
tinuous positive airway pressure (CPAP), the
wedge pressure may not reflect left atrial pressure
(Fig. 7-38Z?).
145
When the pulmonary artery cath-
eter is in the dependent lung and presumably in
the zone 3 region, wedge pressure should accu-
rately reflect left atrial pressure even when PEEP
is applied to the dependent lung.
145
In summary, the lateral decubitus position is
important with regard to pulmonary artery catheter
monitoring in three situations. First, when the non-
dependent lung is collapsed and the catheter is in
the nondependent lung, the measured cardiac out-
put and P^0
2
may be decreased compared with
more normal conditions or the "real " value. Sec-
ond, when the nondependent lung is ventilated
with PEEP and the catheter is in the nondependent
lung, P
pao
may not equal P
]a
. Third, when the cath-
eter is in the dependent lung, P
pao
will be a faithful
index of P
la
even if PEEP is used. In spontaneously
ventilating critically ill patients, there is no signif-
icant difference in P
pao
or cardiac output in the
supine versus either lateral decubitus position.
196
284 Monitoring
Conditions During Thoracotomy in Lateral Decubitus
Position When Pulmonary Artery Catheter Data
May be Inaccurate
Figure 7-38 Conditions during thoracotomy in the lateral decubitus position when pulmonary artery catheter data may be
inaccurate. A, During right thoracotomy with a pulmonary artery (PA) catheter located in the collapsed right lung (one-lung
ventilation [1LV]), the cardiac output (CO) may be lower than when the right lung is ventilated. The thermistor in the collapsed
lung may be exposed to abnormal flow patterns or vascular wall interference. B, When the pulmonary artery catheter is in the
nondependent lung and the nondependent lung is exposed to continuous positive airway pressure (CPAP) or positive end-expiratory
pressure (PEEP), the pulmonary artery occluded pressure (P
pao
) may be inaccurate. Nondependent lung CPAP or PEEP may cause
zone 1 conditions in the nondependent lung. The P
pao
is probably always reasonably accurate when the pulmonary artery catheter is
in the dependent lung, even if the dependent lung is exposed to PEEP.
Monitoring 285
6. Risks and Complications of
Pulmonary Vascular Pressure
Monitoring
Pulmonary vascular pressure monitoring is an
invasive and sophisticated procedure and conse-
quently involves numerous potential physical risks
and requires intelligent and informed interpretation
of the results. The physical risks must be known
to make an objective risk/benefit ratio decision to
insert a pulmonary artery catheter. Failure to un-
derstand the rationale for each index of P
LV
ED (Pi
a
)
P
P
ao< P
C
ap>
P
p
a> P
cv
> as previously discussed, may
cause the unwary and unknowing therapist to use
the wrong pressures to arrive at ill-advised, if not
contraindicated, therapeutic decisions. The physi-
cal risks are related to complications of gaining
and maintaining central venous access and actually
passing and floating the pulmonary artery catheter
into the pulmonary circulation. This section briefly
discusses the physical risks.
a. COMPLICATIONS OF GAINING CENTRAL
VENOUS ACCESS
The most obvious and often dramatic complica-
tion associated with pulmonary vascular pressure
monitoring has to do with the fact that it is neces-
sary to gain controlled access to the central venous
circulation (Table 7-7). Frequently, the internal
jugular, external jugular, and subclavian veins are
used for this purpose. Because the needle that finds
and identifies these neck/central veins (which have
various advantages and disadvantages; Table 7-8)
can inadvertently strike nearby vital organs, it is
important to understand these potential complica-
tions.'
97
,98
First, and probably most common, the carotid
and subclavian arteries lie next to and parallel to
the corresponding internal jugular and subclavian
veins, and the incidence of arterial puncture is site
dependent (see Table 7-8) but overall is approxi-
mately 2.0 per cent.
149
Because the risk of arterial
puncture is least with the antecubital veins and it
is difficult to tamponade the subclavian artery ex-
ternally, the antecubital vein should be used and
the subclavian vein avoided in patients with bleed-
ing diatheses. The potential for vascular injury is
greatly minimized if the finder needle is of small
bore (approximately 22 gauge).
Nerves can be damaged at multiple sites, but
brachial plexus, stellate ganglion, and phrenic
nerve injuries, in particular, have been reported. I
have observed the withdrawal of cerebrospinal
fluid on aspiration of an internal jugular vein
finder needle.
The pleural cavity can be invaded by the prob-
ing needle, causing either pneumo-, hemo-, or chy-
lothorax. The incidence of pneumothorax has been
reported to range from 1 to 6 per cent with the
subclavian vein approach to catheterization, but
use of the internal jugular vein approach for inser-
tion of the catheter has dramatically reduced this
complication. The pericardium and vessels in the
mediastinum can be invaded by the probing
needle, causing cardiac tamponade.
If the syringe used for venous blood aspiration
is disengaged from the finder needle during a
spontaneous inspiration, it is possible for environ-
mental air to be sucked into the central venous
circulation and result in air embolism.
199
Air em-
bolism has also been reported with the use of a
pulmonary artery catheter-introducer kit that did
not provide a self-sealing introducer port
200
as well
with the luminal dilators that pass through the self-
sealing valves.
201
It is estimated that, during 1 sec,
100 ml of air can flow through a 14-gauge needle
with a 5-cm H
2
0 pressure drop across it,
144
and
100 to 300 ml of air can be fatal. If this occurs,
the source of air entry must be occluded, the F, 0
:
increased to 1.0 (to decrease the size of the em-
bolus by causing resorption of nitrogen in the air).
and the patient turned on the left side and to the
Trendelenburg position, which facilitates moving
an air bubble to the apex of the right ventricle
(away from the right ventricular outlet) and in-
creases CVP. Aspiration of air through a catheter
in the right ventricle can be effective.
Pulmonary artery catheters have been fractured
in half
202
and have been cut in two at the time of
surgery.
203
Catheter fragmentation is particularly
likely to occur with a catheter through-needle de-
vice. In this case, retraction of the catheter may
cause it to shear off at the tip of the needle. Thus,
if a pulmonary artery catheter needs to be removed
(e.g., ventricular arrhythmia has occurred) and the
introducer needle is still in the vein, it is impera-
tive to withdraw the catheter and the needle simul-
taneously.
Application of a plastic drape over the patient's
face for pulmonary artery catheterization while the
patient receives oxygen, 3 L/min via nasal cannu-
las, is associated with a mild amount of C0
2
re-
breathing (P,C0
2
= 6 2 mm Hg; PETCO, in-
creased from 37 to 39 mm Hg and P
a
C(X increased
from 41 to 43 mm Hg).
204
b. COMPLICATIONS OF ACTUALLY PASSING
AND FLOATING THE PULMONARY ARTERY
CATHETER
Rupture of an inflated pulmonary artery catheter
balloon has been reported (see Table 7-7), but no
serious embolic sequelae from 1.5 ml of air have
been noted.
205
Rupture of the balloon of a pulmo-
nary artery catheter is partly related to the duration
of catheterization, because the balloon loses elas-
ticity with exposure to blood. Rupture of the bal-
286
loon is detected by noting a decreased resistance
when inflation is attempted. Also, blood may be
aspirated from the balloon port.
Arrhythmias are not uncommon during passage
of the catheter through the right side of the heart
(approximately 15 per cent of pulmonary artery
catheter passages cause premature ventricular con-
tractions), although persistent atrial arrhythmias,
206
transient RBBB, complete heart block,
207
and ven-
tricular fibrillation have occurred.
208
In patients
with pre-existing left bundle branch block, tran-
sient interruption of conduction via the right bun-
dle branch may occur during insertion of a balloon
flotation catheter, which may result in complete
atrioventricular block and asystole. However, pre-
vious insertion of a pacing catheter or the use of a
pacing-port pulmonary artery catheter can prevent
this complication.
183
The major risk factors for the
Table 7-7 CLINICALLY SIGNIFICANT COMPLICATIONS OF GAINING CENTRAL VENOUS ACCESS,
INSERTING AND USING A PULMONARY ARTERY CATHETER, AND MAINTAINING
CENTRAL VENOUS ACCESS*
Pulmonary
Catheterization
Time/Event Complication
Approximate
Incidence
(%) Prevention Treatment
Monitoring 287
Table 7-7 CLINICALLY SIGNIFICANT COMPLICATIONS OF GAINING CENTRAL VENOUS ACCESS,
INSERTING AND USING A PULMONARY ARTERY CATHETER, AND MAINTAINING
CENTRAL VENOUS ACCESS* Continued
Pulmonary Approximate
Catheterization Incidence
Time/Event Complication (%) Prevention Treatment
Pulmonary artery 0.2 Use pulmonary artery Pull back catheter 1-2 cm.
rupture catheter position inflate balloon until P
pa
monitoring catheter (right trace becomes damped,
ventricular port), avoid single-lumen tube/positive-
distal migration (pull pressure ventilation,
pulmonary artery catheter double-lumen tube/high
tip back to main CPAP to bleeding lung,
pulmonary after each P
pao
lobectomy, transfusion,
determination, especially if systemic support
during cardiopulmonary
by-pass), minimize
inflation time, avoid
hyperinflation of balloon,
inflate slowly with
continuous waveform
monitoring
Maintaining central Thrombosis >50% by Increase cardiac output and None
venous access venography blood pressure, use
heparin-bonded catheters,
minimize indwelling/
catheterization time
Sepsis secondary to 0-1 Sterile technique, change See chapters 18, 19; remove
catheter catheters, minimize catheter; administer
catheterization time antibiotics, systemic
support
Abbreviations: PVC
airway pressure.
premature ventricular contractions; RBBB = right bundle branch block; CPAP = continuous positive
development of arrhythmias include myocardial is-
chemia, hypoxemia, electrolyte disorders, and aci-
dosis.'
49
Prolonged catheterization time is another
risk factor. Although this may indicate a less skill-
ful operator, longer catheterization time may also
reflect increased difficulty of catheterization in pa-
tients with severe shock, large right ventricles, di-
lated pulmonary arteries, and marked pulmonary
hypertension.
200
Occasionally, it may be difficult to pass the pul-
monary artery catheter out of the right ventricle,
especially if the right ventricle is dilated (a circum-
stance that also promotes premature ventricle con-
tractions). The head-up and right lateral tilt posi-
tion appears superior to Trendelenburg's position
for passage of pulmonary artery catheter in the
awake patient
209
and the anesthetized patient
210
;
this may be due to the fact that balloon flotation
(tendency of the balloon to seek a nondependent
position) is more important than blood flow.
192
Very occasionally, a catheter may not pass out
of the right atrium; tricuspid regurgitation creates
a whirlpool type of blood flow in the right atrium,
denying entry of the air-inflated balloon into the
right ventricle.
2
" Catheters have knotted on
themselves,
212
2I3
with other indwelling cathe-
ters,
214
and around papillary muscles.
215
-
216
A num-
ber of nonsurgical techniques for removing knot-
ted catheters have been described.
212-218
Passage of pulmonary artery catheters has
Table 7-8 COMPLICATIONS OF CENTRAL VENOUS CATHETERIZATION IN RELATION TO SITE*
*From Tobin M: Pulmonary artery catheter problems. Appl Cardiopulm Pathophysiol 3:279-285, 1990. Used with permission.
288 Monitoring
caused damage to the tricuspid
219
and pulmonary
valves
220
as well as endothelial trauma along the
entire catheter path.
200
Catheters have permanently wedged in distal
pulmonary vessels with subsequent pulmonary in-
farction in the distribution of the vessel containing
| the catheter.
221
222
Infarction is most likely to occur
within 12 to 24 hours of insertion. The risk of
infarction can be decreased by careful monitoring
of the vascular waveform and probably also by the
use of continuous flush with heparin solution.
Pulmonary artery catheters have also perforated
pulmonary vessels,
223
causing massive pulmonary
hemorrhage
224,225
and hemoptysis,
226
and have per-
forated the right ventricle, causing pericardial tam-
ponade.
197
The mortality rate after rupture of a
pulmonary artery by a pulmonary catheter is
greater than 50 per cent and is still higher in hep-
arinized patients. Associated factors include pul-
monary artery hypertension, advanced age, fragil-
ity of tissues, peripheral catheter-tip locations,
anticoagulation (which is unlikely to predispose to
rupture but probably contributes to morbidity once
it occurs), and cardiopulmonary by-pass. Retrac-
tion of the heart, especially to expose distal cir-
cumflex vessels, may push the catheter peripher-
ally, and hypothermia stiffens the catheter,
increasing the risk of perforation.
Although the exact mechanisms of catheter-as-
sociated pulmonary artery rupture have not been
identified, the common factor in most reported
cases appears to be the location of the catheter tip
in distal pulmonary artery branches outside the
mediastinal shadow on the chest X-ray. Smaller
branches may be lacerated by spearing of the ves-
sel wall by the catheter tip propelled by contrac-
tions of the heart or by overdistention of the vessel
at the time of balloon inflation.
Patients typically present with hemoptysis,
which may occur up to 20 hours after inserting the
catheter. In about one third of patients, rupture and
hemoptysis occur almost immediately after cathe-
ter introduction, and typically there is a chest X-
ray infiltrate near the tip of the catheter.
Several important preventive procedures can be
used. First and foremost, with respect to distal
migration of the catheter tip, phasic P
pa
waveform
should always be displayed. There is now avail-
able a pulmonary artery catheter position-monitor-
ing catheter (PA Watch Catheter; Baxter Health-
care, Santa Ana, CA), which has a right ventricular
port 10 cm from the tip of the pulmonary artery
catheter.
227
228
This catheter is considered to be in
proper position when the middle-lumen port,
which is located 10 cm from the tip, transmits a
right ventricular pressure waveform. Using this
pulmonary artery catheter, it is very clear that pul-
monary artery catheters migrate distally (right ven-
tricular port enters pulmonary artery) very fre-
quently (i.e., 50
227
-100 per cent
228
of patient^
Withdrawal distances (to get the right ventricul
port out of the pulmonary artery) ranged from 1 io
6 cm.
227
Thus, this new catheter may add a margin
of safety to pulmonary artery monitoring a
lower its overall cost by eliminating the need L*
chest radiographs ordered solely to confirm cathe-
ter-tip location.
Second, balloon hyperinflation should I
avoided. The balloon should always be inflatec
slowly, with continuous waveform monitori
(P
pao
position should not be detected by palpatij
of the pilot balloon).
229
Third, relocating the catheter tip in the proxirr**
pulmonary artery (especially during cardiopulm
nary by-pass) and readvancing it for each wedgT
pressure determination significantly reduces *
risk of catheter-induced pulmonary hemorrhage.
Management of a ruptured pulmonary arter
should be progressively aggressive as necessan
The pulmonary artery catheter should be pull
back 1 to 2 cm and the balloon inflated until tk-
phasic P
pa
trace dampens; this partial unilater;
occlusion will greatly decrease the blood flow
the bleeding area; infusion of epinephrine throu[__
the distal port may also be helpful. AnticoaguL
tion should be reversed if possible. To avoid co
tamination of the nonbleeding lung, the patie_
may need to be placed in a decubitus position wi
the affected side down. Tracheal toilet and fibe
optic bronchoscopy (perhaps with placement of
bronchial blocker) should follow. If respirato~
distress transpires, endotracheal tube intubatl
and positive-pressure ventilation are indicate
Double-lumen tube intubation is indicated if t
bleeding is massive or if bilateral contamination
a real threat. With double-lumen tube intubatk
differential lung ventilation is possible, and 1
bleeding lung may be tamponaded and held s
from the airway side by high (40-50 cm H
2
CPAP (but not ventilated). With all these man<
vers, systemic support should be provided as n<
essary (transfusion, vasopressure, and so on),
nally, if all else fails to control the bleedi
surgical resection may be necessary.
c. COMPLICATIONS OF MAINTAINING CENTRAL
VENOUS ACCESS
Thrombophlebitis of the veins in question is
ways possible (see Table 7-7), and indwelling
cular catheters and sheath introducers can alw
serve as a source of infection and sepsis.
205
231
Thrombi commonly form on the surface of \
monary artery catheters soon after their insert
as a result of an interaction between blood and
physiochemical properties of the catheter's suri
as well as secondary to catheter-induced endoi
Monitoring 289
Hal damage (including heart valve damage, pre-
dominantly the pulmonary valve). In addition to
thrombus formation on the catheter, venous throm-
bosis has been found by venography in two thirds
of patients with internal jugular vein catheters.
217
Such catheter-related thrombosis has been shown
to be associated with an increased risk of proximal
pulmonary artery thromboembolic
38
Thrombotic
complications appear to have been reduced by
heparin bonding of catheters and by minimizing
the time the catheter is indwelling.
239
240
Catheter-related infection is thought to arise in
the fibrin sheath that forms around the catheter
within 24 to 48 hours of insertion.
236
Most can-
nula-related infections arise from organisms that
migrate from the patient's skin into the cannula
tract. These originate from the patient's own skin
flora or from the hands of medical personnel. The
tip may also be colonized hematogenously by or-
ganisms from other sites. Prevention and treatment
of sepsis are standard and are discussed to some
extent in chapter 19.
These complications of central venous access
and catheter flotation are admittedly rare events;
therefore, many of the references to these compli-
cations are case reports (which involve only one
or a few patients) rather than studies of series of
patients. Not surprisingly, therefore, one large se-
ries of 1400 patients concluded that, taking every-
thing into consideration, pulmonary artery cath-
eterization is associated with a very acceptable low
incidence of morbidity and mortality.
198
Neverthe-
less, to make an intelligent judgment about
whether a pulmonary artery catheter should be
used in a given patient, it is necessary to enter this
information into the risk-benefit equation or bal-
ance.
7. Transesophageal Echocardiography
Although transesophageal echocardiography
(TEE) can reveal information about atrial and ven-
tricular septal, valvular, and aortic function, intra-
cardiac masses, and ventricular volumes, the main
purpose of intraoperative TEE for noncardiac tho-
racic surgery is to monitor ventricular ischemia.
The main contraindications to the use of TEE in-
volve esophageal diseases such as strictures, var-
ices, scleroderma, esophagitis, upper gastrointes-
tinal bleeding, dysphagia, history of esophageal
surgery, and chest wall radiation therapy.
241
Com-
plications from the insertion and use of the TEE
probe are not a common occurrence (~1 per cent)
and consist of arrhythmias, bronchospasm, unsuc-
cessful insertion, and esophageal perforation (0.02
per cent).
241
At present, the standard TEE images are all
transverse (cross sectional). There are four major
transverse views: the basal short-axis view (Fig.
7-39A), frontal long-axis or four-chamber view
(Fig. 7-395), the short-axis view of the left ventri-
cle at midpapillary region (Fig. 7-39C), and the
long-axis view of the descending thoracic aorta
(Fig. 7-39D).
242
Semiquantitative and quantitative
tools like color-flow mapping (CFM) and Doppler
echocardiography have further enhanced the utility
of TEE. CFM simultaneously presents real-time
images of intracardiac blood flow and structure in
two dimensions. The blood flow going toward the
transducer is depicted in various hues of red and
away from the transducer in various hues of blue.
CFM facilitates evaluation of valvular and congen-
ital cardiac lesions. Doppler echocardiography
uses the Doppler principle; in valvular heart dis-
ease, by accurately measuring the flow velocity by
Doppler shift, the valve cross-sectional area and
the regurgitant flow can be reliably estimated.
24.
The identification of regional wall abnormalities
and their association with coronary artery ischemia
has played a major role in the development of
TEE. Ischemia may be supply ischemia, demand
ischemia, or mixed ischemia. The earliest sign of
ischemia is impaired ventricular relaxation. This
sign is followed (in order of decreasing sensitivity )
by regional wall-motion abnormalities, impaired
global systolic function, ventricular pressure-vol-
ume (compliance) changes, electrocardiographic
changes of the ST segment, and only then by chest
pain.
In one study of 98 patients anesthetized for cor-
onary artery bypass surgery, myocardial ischemia
was identified by TEE (wall-motion abnormalities)
in 14 patients; in 10 of these, it was associated
with concomitant ST-segment depression of at
least 1 mm. The onset of ischemia, as defined by
TEE, was accompanied by a mean increase in pul-
monary capillary wedge pressure of 3.5 4.8 mm
Hg compared with a mean change of 0 2.2 mm
Hg between observations not associated with the
onset of ischemia (p < .01). An increase in P
pao
of
at least 3 mm Hg tested as an indicator of ische-
mia, but the sensitivity of this indicator was only
33 per cent and its positive predictive value was
only 16 per cent.
121
The left ventricle short-axis view at the level o
the papillary muscles is used to determine ische-
mia (this area represents perfusion by all three
coronary vessels), and the ventricular wall may be
divided into four segments (Fig. 7-40).
I21
Each
myocardial segment (anterior, posterior, septal, lat-
eral) may be examined separately. Normal seg-
mental contraction of the heart may be defined as
shortening of the radius of the left ventricle by
more than 30 per cent with wall thickening. One
of the representative scoring schemes for regional
wall-motion abnormalities defines wall-motion ab-
290 Monitoring
Basal short axis
Basal 4-chamber
Figure 7-39 Schematic representa-
tion of different transesophageal echo-
cardiography (TEE) scan planes and the
corresponding images. The at the apex
of the sector indicates that the transducer
lies posteriorly to the imaged structure.
A, Basal short-axis views: (1) main pul-
monary artery, (2) pulmonic valve, (3)
aortic root and coronanes, (4) aortic
valve. B, Basal four-chamber views: (5)
left ventricle (LV) outflow tract, (6) clas-
sic four chamber, (7) coronary sinus.
Monitoring 291
LV short axis
Figure 7-39 Continued C, LV
short-axis views: (8) mitral valve, (9)
midpapillary muscle, (10) apical. D,
LV long-axis views: (11) apical, (12)
obtuse angle, (13) short-axis view of
the descending thoracic aorta. A =
anterior; Ao = aorta; AL = antero-
lateral papillary muscle; CS = coro-
nary sinus: FO = fossa ovalis; IVC
= inferior vena cava; LA = left
atrium; LAA = left atrial append-
age; LCA = left coronary artery;
MPA = main pulmonary artery; PM
= posteromedial papillary muscle;
PV = pulmonic valve or pulmonary
vein; RA = right atrium; RAA =
right atrial appendage; RCA =
right coronary artery; RLPV = right
lower pulmonary vein; RPA =
right pulmonary artery; RUPV
= right upper pulmonary vein; RV
= right ventricle; SVC = superior
vena cava. (From Thys DM, Hillel
Z: Echocardiography. In Benumof JL
(ed): Clinical Procedures in Anesthe-
sia and Intensive Care. Philadelphia,
J. B. Lippincott, 1992, pp 459^196.
LV long axis and descending aorta
292 Monitoring
Figure 740 Schematic representation of echocardiographic
short-axis view of left ventricle at the level of papillary mus-
cles, indicating division into four segments of myocardium
used for grading wall motion per segment (see text). A =
anterior; S = septal; L = lateral; = posterior; LV = left
ventricle; RV = right ventricle. (From Van Daele MERM,
Sutherland GR, Mitchell MM, et al: Do changes in pulmonary
capillary wedge pressure adequately reflect myocardial ische-
mia during anesthesia? Circulation 81:865-871, 1990. Used
with permission.)
normalities as follows: mild hypokinesis, evi-
denced by shortening of the radius of the ventricle
by less than 30 per cent but with more than 10 per
cent wall thickening; severe hypokinesis, evi-
denced by less than 10 per cent radial shortening
and minimal wall thickening; akinetic segment,
evidenced by no wall thickening during systole;
and, finally, a dyskinetic segment, evidenced by
left ventricular wall bulging and thinning during
systole. Regional wall-motion abnormalities can
occur with ischemia, but it is important to remem-
ber that they can also occur when there is prior
myocardial infarction with tissue fibrosis, an intra-
ventricular conduction abnormality caused by bun-
dle branch block, premature ventricular contrac-
tions, and ventricular pacing or during thoracotomy
because of shifting of cardiac structures.
241
Thus,
onset of new wall-motion abnormality in patients
with baseline impairment of systolic left ventricu-
lar function may be difficult to detect.
243
In human subjects undergoing coronary angio-
plasty, visible wall-motion abnormality may occur
within 10 to 15 sec of coronary occlusion.
244
Onset
of hypokinesia occurred within 15 sec of balloon
occlusion.
244
Hypokinesia develops at 50 per cent
reduction of coronary blood flow, whereas, with
90 per cent reduction in flow, dyskinesia takes
place.
243
New wall-motion abnormalities observed
intraoperatively"may be reversible when associated
with a short period of ischemia or may be irrevers-
ible when associated with infarction.
243
TEE is valuable in the diagnosis of aortic dis-
section; in one study, the sensitivity was 99 per
cent and the specificity 98 per cent.
245
TEE is very
sensitive in detecting as little as 0.01 ml/kg of air
(microbubbles) and is therefore extremely sensi-
tive in detecting a patent foramen ovale and intra-
cardiac air. Valsalva and cough maneuvers are
used to augment a right-to-left shunt in patients to
rule out a patent foramen ovale with contrast and
Doppler echocardiography. TEE is also helpful in
patients suspected of having an intracardiac mass
(thrombus, tumor, and/or vegetations), although
the sensitivity and specificity are not nearly as
good as for aortic dissections and intracardiac air.
TEE is very beneficial in visualizing cardiac val-
vular function. TEE can determine whether left
ventricular end-diastolic volume is very high or
very low but is moderately imprecise for follow-
ing, on-line, changes in left ventricular end-dia-
stolic volume of intermediate values.
8. Lung Water Measurements
The preceding pulmonary artery catheter meas-
urements greatly increase the understanding of
how the respiratory and cardiovascular systems
work together. Recently, with a combination of
these techniques and methodology, extravascular
lung water measurements have become available
for clinical use (American Edwards Laboratories
9310 Computer). Extravascular lung water is
measured using a thermal change-green dye con-
centration change double-indicator dilution tech-
nique. The two indicators are injected simultane-
ously into the central venous circulation and are
detected by a thermistor-tipped arterial catheter.
One indicator, indocyanine green dye, binds to
serum albumin and remains intravascular as it
passes through the lung. The other indicator, a
cold bolus of dextrose solution, diffuses through-
out the extravascular space at the same time.
Time-related dilution curves are determined and
analyzed by the computer. Cardiac output and
mean transient time are automatically calculated
for each indicator. The product of cardiac output
and mean transit time for a given indicator yields
the volume distribution for that indicator. The dif-
ference between the volume distribution of the two
indicators represents the volume of extravascular
water in the lungs. The correlation between the
double-indicator dilution technique and gravimt-
ries, the accepted standard, has been good (r =
0.96),
246-249
and it has been possible to perform
multiple determinations rapidly and reproducibly.
However, it is important to note that the correla-
tion between extravascular lung water measure-
ments and other observable parameters of respira-
tory function and the effect of various therapeutic
modalities have not yet been established.
Monitoring 293
9. Computed Tomographic Pulmonary
Scan
Computed tomographic (CT) scanning may
prove to be a new useful demonstration/monitor of
lung disease.
250
One study of adult respiratory dis-
tress syndrome patients showed that pulmonary
pressure increases in concert with lung weight and
that, as the mass of noninflated lung tissue in-
creases, so do venous admixture and V
[:
/V
T
, while
P
a
0
2
decreases.
251
Without the CT scanner, no
such correlations would have been possible, be-
cause there are no other means for measuring reli-
ably at the bedside either lung weight or nonin-
flated lung tissue mass. Earlier inhalation and
indicator dilution techniques that measured extra-
vascular lung water (see section IV.G.8.) were re-
stricted to assessing those areas within the reach
of inhaled or perfused agents, whereas X-rays are
not.
The correlation in early adult respiratory distress
syndrome of increased pulmonary artery pressure
with increased lung weight as estimated by CT
corroborates our current understanding of fluid and
solute transport in the injured lung. It is expected
that increasing P
p;i
(caused by hypoxic and media-
tor-included vasoconstriction, thrombosis, and
obliteration of pulmonary microvasculature; see
Fig. 7-27) in the presence of lung injury will cause
the lung to gain weight by directly forcing plasma
into the interstitial and alveolar spaces. As in other
studies, there was a marked posterior distribution
of regions of high density probably resulting from
the effect of gravity, causing airway and interstitial
fluid to migrate to dependent lung regions, thereby
filling airspaces with fluid and promoting consoli-
dation. PEEP augmented the recruitment of lung
volume during adult respiratory distress syndrome
by increasing low-density, presumably ventilated
lung regions at the expense of high-density, pre-
sumably nonventilated (shunted) regions.
REFERENCES
1. Cooper JB, Newbower BS, Kitz RJ: An analysis of major
errors. Anesthesiology 60:34-42, 1984.
2. Paulus DA, Basta JW, Klie H, Radson EA: Preanesthetic
checklist. Anesth Analg 64:264, 1985.
3. Debban DG, Bedford RF: Overdistension of the rebreath-
ing bag: A hazardous test for circle system integrity.
Anesthesiology 42:365-366, 1975.
4. Ward CS: The prevention of accidents associated with
anesthetic apparatus. Br J Anaesth 40:692-701, 1968.
5. Westenskow DR, Jordan WS, Jordan R, Gillmore ST:
Evaluation of oxygen monitors for use during anesthesia.
Anesth Analg 60:53-56, 1981.
6. Meyer RM: A case for monitoring oxygen in the expira-
tory limb of the circle. Anesthesiology 61:347, 1984.
7. Lee E: Of stethoscopes and stopcocks. Anesth Analg
73:98-99, 1991.
8. Schwartz N: Monitoring bilateral breath sounds. Anesthe-
siology 66:711-712, 1987.
9. Scheller M, Jones B, Benumof JL: Effect of changes in
I:E ratio and flow rate with constant tidal volume on
minute ventilation and P
a
COo. J Cardiothorac Anesthesiol
3:564-567, 1989.
10. Saklad M, Paliotta J, Weyerhauser A: On line monitoring
of ventilatory parameters. Clin Anesth 9:335-362. 1973.
11. Egbert LD, Biano D: The educated hand of the anesthe-
siologist. Anesth Analg 46:195-200, 1967.
12. Silverman MS, Bishop MJ: Manual vs mechanical venti-
lation in a model of bronchospasm: Does the "educated
hand" exist? Anesthesiology 71:A441. 1989.
13. Knill RL: Secondary prevention of hypoxemia. Anesthe-
siology 69:438-439, 1988.
14. Knill RL: Evaluation of arterial oxygenation during an-
aesthesia. Can Anaesth Soc J 32:S16-S19, 1985.
15. Comroe JH Jr, Botelho S: The unreliability of cyanosis in
the recognition of arterial anoxemia. Am J Med Sci
214:1-6, 1947.
16. Medd WE, French EB, Wylie VM: Cyanosis as a guide
to arterial oxygen desaturation. Thorax 14:247-250.
1959.
17. Kelman GR, Nunn JF: Clinical recognition of hypoxae-
mia under fluorescent lamps. Lancet 1:1400, 1966.
18. Cot C, Goldstein E. Cot M, Hoaglin DC, Ryan J: A
single blind study of pulse oximetry in children. Anesthe-
siology 68:184-188, 1988.
19. Manninen PH, Knill RL: Cardiovascular signs of acute
hypoxaemia and hypercarbia during enflurane and halo-
thane anaesthesia in man. Can Anaesth Soc J 26:282-
287, 1979.
20. Knill RL, Gelb AW: Ventilatory responses to hypoxia
and hypercapnia during halothane sedation and anesthesia
in man. Anesthesiology 40:244-251, 1978.
21. Knill RL, Kieraszewicz HT, Dodgson BG: Chemical reg-
ulation of ventilation during isoflurane sedation and an-
aesthesia in humans. Can Anaesth Soc J 30:607-614.
1983.
22. Landsgaard C, Van Slyke DD: Cyanosis. Baltimore, Wil-
liams & WilkinsCo, 1923.
23. Alexander CM. Teller LE, Gross JB: Principles of pulse
oximetry: Theoretical and practical considerations.
Anesth Analg 68:368-376, 1989.
24. Jobes DR, Nicolson SC: Monitoring of arterial hemoglo-
bin oxygen saturation using a tongue sensor. Anesth An-
alg 67:186-188, 1988.
25. Cot CJ, Daniels AL, Connolly M. Szyfelbein SK, Wick-
ens CD: Tongue oximetry in children with extensive ther-
mal injury: Comparison with peripheral oximetry. Can J
Anaesth 39:454^157, 1992.
26. Ezri T, Lurie S. Konichezky S, Soroker D: Pulse oximetry
from the nasal septum. J Clin Anesthesiol 3:447-450.
1991.
27. O'Leary RL, Landon M, Benumof JL: Buccal S
P
0
:
is
more accurate than finger S
P
0,. Anesth Analg 75:495-
498, 1992.
28. Kagle DM, Alexander CM, Berko RS, Guiffre M, Gross
JB: Evaluation of the Ohmeda Biox 3700 pulse oximeter:
Steady state and transient response characteristics. Anes-
thesiology 66:376-380, 1987.
29. Severinghaus JW, Naifeh KH: Accuracy of response of
six pulse oximeters to profound hypoxia. Anesthesiology
67:551-558, 1987.
30. Broome IJ, Harris RW, Reilly CS: The response times
during anaesthesia of pulse oximeters measuring oxygen
saturations during hypoxaemic events. Anaesthesia
47:17-19, 1992.
31. Guggenberger H, Lenz G, Federle R: Early detection of
inadvertent oesophageal intubation: Pulse oximetry vs
CHAPTER 8
Choice of Anesthetic Drugs and
Techniques
I. Introduction
II. Most Common and Important
Cardiopulmonary Considerations for
Patients Undergoing Thoracic
Surgery
A. Pulmonary Considerations (Two-
Lung Ventilation)
1. Reactive Airways
a. Effect of Anesthetic Drugs on
Airway Reactivity
2. Gas-Exchange Impairment
B. Cardiac Considerations
1. Coronary Artery Disease
a. Effect of Anesthetic Drugs on
Cardiovascular Function
III. Choice of Anesthesia and Arterial
Oxygenation During One-Lung
Ventilation
A. Effect of Anesthetics on Hypoxic
Pulmonary Vasoconstriction
B. Effect of Anesthetics on Arterial
Ventilation
1. Two-Lung Ventilation: Blood Flow
Distribution
2. One-Lung Ventilation: Blood Flow
Distribution, Shunt Flow, and
Arterial Oxygen Tension
3. Effect of Isoflurane on the One-
Lung Ventilation Blood Flow
Distribution, Shunt Flow, and
Arterial Oxygen Tension
IV. Recommended Anesthesia Induction
and Maintenance Drugs and
Techniques
A. Summary of Advantages of
Anesthetic Drugs
1. Inhalational Anesthetics
2. Intravenous Anesthetics
B. Recommended Anesthetic Drugs
and Technique
1. Insertion of an Epidural Catheter
Before Induction of Anesthesia
2. Induction of Anesthesia
3. Maintenance of Anesthesia
4. Total Intravenous Anesthesia
Technique
300
I. INTRODUCTION
The choice of anesthetic drug and technique in
the vast majority of thoracic surgery cases is based
on the preoperative cardiopulmonary evaluation.
Some drugs and techniques may favor the overall
perioperative function of one organ at the expense
of another. For example, the halogenated drugs
may prevent or minimize bronchospasm but at the
same time may decrease myocardial contractility,
whereas the narcotics may preserve myocardial
contractility but may not prevent bronchospasm in
patients with reactive airways and may cause post-
operative respiratory depression. Because different
patients will have varying degrees of dysfunction
of different organs and anesthetic drugs differen-
tially affect the various organs, the most appropri-
ate anesthetic will depend on the patient.
This chapter first briefly considers the usual and
major pulmonary and cardiac problems that pa-
tients undergoing thoracic surgery may have (with
both lungs ventilated) as well as the most impor-
tant effect anesthetic drugs might have on these
problems. Next, the specific effect of anesthetic
drug and technique on gas exchange for the special
one-lung ventilation situation (in particular, on hy-
poxic pulmonary vasoconstriction [HPV]) is cov-
ered. Finally, the chapter recommends anesthetic
drugs and techniques that should simultaneously
minimize the pulmonary and cardiac problems but
yet incorporate enough flexibility to emphasize the
function of one organ over another, should that be
considered necessary. As with most of the other
chapters in this book, respiratory considerations
are emphasized more than cardiac considerations.
II. MOST COMMON AND
IMPORTANT CARDIOPULMONARY
CONSIDERATIONS FOR PATIENTS
UNDERGOING THORACIC
SURGERY
A. Pulmonary Considerations (Two-
Lung Ventilation)
1. Reactive Airways
The mechanisms by which increases in airway
resistance may be stimulated have been summa-
rized' and include mechanical and chemical mu-
cosal stimulation (causing various neural reflexes
that are mediated by medullary centers, local arcs,
and the autonomic nervous system), anaphylactoid
bronchoconstriction, histamine-induced broncho-
constriction, alpha-adrenergic predominance, va-
gal (cholinergic) predominance, and exercise-in-
duced bronchoconstriction. In this discussion, the
term "reactive airways" applies to patients who
come to surgery with pre-existing bronchospasm
or will likely react to the mechanical stimulation
of endotracheal tube insertion and/or tracheobron-
chial tree manipulation with bronchospasm.
Patients undergoing thoracic surgery are likely
to have an increased incidence of reactive airways
for two reasons. First, the vast majority of patients
have a long and significant smoking history and
therefore have varying degrees of chronic obstruc-
tive pulmonary disease, excess secretions, and in-
creased reversible airway resistance (bronchocon-
striction). The presence of the increased airway
resistance can often be readily demonstrated by
improvement in expiratory airflow rates following
the preoperative administration of bronchodilating
drugs. Indeed, there is a strong dose-response re-
lation between the amount of secretions, cough,
bronchoconstriction, severity of chronic obstruc-
tive pulmonary disease, and the risk of mortality
from chronic obstructive pulmonary disease and
the number of cigarettes smoked per day, the num-
ber of years of smoking, and the depth of smoke
inhalation.
2
Second, thoracic surgery often de-
mands direct manipulation of the tracheobronchial
tree. Although much of the direct surgical contact
is external (on the adventitia), the manipulations
(clamping, compression between the fingers)
usually cause mucosal stimulation, as does endo-
tracheal intubation, and can therefore cause bron-
chospasm. Two studies suggest that, just as endo-
thelial cells are important modulators of vascular
tone, the epithelium of the airway is a metaboli-
cally active tissue that can modulate airway
smooth-muscle tone/function by the production
and destruction of inflammatory mediators such as
prostaglandins and perhaps relaxing and constrict-
ing factors.
3
4
Epithelial damage may result in loss
of an epithelium-derived relaxant factor and loss
of enzymes that degrade constrictor neuropep-
tides.
5
In addition, damage to the epithelium in-
creases access of luminal substances to muscle and
to afferent nerve fibers that elicit the irritant bron-
choconstrictor responses seen so often with airway
instrumentation during anesthesia.
5
In fact, even
normal, healthy patients undergoing surgical stim-
ulation of the lung parenchyma and airways can
experience bronchospasm,
6
especially if they are
too lightly anesthetized.
a. EFFECT OF ANESTHETIC DRUGS ON AIRWAY
REACTIVITY (Table 8-1)
The following discussion of the pharmacology
of anesthetic drugs is limited to those aspects that
suggest their use or avoidance in patients with
reactive airways. Some of the drugs used in anes-
thetic practice (especially the halogenated drugs)
decrease the reactivity of airways. However, there
302 Choice of Anesthetic Drugs and Techniques
Table 8-1 EFFECT OF ANESTHETIC DRUGS ON AIRWAY REACTIVITY
Anesthetic Drug
Class Drug Bronchodilation Bronchoconstriction Comment
Inhalation Isofiurane + + 0 Drug of choice
Halothane + + 0 See text
Enflurane + + 0 See text
N
2
0 0 0 Light anesthesia
Narcotics Fentanyl 0 0 Drug of choice
Meperidine 0 + Releases histamine
Morphine 0 + + Releases histamine
Induction Ketamine + 0 Drug of choice for
asthmatic
Thiopental 0 + ? Light anesthesia?
Relaxants Vecuronium 0 0 Drug of choice
Pancuronium 0 0 Drug of choice
Succinylcholine 0 0 Drug of choice
Atracurium 0 + ? Releases histamine, high
doses
Metocurine 0 + Releases histamine,
highmoderate doses
i/-tubocurarine 0 + + Releases histamine,
normal doses
Adjuncts Lidocaine + 0 Useful pre- and
intraoperatively
Neostigmine 0 + + + Must use atropine
concomitantly
Atropine + 0 See text
Key: + , + * - , + + + = mild, moderate, severe effect, respectively.
must be some initial smooth-muscle constriction in
order for a bronchodilating anesthetic to have any
effect on bronchomotor tone. Thus, measurements
of airway resistances in normal men have gener-
ally failed to show a bronchodilatory effect of hal-
othane.
7
Similarly, in patients with normal bron-
chomotor tone on cardiopulmonary by-pass, the
administration of halothane both by the airway and
systemically (via the pump circuit) does not alter
the resistive work of breathing (as it does in pa-
tients with increased bronchomotor tone; see the
following).
8
When bronchoconstriction is provoked in pa-
tients by either hypocapnia
9
or inhalation of ultra-
sonic aerosols,
10
halothane and enflurane reliably
decrease bronchomotor tone.
There are two important mechanisms of halo-
genated drug inhibition of bronchomotor tone. The
first, and most important, the mechanism of bron-
chodilation by the halogentated drugs is by direct
action on the airway musculature and/or local re-
flex arcs rather than, or at least in addition to, an
effect on centrally mediated reflex pathways. This
contention is best supported by the fact that sys-
temic (intravenous) administration of halothane
via cardiopulmonary by-pass pump does not de-
crease hypocapnia-induced increased airway resis-
tance, whereas inhaled halothane (during cardio-
pulmonary by-pass) does.
8,
" Because halothane
has a direct relaxant effect on bronchial smooth
muscle, it is not surprising that in animals it can
block acetylcholine-, histamine-, alpha-adrenergic-,
and antigen-induced increases in bronchial muscle
tension.
12
Second, halothane also blocks central reflex
pathways. Clinically relevant concentrations of
halothane reduced the amount of acetylcholine re-
leased from nerve terminals in response to nerve
stimulation and reduced the size of the muscle
contraction. Thus, inhalation anesthetics such as
halothane suppress the final step of the reflex path-
way, particularly vagal reflexes.
5
The fact that the
direct dilating effects are more important than the
effects on reflexes is highlighted by one study that
showed that at 1.2 minimum alveolar concentra-
tion (MAC), 60 per cent of the decrease in re-
sponse to vagal nerve stimulation caused by halo-
thane could be attributed to a direct effect on
smooth muscle.
13
Isoflurane is probably just as ef-
ficacious as halothane or enflurane in decreasing
elevated bronchomotor tone.
5
I2
l4
Because the halogenated drugs are the most po-
tent bronchodilating anesthetic drugs used today,
they must be considered the anesthetic drug of
choice for patients with reactive airways. If the
halogenated drug is used as a primary induction
agent for a patient with reactive airways, then hal-
othane might be considered the induction drug of
choice because it is less pungent than isoflurane.
However, isoflurane is a better choice for the
maintenance of anesthesia for four reasons.
First, isoflurane has a high arrhythmogenic
threshold (in contradistinction to halothane, which
sensitizes the myocardium to catecholamines and
is commonly associated with ventricular arrhyth-
mias). This consideration has increased importance
in patients with reactive airways because they may
more likely receive aminophylline and
2
^8 8
and may more likely become acidotic (all of which
may cause arrhythmias).
Second, isoflurane is not metabolized as much
as halothane (it has less or no hepatic toxicity).
Third, isoflurane provides much more cardio-
vascular stability and potency than enflurane.
Fourth, isoflurane is efficacious in treating very
severe bronchospasm (status asthmaticus).
15-18
Consequently, isoflurane should be regarded as the
halogenated drug of choice for maintenance anes-
thesia in patients with reactive airways.
Fentanyl does not have any effect on broncho-
motor tone, which is consistent with the fact that,
in humans, fentanyl does not change plasma his-
tamine concentrations.
19
In contrast, morphine is
known to release histamine and to increase central
vagal tone. In dogs, bronchoconstriction caused by
morphine has been shown to be reduced either by
administration of an antihistamine intravenously or
by bilateral vagotomy.
20
Similarly, meperidine has
been shown in dogs to have a bronchoconstricting
effect similar to that of morphine.
20
Consequently,
in patients with reactive airways for whom a nar-
cotic supplement to nitrous oxide (which has no
effect on bronchomotor tone) or halogenated drug
anesthesia is desired, or to use a narcotic alone in
high doses, fentanyl is the preferred choice. How-
ever, it should be remembered that an N
2
0 nar-
cotic-relaxant anesthetic, as ordinarily adminis-
tered with low to moderate doses of narcotic,
produces a light anesthesia and will not prevent
bronchospasm in patients with reactive airways.
Ketamine and thiopental are intravenous anes-
thetic drugs that are mainly associated with the
induction of anesthesia. For the patient with reac-
tive airways, ketamine is believed to be more ad-
vantageous than most other anesthetic drugs used
for induction. In dogs, ketamine protects against
antigen-induced bronchospasm, while thiopental
does not; the protective effect can be blocked by
propanolol, suggesting that the mechanism of ac-
tion of ketamine is, perhaps, due to beta-adrener-
gic stimulation
21
(although the drug may have a
direct bronchodilating effect on airway smooth
muscle).
22
Thus, ketamine has been used very suc-
cessfully in patients with a history of asthma
23
and
in treating patients with bronchospasm while on
mechanical ventilation (who were refractory to
maximum broncholytic therapy),
24
whereas thio-
pental has clinically been associated with more
bronchospasm than any other anesthetic drug (per-
haps because of light levels of anesthesia that
leave airway reflexes relatively intact).
25
Conse-
quently, ketamine (1 to 2 mg/kg) may be the drug
of choice for the intravenous induction of anesthe-
sia in patients with bronchospastic disease requir-
ing a rapid induction of anesthesia. Ketamine has
been successfully used as the sole anesthetic for
thoracic surgery.
26
With the exception of d-tubocurarine, all of the
muscle relaxants may be used in patients with re-
active airways. In humans, i/-tubocurarine releases
histamine and increases airway resistance.
27
Meto-
curine can also release histamine in ordinary clin-
ical doses but not nearly as much as d-tubocura-
rine. Atracurium may release histamine but only
in very high doses. Succinylcholine is structurally
related to acetylcholine and could theoretically re-
lease histamine, but clinically it is unassociated
with bronchospasm, even in asthmatics. Pancuro-
nium and vecuronium do not release histamine.
Lidocaine, given immediately before intubation
(1-2 mg/kg intravenously), is a useful drug in the
prevention of reflex bronchoconstriction and laryn-
gospasm provoked by instrumentation of the
airway. Similarly, lidocaine by infusion (1-3
mg/kg/hour) may be useful in diminishing the
reactivity of airways throughout surgery in patients
with limited cardiac reserve who would not he-
modynamically tolerate the usual doses of the
usual anesthetics.
Intravenous administration of lidocaine has also
been successfully used to treat bronchospasm dur-
ing anesthesia.
28
Intravenously administered lido-
caine also significantly lowers the incidence of
induced laryngospasm (laryngospasm quantita-
tively defined/measured as an increase in intraglot-
tic pressure of 45 15 mm Hg for 61 19 sec);
a serum lidocaine level of 0.4 g/ml appears to be
the lower limit of the therapeutic range, although
further study is necessary to construct a dose-re-
sponse curve.
29
Lidocaine, inhaled in an ultrasonic aerosol, has
been shown to produce mild bronchodilation in
healthy subjects, whereas normal saline induced
mild bronchoconstriction.
30
In a similar study, li-
docaine administered by aerosol both prevented
and reversed increases in airway resistance pro-
voked by ultrasonically nebulized water.
31
The
lidocaine was thought to be acting directly on
bronchial smooth muscle. Thus, it appears that ad-
ministration of lidocaine either via the airway or
intravenously has a role in preventing and protect-
ing against the development of bronchospasm in
patients with reactive airways.
Neostigmine, physostigmine, and pyridostig-
mine are cholinesterase-blocking drugs that can be
expected to produce an increase in airway resis-
tance by increasing cholinergic activity. In clinical
practice, of course, atropine is used to block this
effect when anticholinesterases are administered to
reverse the effects of neuromuscular blocking
drugs.
Atropine, the classic cholinergic blocker, not
only reverses effects of anticholinesterase drugs
but can also directly dilate the airways. A decrease
in airway resistance was found following intrave-
nous administration of atropine (0.84 mg/70 kg) to
patients anesthetized with 75 per cent nitrous ox-
ide and oxygen.
32
In addition to its potential effect
on airway resistance, atropine has been shown to
increase respiratory dead space in both dogs and
humans, presumably by dilating larger bronchi.
33
Table 8-1 summarizes the effect of inhalation
and intravenous anesthetic drugs on bronchomotor
tone and, in view of the preceding discussion,
ranks the various drugs of each class in terms of
their efficacy for the patient with reactive airways.
Ketamine appears to be the intravenous induction
drug of choice, with an adequate dose of thiopental
as a close second. Intravenous lidocaine is proba-
bly a useful adjunct in the peri-induction period.
Isoflurane appears to be the halogenated drug of
choice for maintenance anesthesia. Fentanyl is the
narcotic of choice. Relaxation can be facilitated by
succinylcholine, but if a rapid intubation is not
necessary, vecuronium or pancuronium may be the
relaxants of choice, with atracurium a reasonably
close second choice. Nitrous oxide is a benign
drug for the patient with reactive airways (if anes-
thesia caused by other drugs is adequate), but in
view of the fact that a large number of patients
undergoing thoracic surgery will have one-lung
ventilation and therefore require a high F,0
2
, use
of nitrous oxide is limited in these patients.
2. Gas-Exchange Impairment
Chapter 4 describes in great detail the patho-
physiology of two-lung ventilation in the open-
chest paralyzed patient in the lateral decubitus po-
sition. In summary, the nondependent lung may be
well ventilated owing to an increase in compliance
but may be poorly perfused owing to gravitational
effects. The dependent lung may be poorly venti-
lated owing to a decrease in compliance but may
be well perfused. Consequently, there may be mis-
matching of ventilation and perfusion in the open-
chest, paralyzed patient in the lateral decubitus
position. The low ventilation-perfusion ratio in the
dependent lung can be improved with selective
dependent-lung positive end-expiratory pressure
(PEEP) while the nondependent lung is ventilated
with zero end-expiratory pressure (ZEEP). The
differential lung ventilation combination of depen-
dent-lung PEEP and nondependent-lung ZEEP
provides better arterial oxygenation compared with
ventilation of both the nondependent and depen-
dent lungs with ZEEP.
Chapter 4 also describes in great detail the
pathophysiology of one-lung ventilation in the lat-
eral decubitus position. The one-lung ventilation
condition creates a large obligatory right-to-left
shunt that is not present during the two-lung ven-
tilation condition. This right-to-left shunt during
one-lung ventilation, however, is minimized by the
presence of HPV in the nondependent, nonventi-
lated lung. Section III of this chapter considers in
great detail the effects of anesthetic drugs on HPV
in the nonventilated, nondependent lung.
B. Cardiac Considerations
1. Coronary Artery Disease
Patients undergoing thoracic surgery, especially
those beyond their fourth decade, should be ap-
proached with an increased index of suspicion that
coronary artery disease may be present. Ten major
cohort studies, accounting for more than 20 mil-
lion person years of observation in several coun-
tries, support the statement in the 1983 Surgeon
General's report that "cigarette smoking should be
considered the most important of the known mod-
ifiable risk factors for coronary heart disease in the
United States."
34
These studies showed that men
40 to 59 years of age who were smoking a pack or
more per day at the time of initial examination had
a risk for a first major coronary event that was 2.5
times as great as that of nonsmokers, with a strong
dose-response relation.
35
Studies both in the
United States and abroad have demonstrated con-
sistently that women whose smoking patterns are
similar to those of men have a similar increased
risk of death from coronary heart disease and for
common morbidity from the disease, such as an-
gina pectoris, compared with nonsmokers.
34
36
The
risk of death from coronary heart disease among
both male and female smokers is increased by
early initiation of smoking, long total exposure to
smoking, and deep smoke inhalation.
Although smoking, hypertension, and hypercho-
lesterolemia confer approximately the same aver-
age increase in the risk of coronary heart disease
in populations, smoking in the presence of other
risk factors for coronary heart disease appears to
create a synergistic effect on mortality from the
disease.
34
The lifestyle of many smokers also in-
cludes caffeine and nicotine addiction and obses-
sive-compulsive type A behavior and results in an
increased incidence of systemic hypertension; the
combination of smoking and systemic hyperten-
sion increases the risk of coronary artery disease.
Pipe and cigar smokers have a risk of experiencing
a major coronary event and subsequent morbidity
from chronic heart disease that is intermediate be-
tween that for nonsmokers and cigarette smokers.
34
Smoking cessation results in a decreased risk of
mortality from coronary heart disease, and the de-
gree of risk reduction is determined by the length
of time after cessation, the amount smoked, and
the duration of smoking before cessation. Al-
though the risk of coronary heart disease attribut-
able to smoking declines by approximately 50 per
cent 1 year after cessation, only after a decade or
more does it approach that of a person who has
never smoked.
34
The anesthetic management and choice of anes-
thesia for patients with coronary artery disease are
based on the factors that determine myocardial
oxygen supply and demand (Fig. 8-1). Fortu-
nately, the reserve on the supply side is large
enough to make ischemia impossible in normal
hearts even with the most vigorous level of exer-
cise/stress. However, in the presence of coronary
artery disease or abnormal left ventricular hyper-
trophy, an imbalance between supply and demand
may occur with minimal stress/exercise or even at
rest and during anesthesia.
37
Myocardial oxygen supply is determined by the
product of coronary artery blood flow (most im-
portant, by far, in most clinical situations) and the
oxygen content of coronary artery blood. For a
given coronary vascular resistance, coronary artery
blood flow is increased by an increased diastolic
filling time (slow heart rate, most important oxy-
gen-supply factor; Q in Fig. 8-1), and the coro-
nary artery perfusion pressure (the second most
important oxygen-supply factor; (2) in Fig. 8-1).
The coronary perfusion pressure is equal to coro-
nary artery diastolic pressure (P
sad
) (there is no
perfusion during systole) minus the preload (which
is ventricular end-diastolic pressure or the outlet
DETERMINANTS OF MYOCARDIAL OXYGEN SUPPLY AND DEMAND: THE BALANCE
Figure 81 The determinants of myocardial oxygen supply and demand. See text for full explanation. (SVR = systemic vascular
resistance; EDRF = endothelial relaxing factor; P
LVBD
= left ventricular end-diastolic pressure; P
SAD
= systemic arterial diastolic
pressure.
or back pressure). Coronary vascular resistance is
(1) inversely proportional to myocardial metabolic
rate (this coupling mechanism acts within one car-
diac cycle, and maximal coronary dilation or con-
striction can be elicited within 15 to 20 sec), (2)
directly proportional to perfusion pressure so that
blood flow is kept constant between 60 and 160
mm Hg (at less than 60 mm Hg, blood flow be-
comes pressure dependent), (3) directly propor-
tional to systolic compression (the magnitude of
this effect decreases from endocardium to epicar-
dium, and, in fact, left ventricular subendocardial
blood flow ceases during systole), (4) decreased in
response to humoral factors (angiotensin II, sero-
tonin, thromboxane, prostacyclin, and bradykinin)
that can promote or inhibit release of endothe-
lium-derived relaxing factor, and (5) under neural
influence (alpha-adrenergic stimulation induces
coronary constriction; beta-adrenergic and para-
sympathetic stimulation induces coronary vaso-
dilation).
37
The oxygen content of arterial blood is
a function of the amount of hemoglobin (most
important), position of the oxygen-hemoglobin
dissociation curve (P
50
), and the ventilation-perfu-
sion relationships within the lung (which deter-
mines P
a
0
2
).
Myocardial oxygen demand is raised by tachy-
cardia (the most oxygen-expensive factor; in
Fig. 8-1), increased diastolic and systolic myocar-
dial wall tension, and increased contractile state of
the heart (as caused degree of sympathetic nervous
system activity). Diastolic myocardial wall tension
is determined by preload (left ventricular end-dia-
stolic [LVED] pressure; the second most oxygen-
expensive factor; (5) in Fig. 8-1), and systolic
myocardial wall tension is determined by afterload
(systemic systolic blood pressure and vascular re-
sistance).
Coronary obstruction causes a decrease in per-
fusion pressure. The pressure decrease across a
plaque is proportional to the radius (fourth power),
the length of stenosis, and the magnitude of flow.
As a stenosis increases, a larger pressure decrease
develops. This reduction in perfusion pressure (P
sad
P
L
VED) i
s
compensated for by a distal bed vaso-
dilation until the critical lower threshold of coro-
nary perfusion pressure (P
sad
P
LVE
D) *
S
reached
(60 mm Hg). At this point, vasodilation is max-
imal, further autoregulation is not possible, and
flow becomes dependent on perfusion pressure in
a linear manner. Therefore, the significance of cor-
onary artery disease becomes clinically apparent
when the perfusion pressure of coronary vascula-
ture falls below this threshold.
The area of myocardium that is most vulnerable
to the effects of coronary artery disease and re-
duced perfusion pressure is subendocardium (inner
one third or one fourth of the myocardium). The
subendocardium is most susceptible because it has
the highest metabolic rate (resulting from an in-
creased systolic and diastolic wall tension) com-
pared with epicardium; consequently, the ratio of
endocardial to epicardial blood flow is approxi-
mately 1.25:1.0 in conscious dogs.
38
From the foregoing discussion of myocardial
oxygen supply and demand, one can see that an
increase in the diastolic phase (decreased heart
rate, decreased P
LVE
D)
m a v
have a dramatic impact
on subendocardial perfusion. On the demand side,
the importance of controlling heart rate is again
evident as is a reduction in wall tension through
decreasing preload and afterload. The medicophar-
macologic armamentarium to achieve these end
points is made up of three classes of agents (ex-
cluding antiplatelet agents): (1) nitrates, (2) beta-
adrenergic receptor blockers, and (3) calcium
channel blockers.
Nitrates relax the vascular smooth muscle of the
venous and coronary arterial circulation. The vaso-
dilatory effect of nitrates on coronary arteries
yields increased supply of oxygen to the myocar-
dium. Through venous and arterial vasodilation,
nitrates decrease preload and afterload (diastolic
and systolic wall tension, respectively) and there-
fore myocardial oxygen consumption. The in-
creased supply of and decreased demand for oxy-
gen caused by nitrates favorably affect the
subendocardium; the net result is a preferential
redistribution of blood to the subendocardium as a
result of treatment with nitrates.
39
By controlling the heart rate and contractility,
beta-adrenoreceptor blocking agents decrease the
myocardial oxygen requirement at times of in-
creased sympathetic output. Slower heart rate not
only reduces myocardial oxygen consumption but
also leads to a prolonged diastolic phase, which
has a favorable effect on subendocardial blood
flow.
The three commonly used calcium channel-
blocking agents in the United States are nifedipine,
diltiazem, and verapamil. They all inhibit calcium
influx into smooth-muscle cells and cardiac cells,
causing vasodilation and negative inotropy,
thereby reducing myocardial oxygen consumption
as well as enhancing myocardial oxygen delivery.
In summary, common precipitating factors for
myocardial ischemia are tachycardia (increases ox-
ygen demand and decreases oxygen supply), in-
creased preload (increases oxygen demand and de-
creases oxygen supply), hypertension (increases
systolic wall tension more than increasing oxygen
supply from the increase in perfusion pressure),
and hypotension (decreases oxygen supply more
than oxygen consumption). It follows that anes-
thetic management should minimize myocardial
oxygen demand by continuing preoperative beta
blockers until the time of surgery, minimize prein-
duction anxiety, keep heart rate low, maintain ad-
equate levels of anesthesia, and use myocardial
depressant drugs when indicated. Anesthetic man-
agement should also maximize myocardial oxygen
supply by keeping diastolic blood pressure normal
or increased, keeping heart rate low, ensuring ad-
equate arterial oxygenation, and using venous and
arterial vasodilators to reduce preload and after-
load, respectively, and to relieve coronary artery
spasm.
a. EFFECT OF ANESTHETIC DRUGS ON
CARDIOVASCULAR FUNCTION (Table 8-2)
This discussion is not intended as a review of
the general pharmacology of anesthetic drugs;
rather, it discusses those features of anesthetic
drugs that suggest their use or avoidance in pa-
tients with coronary artery disease. Depending on
the drug and the amount used, anesthetic technique
may alter all of the determinants of myocardial
oxygen supply and demand.
The narcotics, especially fentanyl, have minimal
primary hemodynamic effects if adequate doses
are used (or are supplemented by other drugs).
High doses of fentanyl (greater than 60 g/kg)
provide remarkable hemodynamic stability for cor-
onary artery by-pass surgery and, in this group of
patients, low to moderate doses of fentanyl (15
g/kg) will also provide stable anesthesia for in-
duction and intubation, provided the fentanyl is
administered rapidly (within 12 sec).
40
The de-
crease in heart rate with fentanyl is a desirable
characteristic in patients with coronary artery dis-
ease. Meperidine has a negative inotropic effect
and positive chronotropic effect.
Propofol appears to affect the cardiovascular
system in a biphasic manner.
41
First, there is a
marked reduction in systemic vascular resistance
with an associated tachycardia, an increase in car-
diac output, and a decrease in arterial pressure.
Second, as the systemic vascular resistance in-
creases toward normal, there is a decrease in heart
rate and cardiac output, and a further slight de-
crease in arterial pressure.
At 1 to 2 MAC, the halogenated drugs have a
number of unfavorable cardiovascular effects. A
major cardiovascular disadvantage of the halogen-
ated drugs is a 20 to 40 per cent decrease in sys-
temic blood pressure. Since cardiac output is de-
creased 20 to 40 per cent by halothane and
enfiurane, and not at all by isoflurane, and in the
context of a 20 to 40 per cent decrease in systemic
blood pressure caused by all drugs, systemic vas-
cular resistance is only minimally affected by hal-
othane and enfiurane but is decreased by isoflur-
ane. Since filling pressures are increased by
halothane and enfiurane, and not at all by isoflur-
ane, and in the context of a decrease in cardiac
output caused by halothane and enfiurane, but not
Table 8-2 EFFECT OF ANESTHETIC DRUGS ON CARDIOVASCULAR FUNCTION
*Ketamine has a direct myocardial depressant action that becomes evident when its sympathetic stimulating effects are blocked
or when the sympathetic nerves are maximally stimulated, as in severe hypotension.
Key: 0 = no change; 1 and 11 = 10 to 20 per cent and 20 to 40 per cent decrease, respectively; f , \ , ] =
progressively greater increases; = small changes that depend on circumstances and reflex activity.
by isoflurane, cardiac contractility is decreased by
halothane and enflurane and not at all by isoflur-
ane. Halothane greatly sensitizes the myocardium
to catecholamines, and arrhythmias may be prom-
inent in patients with irritable ventricular foci
(which may be due to ischemia caused by coro-
nary artery disease). Enflurane causes a 20 to 40
per cent and isoflurane a 10 to 20 per cent increase
in heart rate, which may be very and moderately
disadvantageous, respectively, to the patient with
coronary artery disease. Isoflurane may act as a
nonspecific coronary vasodilator, which can theo-
retically result in a "coronary steal"; that is, vaso-
dilation and an increase in blood flow in normal
areas can occur at the expense of blood flow to
ischemic areas that have a fixed vascular resis-
tance.
42
Isoflurane may also induce myocardial is-
chemia in patients with coronary artery disease by
causing systemic hypotension and tachycardia (in
addition to a coronary steal).
One study of anesthetized patients just before
undergoing coronary artery by-pass grafting sur-
gery determined the effect of equipotent doses of
halothane and isoflurane on coronary blood flow
and myocardial ischemia and thereby provided an
objective basis for analyzing the contention that
isoflurane predisposes to the development of is-
chemia by "stealing" blood flow.
43
Global and
regional myocardial blood flow and metabolism
were examined in 20 patients with coronary artery
disease before surgical stimulation. Half were an-
esthetized with halothane (0.8 per cent) and half
with isoflurane (1.2 per cent) (these concentrations
are equipotent). Coronary perfusion pressure de-
creased similarly in both groups. During halothane
anesthesia, coronary sinus blood flow, an index of
global perfusion, decreased from an awake value
of 129 7 to 97 7 ml/min (p < .05), and great
cardiac vein blood flow, an index of regional per-
fusion (drains the area supplied by the left anterior
descending coronary artery), decreased from 60
8 to 44 5 ml/min (p < .05).
In contrast, during isoflurane anesthesia, global
coronary blood flow increased from 131 13 to
153 16 ml/min (p < .05), whereas regional
blood flow decreased from 68 7 to 56 6 ml/
min (p < .05). Thus, the ratio of great cardiac vein
blood flow to coronary sinus blood flow was un-
changed during halothane anesthesia but decreased
significantly during isoflurane. Neither global nor
regional coronary vascular resistance was altered
by halothane, whereas isoflurane decreased global
coronary vascular resistance without affecting re-
gional coronary vascular resistance. All patients
receiving halothane had net myocardial lactate ex-
traction. In the isoflurane group, four patients
showed global lactate production and three re-
gional lactate production. All patients demonstrat-
ing lactate production also developed electrocar-
diographic evidence of myocardial ischemia,
which was not present before induction.
The authors concluded that halothane is a pref-
erable anesthetic to isoflurane in patients with cor-
onary artery disease because the latter has the pro-
pensity to induce maldistribution of the coronary
circulation (steal) and myocardial ischemia. The
maldistribution of coronary artery blood flow
(steal) caused by isoflurane is attributed to vasodi-
lation in normal coronary arteries while stenotic
ones must remain constant in size. Indeed, one
case report documented the development of ische-
mia in an isoflurane-anesthetized patient that was
successfully treated by switching to halothane.
44
The neuromuscular blocking drugs have hemo-
dynamic effects, but they are not generally of large
magnitude. Pancuronium has the most undesirable
effects for the patient with coronary artery disease
because it can stimulate the sympathetic nervous
system, causing tachycardia and increased blood
pressure and cardiac output. i/-tubocurarine also
has undesirable effects for the patient with coro-
nary artery disease because of the development of
hypotension and tachycardia. Succinylcholine may
mimic acetylcholine at nicotinic and muscurinic
receptors; consequently, a dose-related tachycardia
followed by bradycardia may be seen owing to
sequential stimulation of these receptors, respec-
tively. The other relaxants have only minimal ef-
fects on cardiovascular function.
Ketamine can cause a significant stimulation of
the sympathetic nervous system that results in an
increase in contractility, tachycardia, and hyper-
tension, whereas thiopental can cause significant
depression of myocardial contractility and hypo-
tension. Consequently, both of these intravenous
anesthesia-inducing drugs have major disadvan-
tages in patients with coronary artery disease.
However, the sympathomimetic effect of ketamine
may be used to good advantage in hypovolemic
patients.
III. CHOICE OF ANESTHESIA AND
ARTERIAL OXYGENATION DURING
ONE-LUNG VENTILATION
A. Effect of Anesthetics on Hypoxic
Pulmonary Vasoconstriction
As discussed in chapters 3 and 4, an undesirable
property of general anesthesia is inhibition of HPV
in the nonventilated, nondependent lung by the
anesthetic drug. All of the inhalation anesthetics
and many of the injectable anesthetics have been
studied with regard to their effect on HPV. Halo-
thane has been the most extensively studied agent
Table 8-3 EFFECT OF HALOTHANE ON HYPOXIC PULMONARY VASOCONSTRICTION (HPV) IN
VARIOUS EXPERIMENTAL PREPARATIONS
Key: j = decrease; f = increase; J /Dr-to % = HPV was progressively decreased to the maximum shown in column 5 over
the concentration range shown in column 4.
(Table 8-3).
45-
*
7
The experimental preparation
used may be divided into four basic categories: (1)
in vitro, (2) in vivo-not intact (pumped perfused
lungs, no systemic circulation or neural function),
(3) in vivo-intact (normally perfused lungs, nor-
mal systemic circulation), and (4) humans (volun-
teers or patients). It appears, according to this
breakdown of experimental preparation, that inhi-
bition of HPV by halothane is a universal finding
in the in vitro and in vivo-not intact preparations.
The site of inhibition is precisely where one would
expect it to be; namely, at the arteriolar level
(middle segment of an inflow/outflow occlusion
model), which is the same site of action of hy-
poxia.
55
56
68
However, in the more normal or
physiologic in vivo-intact and human studies, hal-
othane has caused no or only a very slight decrease
in HPV response. Thus, it appears that a funda-
mental property of halothane is its inhibition of
HPV in experimental preparations, which can be
controlled for other physiologic influences (e.g.,
pulmonary vascular pressure, cardiac output,
mixed venous oxygen tension, C0
2
level, and tem-
perature) that can have an effect on the HPV re-
sponse.
In the more biologically complex in vivo mod-
els, other factors seem to be involved that greatly
diminish the inhibitory effect of halothane on
HPV. Important mthodologie differences between
the in vitro and in vivo-not intact preparations and
the in vivo-intact and human models that could
account for the observed differences in halothane
effect on HPV are presence (or absence) of perfu-
sion pulsations, perfusion fluid composition, size
of perfusion circuit
61
; baroreceptor influences, ab-
sence of bronchial blood flow (which abolishes all
central and autonomic nervous activity in the
lung)
69
, chemical influences (i.e., pH, Po,). hu-
moral influences (i.e., histamine and prostaglandin
release from body tissues), lymph flow influences,
and, very importantly, unaccounted for or uncon-
trolled changes in physiologic variables, such as
cardiac output, mixed venous oxygen tension, and
pulmonary vascular pressures, which might have
directionally opposite effects on HPV, and the use
of different species.
70-72
Ether has been the next most studied drug, and
it appears that the quantitative effect of ether on
HPV is also dependent on the type of experimental
preparation used. Thus, the in vitro and in vivo-
not intact models show much more inhibition of
HPV by anesthetic drug (ether) than the in vivo-
intact and human models (Table 8-4).
46
47
5()
-
5I
54

64
"
74
Although the number of studies involving
Table 8-4
EFFECT OF ETHER, ISOFLURANE, ENFLURANE, METHOXYFLURANE, FLUROXENE, TRICHLOROETHYLENE,
NITROUS OXIDE, AND INJECTABLE ANESTHETICS ON HYPOXIC PULMONARY VASOCONSTRICTION (HPV)
Anesthetic Drug Experimental Preparation Species
Ether
Isoflurane
Enflurane
In vitro: Heart-lung
Heart-lung
Heart-lung
Lung
Lung
In vivo: Not intact,
pump perfused
In vivo: Intact,
normally perfused
Human
In vivo: Not intact, pump
perfused
In vivo: Intact, normally
perfused
Human
Heart-lung
Heart-lung
Human
Rat
Rat
Rat
Cat
Cat
Cat
Dog
Human
Rat
Dog
Dog
Dog
Dog
Dog
Dog
Dog
Rabbit
Human
Human
Human
Rat
Rat
Rat
Human
Regional
(R) vs.
Whole
(W)Lung
Hypoxia
w
w
w
w
w
w
w
R
R
R
R
R
R
W
R
R
R
R
W
W
W
R
Dose
(Converted
to MAC)
0.5-1.0
1-2
4-6
0.5-5.0
1
2.5-5.0
1.5-3.0
1-2
0.2
0-2
1-3
1-2
1-2
0-2.4
1.3
1.0
1.2
1.0-1.5
1.0
1.0
0-2
1-3
1-3
1-2
Effect on
HPV/Magnitude of
Change
1/Dr-to 100%
1/Dr-to 100%
160 + 70%
I 90-95%
185%
1 /Dr-to 95%
1 /55%
1 /33%
1 /Dr-to 90%
0
1 /Dr-to 60%
1 /Dr-to 50%
0
1 /Dr-to 50%
0
0
0
0
0
0
1 /Dr-to 90%
1 /60%
1/Dr-to 100%
0
Study
Bredesen et al.
41
Bjertnaes
47
Bjertnaes et al.
5
Sykes et al.
51
Hurtig et al."
Loh et al.
54
Sykes et al.
74
Pavlin et al.
64
Marshall et al.
49
Gardaz et al.
76
Benumof & Wahrenbrock
62
Mathers et al.
63
Saidman & Trousdale
75
Domino et al.
77
Chuda et al.
78
Naeije et al.
79
Groh et al.
8
"
Jolin-Carlsson et al.
81
Benumof et al.
67
Carlsson et al.
82
Marshall et al.
49
Bjertnaes et al.
M
Bjertnaes & Mundal
8
'
Carlsson et al.
84
Key: \ = decrease; \ = increase; Dr-to % = HPV was progressively decreased to the maximum shown in column 6 over the concentration range shown in
column 5.
*Drugs used in these experiments were fentanyl, propofol, meperidine, morphine, thiopental, pentobarbital, hexobarbital, droperidol, diazepam, chlorpromazine,
ketamine, pentazocine, lidocaine, buprenorphine. For doses and blood levels, see the references in the far right column.
halogenated drugs other than halothanenamely,
isoflurane,
49
62
^
67
75
"
82
enflurane,
49
5a 8 3
84
me-
thoxyflurane,
45
^
7
85
86
fluroxene,
62
63
and trichloro-
ethylene
51
have been too small to permit recog-
nition of a clear experimental preparation result
pattern, most of these anesthetics have demon-
strated inhibition of HPV in the in vitro models
(see Table 8-4). However, both isoflurane and en-
flurane show no inhibition of HPV in the only
studies performed in humans
81
84
; because of the
importance of these studies, they are discussed at
some length presently. Nitrous oxide seems to
cause a small, somewhat consistent inhibition of
HPV (see Table 8-4).
47
58
62
63
87
-
89
All injectable
anesthetics studied to date have no effect on HPV
(see Table 8-4).
46
47
59
62
90
-
96
The influence of isoflurane on HPV was studied
in eight subjects before undergoing elective sur-
gery.
82
The lungs were ventilated separately with a
double-lumen endobronchial catheter. After oxy-
gen ventilation of both lungs for 30 min during
intravenous barbiturate anesthesia, the test lung
was rendered hypoxic by ventilation with 8 per
cent 0
2
in nitrogen. The control lung was venti-
lated continuously with 100 per cent 0
2
. Isoflurane
was added to the inspired gas, so that end-tidal
concentrations of 1 per cent and 1.5 per cent were
obtained. Cardiac output (Q
t
) was determined by
thermodilution, and the distribution of blood flow
between the lungs was assessed from the excretion
of a continuously infused, poorly soluble gas
(SF
6
).
The hypoxic challenge during intravenous an-
esthesia resulted in a reduction in the fractional
perfusion of the test lung from 54 per cent to 41
per cent of Q, (Fig. 8-2). Mean pulmonary arterial
pressure increased by 46 per cent, and pulmonary
vascular resistance (PVR) of the test lung more
than doubled. Arterial oxygen tension fell from
375 mm Hg (50 kPa) to 101 mm Hg (13.5 kPa).
Adding isoflurane to the inhalation gas, first at a
concentration of 1 per cent and then 1.5 per cent,
caused no further significant change in the distri-
bution of pulmonary blood flow, although six of
the eight subjects showed a small increase in test
lung blood flow at isoflurane 1.5 per cent (see Fig.
8-2). There was no change in PVR or in any other
circulatory variable. Arterial blood gases remained
essentially unaltered. When the hypoxic challenge
was discontinued, all variables returned to control
values. It is possible that higher isoflurane concen-
trations would have caused a clear change in the
blood flow distribution, but, at clinical concentra-
tions, the effect of HPV in the human is all but
unmeasurable.
Similarly, the degree of HPV was also studied
in eight subjects during enflurane anesthesia and
was compared with that during intravenous pento-
barbital anesthesia in the same subjects.
84
The
lungs were ventilated separately with the aid of a
double-lumen endobronchial catheter. After preox-
ygenation of both lungs for 30 min, during intra-
venous anesthesia, the right lung (test lung) was
rendered hypoxic by ventilation with 6 per cent 0
2
in nitrogen. The left lung (control lung) was ven-
tilated continuously with 100 per cent 0
2
. Cardiac
output and the distribution of blood flow between
the two lungs was assessed as previously described
Figure 8-2 The perfusion of the test lung (Q^)
and the control lung (Q
CL
) in the control period and
during hypoxia without and with isoflurane. Note
the redistribution of pulmonary blood flow during
hypoxia, indicating functioning hypoxic vasocon-
striction before and during isoflurane anesthesia.
(Reproduced with permission from Carlsson AJ,
Bindslev L, Hedenstierna G: Hypoxia-induced pul-
monary vasoconstriction in the human lung. The
effect of isoflurane anesthesia. Anesthesiology
66:312-316, 1987.)
in the isoflurane study. The hypoxic challenge re-
sulted in a reduction of the fractional perfusion to
the test lung from 57 per cent to 36 per cent of Q
t
.
Mean pulmonary arterial pressure increased by 37
per cent, and PVR of the test lung doubled. Arte-
rial oxygen tension decreased from 45.9 to 9.5
kPa. Inhalation of enflurane to an end-tidal con-
centration of 2 per cent to both lungs caused no
significant change in the distribution of the pul-
monary blood flow, PVR, or any other circulatory
variables (Fig. 8-3). The arterial blood gases re-
mained unaltered.
When the hypoxic challenge was discontinued,
all variables returned toward control values. The
findings suggest that the inhalation anesthetic en-
flurane does not reduce the hypoxic vasoconstric-
tor response in the human lung.
To summarize previous animal studies, it ap-
pears that a fundamental property of inhalational
anesthetics is to decrease HPV. However, in intact
animal preparations, some biologic or physiologic
property seems to remove or greatly lessen the
inhibitory effect of anesthetic drugs on HPV. It
may be that the cause(s) of the difference in effect
of anesthetic drugs on regional HPV from prepa-
ration to preparation, anesthetic to anesthetic, and
species to species
97
(see the following) is closely
related to the mechanism of HPV, which is still
unknown.
B. Effect of Anesthetics on Arterial
Ventilation
An often-made extrapolation of the much more
numerous in vitro and in vivo-not intact HPV
studies is that anesthetic drugs might impair arte-
rial oxygenation during one-lung anesthesia by in-
hibiting HPV in the nonventilated lung. One of the
previously mentioned studies on the effect of iso-
flurane on regional canine HPV was especially
well controlled and showed that, when all non-
anesthetic drug variables that might change re-
gional HPV are kept constant, isoflurane inhibits
single-lung HPV in a dose-dependent manner.
77
Additionally, the study is valuable because the au-
thors offer the reader an easily comprehensible
quantitative summary of the relationship between
dose of isoflurane administered and degree of in-
hibition of the single-lung canine HPV response.
Figure 8-3 The fractional perfusion of the test (right) lung and control (left) lung during the control situations and during
hypoxia without and with enflurane (Enfl.) A functioning hypoxic vasoconstriction was evident from the diversion of pulmonary
blood flow from the test lung to the control lung during hypoxia and enflurane inhalation. (Reproduced with permission from
Carlsson AJ, Hedenstierna G, Bindslev L: Hypoxia-induced vasoconstriction in human lung exposed to enflurane anaesthesia. Acta
Anaesthesiol Scand 31:57-62, 1987.)
If the summary can be extrapolated or applied
to the clinical one-lung ventilation situation (at
least as an approximation), insights can be gained
into what might be expected with regard to arterial
oxygenation when such patients are anesthetized
with isoflurane. In order to put these insights into
sharp clinical focus, it is necessary to first under-
stand what should happen to blood flow, shunt
flow, and arterial oxygenation, as a function of a
normal amount of HPV, when two-lung ventilation
is changed to one-lung ventilation in the lateral
decubitus position. Once the stable one-lung ven-
tilation condition has been described, it is then
possible, using the data from the previously men-
tioned study, to see how isoflurane administration
affects the one-lung ventilation blood flow distri-
bution, shunt flow, and arterial oxygen tension.
1. Two-Lung Ventilation: Blood Flow
Distribution
Gravity causes a vertical gradient in the distri-
bution of pulmonary blood flow in the lateral de-
cubitus position for the same reason that it does in
the upright position. Consequently, blood flow to
the dependent lung is significantly greater than
blood flow to the nondependent lung. When the
right lung is nondependent, it should receive ap-
proximately 45 per cent of total blood flow as
opposed to the 55 per cent that it received in the
upright and supine positions. When the left lung is
nondependent, it should receive approximately 35
per cent of total blood flow as opposed to the 45
per cent that it received in the upright and supine
positions (closed-chest data with normal pulmo-
nary artery pressure) (Fig. 8^l).
98
99
If these blood
flow distributions are combined (both the right and
left lungs being nondependent an equal number of
times), average two-lung ventilation blood flow
distribution in the lateral decubitus position would
consist of 40 per cent of total blood flow perfusing
the nondependent lung and 60 per cent of total
blood flow perfusing the dependent lung (Fig. 8-
4, right-hand panel and Figs. 8-3 and 8-4, left-
hand panels).
It is possible that nondependent-lung blood flow
may increase slightly when the nondependent
hemithorax is open for two reasons.
100
First, if the
compliance of the nondependent lung increases so
much that nondependent-lung alveolar pressure
decreases significantly, nondependent-lung blood
flow may increase relative to dependent-lung
blood flow. Second, if the nondependent lung falls
away from the open chest wall, the vertical dis-
tance between the heart and the nondependent lung
may decrease, which in the face of a constant
pulmonary artery pressure might result in an in-
creased perfusion of the nondependent lung. Con-
sequently, the 40/60 per cent nondependent-/de-
pendent-lung blood flow ratio during closed-chest
two-lung ventilation may be a slight underestima-
tion of the ratio during open-chest two-lung venti-
lation.
2. One-Lung Ventilation: Blood Flow
Distribution, Shunt Flow, and Arterial
Oxygen Tension
When the nondependent lung is nonventilated
(made atelectatic), HPV in the nondependent lung
Blood Flow Distribution: Two Lung Ventilation
Left Lung
Nondependent
Right Lung
Nondependent
Average of Both Lungs
Being Nondependent
Figure 84 This schematic diagram shows that when the left lung is the nondependent lung, the distribution of blood flow
between the nondependent and dependent lungs is 35/65 per cent. When the right lung is the nondependent lung, the blood flow
distribution between the nondependent and dependent lungs is 45/55 per cent. When the left and right lungs are nondependent an
equal number of times, then the average one-lung ventilation blood flow distribution would consist of a nondependent and dependent
lung blood flow ratio of 40/60 per cent.
will increase nondependent-lung PVR and de-
crease nondependent lung blood flow. In the ab-
sence of any confounding or inhibiting factors to
the HPV response, a single-lung HPV response
should decrease the blood flow to that lung by 50
per cent (Figs. 8-5, 8-6, and 8-7).
101
Conse-
quently, the nondependent lung should be able to
reduce its blood flow from 40 to 20 per cent of
total blood flow and the nondependent-/dependent-
lung blood flow ratio during one-lung ventilation
should be 20/80 per cent (see Fig. 8-7, middle
panel).
All the blood flow to the nonventilated nonde-
pendent lung is shunt flow, and therefore one-lung
ventilation creates an obligatory right-to-left trans-
pulmonary shunt flow that was not present during
two-lung ventilation. If no shunt existed during
two-lung ventilation conditions (ignoring the nor-
mal 1 to 3 per cent shunt flow due to the bronchial,
pleural, and thebesian circulations), we would ex-
pect the ideal total shunt flow during one-lung
ventilation to be a minimal 20 per cent of total
blood flow. With a normal hemodynamic and met-
abolic state, the arterial oxygen tension should be
approximately 280 mm Hg (Fig. 8-8)."
,2
Table 4-1 shows a quantitative example of a
model of blood flow to each lung during two-lung
and one-lung ventilation, with an increasing initial
two-lung ventilation shunt through both lungs. As
the initial nondependent-lung shunt flow increases
(i.e., during two-lung ventilation), the amount of
nondependent-lung blood flow able to participate
in nondependent-lung HPV and the amount of
nondependent-lung HPV blood flow diversion de-
crease, and the one-lung ventilation shunt in-
creases. In addition, as the fraction of the Q, that
normally perfuses the operative lung increases
(e.g., in a patient who has little or no lung disease,
i.e., undergoing esophageal surgery), the shunt
flow through the operative lung during one-lung
ventilation increases and the P
a
O
:
decreases.
103
As
the initial dependent-lung shunt flow increases
(i.e., during two-lung ventilation), the amount of
one-lung ventilation shunt also increases irrespec-
tive of nondependent-lung HPV.
3. Effect of Isoflurane on the One-Lung
Ventilation Blood Flow Distribution,
Shunt Flow, and Arterial Oxygen
Tension
Domino et al.
77
found that per cent inhibition of
regional HPV response equals 22.8 (per cent al-
veolar isoflurane) minus 5.3. As previously de-
Amount of Lung That is Hypoxic, %
Figure 8-5 The x-axis shows the amount of lung that is hypoxic. The left-hand y-axis shows the expected amount of blood flow
reduction to the hypoxic lung as a result of hypoxic pulmonary vasoconstriction (HPV). The right-hand y-axis shows the amount of
perfusion pressure increase expected because of HPV. When 40 per cent of the lung is hypoxic, the blood flow reduction to the
hypoxic lung should be very near 50 per cent. (Redrawn with permission from Marshall BE, Marshall C: Continuity of response to
hypoxic pulmonary vasoconstriction. J Appl Physiol 49:189-196, 1980.)
316
Choice of Anesthetic Drugs and Techniques
Conversion of Two-Lung to One-Lung Ventilation:
Blood Flow Distributions
Figure 86 This schematic diagram shows that the two-lung ventilation nondependent/dependent lung blood flow ratio is 40/60
per cent (left-hand side). When two-lung ventilation is converted to one-lung ventilation (l LV) (as indicated by atelectasis of the
nondependent lung), the hypoxic pulmonary vasoconstriction (HPV) response decreases the blood flow to the nondependent lung
by 50 per cent so that the nondependent/dependent lung blood flow ratio is now 20/80 per cent (right-hand side).
Effect of 1 MAC Isoflurane Anesthesia on Shunt
During One Lung Ventilation (1LV) of Normal Lungs
Figure 8-7 This schematic diagram shows that the two-lung ventilation nondependent/dependent lung blood flow ratio is 40/60
per cent (left-hand side). When two-lung ventilation is converted to one-lung ventilation (1 LV) (as indicated by atelectasis of the
nondependent lung), the hypoxic pulmonary ventilation (HPV) response decreases the blood flow to the nondependent lung by 50
per cent, so that the nondependent/dependent lung blood flow ratio is now 20/80 per cent (middle). According to the data of
Domino et al," administration of 1 minimum alveolar concentration (MAC) isoflurane anesthesia should cause a 21 per cent
decrease in the HPV response, which would decrease the 50 per cent blood flow reduction to a 40 per cent blood flow reduction
HPV response. Consequently, the nondependent/dependent lung blood flow ratio would now become 24/76 per cent, representing a
4 per cent increase in the total shunt across the lungs (right-hand side). (Reproduced with permission from Benumof JL: Isoflurane
anesthesia and arterial oxygenation during one-lung ventilation. Anesthesiology 64:419-422, 1986.)
Figure 8-8 The x-axis shows the inspired oxygen concentration. The left-hand y-axis presents the expected arterial Po, for a
family of intrapulmonary shunts. The tick intervals on the right-hand y-axis are the same as those for the left-hand y-axis. As shunt
increases, the family of isoshunt lines becomes flatter and closer together, and for a given F,0
2
an increase in shunt has a decreasing
effect on decreasing arterial Po
:
. The model assumes relatively normal hemoglobin, P
a
C0
2
, and a-v 0
2
content difference (upper
left-hand corner). (Redrawn with permission from Lawler PGP. Nunn JF: A reassessment of the validity of the isoshunt graph. Br I
Anaesth 56:1325-1335, 1984.)
318
Choice of Anesthetic Drugs and Techniques
scribed, under normal conditions, collapse of the
nondependent lung in the lateral decubitus position
causes a nondependent-lung HPV response to de-
crease nondependent-lung blood flow by 50 per
cent; that is, from 40 to 20 per cent of total flow
(see Figs. 8-5, 8-6, and 8-7). Using these values
as a model of the normal two-lung to one-lung
ventilation conversion process, we can construct a
table (Table 8-5) that sequentially relates per cent
alveolar isoflurane to per cent inhibition of the
nondependent-lung HPV response to the resultant
nondependent-lung HPV response (expressed as a
per cent decrease in nondependent-lung blood
flow), to the resultant increase in atelectatic non-
dependent-lung blood flow (which is the shunt
during one-lung ventilation), to an absolute in-
crease in shunt, to a decrease in arterial oxygena-
tion during one-lung ventilation (from 280 mm Hg
[F,0
2
= 1.0] to some lower value).
Table 8-5 and the right-hand panel of Figure 8-
7 show that 1 MAC isoflurane anesthesia would
inhibit the nondependent-lung HPV response by
approximately 21 per cent, which would decrease
the nondependent-lung HPV response from a 50
to 40 per cent nondependent-lung blood flow re-
duction, which would increase nondependent-lung
blood flow from 20 to 24 per cent of total blood
flow, causing shunt to increase by 4 per cent of
the cardiac output and P
a
0
2
to decrease a moderate
amount to 205 mm Hg (F,0
2
= 1.0) (see Fig. 8-
8). Table 8-5 shows that one-half MAC isoflurane
anesthesia would cause a very small increase in
the total one-lung ventilation shunt and a small
decrease in P
a
0
2
, whereas 2 MAC isoflurane an-
esthesia would cause a moderate increase in the
total one-lung ventilation shunt and a large de-
crease in P
a
0
2
. Since isoflurane causes undesirable
hemodynamic effects at high doses (greater than 1
MAC), and moderate doses of fentanyl (20 g/kg)
have a relative absence of hemodynamic effect,
isoflurane anesthesia is usually administered in a 1
MAC or less concentration and is often supple-
mented with moderate doses of narcotics (or vice
versa) (see Recommended Anesthesia Induction
and Maintenance Drugs and Techniques).
A number of important nonanesthetic drug fac-
tors might make the administration of isoflurane
anesthesia have less of an effect on shunting and
arterial oxygenation during one-lung ventilation
than the preceding analysis suggests. First, and
most important, the absolute level of shunt is al-
most always higher in surgical patients than the
minimal 20 per cent used in the preceding analysis
of one-lung ventilation (see Table 4-1). The effect
of a given increase in shunt on P
a
0
2
depends on
the absolute level of the initial shunt and the in-
spired oxygen concentration (see Fig. 8-5). '
2
With an F,0
2
of 1.0, an increase in shunt from 20
to 24 per cent of the cardiac output decreases the
P
a
0
2
a moderate amount. However, if the two-lung
and one-lung ventilation shunt is increased, per-
haps owing to pre-existing or anesthesia-induced
lung disease, the same isoflurane-induced increase
in shunt will cause much less of a decrease in P
a
O
;
(the larger isoshunt lines of Fig. 8-8 are much
flatter and closer together). For example, if the
one-lung ventilation shunt is 30 per cent without
isoflurane and 34 per cent with isoflurane, the de-
crease in P
a
0
2
will be very small and perhaps not
detectable, given the usual accuracy of clinical
methodology.
In fact, in clinical one-lung ventilation studies
involving intravenously anesthetized patients with
this level of shunting, administration of 1 MAC
isoflurane (and halothane) anesthesia during stable
one-lung ventilation conditions causes no detect-
able decrease in P
a
0
2
(Figs. 8-9 and 8-10).
65
67
In
one of these clinical studies,
65
stable one-lung ven-
tilation conditions in the lateral decubitus position
were established in patients who were anesthetized
with only intravenous drugs (see Fig. 8-9). While
stable one-lung ventilation was maintained, inha-
Abbreviations: HPV = hypoxic pulmonary vasoconstriction; MAC = minimum alveolar concentration.
HALOTHANE GROUP
ISOFLURANE GROUP
Figure 8-9 Changes in P
a
O
:
throughout the entire experimental sequence (experimental steps 1-5). Patients are divided according
to the inhalational anesthetic drug they received (halothane or isoflurane). (2-LV = two-lung ventilation; 1-LV = one-lung
ventilation; IV = intravenous anesthesia; IH = inhalational anesthesia. Open circles = individual patient data; closed circles =
mean standard deviations for each group.) (From Rogers SN, Benumof JL: Halothane and isoflurane do not decrease P
a
O
:
during
one-lung ventilation in intravenously anesthetized patients. Anesth Analg 64:946-954, 1985. Used with permission.)
lational anesthetics were administered (halothane
and isoflurane end-tidal concentrations were
greater than 1 MAC for at least 15 min) and then
discontinued (halothane and isoflurane end-tidal
concentrations decreased to near zero) (see Fig.
8-9).
In the other clinical study,
67
steady-state one-
lung ventilation conditions in the lateral decubitus
position were established in patients who were an-
esthetized with only inhalational drugs (halothane
and isoflurane end-tidal concentrations were con-
stantly greater than 1 MAC for more than 40 min)
(see Fig. 8-10). While one-lung ventilation was
continued, an intravenous anesthesia was adminis-
tered (halothane and isoflurane end-tidal concen-
trations decreased to near zero) (see Fig. 8-10).
There was no significant difference in P
a
O
:
during
inhalation anesthesia with either halothane or iso-
flurane compared with intravenous anesthesia dur-
ing one-lung ventilation in either of the two exper-
imental sequences (see Figs. 8-9 and 8-10). In
addition, there were no significant changes in
physiologic variables, such as cardiac output, pul-
monary vascular pressure, and mixed venous oxy-
gen tension, that might secondarily alter nonde-
pendent-lung HPV. Thus, irrespective of whether
inhalational anesthesia is administered before (see
Fig. 8-10) or after (see Fig. 8-9) intravenous an-
esthesia during one-lung ventilation, inhalation an-
esthesia does not further impair arterial oxygena-
tion. These findings are consistent with the
interpretation that 1 MAC halothane and isoflurane
do not inhibit HPV in patients with a moderate
level of shunting enough to cause a significant
decrease in P
a
0
2
during one-lung ventilation in the
lateral decubitus position.
Further considering the implications of Figure
8-8, some anesthesiologists use an F,0
2
less than
1.0 during one-lung ventilation (which I do not
recommend). As can be seen from Figure 8-8, the
family of isoshunt lines is much closer together at
F,0
2
= 0.5 than at F,0
2
= 1.0, and the decrease
in P
a
0
2
with a given increase in shunt is much less
(but the absolute level of P
a
0
2
is uncomfortably
low).
Second, as pointed out by Domino et al.,
77
the
Figure 8-10 P
a
0
2
and Q
s
/Q
(
values during the four experi-
mental sequence steps. Conversion from two-lung ventilation
(2-LV) to one-lung ventilation (1-LV) during inhalation anes-
thesia (IH) in both groups (step 1 to step 2) caused a very large
and significant decrease in P
a
O, and increase in Q
s
/Q,. Conver-
sion from IH to intravenous anesthesia (IV) during 1-LV (step
2 to step 3) caused a slight but significant increase in P^CX and
decrease in Q
s
/Q, in the halothane group and a very slight and
nonsignificant increase in P
a
0
2
and decrease in Q
s
/Q, in the
isoflurane group. Conversion from 1-LV to 2-LV during IV
anesthesia (step 3 to step 4) caused a very large and significant
increase in P
a
0
2
and decrease Q
s
/Q, in both the halothane and
isoflurane groups. The small numbers to the left of the first
data point refer to the patient number. (From Benumof JL,
Augustine SD, Gibbons JA: Halothane and isoflurane only
slightly impair arterial oxygenation during one lung ventilation
in patients undergoing thoracotomy. Anesthesiology 67:910
secondary effects of anesthesia with isoflurane
may counteract the direct HPV-inhibitory effect of
the drug. Thus, a decrease in cardiac output, mixed
venous oxygen tension, and pulmonary artery
pressure, all of which may accompany isoflurane
anesthesia, would intensify nondependent-lung
HPV at the same time isoflurane was decreasing
it.
Third, the presence of chronic irreversible dis-
ease in the vessels of the nondependent lung may
render these vessels incapable of an HPV
response.
104
'
5
Fourth, the presence of disease in the dependent
lung (either pre-existing or anesthesia-induced),
which increases dependent-lung vascular resis-
tance, will make the dependent lung less able to
accept redistributed blood flow and thereby de-
crease the nondependent-lung HPV response.
101,
106-108 Y^g
s m a
u
e r me
j-jVP response, the less of
an effect isoflurane anesthesia can have on the
HPV response.
Fifth, surgical interference with blood flow to
the nondependent lung will also decrease the effect
that isoflurane anesthesia can have on the one-lung
ventilation shunt.
Sixth, species differences
70-72
97
and differences
in the study and clinical one-lung ventilation meth-
odology (nitrogen ventilation vs. atelectasis, ad-
ministration of isoflurane to the hypoxic lung vs.
the normoxic or hyperoxic lung, and large vs.
small alveolar-to-mixed venous isoflurane tension
gradients, respectively) may alter the precise rela-
tionship between per cent inhibition of single-lung
HPV and the alveolar concentration of isoflurane.
In summary, as demonstrated by Domino et al.,
isoflurane anesthesia has a direct inhibiting effect
on regional HPV in dogs.
77
In the simple case, in
which physiologic variables (cardiac output,
mixed venous oxygen tension, pulmonary vascular
pressures, and carbon dioxide tension) are normal
and the amount of lung disease is minimal, the
effect of isoflurane on shunting during one-lung
ventilation is reasonably predictable and moder-
ately small. In the complex case, in which physio-
logic variables are abnormal and/or the amount of
lung disease is extensive, the effect of isoflurane
on shunting during one-lung ventilation is much
less predictable but almost certainly still small.
Nevertheless, it should be remembered that it is
the compromised patient who will be most intol-
erant of any further anesthesia-induced inhibition
of HPV. For this kind of patient, the effect of
isoflurane anesthesia on shunting must be carefully
considered, arterial oxygenation must be closely
monitored, and therapeutic measures to decrease
shunting, such as nondependent-lung, continuous
positive airway pressure (CPAP) and return to
two-lung ventilation, should be quickly instituted,
if necessary.
IV. RECOMMENDED ANESTHESIA
INDUCTION AND MAINTENANCE
DRUGS AND TECHNIQUES
A. Summary of Advantages of
Anesthetic Drugs
1. Inhalational Anesthetics
General anesthesia with controlled ventilation is
the safest method of anesthetizing patients for the
.
vast majority of elective thoracic procedures. Al-
though a variety of general anesthesia techniques
can be used, the volatile halogenated anesthetic
drugs are good choices for several reasons. First,
the halogenated drugs have a salutary effect on
airway irritability. The mechanism of this action is
controversial, but, as previously discussed, there is
evidence that these drugs can block specific forms
of bronchoconstriction
12
l09
as well as have a non-
specific bronchodilating effect that is related to the
depth of anesthesia.
8
Obtundation of airway re-
flexes in patients who have reactive airways (i.e.,
smokers)"
0-
"
2
and who may have their airways
directly manipulated by the surgeon is a highly
desirable property of the general anesthesia pro-
duced by these drugs.
Second, the use of volatile halogenated drugs
allows for delivery of a high inspired oxygen con-
centration without loss of anesthesia. Although a
nitrous oxide-oxygen-narcotic-relaxant anesthesia
technique can be used, nitrous oxide necessitates a
significant decrease in the inspired oxygen concen-
tration and increases the chance of developing hy-
poxemia (especially if one-lung ventilation is em-
ployed)."
1
Unless very high doses of narcotics are
used, airway reflexes and reactivity may remain at
a high level.
Third, since the volatile halogenated drugs can
be rapidly eliminated, concern over postoperative
hypoventilation in extubated patients may be di-
minished. Doses of intravenous anesthetics, such
as the narcotics, ketamine, and the barbiturates,
which render the patient areflexic to surgical stim-
ulation, may cause the patient to require a period
of postoperative ventilation.
Fourth, in the usual clinical doses (near 1
MAC), the halogenated anesthetic drugs provide a
reasonable degree of cardiovascular stability. This
may be of particular importance in patients with
coronary artery disease and systemic hypertension.
Fifth, the halogenated drugs do not appear to
decrease P
a
0
2
any more than intravenous anes-
thetics during one-lung ventilation (see the
following).
65
67
2. Intravenous Anesthetics
The concept of balanced intravenous anesthesia
is not new. Specific anesthetic drugs are used to
produce specific effects that make up the anes-
thetic state, namely hypnosis, obtundation of au-
tonomic reflexes, and muscular relaxation. For one
example of total intravenous anesthesia, hypnosis
can be induced with a loading dose of propofol (2
mg/kg), the pressor response to intubation atten-
uated with a loading dose of alfentanil (10-15
g/kg or bolus 10 g/kg/min for 10 min), paraly-
sis and tracheal intubation facilitated with vecu-
ronium (0.1 mg/kg), and the anesthetic state main-
tained with titratable infusions of propofol (3-4
mg/kg/hour), alfentanil (1.0 g/kg/min), and ve-
curonium."
4
The lungs may be ventilated with ox-
ygen or an air/oxygen mixture to avoid all inhala-
tional anesthetic agents. At the termination of
surgery, all infusions are discontinued, residual
muscle relaxation is reversed, and arousal occurs
promptly, usually within 10 min."
5
Total intravenous anesthesia has several acute
intraoperative benefits, and most of these benefits
have to do with the elimination of nitrous oxide.
Intraoperative risks attendant to nitrous oxide use
include the limitation of inspired oxygen concen-
tration, exacerbation of air embolus or pneumo-
thorax, diffusion into closed body cavities such as
the inner ear, gastrointestinal tract, or emphyse-
matous bullae, elevation of PVR and pulmonary
artery pressure in predisposed patients, bone mar-
row suppression after prolonged administration,
and elevation of intracranial pressure in patients
with pre-existing intracranial hypertension.
Total intravenous anesthesia also has several
disadvantages. Whenever fixed doses of intrave-
nous agents are used, their pharmacokinetic and
pharmacodynamic effects vary according to the
patient's physiology. A given dose of an anesthetic
will affect a patient in hemorrhagic shock differ-
ently than a hemodynamically stable patient. The
hemodynamic state may vary greatly during any
surgical procedure (as in thoracic surgery), espe-
cially those in which large fluid shifts may occur.
Unlike a volatile agent, once an intravenous drug
is given, it cannot be recovered. With volatile
agents, their partial pressure in the blood can be
estimated by monitoring end-tidal concentrations;
no equivalent real-time measures of intravenous
drug blood concentrations exist. Should hidden
blood loss occur, profound circulatory depression
may occur as a result of deepening of anesthesia.
Because the anesthetic state depends on the intra-
venous infusion of drugs, the technique seems to
require a second dedicated intravenous catheter.
Finally, intraoperative awareness is a very impor-
tant problem.
New short-acting hypnotics and analgesics are
available with a wide therapeutic ratio that show
little accumulation and that are suitable for contin-
uous infusion. Propofol and alfentanil are both
suitable to be given by infusion because their phar-
macokinetic profiles predict little cumulative ef-
fects within therapeutic dosage ranges. Propofol
has a rapid onset of action; permits rapid awaken-
ing, resulting in clear-headed, alert, oriented pa-
tients; provides smooth maintenance anesthesia
that is easy to control, provides stable hemody-
namics, and does not inhibit HPV
96
; therefore, it
provides excellent P
a
0
2
during one-lung ventila-
tion."
6
An example of an infusion strategy for
general surgery was given above previously, and
slightly different strategies for thoracic surgery are
given next.
The narcotics, especially alfentanil and fentanyl,
have a number of desirable properties that could
be used to advantage for patients undergoing tho-
racic surgery. First, fentanyl and alfentanil have no
adverse hemodynamic effects and, therefore, are
useful in patients who have significant coronary
artery disease.
Second, if significant blood levels exist at the
end of surgery, the narcotics can allow an intu-
bated patient to have a smooth transition from
surgery into the postoperative period.
Third, the narcotics, if used in moderate dosage,
greatly diminish the amount of volatile halogen-
ated drug anesthesia required to achieve surgical
levels of anesthesia."
7
Fourth, high doses of narcotics or moderate
doses in conjunction with halogenated drugs allow
for the use of a high inspired oxygen concentration
without loss of anesthesia.
Fifth, the narcotics are thought not to diminish
regional HPV and, therefore, should permit opti-
mal oxygenation during one-lung ventilation.
Ketamine, in combination with nitrous oxide
and a muscle relaxant, has also been used for an-
esthesia in thoracic surgery."
8
Although we do not
ordinarily use ketamine for elective thoracic pro-
cedures, the drug is very useful for the induction
of general anesthesia in critically ill patients
undergoing emergency thoracic surgery for several
reasons. First, ketamine has sympathomimetic
properties"
9
that are highly desirable because
many emergency thoracic procedures are associ-
ated with hypovolemia (gunshot and stab wounds
of the chest, blunt trauma, and massive hemopty-
sis). However, it should be remembered that keta-
mine will depress cardiovascular function (sys-
temic blood pressure, contractility) if the degree of
hypovolemia is severe and the patients are sympa-
thetically exhausted.
Second, ketamine has a rapid onset of action
and can be used safely, along with cricoid pres-
sure, to induce anesthesia in patients with full
stomachs.
Third, ketamine may reduce bronchospasm in
asthmatic patients
21-24
26
l2
; in view of the number
of reports describing decreased airway resistance
in patients with bronchospasm with ketamine in-
fusion, it seems reasonable to extrapolate these
findings to thoracic surgery patients.
Fourth, ketamine does not impair arterial oxy-
genation during one-lung ventilation (perhaps ow-
ing to lack of effect on HPV).
95
Rapid sequence induction is used to secure the
airway in emergency surgical procedures to avoid
aspiration of gastric contents. Succinylcholine is
ordinarily the relaxant of choice because of its
rapid onset of action (1 min to full relaxation).
However, succinylcholine may produce muscle
fasciculations, myalgia, and increases in intragas-
tric, intraocular, and intracranial pressures. Succi-
nylcholine should also be avoided in patients with
burns, crush injuries, and renal failure, in patients
who have chronic lack of use of muscles, and in
patients with a history of malignant hyperthermia
or pseudocholinesterase deficiency.
Fortunately, there have been several new devel-
opments that have facilitated the speed of onset
and safety of inducing and maintaining neuromus-
cular blockade with nondepolarizing muscle relax-
ants. First, but least important because of unpre-
dictability, is use of the priming principle. For
example, a priming dose of 15 g/kg of vecuron-
ium with 100 g/kg total dose, produces excellent
intubating conditions in a mean onset time (75
10 sec) not significantly different from that of suc-
cinylcholine, 1.5 mg/kg (58 4 sec).
121
However,
it should be remembered that occasionally the
priming dose of a nondepolarizing muscle relaxant
may produce an undesirable degree of blockade
and attendant problems (inadequate respiration, as-
piration).
Second, the administration of a small dose of
alfentanil (30 g/kg) decreases the incidence of
unacceptable intubating conditions to 4 per cent
when intubation is performed 1 min after admin-
istration of alfentanil and an intubating dose of
vecuronium (priming dose, priming interval, and
intubating dose of vecuronium were 0.01 mg/kg, 4
min, and 0.1 mg/kg, respectively).
122
Preintubation
administration of lidocaine may also be useful in
this regard.
123
Third, use of triple to quadruple the usual rec-
ommended intubating dose of a nondepolarizing
relaxant can increase the speed of onset of com-
plete blockade. For example, the ED
95
(i.e., the
effective dose for 95 per cent of patients) dose of
mivacurium is 0.7 mg/kg, which yields good to
excellent intubating conditions in 3.3 min, whereas
0.15 mg/kg and 0.2 to 0.25 mg/kg yields good to
excellent intubating conditions in 2.5 and 2.0 min,
respectively.
124-126
Fourth, the new short- and intermediate-acting
nondepolarizing muscle relaxants (mivacurium,
vecuronium, and atracurium) can be easily admin-
istered by continuous infusion, which results in
easy and complete reversal. Average initial bolus
and continuous infusion doses for mivacurium, ve-
curonium, and atracurium are 0.15 mg/kg and 6 to
7 g/kg/min, 80 g/kg and 1.0 g/kg/min, and
0.5 mg/kg and 6 g/kg/min, respectively; in all
cases with all drugs, the continuous infusion rate
should be adjusted according to the findings with
a neuromuscular blockade monitor.
B. Recommended Anesthetic Drugs
and Technique
It should be obvious from the foregoing discus-
sions that there are advantages and disadvantages
to both the inhalation and intravenous anesthetic
drugs. One study made this point very well by
examining the respiratory and cardiac differences
between various anesthetic regimens (intravenous
based, inhalation based, epidural based) before and
during one-lung ventilation.
127
The patients (n =
36) were randomly allocated to one of the follow-
ing three groups: group A: propofol, 10 mg/
kg/hour, fentanyl; group B: 1 MAC enflurane, fen-
tanyl; group C: thoracic epidural anesthesia, 0.4
per cent enflurane. Before induction of anesthesia,
significant differences were not found. During
two-lung ventilation, cardiac index was signifi-
cantly decreased in group patients in comparison
with those in group C (p < .01). During one-lung
ventilation, significant differences were not found
except for an increased shunt fraction (Q
s
/Q
t
) in
group patients (groups A-B: < .05; groups
B-C: < .05; groups A-C: not significant). Be-
cause Q
s
/Q, was significantly increased and hypox-
emia occurred in group patients, regimen A or
C might be preferred in patients at high risk for
hypoxia or when the application of CPAP to the
nondependent lung is not possible (see chapter
11). Cardiac index was best maintained in group
C patients. Regimen C might be of value in pa-
tients at high risk for poor perfusion, taking into
account both the possible complications of the ep-
idural block and the reduced need for postopera-
tive analgesic agents.
The following recommended anesthetic tech-
nique takes advantage of the desirable properties
and minimizes the undesirable properties of these
drugs. Thus, the halogenated drugs are used for
their effect on bronchomotor tone, to administer
100 per cent oxygen, and to allow for early extu-
bation while not decreasing hemodynamic func-
tion and arterial oxygenation, whereas fentanyl is
used to ensure hemodynamic stability while not
jeopardizing early extubation if desired. If it is
thought that the patient will not be extubated early
or if greater hemodynamic stability is desired, an-
esthesia consisting of more fentanyl and less halo-
genated drug can be used.
1. Insertion of an Epidural Catheter
Before Induction of Anesthesia
Epidural anesthesia has been reported to exert
beneficial effects in various surgical procedures,
including thoracic operations.
128
The definite ad-
vantages (major and consistently reported) are a
great reduction in the intraoperative requirements
for general anesthetics; a great decrease, if not
complete elimination, in the postoperative require-
ment for systemic analgesics and improved post-
operative ventilatory function (if the catheter is
used postoperatively); and improved cardiac per-
formance if local anesthetic is used (causes de-
creased afterload because of sympathetic block-
ade). Other possible benefits include decreased
incidence of venous thromboembolism as well as
blood loss, better suppression of stress responses,
and a positive effect on postoperative nitrogen bal-
ance. Potential disadvantages include the time re-
quired to establish the epidural anesthesia, intra-
vascular volume needed to avoid a decrease in
blood pressure, and technical complications such
as epidural hematoma. Reported levels of insertion
and preoperative drugs and dosages of drugs suc-
cessfully used are T
I2
_
M
and 12 to 15 ml of 0.5 per
cent bupivacaine,
128
L
2
_
3
and fentanyl 4 g/kg in
20 ml saline 40 min before incision,
129
T
6
_
7
and
fentanyl 250 g,
l 30
T
4
_
5
and 50 g sufentanil in 7
ml saline or 8 ml 0.5 per cent bupivacaine.
131
2. Induction of Anesthesia (Fig. 8-11)
The patient is preoxygenated by spontaneously
breathing 100 per cent oxygen through a black
rubber anesthesia mask that is connected to an
anesthesia circle system. The preoxygenation or
denitrogenation process with tidal breathing (typi-
cally 3-5 min) may be greatly speeded up by hav-
ing the patient take several (four to eight) deep or
maximal breaths. The exact comparison/result of a
few deep breaths with several minutes of tidal
breathing with respect to preinduction P
a
0
2
or
S
a
0
2
and time to various desaturation levels will
obviously depend on how many deep breaths were
taken and the duration of the tidal breathing.
132
"
136
Fentanyl is administered intravenously until the
respiratory rate is approximately 8 to 10
breaths/min. This usually corresponds to a dose of
10 to 15 g/kg and is usually administered over 3
min. When the respiratory rate is relatively slow
and deep, and response to commands are becom-
ing sluggish, a small dose of thiopental (Pentothal;
2 to 3 mg/kg) or ketamine (1 to 2 mg/kg) (if the
patient is thought to have an especially reactive
airway or to be minimally to moderately hypovo-
lemic) is administered, which renders the patient
unconscious and usually apneic. The airway is
then established, and ventilation is controlled with
intermittent positive-pressure oxygen via the black
rubber mask. While the patient is being ventilated
with positive pressure, concentrations of 2.5 to 0.5
per cent isoflurane are administered. The higher
isoflurane concentration is used initially for a short
period of time (overpressure, 1 to 2 min), and as
the patient demonstrates signs of deepening anes-
thesia, the inspired isoflurane concentration is de-
creased. In view of the fact that general anesthetics
Anesthetic Technique for Typical
One Lung Ventilation Thoracic Surgery Cases
Airway
ACTION ENDPOINT
Drug
Laryngoscopy
Li docai ne I nt r at r acheal ^
Double Lumen Tube Intubati on
Admi ni st er
Mai ntenance Isofl urane,
Fentanyl , Paralysis
Figure 8-11 This action-end point flow diagram describes the anesthetic technique used for a typical one-lung ventilation
thoracic surgery case. The actions are divided into those that are primarily concerned with airway management versus those that
involve administration of a" drug. All drug administrations are associated with an end point. See text for full explanation. (IPPB =
intermittent positive-pressure breathing.)
significantly decrease the ventilatory response to
C0
2
(to a much greater degree in patients with
mechanical ventilatory impairment compared with
normal patients), patients are not allowed to
breathe spontaneously until the end of the proce-
dure; alarming degrees of hypercapnia have been
observed in similar circumstances when sponta-
neous ventilation was allowed.
137
Early during the period of positive-pressure
ventilation with isoflurane, paralysis is induced
with either pancuronium 0.02 mg/kg and metocu-
rine 0.08 mg/kg, or vecuronium 0.1 mg/kg, or
atracurium 0.5 mg/kg. The development of full
paralysis is monitored with a neuromuscular
blockade monitor. During the period of deepening
isoflurane anesthesia and paralysis, blood pressure
is supported with an infusion of approximately 10
ml/kg crystalloid. If more cardiovascular support
is required, the first drugs used (while fluid is
being infused) are ephedrine, 0.05 to 0.1 mg/kg,
atropine, 0.02 mg/kg, and calcium, 5 mg/kg. In
patients with stable angina undergoing operations
of short duration, the use of nitroglycerin infusion
(0.9 g/kg) and low-dose fentanyl (3 g/kg) be-
fore laryngoscopy and intubation significantly de-
creases the incidence of myocardial ischemia as-
sociated with induction of anesthesia and tracheal
intubation (compared with fentanyl, 8 g/kg
alone).
138
When the patient has been judged to be ade-
quately anesthetized (surgical stage as judged by
changes in blood pressure, heart rate, and eye signs
[the eyes should be central, conjugate, fixed, star-
ing, without tears and with nondilated pupils]) and
paralyzed in the previously described manner, li-
docaine, 1.5 mg/kg,
139
, 4 0
is administered intrave-
nously (optimal time of injection is 3 min before
tracheal intubation),
141
laryngoscopy is performed,
the tracheobronchial tree sprayed with a laryngo-
tracheobronchial spray system, and the trachea in-
tubated with a double-lumen tube (see chapter 9).
The intravenous and intratracheal lidocaine should
diminish both the airway and cardiovascular re-
sponse to endotracheal intubation.
139-142
If the pa-
tients have received a higher dose of fentanyl (25
g/kg), then intratracheal lidocaine alone is suffi-
cient and preferable.
143
Although lidocaine and fentanyl protect against
increases in blood pressure, intravenous esmolol
(5-10 mg) will also protect against increases in
heart rate in addition to increases in blood pressure
(half-life = 9 min).
144-147
This is an important con-
sideration because tachycardia may be more stren-
uous on the heart than an increase in afterload.
148
Esmolol also decreases the incidence of arrhyth-
mias attributable to tracheal intubation.
147
Conse-
quently, rapid intravenous injection of esmolol
before rapid sequence induction is a simple, con-
venient, and effective technique that should be
considered for blunting hemodynamic responses to
intubation. After intubation, the patient is venti-
lated with maintenance doses of isoflurane and
administered maintenance doses of narcotics and
relaxants. Use of maintenance paralysis decreases
isoflurane requirements, possibly allowing for a
more rapid emergence from anesthesia.
149
3. Maintenance of Anesthesia
Anesthesia is maintained with both isoflurane
(concentration approximately 0.5 to 1.0 MAC) and
narcotics. Isoflurane is primarily used if the patient
is thought to stand a reasonable chance of being
extubated within the first couple of hours postop-
eratively. Narcotics (fentanyl) are primarily used if
the patient is thought not to have a reasonable
chance of being extubated in the immediate post-
operative period and will require a significant pe-
riod of postoperative ventilation. Relaxants (pri-
marily pancuronium) are administered in small
doses to keep the level of neuromuscular blockade,
as judged by a neuromuscular blockade monitor,
near the 90 per cent paralysis level.
If the patient is thought to have a reasonable
chance of being extubated in the first postoperative
hour, the patient is turned supine, the double-lu-
men tube changed to a single-lumen tube, paraly-
sis reversed, and spontaneous ventilation allowed
to recur. Fentanyl is administered in extremely
small increments (0.3 g/kg) while the patient is
breathing spontaneously. The goal of the fentanyl
administration is to have the patient breathing
relatively slowly (approximately 10 to 12
breaths/min) and deeply when surgery is com-
pleted. The presence of a moderate narcotic base
allows the patient to be returned to the recovery
room for a short period of mechanical ventilatory
support (if needed) and weaned and extubated in a
relatively smooth manner.
4. Total Intravenous Anesthesia
Technique
Propofol's well-known pharmacokinetic prop-
erties make this drug most suitable for infusion
techniques. The same is true of alfentanil; these
two drugs together form an ideal combination for
total intravenous anesthesia inasmuch as one pro-
vides anesthesia and the other hypnosis. There
have been two studies using total intravenous an-
esthesia for thoracic surgery. In one study,"
6
an
alfentanil bolus of 1 mg plus lidocaine, 1 mg/kg,
was given intravenously. This was followed by
propofol, 2 to 2.5 mg/kg, and atracurium, 0.5
mg/kg. After intubation, two infusions, one with
propofol and one with alfentanil, were started im-
mediately thereafter by way of two separate infu-
sion pumps. Alfentanil was given at a rate of 3
g/kg/min for 10 min followed by a constant in-
fusion of 1 g/kg/min. The infusion was discon-
tinued approximately 20 min before the end of the
operation. The propofol infusion was started at 0.2
mg/kg/min (12 mg/kg/hour). Thereafter, it was
varied to adjust the depth of anesthesia. Between
0.1 and 0.15 mg/kg/min (6-9 mg/kg/hour) was
required, and the infusion was discontinued at the
time of skin closure. Extra boluses of 20 mg of
propofol were given in addition when blood pres-
sure and pulse rate increases made them necessary,
together with repeat doses of the muscle relaxant
when slight movements occurred. Not surpris-
ingly, the other important total intravenous anes-
thesia for thoracic surgery study (propofol with or
without nitrous oxide) showed that the total
amount of propofol infused could be decreased
from 7.2 2.7 mg/kg/hour with air oxygen to 5.7
2.0 mg/kg/hour with nitrous oxide oxygen.
150
REFERENCES
1. Aviado DM: Regulation of bronchomotor tone during
anesthesia. Anesthesiology 42:68, 1975.
2. Fielding JE: Smoking: Health effects and control. Engl
J Med 313:491^198, 1985.
3. Furchgott RF, Vanhoutte PM: Endothelium-derived re-
laxing and contracting factors. Fed Am Soc Exp Biol J
3:2007-2018, 1988.
4. Munakata M, Huang I, Mitzer W, Menkes H: Protective
role of epithelium in the guinea pig airway. J Appl Phys-
iol 66:1547-1552, 1989.
5. Hirschmann CA, Bergman NA: Factors influencing intra-
pulmonary airway calibre during anaesthesia. Br J An-
aesth 65:30-42, 1990.
6. Bennett DJ, Torda TA, Horton DA, et al: Severe broncho-
spasm complicating thoracotomy. Arch Surg 101:555,
1970.
7. Brakensiek AL, Bergman NA: The effects of halothane
and atropine on total respiratory resistance in anesthetized
man. Anesthesiology 53:341, 1970.
8. Patterson RW, Sullivan SF, Malm JR, et al: The effects
of halothane on human airway mechanics. Anesthesiol-
ogy 29:900, 1968.
9. McAslan C, Mima M, Norden I, et al: Effect of halothane
and methoxyflurane on pulmonary resistance to gas flow
during lung bypass. Scand J Thorac Cardiovasc Surg
5:193, 1971.
10. Waltemath CL, Bergman NA: Effect of ketamine and
halothane on increased respiratory resistance provoked by
ultrasonic aerosols. Anesthesiology 41:473, 1974.
11. Meloche R, Norlander O, Norden I, Herzog P: Effects of
carbon dioxide and halothane on compliance in pulmo-
nary resistance during cardiopulmonary bypass. Scand J
Thorac Cardiovasc Surg 3:69, 1969.
12. Hirshman CA, Edelstein G, Peetz S, Wayne R, Downes
H: Mechanism of action of inhalational anesthesia on
airways. Anesthesiology 56:107, 1982.
13. Vettermann J, Warner DO, Brichant JF, Rehder K: Halo-
thane directly relaxes airway smooth muscle. Anesthe-
siology 71 :A1070, 1989.
14. Vettermann J, Beck KC, Lindahl SHE, Brichant JF, Reh-
der K: Actions of enflurane, isoflurane, vecuronium, atra-
curium and pancuronium on pulmonary resistance in
dogs. Anesthesiology 69:688-695, 1988.
15. Revell S, Greenhalgh D, Absalom SR, Soni N: Isoflurane
in the treatment of asthma. Anaesthesia 43:477^179,
1988.
16. Parnass SM, Feld JM, Chamberlin WH, Segil LJ: Status
asthmaticus treated with isoflurane and enflurane. Anesth
Analg 66:193-195, 1987.
17. Bierman MI, Brown M, Muren O, Keenan RL, Glauser
FL: Prolonged isoflurane anesthesia in status asthmaticus.
Crit Care Med 14:832-833, 1986.
18. Johnston RG, Noseworthy TW, Friessem EG, Yule HA,
Shustock A: Isoflurane therapy for status asthmaticus in
children and adults. Chest 97:698-701, 1990.
19. Moss J, Rosow CE, Savarese JJ, et al: Role of histamine
in the hypotensive action of d-tubocurarine in humans.
20. Shemano I, Wendel H: Effects of meperidine hydrochlo-
ride and morphine sulfate on the lung capacity of intact
dogs. J Pharmacol Exp Ther 149:379-384, 1965.
21. Hirshman CA, Downes H, Farbood A, et al: Ketamine
block of bronchospasm in experimental canine asthma.
BrJ Anaesth 51:713-718, 1979.
22. Ben-Harari RR, Bansinath M: Reinvestigation of the
mechanism of bronchodilation by ketamine. Anesthesiol-
ogy 75: A627, 1991.
23. Corssen G, Gutierrez J, Reeves JG, et al: Ketamine in the
anesthetic management of asthmatic patients. Anesth An-
alg 51:588-596, 1972.
24. Hemmingsen C, Nielsen PK, Odorico J: Ketamine in the
treatment of bronchospasm during mechanical ventila-
tion. Anesth Analg 74:S135, 1992.
25. Clarke RSJ, Dundee JW, Garrett RT, McArdle GK, Sut-
ton JA: Adverse reactions to intravenous anesthesia. Br J
Anaesth 47:575, 1975.
26. Rees DI, Howell ML: Ketamine-atracurium by continu-
ous infusion as the sole anesthetic for pulmonary surgery.
Anesth Analg 65:860-864, 1988.
27. Crago RR, Bryan AC, Laws AIC, Winestock AE: Respi-
ratory flow resistance after curare and pancuronium meas-
ured by forced oscillations. Can Anaesth Soc J 19:607-
614, 1972.
28. Brandus V, Joffe S, Benoit CV, Wolff WI: Bronchial
spasm during general anesthesia. Can Anaesth Soc J
17:269-274, 1970.
29. Jarnberg PO, Andersen P, Cohen J, Jamond M, Smith J:
Intravenous lidocaine inhibits induced laryngospasm in
dogs during halothane anesthesia. Anesthesiology
75:A971, 1991.
30. Gal TJ: Airway responses in normal subjects following
topical anesthesia with ultrasonic aerosols with 4 per cent
lidocaine. Anesth Analg 59:123-129, 1980.
31. Loehning RW, Waltemath CL, Bergman NA: Lidocaine
and increased respiratory resistance produced by ultra-
sonic aerosols. Anesthesiology 44:306-310, 1976.
32. Don HF, Robson JG: The mechanics of the respiratory
system during anesthesia: The effects of atropine and
carbon dioxide. Anesthesiology 26:168-178, 1965.
33. Severinghaus JW, Stupfel M: Respiratory dead space in-
crease following atropine in man and atropine, vagal or
ganglionic blockade and hypothermia in dogs. J Appl
Physiol 8:81-87, 1955.
34. U.S. Department of Health and Human Services: The
health consequences of smoking: Cardiovascular disease.
A report of the Surgeon General. Rockville, MD, U.S.
Department of Health and Human Services, 1983.
35. The Pooling Project Research Group: Relationship of
blood pressure, serum cholesterol, smoking habit, relative
weight, and ECG abnormalities to incidence of major
coronary events: Final report of the Pooling Project. J
Chronic Dis 31:201-306, 1978.
CHAPTER 9
Separation of the Two Lungs
(Double-Lumen Tube and
Bronchial Blocker Intubation)
I. Introduction
II. Indications for Separation of the
Two Lungs
A. Absolute Indications
B. Relative Indications
III. Double-Lumen Tube Intubation
A. Various Double-Lumen
Endotracheal Tubes
1. Robertshaw Double-Lumen
Endotracheal Tube
a. Original Red Rubber
Robertshaw Double-Lumen
Endotracheal Tube
b. Modern Plastic Disposable
Robertshaw Double-Lumen
Endotracheal Tube
B. Conventional (Nonfiberoptic)
Double-Lumen Tube Intubation
Procedure
1. Choice of Left- Versus Right-
Sided Double-Lumen Tube
2. Choice of Double-Lumen Tube
Size
3. Double-Lumen Tube Intubation
Procedure
a. Preintubation Procedures
b. Intubation
c. Routine Checking of Double-
Lumen Tube Position
d. Auscultation and Unilateral
Clamping Maneuvers to
Diagnose Double-Lumen
Tube Malposition
e. Nonfiberoptic Method for
Intubation of Left Main-Stem
Bronchus When a Left-Sided
Tube Will Not Pass Routinely
C. Use of Fiberoptic Bronchoscope to
Insert the Bronchial Lumen of a
Double-Lumen Tube Into a Main-
Stem Bronchus
D. Use of Fiberoptic Bronchoscope to
Determine Precise Double-Lumen
Tube Position
1. Margin of Safety in Positioning
Double-Lumen Tubes
a. Left-Sided Double-Lumen
Tubes
(1) Definition of Margin of
Safety in Positioning
Left-Sided Double-
Lumen Tubes
(2) Measurement of Length
of Left Main-Stem
Bronchus
(3) Measurement of Margin
of Safety for Left-Sided
Double-Lumen Tubes
(4) Clinical Examples/
Implications of
Exceeding Margin of
Safety for Left-Sided
Double-Lumen Tubes
b. Right-Sided Double-Lumen
Tubes
(1) Definition of Margin of
Safety in Positioning
Right-Sided Double-
Lumen Tubes
(2) Measurement of Length
of Right Main-Stem
Bronchus and Right
Upper Lobe Bronchial
Orifice
(3) Measurement of Margin
of Safety for Right-Sided
Double-Lumen Tubes
(4) Clinical Examples/
Implications of
Exceeding Margin of
Safety for Right-Sided
Double-Lumen Tubes
2. Relationship of Fiberoptic
Bronchoscope Size to Double-
Lumen Tube Size
Use of Chest X-Ray to Determine
Double-Lumen Tube Position
Other Methods to Determine
Securing the Double-Lumen Tube
Quantitative Determination of Cuff-
Seal Pressure Hold
Complications of Double-Lumen
Endotracheal Tubes
Relative Contraindications to the
Use of Double-Lumen Endotracheal
Tubes
330
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation) 331
IV. Bronchial Blockers (With Single-
Lumen Endotracheal Tubes)
A. Univent Bronchial Blocker Tube
1. Description of Univent Bronchial
Blocker Tube
2. Insertion of Univent Tube and
Positioning of Bronchial Blocker
3. Advantages/Noteworthy Positive
Attributes of Univent Bronchial
Blocker Tube System
4. Potential Limitations of Univent
Bronchial Blocker Tube System
and Solutions
I. INTRODUCTION
The complete functional separation of the two
lungs is often the most important anesthetic con-
sideration for patients undergoing thoracic surgery.
The procedure can occasionally be life-saving, and
very frequently it greatly facilitates the conduct of
surgery. Newly introduced disposable plastic dou-
ble-lumen tubes (DLTs), which are relatively non-
traumatic and easy to insert, and the advent of
fiberoptic bronchoscopy (FOB), which makes lo-
cation of a DLT under direct vision possible and
therefore a precise, repeatable, low-risk maneuver,
have greatly increased the efficacy and use of
DLTs. In addition, the new Univent bronchial
blocker tube (Fuji Corp., Tokyo, Japan) makes the
use of a bronchial blocker much easier and simpler
by coupling the insertion of the main single-lumen
tube with the insertion of the bronchial blocker;
because the single-lumen tube and bronchial
blocker are inserted together, the only new skill
requirement is to push, under fiberoptic guidance
and during continuous ventilation, the bronchial
blocker into the desired main-stem bronchus. This
chapter sequentially discusses the indications for
separation of the two lungs, conventional tech-
niques of DLT insertion, and determination of pre-
cise DLT position by FOB. The chapter then con-
siders how to position properly the bronchial
blocker of the Univent tube. Next, endobronchial
intubation with a single-lumen tube is briefly con-
sidered, as are a variety of other ways to separate
the two lungs (coaxial tubes, the Wilson tube, T-
tube ventilating system).
II. INDICATIONS FOR SEPARATION
OF THE TWO LUNGS
There are several absolute and relative indica-
tions for separation of the two lungs during tho-
racic operations or procedures (Table 9-1).
5. Methods to Obtain a Just-Seal
Volume in Bronchial Blocker
Cuff
6. Firm Clinical Indications for Use
of Univent Bronchial Blocker
System
B. Bronchial Blockers That Are
Independent of a Single-Lumen
Tube
V. Endobronchial Intubation With
Single-Lumen Tubes
A. Absolute Indications
Separation of the two lungs for any of the ab-
solute indications discussed here should be consid-
ered a life-saving maneuver because failure to sep-
arate the two lungs under any of these conditions
could result in a life-threatening complication or
situation. There are three absolute indications for
separating the two lungs (Table 9-1 and Fig. 9-1 ).
First, separation of one lung from the other is
Table 9-1 INDICATIONS FOR SEPARATION
OF THE TWO LUNGS (DOUBLE-
LUMEN TUBE INTUBATION) AND/
OR ONE-LUNG VENTILATION
Absolute
1. Isolation of one lung from the other to avoid spillage or
contamination
A. Infection
B. Massive hemorrhage
2. Control of the distribution of ventilation
A. Bronchopleural fistula
B. Bronchopleural cutaneous fistula
C. Surgical opening of a major conducting airway
D. Giant unilateral lung cyst or bulla
E. Tracheobronchial tree disruption
F. Life-threatenng hypoxemia caused by unilateral lung
disease
3. Unilateral bronchopulmonary lavage
A. Pulmonary alveolar proteinosis
Relative
1. Surgical exposurehigh priority
A. Thoracic aortic aneurysm
B. Pneumonectomy
C. Thoracoscopy
D. Pulmonary resection via median sternotomy
E. Upper lobectomy
F. Mediastinal exposure
2. Surgical exposuremedium (lower) priority
A. Middle and lower lobectomies and subsegmental resec-
tions
B. Esophageal resection
C. Procedures on the thoracic spine
3. Postcardiopulmonary by-pass status after removal of totally
occluding chronic unilateral pulmonary emboli
4. Severe hpyoxemia caused by unilateral lung disease
332 Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation)
Absolute Indications for
Lung Separation/One Lung Ventilation
Figure 9-1 This schematic diagram shows the absolute indications for lung separation and/or one-lung ventilation. Hemorrhage
and infection in one lung can contaminate and soil the other lung. A low-resistance ventilation pathway, such as a bronchopleural
fistula (BPF) or a surgically opened major airway, can make positive-pressure ventilation of the other lung impossible. A giant
unilateral cyst or bulla can rupture if exposed to positive-pressure ventilation and can result in a tension pneumothorax. Unilateral
lung lavage requires the instillation of large amounts of saline into one lung while the other lung is being ventilated. Major
tracheobronchial disruption can lead to mediastinal and pulmonary interstitial emphysema if exposed to positive pressure.
absolutely necessary to prevent spillage of pus or
blood from an infected (abscessed) lung or bleed-
ing lung, respectively, to a noninvolved lung.
Acute contamination of a lung with either blood
or pus from the other lung usually results in severe
massive (bilateral) atelectasis, pneumonia, and
sepsis. Second, a number of unilateral lung prob-
lems can prevent adequate ventilation of the other
noninvolved side. A large bronchopleural or bron-
chopleural cutaneous fistula or a surgically opened
conducting airway has such a low resistance to gas
flow that a tidal inspiration delivered by positive
pressure will exit via the low-resistance pathway,
and it will become impossible to ventilate the
other, more normal lung adequately. A giant uni-
lateral bulla or cyst may rupture if exposed to
positive-pressure ventilation and result in a tension
pneumothorax or pneumomediastinum. Very se-
vere or life-threatening hypoxemia caused by uni-
lateral lung disease may require differential lung
ventilation and positive end-expiratory pressure
(PEEP).
1
Positive-pressure ventilation of a lung
with a tracheobronchial tree disruption can result
in dissection of gas into the pulmonary interstitial
space or mediastinum, resulting in a tension pneu-
momediastinum. Third, separation of the two
lungs is absolutely necessary to perform unilateral
bronchopulmonary lavage in patients with pulmo-
nary alveolar proteinosis (and, rarely, asthma and
cystic fibrosis).
B. Relative Indications
There are a large number of relative indications
for separation of the two lungs, and they are all
for the purpose of facilitating surgical exposure by
collapsing the lung in the operative hemithorax.
These relative indications can be divided into
high-priority and medium-, or lower, priority cat-
egories (Table 9-1 and Fig. 9-2). Of the relative
indications, repair of a thoracic aortic aneurysm is
usually the highest priority because it requires ex-
posure of the thoracic aorta as it runs the entire
length of the left hemithorax. A pneumonectomy,
especially if performed through a median sternot-
omy,
2
is greatly aided by the wide exposure of the
lung hilum, which is afforded by collapse of the
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker Intubation) 3 3 3
operative lung. Examination of the pleural space
(thoracoscopy) and pulmonary resections through
a thoracoscope are considerably aided by collapse
of the ipsilateral lung. Similarly, an upper lobec-
tomy, which is technically the most difficult lobec-
tomy, and many mediastinal exposures may be
made much easier by eliminating ventilation to the
lung on the side of the procedure. The surgical
items in the medium-priority category do not rou-
tinely require collapse of the lung on the operative
side but still significantly aid surgical exposure
and eliminate the need for the surgeon to handle
(retract, compress, pack away) the operative lung.
Severe intraoperative retraction of the lung on the
operated side can traumatize the operative lung
and impair gas exchange both intra-
3 4
and post-
operatively.
5
6
The lower priority items consist of
middle and lower lobectomies, less extensive pul-
monary resections, thoracic spinal procedures that
are approached anteriorly through the chest,
7
and
esophageal surgery. However, even relatively
small operations such as wedge and segmental re-
sections benefit by DLT insertion because of the
ability to alternate easily and quickly between lung
collapse and inflation, which is sometimes re-
quired to visualize lung morphology better and to
facilitate identification and separation of important
planes and fissures. Finally, the separation of the
lungs after removal of totally occluding and pre-
dominantly unilateral chronic pulmonary emboli
(postcardiopulmonary by-pass) can be very helpful
if significant transudation of hemorrhagic fluid
Relative Indications for Lung Separation/
One Lung Ventilation
Significant Hypoxemia due to Unilateral Lung Disease
Figure 92 This schematic diagram depicts the relative indications for lung separation and/or one-lung ventilation. The indica-
tions are all for facilitating surgical exposure and can be divided into high-priority and low-priority categories. The high-priority
items for lung collapse are thoracic aortic artery aneurysm repair (requires exposure of the entire thoracic aorta, which is greatly
facilitated by collapsing the left lung), pulmonary resection via median sternotomy pneumonectomy (requires wide exposure of the
lung hilum), thoracoscopy (especially when therapeutic and video-assisted) and upper lobe lobectomy (technically the most difficult
lobe to expose). The lower priority items for lung collapse are middle and lower lobe lobectomy and esophageal surgery. Procedures
on the thoracic spine that are approached anteriorly via the chest are facilitated by collapse of the lung on the operative side. Lung
separation is useful if unilateral pulmonary edema occurs following removal of a chronic, totally occluding, unilateral pulmonary
embolus (postcardiopulmonary by-pass).
across the alveolar capillary membrane in the re-
gion of the lung supplied by the previously oc-
cluded vessel occurs after cardiopulmonary by-
pass (reperfusion of a previously and chronically
nonperfused vascular bed). Should significant and
predominantly unilateral postthromboembolec-
tomy postcardiopulmonary by-pass pulmonary
edema occur, the patient should be returned to
cardiopulmonary by-pass, and a double-lumen en-
dotracheal tube should be inserted so that differ-
ential lung ventilation may be used (see chapters
11 and 20). Finally, significant hypoxemia result-
ing from unilateral lung disease may be treated
more easily by differential lung ventilation and
PEEP.
1
III. DOUBLE-LUMEN TUBE
INTUBATION
Double-lumen endotracheal tubes have evolved
to become considered the lung-separation tech-
nique of choice for the majority of thoracic surgery
cases and are discussed here in great detail. Bron-
chial blockade with the Univent tube in adults is
greatly increasing and is also described at length.
Endobronchial tubes are not often used today and
are only briefly described at the end of the chapter.
DLTs are favored over bronchial blockers and en-
dobronchial tubes for lung separation primarily be-
cause they are more versatile than bronchial block-
ers or endobronchial tubes. The most important
double-lumen function not available with a bron-
chial blocker is independent bilateral suctioning.
In addition, it is easier to apply continuous positive
airway pressure (CPAP) to the nonventilated
operative lung with a DLT compared with a bron-
chial blocker. Endobronchial tubes are very lim-
ited in function and allow only one-lung ventila-
tion.
There are two firm disadvantages/contraindica-
tions to the use of a DLT compared with a bron-
chial blocker. First, very distorted tracheobron-
chial tree anatomy, including exophytic and
stenotic lesions, as well as tortuosity may preclude
successful correct placement/positioning of a DLT.
Second, changing from a DLT to a single-lumen
tube during or at the end of a case can be expected
to be a difficult/risky procedure on occasion. Ex-
amples of situations in which a change is required
from DLT to a single-lumen tube during the oper-
ative period, and the need for this change is known
to exist preoperatively, are (1) turning the patient
from the supine to the prone position (45 per cent
of cases involving an anterior approach to surgery
on the vertebral bodies
7
) and (2) for postoperative
ventilation in the intensive care unit (ICU). There
are many known anatomic causes of difficult lar-
yngoscopy and intubation,
8
but the most important
intraoperative/postoperative risk factor for making
laryngoscopy and intubation difficult is the infu-
sion of massive amounts of blood and fluids (the
upper airway becomes edematous).
There are two relatively minor in situ disadvan-
tages to DLTs related to the fact that the lumens
of a DLT may be narrow. First, suctioning may be
more difficult down a narrow lumen, but this is
usually not a problem with the new disposable
Robertshaw-type DLTs, which have nonadhering
suction catheters that slide easily down the lumens
of the DLT. Second, although airway resistance
may be increased with a narrow lumen, the in-
creased airway resistance can be easily overcome
by positive-pressure ventilation.
9
A. Various Double-Lumen Endotracheal
Tubes
DLTs are essentially two catheters bonded to-
gether, and each lumen is intended to ventilate one
of the two lungs. DLTs are made as left- and right-
sided tubes. A left-sided tube means that the left
lung catheter is placed into the left main-stem
bronchus, whereas the right lung catheter ends in
the trachea; therefore, for a left-sided tube, the left
lung catheter is longer than the right lung catheter
(Fig. 9-3). A right-sided tube means that the right
lung catheter is placed into the right main-stem
bronchus, whereas the left lung catheter ends in
the trachea; therefore, for a right-sided tube, the
right lung catheter is longer than the left lung
catheter (Fig. 9-3). All the DLTs have a proximal
cuff for the trachea and a distal cuff for a main-
stem bronchus; the endobronchial cuff causes sep-
aration and sealing off of the two lungs from each
other, and the tracheal cuff causes separation and
sealing off of the lungs from the environment. The
part of the right lung catheter of the right-sided
DLT that is in the right main-stem bronchus must
be slotted to allow for ventilation of the right up-
per lobe (Fig. 9-3) because the right main-stem
bronchus is too short to accommodate both the
right lumen tip and the right endo-bronchial cuff.
All the double-lumen endotracheal tubes have
two curves that lie in planes approximately 90
degrees apart from one another. The distal curve
is designed to facilitate placement of the distal
catheter tip into the appropriate main-stem bron-
chus, and the proximal curve is designed to ap-
proximate the oropharyngolaryngeal curve.
The DLTs that have been used for lung separa-
tion and one-lung ventilation include the Carlens,
White, Bryce-Smith, and Robertshaw types. The
Robertshaw double-lumen endotracheal tube is by
far the most commonly used, and the disposable
polyvinylchloride Robertshaw tube has almost
completely replaced the red rubber Robertshaw
Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker intubation) 335
Double-Lumen Tubes
tube (the former is easier to pass and suction
through, is positioned more quickly, has a lower
airflow resistance, and causes less mucosal dam-
age).
10
"
12
Consequently, the modern polyvinyl-
chloride tube is described in great detail. The other
DLTs (the first three mentioned) are seldom used;
however, some anesthesiologists still routinely use
the Carlens tube, and some prefer the right-sided
red rubber Robertshaw type (when a right-sided
tube is used) because the right upper lobe ventila-
tion slot is longer than it is in the polyvinylchlor-
ide tube
13
and may better locate opposite/ventilate
the right upper lobe (see Margin of Safety in Po-
sitioning Double Lumen Tubes). The remaining
DLTs are largely of historical interest and are only
briefly described here.
The left-sided Carlens tube (Fig. 9-4) was the
first double-lumen endotracheal tube utilized for
one-lung ventilation.
14
The tube has a carinal hook
to aid in its proper placement and to minimize tube
movement after placement. Potential problems
with carinal hooks include increased difficulty
(more rotations) and laryngeal trauma during intu-
bation, amputation of the hook during passage,
malpositioning of the tube due to the hook (see
Margin of Safety in Positioning Double-Lumen
Tubes), and physical interference when performing
a pneumonectomy.
15
Therefore, some anesthesiol-
ogists prefer to use the tube with the hook cut off.
The tube is available in four sizes: 41, 39, 37, and
35 French (which correspond to an internal diam-
eter for each lumen of approximately 7.0, 6.5, 6.0,
and 5.5 mm, respectively). The cross-sectional
shape of each lumen is oval, and this accounts for
the occasional difficulty in passing a suction cath-
eter down the lumen (see Table 9-2).
The White tube was essentially a modified right-
sided Carlens tube and was used for right main-
stem bronchus intubation.
16
The right main-stem
bronchial cuff is slotted to provide for ventilation
of the right upper lobe. As with the Carlens tube.
suctioning may occasionally prove to be difficult.
and the carinal hook can cause a variety of physi-
cal problems.
The Bryce-Smith tube (Fig. 9-5) represents an-
other modification of the Carlens tube and was
Placement at the Carina
A. Carlens Tube
Figure 9- 4 A, Sketch of the red rubber Carlens double-
lumen tube. B. Close-up of the placement of the red rubber
Carlens double-lumen tube at the carina. Note that the left
endobronchial lumen and carinal hook straddle the carina.
Table 9-2 PHYSICAL CHARACTERISTICS OF VARIOUS DOUBLE-LUMEN TUBES*
*Based in part on Brodsky.'t The 37, 39, and 41 F correspond to small, medium, and large, respectively, in the red rubber
tubes.
tAverage, each lumen.
intended to reduce trauma to the larynx and tra-
cheobronchial tree.
17
This tube was originally de-
signed for placement in the left main-stem bron-
chus, although a right-sided tube was soon
developed as well.
18
The cuff for the right main-
stem bronchus is slotted to allow for ventilation of
the right upper lobe. These tubes do not have a
carinal hook, and both lumens (arranged anteriorly
and posteriorly) are round, allowing for greater
ease in passing a suction catheter. Bryce-Smith
double-lumen endotracheal tubes are available in
three sizes according to the internal diameter of
the lumen: 7, 6.5, and 6 mm.
A new coaxial double-tube system for lung sep-
aration has been described.
19
Separate ventilation
is achieved by means of an appropriate cuffed
bronchial tube whose tip is positioned in a main
bronchus by passing it through a standard orotra-
cheal tube (Fig. 9-6). The tubes assembled in this
way make up two independent channels, each in
communication with one of the main bronchi.
Ventilation is then carried out separately through
B. Placement at the Carina D. Placement at the Carina
A. Left Bryce-Smith Tube C. Right Bryce-Smith Tube
Figure 9-5 A, Sketch of the left Bryce-Smith double-lumen tube. B, Close-up of the placement of the left Bryce-Smith double-
lumen tube at the carina. C, Sketch of the right Bryce-Smith tube. D, Close-up of the placement of the right Bryce-Smith double-
lumen tube at the carina.
To the respirotor
Figure 9-6 Schematic representation of the coaxial double
tube with the bronchial tube positioned within the left main-
stem bronchus. (From Nazeri S, Trazzi R, Moncalvo F, et al:
Selective bronchial intubation for one lung anesthesia in tho-
racic surgery. Anaesthesia 41:519-526, 1986. Used with per-
mission.)
the bronchial tube on one side and the residual
tracheal tube lumen on the other side. The bron-
chial tube can move freely in the main lumen, and
the seal between the two channels is airtight. The
bronchial tubes for the right and left sides are
curved differently, and the left bronchial tube is
fitted with a tracheal carinal hook, whereas the
right bronchial tube has a hook for the first main-
stem carina (separating the right upper lobe from
the right middle and right lower lobes). The sizes
of the tubes are such that they supply two channels
with similar airflow resistance: This allows the
anesthestic gases to be delivered through a Y-
shaped connector to both channels (see Fig. 9-6),
supplying the same flow to the two lungs when
they are both being ventilated. Because this new
coaxial double-tube system is not in widespread
use (since its introduction in 1986) and it is not
apparent at present that it will be in widespread
use in the foreseeable future, the reader is referred
to the original articles for the details of inser-
tion/operation.
Finally, passing a conventional DLT through a
tracheostomy presents some special problems that
can be solved by using a modified DLT. Conven-
tional DLTs made of polyvinylchloride can be
used with tracheostomies, but, because of the
shortened length of the upper airway, special atten-
tion is required to prevent kinking and subsequent
luminal obstruction. Fixation of the tube to prevent
movement and displacement may also be a prob-
lem with the standard DLT. These problems are
greatly magnified if the tube is required for ex-
tended periods, as may occur when selective split-
lung ventilation in the ICU is indicated.
A DLT has been designed (left-sided 41 French
Broncho-Trach DLT [Sheridan, Argyle, NY]) that
has been used successfully in patients with tra-
cheostomies (Fig. 9-7).
20
The distance between the
distal tip of the endobronchial lumen and the bi-
furcation of the tracheal and endobronchial lumens
is shortened to 18.5 cm from 32.0 cm to reflect the
markedly reduced length of the upper airway. The
proximal length of both lumens after they bifurcate
is also shortened to 3.0 cm from 7.5 cm to reduce
the chance of kinking. A 90-degree bend is placed
approximately 2.5 cm proximal to the proximal
edge of the tracheal cuff to allow the tube to exit
from the neck at a less awkward angle. The tub-
ings to the pilot balloons are shortened to 11.0 cm
from 23.0 cm for convenience. In all other aspects.
Figure 9-7 Modified left-sided
polyvinyl chloride (PVC) tracheos-
tomy double-lumen endobronchial
tube (DLT) next to a standard left-
sided PVC DLT (Broncho-Trach,
Sheridan, Argyle, NY). (From Brod-
sky JB, Tobler HG, Mark JBD: A
double-lumen endobronchial tube for
tracheostomies. Anesthesiology 74:
387-388, 1991. Used with permis-
sion.)
the tube is identical to a standard 41 French Sher-
idan polyvinylchloride DLT. If necessary, other
(37F and 39F) polyvinylchloride DLTs may be
similarly modified for smaller patients.
1. Robertshaw Double-Lumen
Endotracheal Tube
a. ORIGINAL RED RUBBER ROBERTSHAW
DOUBLE-LUMEN ENDOTRACHEAL TUBE
The original Robertshaw DLT, introduced in
1962, was made as a reusable red rubber tube (Fig.
9-8).
21
This tube was designed to provide the larg-
est possible lumen in order to decrease airway
resistance and to facilitate removal of secretions.
The lumens are D-shaped and lie side by side, like
those of the Carlens tube, but are larger in size. As
with the other double-lumen endotracheal tubes, it
has two curves (in planes 90 degrees apart) that
facilitate intubation and proper endobronchial
placement. Both right- and left-sided tubes are
available, and the absence of a carinal hook allows
for easier tracheal intubation and perhaps correct
positioning. Because of these features, the original
Robertshaw DLT rapidly gained popularity.
22
The right-sided tube has a relatively long, slot-
ted endobronchial cuff (22 mm vs. 11 mm for the
polyvinylchloride tube) to permit ventilation of the
right upper lobe. The long right upper lobe venti-
lation slot results in a much lower incidence of
right upper lobe obstruction after "blind" inser-
tion compared with the polyvinylchloride tube (10
per cent vs. 89 per cent).
13
The endobronchial cuff
has an additional area of inflation on the nonslot-
ted side above the slot to effect a more reliable
seal (in contrast to the endobronchial cuffs of the
other right-sided tubes that do not have this infla-
tion area). On the slotted side, inflation of the
endobronchial cuff is restricted. However, the right
endobronchial cuff design forces the right upper
lobe slot to lie flat against the right upper lobe
orifice, and if the right upper lobe slot is not per-
fectly aligned with the right upper lobe orifice, the
right upper lobe ventilation slot will be perfectly
aligned with the right upper lobe orifice, and the
right upper lobe ventilation slot will be blocked
(obstructed) by the right main-stem bronchial wall
(and vice versa). Unfortunately, the bronchial cuff
on the red rubber cuff is a low-volume-high-pres-
sure cuff; the mean just-seal volume and intracuff
pressure ( standard deviation) are 2.2 1.1 ml
and 130 41 mm Hg, respectively;
23,24
although
only 10 to 20 per cent of intracuff pressure is
transmitted/applied to the bronchial mucosal
wall,
25
mucosal perfusion is probably impaired to
some degree by the sealing of the bronchial cuff
of the red rubber tube. In addition, high-pressure
B. Placement at the Carina D. Placement at the Carina
A. Left Robertshaw Tube C. Right Robertshaw Tube
Figure 9-8 A, Sketch of the left-sided red rubber Robertshaw double-lumen tube. B, Close-up of the placement of the left-sided
red rubber Robertshaw double-lumen tube at the carina. C, Sketch of the right-sided red rubber Robertshaw double-lumen tube. D,
Close-up of the placement of the right-sided red rubber Robertshaw double-lumen tube at the carina.
cuffs such as the bronchial cuff of the red rubber
tube have a tendency to dilate asymmetrically, so
greater volumes may be needed to effect a seal.
b. MODERN PLASTIC DISPOSABLE
ROBERTSHAW DOUBLE-LUMEN
ENDOTRACHEAL TUBE
The Robertshaw-type tube is now made of a
clear, nontoxic, tissue-implantable plastic (denoted
by the marking Z-79) and is disposable (Fig. 9-9).
The tubes are made in sizes 41, 39, 37, 35 and 28
French (internal diameter of each lumen is approx-
imately 7.4, 7.0, 6.5, 6.0, and 4.5 mm, respec-
tively) (see Table 9-2). These tubes are relatively
easy to insert and have appropriate end-of-lumen
and cuff arrangements that minimize lobar ob-
struction. Each lumen is color coded (bronchial is
blue and tracheal is clear colorless) and labeled.
The endobronchial cuff is colored brilliant blue,
which is a very important feature for recognition
when using a fiberoptic bronchoscope. The ends
of both lumens have a black radiopaque line,
which is an essential recognition marker when
viewing a chest X-ray. The tubes have high-vol-
ume-low-pressure tracheal and endobronchial
cuffs; the mean just-seal volume and intracuff
pressure ( standard deviation) are 3.0 2 ml
and 56 21 mm Hg, respectively. Because the
actual pressure transmitted to the bronchial wall is
only approximately 10 to 20 per cent of the meas-
ured intracuff pressure (the difference between
measured intracuff pressure and the pressure ap-
plied to the bronchial wall is the pressure required
simply to distend and maintain the inflation of the
cuff itself), mucosal perfusion is probably main-
tained. The slanted doughnut-shaped endobron-
chial cuff on the Mallinkrodt right-sided DLT al-
lows the right upper lobe ventilation slot to ride
off of (away from) the right upper lobe orifice,
which minimizes the chance of right upper lobe
obstruction by the tube.
The clear color-coded tubing is helpful because
it permits continuous observation of the tidal
movement of respiratory moisture as well as ob-
servation of secretions from each lung. When the
polyvinylchloride material heats to body tempera-
ture, the tube shape changes to approximate the
shape of the airway more closely. The tubes are
packaged with malleable stylets and are relatively
easy to insert and position. These tubes have large
internal to external diameter ratios and gentle cur-
vatures, thereby allowing relatively easy section-
ing, and have a low resistance to airflow. They are
packaged with their own nonadhering suction
catheters. For these reasons, the Robertshaw-type
tubes are now considered by far the double-lumen
endotracheal tube of choice by most anesthesiolo-
gists. As expected, several manufacturers make
these DLTs (Mallinkrodt, Sheridan, Rusch, Por-
tex).
. Conventional (Nonfiberoptic)
Double-Lumen Tube Intubation
Procedure
1. Choice of Left- Versus Right-Sided
Double-Lumen Tube
It has been recommended that the nonoperated
main-stem bronchus be routinely intubated be-
cause intubation of the operative main-stem bron-
chus may interfere with the performance of sur-
gery. Consequently, there is no controversy
regarding using a left-sided double-lumen endotra-
cheal tube for right thoracotomies requiring col-
lapse of the right lung and ventilation of the left
lung (Fig. 9-10). However, there is controversy
regarding using a right-sided DLT for left lung
surgery (see later discussion), and for this reason,
either a left- or right-sided tube may be used for
left thoracotomies requiring collapse of the left
lung and ventilation of the right lung (see Fig. 9-
10). However, because the right upper lobe venti-
lation slot of a right-sided tube has to be closely
apposed to the right upper lobe orifice to ensure
unobstructed right upper lobe ventilation, and be-
cause there is considerable anatomic variation in
the exact position of the right upper lobe orifice
and, therefore, length of the right main-stem bron-
chus (in fact, it is well known that an anomalous
right upper lobe can take off from the trachea; the
incidence is 1:250 in normal patients and 1:50 in
patients who have some other congenital defect
[see later discussion]), use of a right-sided tube for
left lung collapse introduces the risk of inadequate
right upper lobe ventilation. For this reason, a left-
sided tube is preferable for most cases requiring
one-lung ventilation. If clamping of the left main-
stem bronchus is necessary, the tube can be with-
drawn at that time into the trachea and then used
in the same manner as a single-lumen endotracheal
tube (ventilate the right lung with both lumens
with the endobronchial cuff deflated) (see Fig. 9-
10). If after withdrawal the tracheal cuff is supra-
glottic, then the bronchial cuff should be inflated
and the patient ventilated through the bronchial
lumen.
Contraindications to the use of a left-sided DLT
are proximal left main-stem bronchial lesions that
could be traumatized by the passage of a left-sided
tube. These lesions include strictures, endoluminal
tumors, tracheobronchial disruptions, and com-
pression of the airway by an external mass (includ-
ing a thoracic aortic aneurysm).
26
In addition, left
lower lobe and left upper lobe tumors may push
The Advantages of the Modern
Plastic Disposable Robertshaw Double-Lumen
Endobronchial Tubes

Figure 9-9 A, Left- and right-sided disposable Robertshaw double-lumen tubes. B, Schematic diagram depicts the advantages of
left-sided and right-sided modern plastic disposable Robertshaw double-lumen endobronchial tubes. Both lumens of the left-sided
double-lumen tube are shown, whereas only the distal endobronchial lumen of the right-sided double-lumen tube is shown.
340
Right Lung Surgery
and Left-Sided
Double Lumen Tube
Left Lung Surgery
and Right-Sided
Double-Lumen Tube
Left Lung Surgery and
Left-Sided Double-Lumen
Tube Pulled Back

Figure 9-10 Use of left-sided and right-sided double-lumen tubes for left- and right-lung surgery (as indicated by the clamp).
When surgery is going to be performed on the right lung, a left-sided double-lumen tube should be used (A). When surgery is going
to be performed on the left lung, a right-sided double-lumen tube can be used (B). However, because of uncertainty as to the
alignment of the right upper lobe ventilation slot to the right upper lobe orifice, a left-sided double-lumen tube can also be used for
left lung surgery (C). If the left-lung surgery requires a clamp to be placed high on the left main-stem bronchus, the left
endobronchial cuff should be deflated, the left-sided double-lumen tube pulled back into the trachea, and the right lung ventilated
through both of the lumens (use the double-lumen tube as a single-lumen tube).
and pull, respectively, the left main-stem bronchus
off the trachea at a very acute angle. This distor-
tion of the tracheobronchial tree may make it im-
possible to cannulate the left main-stem bronchus,
even over a fiberoptic bronchoscope. Finally,
sleeve resection of the proximal left main-stem
bronchus contraindicates the presence of an endo-
bronchial catheter in the surgical field.
The largest size tube that can comfortably pass
the glottis should be used because a relatively
small DLT may require excessive cuff volume for
an endobronchial cuff seal to be obtained (see the
discussion of endobronchial cuff problems in sec-
tion III. D. and Fig. 9-20), may be inadvertently
inserted too deeply, may not protect against trans-
bronchial spread of pus, blood, or necrotic tumor
if the endobronchial cuff is deflated during two-
lung ventilation (larger space around small lu-
men),
27
and may cause difficulty with suctioning
secretions.
In summary, the new plastic disposable Robert-
shaw-type DLTs are by far the most commonly
used DLTs. Because a right-sided tube incurs the
risk of inadequate right upper lobe ventilation, left-
sided tubes are used more commonly than right-
sided tubes. Consequently, the rest of this chapter
emphasizes the insertion and location of the left-
sided Robertshaw-type DLT.
2. Choice of Double-Lumen Tube Size
The appropriate size of DLT is one that results
in a just-seal volume for the endobronchial cuff
that is greater than l ml but less than 3 ml. If the
endobronchial just-seal cuff volume is less than l
ml, then the outside diameter of the bronchial lu-
men is very near the internal diameter of the bron-
chial lumen, and the risk of bronchial wall damage
is increased, especially if the DLT is inserted too
deeply. If the endobronchial just-seal cuff volume
is greater than 3 ml, then the outside diameter of
the endobronchial tube is too small in relation to
the bronchial lumen, the intracuff pressures may
be excessive (the bronchial cuff will behave as a
high-pressure cuff),
28
and the risk of bronchial wall
damage will be increased, especially if nitrous ox-
ide is used.
In general, as height and weight increase, the
appropriate size of the DLT (as defined previ-
ously) increases, although height is much more
important than weight.
29
Because women are usu-
ally shorter than men, sizes 35 French and 37
French are ordinarily appropriate for women,
whereas sizes 39 French and 41 French are appro-
priate for men. However, choice of DLT by height,
irrespective of gender, makes more sense; Figure
9-11 shows the frequency of choice of DLT size
Figure 911 Larger double-lumen tubes were chosen as pa-
tient height increased. (From Brodsky JB, Benumof JL, Ehren-
werth J, Ozaki GT: Depth of placement of left double-lumen
endobronchial tubes. Anesth Analg 73:570-572, 1991. Used
with permission.)
by grouped interval of height (panels A, B, and C)
in a large series of patients from three institutions
when it was known that the DLT was in proper
position and cuff volume was appropriate.
30
Al-
though there are strong tendencies for short and
tall individuals to receive small and large DLTs,
respectively, the choice of size was still moder-
ately variable.
3. Double-Lumen Tube Intubation
Procedure
a. PREINTUBATION PROCEDURES
Before intubation with a double-lumen endotra-
cheal tube, both cuffs and lumen connections are
checked. A 3-ml syringe should be placed on the
end of the bronchial cuff pilot tube because proper
bronchial cuff inflation rarely requires more than 1
to 2 ml of air; a 5- or 10-ml syringe should be
placed on the tracheal cuff pilot tube. Both cuffs
should be tested for leaks. Because the high-
volume-low-pressure cuffs can be easily torn by
teeth, the distal tube is coated with a lubricating
ointment (preferably containing a local anesthetic)
to minimize this possibility. If a less than optimal
view of the larynx is anticipated, the stylet that
comes packaged with the tube is lubricated, in-
serted into the endobronchial (in this case, the left)
lumen, and appropriately curved.
The patient is then anesthetized and paralyzed
as described in chapter 8. With the induction of
anesthesia, patients who are apneic for more than
the normal amount of time are at risk for hypoxia
before intubation and/or ventilation is achieved;
this may occur despite "preoxygenation." During
this apneic period, oxygen can be drawn from the
pharynx into the trachea and lungs, provided that
the airway is patent; the transfer of oxygen in this
manner should delay the onset of hypoxia.
One randomized, double-blind study proved that
insufflation of oxygen into the nasopharynx of
denitrogenated patients could effectively prolong
the safe apneic period before intubation.
31
Patients
were between 35 and 65 years old (52 4 years,
mean standard error) and had significant histo-
ries of tobacco use (20 6 pack-years). During
pharyngeal oxygen insufflation, S
a
0
2
never fell be-
low 97 per cent during the entire 10 min of apnea
in any subject; the mean minimum saturation
achieved ( standard deviation) was 98 1 per
cent. Conversely, in the absence of oxygen insuf-
flation, the duration of apnea was 6.8 0.6 min
(p < .001), and the mean minimum saturation
observed ( standard deviation) was 91 1 per
cent (p < .0001). These results were independent
of whether preoxygenation was by spontaneous or
controlled ventilation. Thus, insufflation of oxygen
via nasopharyngeal cannula provides at least 10
min of adequate oxygenation in unintubated, deni-
trogenated, apneic patients whose airways are
unobstructed. By significantly delaying the onset
of hypoxia, this technique may be life-saving: Ex-
tra time is allowed to obtain control of the airway
in critical situations. On rare occasions in which
tracheal intubation is known to be very difficult or
respiration is already compromised, it may be ap-
propriate to institute prophylactic transtracheal jet
ventilation before inducing general anesthesia or
attempting conventional or fiberoptic tracheal in-
tubation with the patient either awake or anesthe-
tized.
8
32
b. INTUBATION
A curved open-phalange blade (Macintosh) is
usually preferred for laryngoscopy because it ap-
proximates the curvature of the tube and therefore
provides the largest possible area through which to
pass the tube. However, a straight (Miller) blade
may be a better choice in patients with overriding
teeth or an excessively anterior larynx.
Double-lumen endotracheal tubes with carinal
hooks are first inserted through the vocal cords
with the distal curve concave anteriorly (just like
a single-lumen tube) and the hook facing poste-
riorly. When the tip of the tube has passed the
vocal cords, the tube is rotated 180 degrees, so
that the hook passes anteriorly through the glottis.
After the tube tip and the hook pass the larynx, the
tube is rotated 90 degrees so that the tube tip is
curved toward and enters the appropriate bronchus
and the hook engages the carina.
The Robertshaw DLT is passed with the distal
curvature initially concave anteriorly (Fig. 9-12A)
(just like a single-lumen tube). After the tube tip
passes the larynx, and while anterior force on the
laryngoscope is continued, the stylette (if used) is
removed and the tube is carefully rotated 90 de-
grees (so that the distal curve is now concave
toward the appropriate side and the proximal curve
is concave anteriorly) to allow endobronchial in-
tubation on the appropriate side (Fig. 9-125).
Continued anterior force by the laryngoscope dur-
ing tube rotation prevents hypopharyngeal struc-
tures from falling in around the tube and interfer-
Passage of Left-Sided Double-Lumen Tube
Figure 9-12 This schematic diagram depicts the passage of the left-sided double-lumen tube in a supine patient. A, The tube is
held with the distal curvature concave anteriorly and the proximal curve concave to the right and in a plane parallel to the floor.
The tube is then inserted through the vocal cords until the left cuff passes the vocal cords. The stylet is then removed. B, The tube
is rotated 90 counterclockwise so that the distal curvature is concave anteriorly and the proximal curvature is concave to the left
and in a plane parallel to the floor. C. The tube is inserted until either a moderate resistance to further passage is encountered or the
end of the common molding of the two lumens is at the teeth. Both cuffs are then inflated, and both lungs are ventilated. Finally.
one side is clamped while the other side is ventilated and vice versa (see text for further explanation).
ing with a free 90-degree distal tube tip rotation.
Failure to obtain close to a 90-degree rotation of
the distal tube tip, while the proximal end does
rotate 90 degrees, will cause either a kink or twist
in the shaft of the tube and/or prevent the distal
end of the lumen from lying free in the main-stem
bronchus (i.e., not up against the bronchial wall).
After rotation, the tube is advanced until most of
it is inserted.
30
When the proper depth of insertion is defined as
the cephalad surface of the bronchial cuff being
immediately below the carinal bifurcation, the av-
erage depth of insertion for both male and female
patients 170 cm tall is 29 cm, and for each 10-cm
increase or decrease in height, average placement
depth is increased or decreased 1 cm (Fig. 9-13).
30
The correlation between depth of insertion and
height is highly significant (p < .0001) for both
male and female patients. Nevertheless, it should
be understood that the depth of DLT insertion at
any given height is still normally distributed (see
Fig. 9-13), and correct DLT position should al-
ways be confirmed fiberoptically after initial place-
ment.
If the proximal end of the common or two-
lumen binding mold is near or at the level of the
teeth (i.e., at a depth of 32-33 cm) in a normal-
size person and/or moderate resistance to further
passage is encountered, it usually means that the
tube has been pushed in too far and the tube has
been firmly seated in a distal bronchus (Fig. 9-
12C). Double-lumen endotracheal tubes may also
be passed successfully via tracheostomy, although
it should be remembered that the tracheal cuff may
be at the tracheal stoma or lie partly outside the
trachea in this situation.
33
34
A special tracheos-
tomy DLT has been designed for use with tra-
cheostomies (see Figure 9-7).
20
C. ROUTI NE CHECKI NG OF DOUBLE-LUMEN
TUBE POSI TI ON
Once the tube tip is thought to be in an endo-
bronchial position, the following steps are carried
out to ensure proper functioning of the tube. First,
inflate the tracheal and endobronchial cuffs until
moderate tension is palpated in the external pilot
balloons. The endobronchial cuff should not re-
quire more than 2 to 3 ml of air; if it does, the cuff
Figure 9-13 The depth of insertion for left double-lumen tubes for all patients and for three grouped intervals based on patient
height. At each grouped interval, the depth of insertion was normally distributed. (From Brodsky JB, Benumof JL, Ehrenwerth J,
Ozaki GT: Depth of placement of left double-lumen endobronchial tubes. Anesth Analg 73:570-572, 1991. Used with permission.)
may be herniating out of the main-stem bronchus
(which is, by far, the most likely possibility) or the
main-stem bronchus is malotic, or the DLT size is
inappropriately small. Later, after one has com-
pleted fiberoptic confirmation of proper DLT po-
sition (see later discussion), the tracheal lumen can
be clamped and the endobronchial cuff deflated.
Then, while ventilating through only the endo-
bronchial lumen, its cuff can be slowly inflated
until the minimum amount of air needed to prevent
an air leak is determined (which is the just-seal
volume of air). Other more quantitative methods
of determining the just-seal volume of cuff air are
the positive-pressure ventilation/air bubble under-
water technique (see section III.H.), negative circle
system pressure/collapse of reservoir bag, and use
of capnography (see Figs. 9-39 to 9-41 and Han-
nallah,
35
Hannallah and Benumof,
36
and Essig and
Freeman
37
). Determination of an air leak when the
cuff is deflated is important because it rules out
the possibility that the DLT is too tightly impacted
in the bronchus.
Several positive-pressure ventilations should be
delivered and the chest auscultated and observed
bilaterally to determine that the trachea, rather than
the esophagus, has been intubated and that both
lungs are being ventilated (Fig. 9-12Q.
In addition to seeing the tube go through the
vocal cords, correct intubation position is checked
by feeling and observing the anesthesia reservoir
bag to make sure it has the appropriate compliance
and movement, maintaining normal pulse oxime-
try and end-tidal C0
2
values, and perhaps palpat-
ing the tracheal cuff in the neck. If only unilateral
breath sounds or chest movement are present, it is
likely that both of the lumens of the tube have
entered a main-stem bronchus (if both of the lu-
mens enter the left main-stem bronchus, the find-
ings may mimic an esophageal intubation and vice
versa). In this situation, quickly deflate the cuffs,
withdraw the tube 1 to 2 cm at a time, inflate the
cuffs, and reassess ventilation until bilateral breath
sounds are heard. If bilateral breath sounds are not
heard, and the tube has been withdrawn a signifi-
cant amount, the entire procedure must be re-
peated, beginning with establishing the airway and
oxygen ventilation via mask, laryngoscopy, and
reinsertion of the DLT through the vocal cords. If
bilateral breath sounds are present, then one side
is clamped, and breath sounds and chest move-
ment should disappear on the ipsilateral side and
remain on the contralateral side. Next, the clamped
side should be undamped, and the breath sounds
and chest movement should reappear on that side.
During unilateral clamping, the breath sounds on
the ventilated side should be compared with and
calibrated against unilateral chest wall movements
and the inspiratory disappearance and expiratory
appearance of respiratory gas moisture in the clear
tubing of the ventilated side (Fig. 9-14). In addi-
tion, the compliance of the lung should be gauged
using hand ventilation. The unilateral clamping
and unclamping should then be repeated on the
opposite side to ensure adequate lung separation
and cuff seal.
DLT adaptors have been described that permit
each lumen independently to be either open to the
mechanical ventilator or the atmosphere or com-
pletely blocked (as though it was clamped) by
simply turning a dial/stopcock to the desired set-
ting without the need for airway disconnection
and/or external clamping maneuvers.
38
39
With one
adaptor, the two lungs may receive any combina-
tion of PEEP, CPAP, FOB, and/or suctioning.
17
These features greatly facilitate the testing for lung
separation as well as use of all of the other one-
lung ventilation/anesthesia maneuvers that are de-
manded by modern anesthetic practice (see chapter
11).
In summary, when double-lumen endotracheal
tube position is correct, the breath sounds are nor-
mal and follow the expected unilateral pattern with
unilateral clamping, the chest rises and falls in
accordance with the breath sounds, the ventilated
lung feels reasonably compliant, no leaks are pres-
ent, and respiratory gas moisture appears and dis-
appears with each tidal ventilation (see Fig. 9-14).
When comparing both lumens during two-lung
ventilation and when changing from two-lung ven-
tilation to one-lung ventilation with a known con-
stant tidal volume, the exhaled tidal volume should
not decrease by more than 15 per cent, expiratory
flow rate from either lung should not slow mark-
edly, peak airway pressure should not increase by
more than 60 per cent, and there should not be an
obvious difference in the rate of appearance of the
fog in the two lumens during exhalation. However,
it should be realized that, in the presence of ad-
vanced lung disease, loss of lung tissue, or atelec-
tasis, more exaggerated changes in the just-men-
tioned variables are expected when switching from
two-lung ventilation to diseased-lung ventilation.
The auscultation findings with a DLT, whose
endobronchial cuff has gone past an upper lobe
orifice while still allowing ventilation of the upper
lobe by the tracheal lumen (see Fig. 9-27), may
closely mimic the findings expected for a properly
positioned DLT during two-lung ventilation.
40 4I
Conversely, when the double-lumen endotracheal
tube is malpositioned, any or all of the following
may occur: The breath sounds are poor and corre-
late poorly with unilateral clamping, the chest
movements do not follow the expected pattern, the
ventilated lung feels noncompliant, leaks are pres-
ent, or the respiratory gas moisture in the clear
tubing is relatively stationary. It is very important
Correct Position of Double-Lumen Tube
and Unilateral Clamping
Ipsilateral
Breath
Sounds
Disappear
Ipsilateral
Hemithorax
Does Not
Move
Contralateral
Breath Sounds
Remain and Have
the Expected Quality
Contralateral
Hemithorax
Rises and Falls
Ipsilateral
Respiratory Gas
Moisture is Stationary
Contralateral
Respiratory Gas Moisture
Disappears on Inhalation
and Reappears on
Exhalation
Breathing Bag
Has the Expected
Compliance for
Figure 9-14 This schematic dia-
gram shows the results of unilateral
clamping when the double-lumen tube
is in the correct position.
to realize, however, that even if the DLT is
thought to be properly positioned based on clinical
signs, subsequent FOB will reveal a 40 to 48 per
cent incidence of malpositioning.
42,43
(see section
III.D.). Obviously, the auscultation and fiberoptic
(see later discussion) findings can always be sup-
plemented by direct observation of the operative
lung and mediastinum when the chest is open.
It is controversial as to whether the endobron-
chial cuff should be left inflated or deflated after
confirmation of proper position of the DLT (see
section III.D.) Deflation of the cuff during two-
lung ventilation in the initial stages of surgery
(preparing the chest, draping, opening the chest
wall, etc.) minimizes the chance of pressure dam-
age to the bronchial mucosa. Alternatively, leaving
the cuff inflated (after readjustment to a just-
seal volume) prevents trans-main-stem bronchial
spread of secretions (pus, blood) and necrotic tu-
mor.
22
In cases in which migration of material
from the nondependent lung to the dependent lung
is a possibility, adjustment of the cuff to a just-
seal volume seems to minimize risk and maximize
benefit.
d. AUSCULTATION AND UNILATERAL CLAMPING
MANEUVERS TO DIAGNOSE DOUBLE-LUMEN
TUBE MALPOSITION
When it is felt that the double-lumen endotra-
cheal tube is malpositioned based on clinical signs,
it is theoretically possible to diagnose the malpo-
sition of the tube more precisely by a combination
of several unilateral clampings, chest auscultation,
and left endobronchial cuff inflation/deflation ma-
neuvers (Fig. 9-15). With reference to a left-sided
double-lumen endotracheal tube, there are three
possible gross malpositions: in too far on the left
(both lumens in the left main-stem bronchus), out
too far (both lumens in the trachea), and in or
down the right main-stem bronchus (at least the
left lumen is in the right main-stem bronchus).
When the right (tracheal) side is clamped (breath
sounds should be heard only over the left lung)
and the tube is in too far on the left side, breath
Double-Lumen Tube Malpositions
Procedure Breath Sounds Heard
Clamp Right Lumen
Both Cuffs Inflated
Clamp Left Lumen
Both Cuffs Inflated
Clamp Left Lumen
Deflate Left Cuff
Left
None or Very 11
Left
Left and Right
None or Very 11
Left and Right
Right
None or Very 11
Right
Figure 9-15 There are three major (involving a whole lung) malpositions of a left-sided double-lumen endotracheal tube. The
tube can be in too far on the left (both lumens are in the left main-stem bronchus), out too far (both lumens are in the trachea), or
down the right main-stem bronchus (at least the left lumen is in the right main-stem bronchus). In each of these three malpositions,
the left cuff, when fully inflated, can completely block the right lumen. Inflation and deflation of the left cuff while the left lumen
is clamped creates a breath sound differential diagnosis of tube malposition. (See text for full explanation.) (L = left; R = right:
1 = decreased.)
sounds will be heard only on the left side. When
the tube is out too far and the right side is
clamped, breath sounds will be heard bilaterally
(the tube needs to be advanced further). When the
tube is in or down the right side and the right
lumen is clamped, breath sounds will be heard
only on the right side (the tube needs to be pulled
back, rerotated, readvanced). When the left side is
clamped and the left endobronchial cuff is inflated,
the right lumen is blocked by the left cuff in all
three malpositions. Consequently, with the left
side clamped and the left cuff inflated, no, or very
diminished, breath sounds will be heard bilaterally
in all three malpositions, and there will be marked
resistance to airflow (the right lumen opens be-
tween two inflated cuffs). When the left side is
clamped and the left cuff is deflated, so that the
right lumen is no longer blocked by the left cuff,
breath sounds will be heard only on the left side
when the tube is in too far on the left (the tube
needs to be pulled back), breath sounds will be
heard bilaterally if the tube is out too far (tube
needs to be advanced), and breath sounds will be
heard only on the right side when the tube is in
the right side (tube needs to be pulled back, rero-
tated, and readvanced). The left-cuff infla-
tion/deflation findings provide the key diagnostic
data because they essentially define the position of
the right tracheal lumen by blocking and unblock-
ing it with the left cuff. Another possible diagnos-
tic sequence is presented in Figure 9-1 .
44
There are, however, several situations in which
unilateral clamping, auscultation, and cuff inflation
and deflation maneuvers for determining the integ-
rity of lung separation are either unreliable or im-
possible. First, and most importantly, when the
patient is in the lateral decubitus position, has had
a skin preparation, and is draped, access to the
chest wall is impossible, and the anesthesiologist
cannot listen to the chest. Second, the presence of
unilateral or bilateral lung disease, which either
existed before anesthesia and surgery or was in-
duced by anesthesia, may markedly obscure the
crispness of the chest auscultation end points.
Pulmonary Surgery
Figure 9-16 Decision tree outlin-
ing a sequence of cuff inflation and
positive-pressure ventilation through
a left-sided double-lumen tube that
will ensure proper functioning of the
tube. (From Katz J A, Fairley HB:
Pulmonary Surgery. In Marshall BE,
Longnecker DF, Fairley HB (eds):
Anesthesia for Thoracic Procedures.
Boston, Blackwell Scientific, 1988,
pp 363^413. Used with permission.)
Third, the diagnosis of exactly where the double-
lumen endotracheal tube is located may be con-
fused when the tube is just slightly malpositioned.
Fourth, the tube may have moved because of some
event such as coughing, turning into the lateral
decubitus position, and tracheal manipulation and
hilar retraction by the surgeon. Finally, some com-
bination of these circumstances may culminate in
uncertainty as to where the DLT is located. The
solution to any uncertainty as to the exact position
of the DLT is to determine the position by use of
FOB (see Use of Fiberoptic Bronchoscope to De-
termine Precise Double-Lumen Tube Position).
e. NONFI BEROPTI C METHOD FOR I NTUBATI ON
OF LEFT MAIN-STEM BRONCHUS WHEN A LEFT-
SIDED TUBE WILL NOT PASS ROUTINELY
If a left-sided tube locates in the right main-
stem bronchus and repeated attempts at correct
placement are unsuccessful, rotating the patient's
head and neck to the right before rotating and
advancing the tube may result in proper lateraliza-
tion of the left-sided double-lumen endobronchial
tube.
45
Bronchoscopists have long recognized the
increased difficulty of inserting a rigid broncho-
scope into the left main-stem bronchus because of
the angle it makes with the trachea and because,
in 74 per cent of patients, its orifice is partly cov-
ered with the tracheal carina.
46
The technique recommended for passage of a
rigid bronchoscope into the left main-stem bron-
chus is that the patient's head, face, and neck are
turned to the right. Kubota et al.
47
reported their
experience of selective blind left endobronchial
intubations using a single-lumen endotracheal tube
in 300 adults. The highest success rate (275/300,
or 92 per cent) was achieved when the tube was
rotated 180 degrees (so that the bevel faced toward
the right) and the patient's head was turned to the
right. When the head was not rotated to the right,
the success rate was only 61 per cent (182/300).
The difference in success rates was statistically
significant (p < .01). Turning the head and neck
to the right improves the success of tube placement
most likely because it shifts the larynx to the right
in relation to the carina. This tends to bring the
axis of the left main-stem bronchus more into line
with that of the trachea (i.e., the bronchus would
arise at a smaller angle), and the endobronchial
tube would have a straighter line to the left main-
stem bronchus.
45
It is also possible that turning the
head stretches the trachea and left main-stem bron-
chus, thereby altering the anatomy of the origin of
the left main-stem bronchus to make it wider or
less slit-like, either way rendering it more recep-
tive to the passage of the bronchial catheter of a
left-sided double-lumen endobronchial tube.
C. Use of Fiberoptic Bronchoscope to
Insert the Bronchial Lumen of a Double-
Lumen Tube Into a Main-Stem
Bronchus
The insertion of the bronchial lumen of a DLT
into the appropriate main-stem bronchus may be
aided by FOB (Fig. 9-17). The procedure is indi-
cated whenever the endobronchial lumen will not
locate in the appropriate main-stem bronchus by
blind insertion (e.g., ventilation through only the
endobronchial lumen results in breath sounds over
the wrong lung field or side), when the endobron-
chial lumen needs to be maneuvered past some
sort of disease (e.g., stenosis in the left main-stem
bronchus caused by thoracic aortic aneurysm),
26
or
when intubation is known to be very difficult and
fiberoptic DLT intubation while the patient is
awake (or anesthetized) is indicated
48
49
(as it
would be for a single-lumen tube; of course, the
awake patient requires topical and nerve block an-
esthesia).
48
To use a fiberoptic bronchoscope to insert and
precisely position DLTs and bronchial blockers
successfully, one must be familiar with the normal
anatomy of the tracheobronchial tree.
50
The tra-
chea of an adult is 10 to 14 cm long and is com-
posed of 18 to 24 incomplete, horseshoe-shaped
cartilages that appear as ridges from the four
o'clock position to the eight o'clock position an-
teriorly. These are connected to one another pos-
teriorly by the trachealis muscle (from the four
o'clock position to the eight o'clock position pos-
teriorly). The right side is therefore at three
o'clock and the left side is at nine o'clock. The
distal end of the trachea is displaced to the right
by the aortic arch. A prominent aortic arch and
enlarged paratracheal and bronchial lymph nodes
can cause compression, narrowing, and displace-
ment of the trachea and main-stem bronchi. When
the patient is supine, the carina lies at the level of
the fourth vertebra.
The right main-stem bronchus has a very varia-
ble length with an average of 15 to 18 cm (female
vs. male) before the right upper lobe bronchus
branches off. The right upper lobe bronchus arises
from the lateral aspect (three-o'clock position) of
the main-stem bronchus and divides into three seg-
mental bronchi: the apical, posterior, and anterior
branches. The right middle-lobe bronchus arises
from the anterior wall of the intermediate bronchus
(12 o'clock position), about 2.4 cm beyond the
orifice of the upper lobe bronchus. It divides into
two segmental bronchi: the lateral and the medial
branches. The right lower lobe bronchus is a con-
tinuation of the right main-stem bronchus and
branches into five segmental bronchi.
The left main-stem bronchus is an average of 44
to 49 mm long (female vs. male). It is slightly
more narrow and has a more acute origin from the
carina than does the right main-stem bronchus.
The "secondary carina" lies at the division of the
left upper and left lower lobe bronchi. Although it
may occupy any position between the horizontal
and vertical axes, it usually lies obliquely between
the eight-o'clock and the two-o'clock positions.
The left upper lobe bronchus divides into an upper
division and the lingular bronchus. The upper di-
vision constitutes the upper lobe bronchus and di-
vides into two branches: the apical posterior and
anterior segments. The lingular bronchus divides
into the superior and inferior lingular segments.
The left lower lobe bronchus is a continuation of
the left main-stem bronchus and branches into four
segmental bronchi.
The DLT is usually first placed in the trachea in
a conventional manner (laryngoscopy, manual
tube insertion) (unless the patient is awake and the
fiberscope is used from the outset) until the tra-
cheal cuff just passes the vocal cords. The tracheal
cuff is then inflated, and both lungs are ventilated
with both lumens (use the DLT as if it were a
single-lumen tube). A pediatric fiberoptic broncho-
scope can then be inserted into the bronchial lu-
men through a self-sealing diaphragm in the elbow
connector to the bronchial lumen (which permits
continued positive-pressure ventilation through
that lumen around the fiberoptic bronchoscope)
and passed into the appropriate main-stem
bronchus.
48
50
The tracheal cuff is then deflated,
and the bronchial lumen is passed over the fiber-
optic bronchoscope stylet into the appropriate
main-stem bronchus.
With respect to placing a right-sided DLT cor-
rectly (i.e., getting the right upper lobe ventilation
Use of Fiberoptic Bronchoscope to
Insert Left-Sided Double-Lumen Tube
Pass Fiberoptic Bronchoscope
Down Left Lumen Into
Left Main Stem Bronchus
Push Double-Lumen Tube in
Over Fiberoptic Bronchoscope
Until Left Lumen is in
Left Main Stem Bronchus
Insert Double-Lumen
Tube Into Trachea in
Conventional Manner and
Ventilate Both Lungs
Figure 9-17 This schematic diagram portrays use of the fiberoptic bronchoscope to insert a left-sided double-lumen tube. The
double-lumen tube can be put into the trachea in a conventional manner, and both lungs can be ventilated by both lumens (A). The
fiberoptic bronchoscope may be inserted into the left lumen of the double-lumen tube through a self-sealing diaphragm in the elbow
connector to the left lumen; this allows continued positive-pressure ventilation of both lungs through the right lumen without
creating a leak. After the fiberoptic bronchoscope has been passed into the left main-stem bronchus (B), it is used as a stylet for the
aftercoming left lumen (C). The fiberoptic bronchoscope is then withdrawn. Final precise positioning of the double-lumen tube is
performed with the fiberoptic bronchoscope in the right lumen (see Fig. 9-18).
slot opposite the right upper lobe bronchial ori-
fice), the right upper lobe bronchial orifice may
first be identified fiberoptically and then, while
holding the fiberoptic bronchoscope steady, the en-
dobronchial lumen may be passed over the fiber-
optic bronchoscope into the right main-stem bron-
chus (the tip of the fiberoptic bronchoscope must
be returned to a neutral position during passage of
the DLT over the fiberoptic bronchoscope) until
the tip of the fiberoptic bronchoscope can look-
through the right upper lobe ventilation slot into
the right upper lobe bronchial orifice (Fig. 9-215).
The fiberoptic bronchoscope is then withdrawn
from the bronchial lumen to determine the precise
DLT position (see the following section).
Alternatively, once the DLT is in the trachea,
the fiberoptic bronchoscope can be inserted into
the tracheal lumen through a self-sealing dia-
phragm in the elbow connector to the tracheal lu-
men (which permits continued positive-pressure
ventilation through that lumen around the fiberop-
tic bronchoscope) and passed just proximal to the
tracheal carina.
51
While the carina and the two
main-stem bronchial orifices are in view, the DLT
can be advanced and the degree of lateral rotation
adjusted so that the left lumen enters the left main-
stem bronchus. Final precise positioning (see the
following section) can be done with the fiberoptic
bronchoscope remaining in the tracheal lumen.
D. Use of Fiberoptic Bronchoscope to
Determine Precise Double-Lumen Tube
Position
The deleterious consequences of a malposi-
tioned DLT can be great, even life-threatening.
With almost all DLT malpositions, gas exchange
can be significantly/profoundly impaired, the non-
operative lung can be difficult/impossible to venti-
late, and the operative lung may not collapse on
the initiation of one-lung ventilation. In addition,
failure to separate the lungs in some specific situ-
ations may result in additional catastrophes such
as flooding both lungs with blood, pus, or lavage
fluid and tension pneumothorax.
There is a very high incidence of malpositioned
DLTs, as determined by FOB, when the DLT is
inserted blindly (i.e., without a fiberoptic broncho-
scope). In one study, 48 per cent of all DLTs were
found to be malpositioned.
52
In another study, 83
per cent of disposable right-sided DLTs were mal-
positioned.
53
In a third study, FOB after ausculta-
tion (which led the physician to believe the DLT
was in optimal position) resulted in repositioning
of 78 per cent of left-sided DLTs and 83 per cent
of right-sided DLTs.
54
In a fourth study, 38 per
cent of all DLTs were malpositioned.
55
In a fifth
study, 44 per cent of disposable tubes required
readjustment using the fiberoptic bronchoscope
during the initial intubation, and 30 per cent of
disposable tubes required readjustment using the
fiberoptic bronchoscope during the operation.
56
The authors of the last three studies concluded that
"auscultation is an unreliable method of confirm-
ing the position of DLTs and should be followed
by fiberoptic bronchoscopy";
54
"because auscul-
tation for tube position is unreliable, broncho-
scopic assessment of final position should be per-
formed in every instance";
55
and "in certain
situations the fiberoptic bronchoscope may have
been life-saving."
56
Given the high incidence of malpositioned
DLTs when DLT position is determined by only
auscultation (i.e., "blindly") and the potentially
serious consequences associated with a malposi-
tioned DLT, it is only a matter of simple common
sense to use an FOB routinely to determine easily,
quickly, and precisely the position of the DLT. In
the studies just quoted, it is open to conjecture and
speculation as to what would have happened if the
position of the DLT was not corrected with the aid
of FOB. However, one study has shown that, when
the position of the DLT is checked only by clinical
signs, there will be intraoperative problems 25 per
cent of the time with either deflating the nonde-
pendent lung, ventilating the dependent lung, or
completely separating the two lungs.
57
If one accepts the premise that the position of a
DLT should be confirmed by FOB, then I also
contend that it should be done first in the supine
position, again after the patient has been turned
into the lateral decubitus position, and again when-
ever there is a question about the DLT position (in
conjunction with other maneuvers such as palpa-
tion of the lumen tips by the surgeon, observation
of the operative lung, mediastinum, etc.). There
are several reasons for repeating FOB. First, when
the patient is supine, access to the patient's head
and DLT is optimal, the orientation of the patient
and FOB is most certain, and correlation of the
auscultatory findings is easiest (by whatever uni-
lateral clamping and auscultation technique one
prefers)
44
58
-
70
(Fig. 9-15) with the FOB findings.
The main value of FOB at this time (supine posi-
tion), given that all auscultation algorithms can
rule in or rule out cannulation of the wrong main-
stem bronchus or failure to cannulate either main-
stem bronchus, is to rule in or out upper lobe
obstruction (by determining endobronchial cuff
position or by visualizing the upper lobe bronchial
orifice). I do not believe one can accurately diag-
nose upper lobe obstruction by auscultation alone
because breath sounds are transmitted from the
ipsilateral lower lobe and across the mediastinum
from the contralateral lung. Second, if the DLT is
properly positioned in the supine position with cer-
tainty (i.e., with FOB confirmation), then the
second and more important confirmation of DLT
position after the patient is in the lateral position
is greatly facilitated by the first look in the supine
position. In addition, events that take place be-
tween the supine and lateral decubitus position
may affect the functioning of the DLT. For exam-
ple, placement of the axillary roll has caused al-
most complete obstruction of the right main-stem
bronchus in a 10-year-old child.
71
Similarly, third
and further intraoperative views are facilitated
by the second view in the lateral position. This
series/cascade of FOB views of DLT position
interact/overlap to give the anesthesiologist
much tighter control of vital respiratory functions
throughout the anesthetic with very little risk.
The exact position of a left-sided double-lumen
endotracheal tube can be ascertained at any time,
in less than a minute, by simply passing a pediatric
fiberoptic bronchoscope through the tracheal lu-
men of the DLT. It is rarely necessary also to have
to pass the fiberoptic bronchoscope down the left
endobronchial lumen. With reference to a left-
sided DLT, looking down the right tracheal lumen.
the endoscopist should see a clear, straight-ahead
view of the tracheal carina (the tracheal lumen
should be approximately 1-2 cm above the ca-
rina), the left lumen going off to the left, and the
upper surface of the left endobronchial balloon just
below the tracheal carina (Figs. 9-18 and 9-19).
(The importance of seeing the upper surface of the
left endobronchial cuff below the tracheal carina is
emphasized in the following section.) It is impor-
tant that the volume of air used to fill the left
endobronchial left cuff does not cause the endo-
bronchial cuff to herniate over the tracheal carina
or cause the tracheal carina to deviate to the right
(Figs. 9-19 and 9-20); both cuff herniation and
carinal deviation can be readily appreciated look-
ing down the tracheal lumen. Looking down the
left lumen (which is sometimes done when insert-
ing a left-sided DLT with a fiberoptic broncho-
scope [see section III.C] and in all cases of bron-
chopulmonary lavage in which perfect tube
position and tight cuff seal are extremely critical),
the endoscopist should see a very slight narrowing
of the left lumen (due to endobronchial cuff prs-
Use of Fiberoptic Bronchoscope
Down the Right Lumen to Determine
Precise Left-Sided Double-Lumen
Tube Position
Figure 918 This schematic diagram portrays use of the
fiberoptic bronchoscope down the right lumen to determine
precise left-sided double-lumen tube position. The endoscopist
should see a clear straight-ahead view of the tracheal carina,
the left lumen going off into the left main-stem bronchus, and,
most importantly (in bold print), the upper surface of the blue
left endobronchial cuff just below the tracheal carina.
sure) as well as the bronchial carina distal to the
end of the tube (see Fig. 9-19).
The endoscopist should not see excessive left
luminal narrowing (due to excessive left cuff pres-
sure) (see Fig. 9-20). Thus, aside from gross mal-
position, important undesirable findings on endos-
copy are related to excessive left cuff inflation and
pressure and consist of cuff herniation over the
tracheal carina, carinal deviation to the right (both
of which may block the right main-stem bronchial
orifice and impair right lung ventilation), and ex-
cessive left lumen constriction (invagination),
which may impair left lung ventilation (Fig. 9-
20).
61
In addition, when an inappropriately under-
sized tube is used, the large endobronchial cuff
volume required for endobronchial cuff seal tends
to force the entire DLT cephalad, making a func-
tional bronchial seal more difficult.
62
With reference to a right-sided DLT, looking
down the left (tracheal) lumen the endoscopist
should see a clear straight-ahead view of the tra-
cheal carina, with the right lumen going off to the
right (Fig. 9-21). The upper surface of the right
endobronchial balloon may not be visible below
the tracheal carina. Looking down the right lumen,
the endoscopist should see the right middle-lower
lobe bronchial carina distal to the end of the tube.
Most importantly, the endoscopist should locate
the right upper lobe ventilation slot and be able to
look directly into the right upper lobe orifice
through the right upper lobe ventilation slot by
simply flexing the tip of the fiberoptic broncho-
scope laterally and superiorly (Fig. 9-21B). There
Use of Fiberoptic Bronchoscope to Determine
Precise Left-Sided Double-Lumen Tube Position
View Down Left (Bronchial) Lumen View Down Right (Tracheal) Lumen
Figure 9-19 This schematic diagram depicts the complete fiberoptic bronchoscopy picture of left-sided double-lumen tubes (both
the desired view and the view to be avoided from both of the lumens). When the bronchoscope is passed down the right lumen of
the left-sided tube, the endoscopist should see a clear straight-ahead view of the tracheal carina and the upper surface of the blue
left endobronchial cuff just below the tracheal carina. Excessive pressure in the endobronchial cuff, as manifested by tracheal
carinal deviation to the right and herniation of the endobronchial cuff over the carina, should be avoided. When the bronchoscope
is passed down the left lumen of the left-sided tube, the endoscopist should see a very slight left luminal narrowing and a clear
straight-ahead view of the bronchial carina off in the distance. Excessive left luminal narrowing should be avoided.
.
Excessive Left Cuff Inflation: Problems
Figure 9-20 Excessive inflation of the left
cuff of a left-sided double-lumen tube can
cause impaired ventilation of both the right
and left lungs. Right-lung ventilation may be
impaired by left-cuff herniation over the tra-
cheal carina (as a result of excessive left-cuff
volume) and by tracheal carinal deviation to
the right (because of excessive left-cuff pres-
sure). Left-lung ventilation may be impaired
by invagination of the left lumen caused by
excessive left-cuff pressure.
Use of Fiberoptic Bronchoscope to Determine
Precise Right-Sided Double-Lumen Tube Position
Right Lumen
Going off to
the Right
Right Upper
Lobe Bronchial
Orifice
View Down Left
(Tracheal) Lumen
View Down Right
(Bronchial) Lumen
Figure 9-21 This schematic diagram portrays use of a fiberoptic bronchoscope to determine precise right-sided double-lumen
tube position. A, When the fiberoptic bronchoscope is passed down the left (tracheal) lumen, the endoscopist should see a clear
straight-ahead view of the tracheal carina and the right lumen going off into the right main-stem bronchus. B, When the fiberoptic
bronchoscope is passed down the right (bronchial) lumen, the endoscopist should see the bronchial carina off in the distance; when
the fiberoptic bronchoscope is flexed laterally and cephalad and passed through the right upper lobe ventilation slot, the right upper
lobe bronchial orifice should be visualized.
should be no overriding of the right upper lobe
ventilation slot on the bronchial mucosa, and the
bronchial mucosa should not be covering any of
the right upper lobe ventilation slot. The fact that
there is little room for error in aligning the right
upper lobe ventilation slot with the right upper
lobe orifice is emphasized in the following section.
The right upper lobe ventilation slot may be
easily located by first finding the bronchial lumen
radiopaque marker. The radiopaque marker ap-
pears as either a white or black line on the inside
of the bronchial lumen, which ends at the proximal
end of the right upper lobe ventilation slot. If one
simply follows the radiopaque line, it will lead the
endoscopist to the right upper lobe ventilation slot.
In my experience, in eight of 10 cases, the clin-
ical signs (breath sounds, chest movements, com-
pliance of the lung[s], movement of respiratory gas
moisture) indicate that the lungs are apparently
clearly and without doubt completely separated
when the DLT is first inserted with the patient in
the supine position. However, in view of the find-
ing by many authors
52-56
that between 40 and 50
per cent of DLTs are malpositioned to some extent
in the supine position, even though the clinical
signs may indicate there is no problem, it is
strongly advisable to check the position of the tube
with a fiberoptic bronchoscope in the supine posi-
tion (especially considering that the procedure
takes less than a minute). Even if no problem is
identified, the procedure still allows the endosco-
pist to become familiar with the patient's anatomy
and facilitates the more important endoscopy per-
formed after turning the patient into the lateral
decubitus position. In approximately two of 10
cases, there is doubt about tube location in the
supine position, and in these patients the fiberoptic
bronchoscope is always used to correct the DLT
malposition. The fiberoptic bronchoscope is al-
ways used to determine DLT position after the
patient has been turned into the lateral decubitus
position. Of course, a determined effort is made to
prevent dislodgment of the tube during turning by
holding onto the tube at the level of the incisors
and by keeping the head absolutely immobile in a
neutral or slightly flexed position. Head extension
can cause movement of the tube in a cephalad
direction, which may result in bronchial decannu-
lation; head flexion can cause movement of the
tube in a caudad direction, which may result in an
upper lobe obstruction or in both lumens being in
a main-stem bronchus (see the next section).
63
64
Finally, the fiberoptic bronchoscope is used any-
time during the procedure when there is a question
about DLT position. This is not an infrequent oc-
currence and is usually caused by surgical manip-
ulation and traction on the hilum, carina, or tra-
chea.
There are some risks to the routine use of FOB
to determine DLT position. First, depending on the
relative sizes of the FOB and the DLT, the use of
a self-sealing diaphragm in the elbow connector,
and the position of DLT, ventilation may (need to)
be interrupted during use of the FOB. Second, if
the FOB is inserted through a self-sealing dia-
phragm in the elbow connector to the side being
visualized, one needs to be cognizant of the con-
tinuous presence and status of the self-sealing dia-
phragm (i.e., a torn fragment is not carried/pushed
into the tracheobronchial tree). Third, poor FOB
technique may injure the mucosa of the tracheo-
bronchial tree. Fourth, with poor cleaning tech-
nique, cross infection or direct mucosal injury
(e.g., by Cidex) is possible. Finally, erroneous
interpretation of the view is always possible. How-
ever, all of these complications are avoidable with
proper education, forethought, protocols, and ex-
perience. Use of FOB, especially while the patient
is being ventilated or has been hyperventilated on
100 per cent oxygen, should be no more distract-
ing than inserting a central venous catheter, naso-
gastric tube, and so on, and should also be no more
time consuming than the multiple, and often con-
fusing, unilateral clamping and declamping/aus-
cultation periods that are usually required when
the DLT is not in the right position.
1. Margin of Safety in Positioning
Double-Lumen Tubes
A correctly positioned DLT does not cause ob-
struction of any conducting airway. This section
discusses the length of adult male and female tra-
cheobronchial trees, over which differently sized
and manufactured DLTs can be moved and still be
correctly positioned: This length is defined as the
margin of safety.
65
Because left- and right-sided
DLTs are constructed differently and the left and
right main-stem bronchi have different anatomy,
the definitions of positioning left and right DLTs
are unique.
a. LEFT-SIDED DOUBLE-LUMEN TUBES
(1) DEFINITION OF MARGIN OF SAFETY IN POSI-
TIONING LEFT-SIDED DOUBLE-LUMEN TUBES. The
left-sided tubes made by different manufacturers
have the same basic structural design and, there-
fore, can be considered together. The most proxi-
mal acceptable position of a left-sided tube is
when the left endobronchial cuff is just below the
tracheal carina (Fig. 9-22, top left panel). If the
endobronchial cuff is placed in a progressively
more proximal position, then the endobronchial
cuff would progressively fill the space above the
carina and obstruct the trachea and contralateral
(right) main-stem bronchus. In addition, the posi-
The Margin of Safety (MS) in Positioning
Figure 9-22 This schematic
shows the definitions of most
proximal and most distal accept-
able positions of left- and right-
sided double-lumen tubes and the
margin of safety in positioning
these double-lumen tubes. Top
panel = all left-sided double-lu-
men tubes; middle panel = Mal-
linkrodt right-sided double-lu-
men tube; bottom panel = Rusch
right-sided double-lumen tube;
LMS = length left main-stem
bronchus; RMS = length right
main-stem bronchus; MS = mar-
gin of safety in positioning dou-
ble-lumen tube; LUL = left up-
per lobe; RUL = right upper
lobe. (From Benumof JL, Par-
tridge BL, Salvatierra C, Keating
J: Margin of safety in positioning
modern double-lumen endotra-
cheal tubes. Anesthesiology
67:729-738, 1987. Used with
permission.)
tive-pressure gas and fluid seal between the two
lungs would be lost.
The most distal acceptable position of a left-
sided tube is when the tip of the left lumen is at
the proximal edge of the left upper lobe bronchial
orifice (see Fig. 9-22, top right panel). If the tip
of the left lumen is placed in a progressively more
distal position, then the tip of the left lumen would
progressively obstruct the left upper lobe bronchial
orifice. This definition of the most distal accepta-
ble position for a left-sided tube is valid, provided
that two assumptions inherent in this definition are
correct. First, the distance between the right and
left lumen tips (see Table 9-5) must be greater
than the length of the left main-stem bronchus
(which it is); this means that the tip of the right
lumen will not enter the left main-stem bronchus
before the tip of the left lumen by-passes the left
upper lobe. Second, the outside diameter of the
left lumen tip of a left-sided tube must be nearly
equal to the internal diameter of the second-gen-
eration bronchus to the left lower lobe (which it
is); this means that there is no, or minimal, space
between the tip of the left lumen and the left lower
lobe bronchus for retrograde gas exchange be-
tween the left lumen and left upper lobe.
The length of tracheobronchial tree between the
most distal and most proximal acceptable position
is the margin of safety in positioning a left-sided
tube. This margin of safety is thus the length of
the left main-stem bronchus minus the distance
between the proximal margin of the left endobron-
chial cuff and the tip of the left lumen (see Fig. 9-
22, top panel).
(2) MEASUREMENT OF LENGTH OF LEFT MAIN
STEM BRONCHUS. Table 9-3 shows the averages
and ranges of left main-stem bronchial lengths for
three different measurement techniques (in vivo
FOB, autopsy study, lung cast study).
65
From
study method to study method, and for the separate
sex subgroups and the combined sex groups, the
results are strikingly similar. In all subgroups and
Table 9-3 MAIN-STEM BRONCHIAL LENGTHS FOR MALES, FEMALES, AND SEXES COMBINED IN
THE IN VIVO FIBEROPTIC BRONCHOSCOPY, AUTOPSY AND CAST STUDY GROUPS*
*From Benumof JL, Partridge BL, Salvatierra C, Keating J: Margin of safety in positioning modern double-lumen endotracheal
tubes. Anesthesiology 67:729-738.
tValues are mean standard deviation. Values in parentheses are ranges.
+One right upper lobe takeoff was above tracheal carina.
Abbreviations: RMSB = right main-stem bronchus; LMSB = left main-stem bronchus; = sex of the casts was unknown.
combined groups, the main-stem bronchial lengths
were approximately normally distributed. The in
vivo bronchoscopically determined lengths of left
main-stem bronchus for males and females (
Figure 9-23 In vivo fiberoptic bronchoscopy measured
length of the left main-stem bronchus (LMS) plotted as a func-
tion of the height of the patients. A weak but statistically
significant correlation was found. (From Benumof JL, Partridge
BL, Salvatierra C, Keating J: Margin of safety in positioning
modern double-lumen endotracheal tubes. Anesthesiology
67:729-738, 1987. Used with permission.)
standard deviation) are 50 8, and 45 7 mm,
respectively, with a slight clinically and statisti-
cally significant positive correlation with height
(Fig. 9-23).
65
Considering all groups, the lengths
of the left main-stem bronchus in males is 48 to
49 mm and in females 44 mm.
(3) MEASUREMENT OF MARGIN OF SAFETY FOR
LEFT-SIDED DOUBLE-LUMEN TUBES. The margin
of safety in positioning a left-sided DLT is the
difference between the length of the left main-stem
bronchus and the length of tube between the prox-
imal margin of the left endobronchial cuff and the
left lumen tip (Table 9-4). This average margin of
safety in positioning left-sided DLTs ranged from
16 to 19 mm.
DLTs with a carinal hook have the hook placed
approximately 10 mm proximal to the proximal
surface of the endobronchial cuff. Consequently,
the tip of the left lumen of a left-sided DLT with
a carinal hook enters the left main-stem bronchus
10 mm deeper than a tube without a carinal hook,
66
and the margin of safety is reduced by approxi-
mately 10 mm.
(4) CLINICAL EXAMPLES/IMPLICATIONS OF EX-
CEEDING MARGIN OF SAFETY FOR LEFT-SIDED
DOUBLE-LUMEN TUBES. Figure 9-24, left panel,
shows an example of a common situation using a
37 French clear plastic left-sided DLT in an aver-
age-sized female. With the left endobronchial cuff
placed just below the tracheal carina, and using
average values for DLT and left main-stem bron-
chus lengths, the margin of safety is 21 mm. How-
ever, because the distance between the right and
left lumen tips for the clear plastic tube (69 mm;
see Table 9-5) is longer than the length of the left
main-stem bronchus (50 mm) in this typical ex-
ample, it is possible for the right lumen to be
above the tracheal carina while the left lumen tip
obstructs the left upper lobe (see Fig. 9-24, right
panel). Figure 9-25 shows a postoperative chest
Table 9-4 AVERAGE MARGIN OF SAFETY IN POSITIONING A LEFT-SIDED DOUBLE-LUMEN TUBE
IS THE LENGTH OF THE LEFT MAIN-STEM BRONCHUS MINUS THE PROXIMAL MARGIN
OF LEFT CUFF TO LUMEN TIP LENGTH*
Size
Manufacturer 35-41 -Fr
Average Length of Left Main-Stem
Bronchus, mm
Male Female Combined
Proximal Margin of
Left Cuff to Left
Lumen Tip, mm
Average Margin of Safety, mm
Male Female Combined
Mallinkrodt
Rusch
Sheridan
All
AH
All
49
49
49
44
44
44
48
48
48
29t
32
30
20
17
19
15
12
14
19
16
18
*From Benumof JL, Partridge BL, Salvatierra C, Keating J: Margin of safety in positioning modern double-lumen endotracheal
tubes. Anesthesiology 67:729-738, 1987. Used with permission.
tManufacturer estimates a 2 mm (variation) tolerance for this value.
roentgenogram of a patient who underwent a right
upper lobectomy and demonstrates this problem:
A left-sided DLT tube is in situ, the right lumen
tip is above the tracheal carina, and the remaining
right lung is well aerated, while the left upper lobe
is partially collapsed. The same situation may oc-
cur when a left-sided DLT with a carinal hook is
used (because the hook is proximal to the proximal
surface of the left cuff) (i.e., the distance between
the left tip and hook exceeds the length of the left
main-stem bronchus).
66
In addition to significant variability in length of
the left main-stem bronchus, there is also fairly
significant variability (up to 20 per cent) between
individual left-sided DLTs of different manufac-
turers and same-size and different-size DLTs of
the same manufacturer with respect to right and
left lumen tip and proximal endobronchial cuff to
37 French Left-Sided Double-Lumen Tube and Adult Female
Left Main Stem Bronchial Dimensions and Relationships
Figure 9-24 This schematic diagram shows the relationship of the lumen tips and cuffs of a correctly and an incorrectly
positioned 37 French left-sided double-lumen tube to the tracheobronchial tree of an average-sized adult female. When the left
endobronchial cuff is just below the tracheal carina (correct position), the margin of safety (MS) is 21 mm because the length of
the left main-stem bronchus (LMS) exceeds the distance between the cephalad surface of the left cuff and the tip of the left lumen
(distance A). However, because the distance between the right and left lumens (distance B) exceeds the length of the left main-stem
bronchus, it is possible to still have the right lumen above the tracheal carina while the left lumen tip and the left endobronchial
cuff obstruct the left upper lobe (LUL). Thus, as a left-sided tube is inserted further (so that the left cuff cannot be visualized below
the carina), the first airway likely to be obstructed will be the left upper lobe.
Figure 9-25 This x-ray shows an example of a patient with the problem schematically depicted in Figure 9-22. The right lumen
is above the tracheal carina, and the right lung is well ventilated, while the left lumen tip is deeply inserted into the left lung, and
the left upper lobe is opacified.
Table 9-5 LEFT TO RIGHT LUMEN TIP
DISTANCE FOR VARIOUS SIZED
AND MANUFACTURED DOUBLE-
LUMEN TUBES
Left-Sided
Double-Lumen Tube
Manufacturer
Mallinkrodt
broncho catheter
Carlens
Leyland
French
41
39
37
35
41
39
37
35
44
38
32
Left to Right Tip Distance
(mm)
70
70
69
66
67
65
64
60
90
82
72
left lumen tip lengths. The 20 per cent variation in
these distances between manufactures and sizes in
the clear plastic DLTs is due to slightly different
designs by the manufacturers and the fact that cut-
ting of the proximal lumen and placement of the
endobronchial cuff is done by hand at the end of
the manufacturing process.
The variation in DLT lengths (lumen to lumen
and left tip to proximal endobronchial cuff) and
left main-stem bronchial length greatly reinforces
the need to determine DLT location with a fiber-
optic bronchoscope. The variation in these lengths
has three more important implications. First, a pa-
tient with a small left main-stem bronchus will
have a smaller margin of safety. In fact, if a patient
with a small left main-stem bronchus (two stan-
dard deviations less than the mean presented in
Table 9-3) received a 37 French plastic tube that
had a large left tip to proximal left cuff distance
and a large lumen to lumen distance (20 per cent
greater than the mean value in Table 9-5), there
would be no margin of safety, even if the left
endobronchial cuff is positioned just below the
tracheal carina (Fig. 9-26); this situation has the
smallest margin of safety realistically possible. In-
deed, with a short left main-stem bronchus, it is
possible to have the left endobronchial cuff be-
yond the left upper lobe so that the left lower lobe
is ventilated by the left lumen while the left upper
lobe and right lung are ventilated by the right
lumen (Fig. 9-27).
40
41
Second, and conversely, if a patient with a long
left main-stem bronchus (two standard deviations
greater than the mean presented in Table 9-3) re-
ceived a 37 French plastic DLT with a short lumen
to lumen distance (20 per cent less than the mean
value in Table 9-5), the right lumen could be in
the left main-stem bronchus (right main-stem
bronchus obstructed) while the left lumen tip is
still above the left upper lobe (Fig. 9-28). How-
ever, it must be emphasized that no matter which
manufactured tube or sized tube is used and no
matter how long or short the left main-stem bron-
chus is (within the range of extremes observed in
the cadaver studies), when the upper surface of the
left endobronchial balloon is just below the tra-
cheal carina, it is not possible for the left lumen
tip to obstruct the left upper lobe or the right
(tracheal) lumen to be near a main-stem bronchus.
Third, in view of the fact that head flexion and
extension can move the tip of a left-sided DLT in
or out 27 mm,
64
these considerations show that it
is easily possible to cause left upper lobe obstruc-
tion with head flexion and bronchial decannulation
with head extension (Fig. 9-29).
Finally, larger, rather than smaller, sized left-
sided DLTs should be used. As tube size increases,
proximal margin of left cuff to left lumen tip
length (all manufacturers) remains constant. Al-
though the margin of safety remains constant with
increasing DLT size, airway resistance and diffi-
culty in secretion removal decrease. In view of the
positive correlation between patient height and
Example of No Margin of Safety (MS)
When Left Main Stem Bronchus (LMS)
is Short and Length A is Large
MS=0
Left Cuff Just
Below Tracheal
Carina
Figure 9-26 This schematic diagram shows that there will
be no margin of safety (MS) when a double-lumen tube is used
that has a large length A in a patient who has a short left main-
stem bronchus (LMS), even though the left endobronchial cuff
is positioned just below the tracheal carina. Nevertheless, with
the left endobronchial cuff just below the tracheal carina, the
left upper lobe is unobstructed.
Figure 9-27 Using a left double-lumen tube, after clamping the left (endobronchial) lumen, the left upper lobe and entire right
lung continued to be ventilated. Withdrawing the tube back 2 cm corrected the problem. (From Brodsky JB, Shulman MS, Mark
JBD: Malposition of left-sided double-lumen endobronchial tubes. Anesthesiology 62:667-669, 1985. Used with permission.)
main-stem bronchial length, and with the provision
that an intubation can be atraumatically performed,
use of 39 and 41 French tubes in patients who are
68 inches (170 cm) or taller and 35 to 39 French
tubes in patients who are shorter appears indicated.
b. RIGHT-SIDED DOUBLE-LUMEN TUBES
(1) DEFINITION OF MARGIN OF SAFETY IN POSI-
TIONING RIGHT-SIDED DOUBLE-LUMEN TUBES.
The Mallinkrodt, Sheridan, Rusch, and Leyland
right-sided DLTs are designed differently (the
shapes of the right endobronchial cuffs are very
different); therefore, they must be considered sep-
arately. Although the Sheridan double-balloon
configuration of the right-sided tube is unique, the
positioning margin of safety is similar to that of
the Mallinkrodt right-sided DLT. The most proxi-
mal acceptable position of a Mallinkrodt right-
sided tube is when the endobronchial cuff is just
below the tracheal carina (Fig. 9-22, middle left
panel). If the endobronchial cuff is placed in a
progressively more proximal position, then the
right endobronchial cuff would progressively fill
the space above the carina and obstruct the trachea
and contralateral (left) main-stem bronchus. In ad-
dition, the positive-pressure gas and fluid seal be-
tween the two lungs would be lost.
The most distal acceptable position of the Mal-
linkrodt right-sided tube is when the distal margin
of the right endobronchial cuff is at the proximal
margin of the right upper lobe bronchial orifice
(see Fig. 9-22, middle right panel). If the Mallin-
krodt (or Sheridan) right endobronchial cuff is
placed in a progressively more distal position, then
the Mallinkrodt right endobronchial cuff would
progressively obstruct the right upper lobe bron-
chial orifice. This definition of the most distal ac-
ceptable position for a Mallinkrodt (or Sheridan)
right-sided tube is valid because of the unique
shape of the Mallinkrodt (or Sheridan) right endo-
bronchial cuff (which forces the right upper lobe
ventilation slot to ride off the right main-stem
bronchial wall), which allows gas exchange be-
tween the right upper lobe ventilation slot, right
lumen tip (for the Mallinkrodt tube), and the right
upper lobe (even though the right upper lobe ven-
Example of Right Lumen Tip Being in
Left Main Stem Bronchus when
Left Lumen Tip is Above Left Upper Lobe
When Left Main Stem Bronchus (LMS)
is Long and Length is Short
Figure 9-28 This schematic diagram shows that it is possi-
ble for the right lumen to be in the left main-stem (LMS)
bronchus when the left lumen tip is still above the left upper
lobe if the left main-stem bronchus is long and length is
short.
tilation slot may not be properly aligned with the
right upper lobe bronchial orifice). The margin of
safety in positioning the Mallinkrodt right-sided
tube is, therefore, the length of the right main-stem
bronchus minus the distance between the proximal
and distal margins of the right endobronchial cuff
(i.e., the length of the right endobronchial cuff)
(see Fig. 9-22, middle panel).
The most proximal acceptable position of the
Rusch and Leyland right-sided tube is when the
distal margin of the right endobronchial cuff is at
the distal margin of the right upper lobe bronchial
orifice (see Fig. 9-22, bottom panel). If the Rusch
or Leyland right-sided tube is placed in a progres-
sively more proximal position, then the distal
margin of the endobronchial cuff and the outside
lateral wall of the right lumen tip would progres-
sively obstruct the right upper lobe bronchial ori-
fice. The most distal acceptable position of a
Rusch or Leyland right-sided tube is when the
proximal margin of the right endobronchial cuff is
at the proximal margin of the right upper lobe
bronchial orifice. If the Rusch or Leyland right-
sided tube is placed in a progressively more distal
position, then the proximal margin of the right
endobronchial cuff and outside lateral wall of the
right endobronchial lumen would progressively
obstruct the right upper lobe bronchial orifice.
These definitions of the most proximal and most
distal acceptable position for a Rusch and Leyland
right-sided tube are valid because the shape of the
Rusch and Leyland right endobronchial cuff forces
the right upper lobe ventilation slot and outside
lateral wall of the right endobronchial tube to be
in apposition to the right upper lobe bronchial ori-
fice and lateral bronchial mucosa. The margin of
safety in positioning the Rusch and Leyland right-
sided tube is, therefore, the distance between the
distal and proximal margins of the right endobron-
chial cuff (i.e., the length of the right upper lobe
ventilation slot) minus the diameter of the right
upper lobe bronchial orifice (see Fig. 9-22, bottom
panel).
(2) MEASUREMENT OF LENGTH OF RIGHT
MAIN-STEM BRONCHUS AND RIGHT UPPER LOBE
BRONCHIAL ORIFICE. Table 9-3 shows the aver-
ages and ranges of right main-stem bronchial
lengths for three different measurement techniques
(in vivo FOB, autopsy study, cast study).
65
From
study method to study method and for the separate
sex subgroups and the combined sex groups, the
results are very similar. In all subgroups and com-
bined groups, the main-stem bronchial lengths are
approximately normally distributed. Considering
all groups, the length of the right main-stem bron-
chus is 18 to 19 mm in males and 13 to 14 mm in
females. The average diameters ( standard de-
viation) of the right and left upper lobe bronchial
orifices in the lung casts were 10.7 2.1 mm and
10.3 1.9 mm, respectively, with a range of
values of 6 to 18 mm. The distribution of values
was approximately normal. These findings are
nearly identical to those of a previous and more
comprehensive, but not as quantitative, study (Fig.
9-30).
67
(3) MEASUREMENT OF MARGIN OF SAFETY FOR
RIGHT-SIDED DOUBLE-LUMEN TUBES. The margin
of safety in positioning right-sided tubes is shown
in Table 9-6. The margin of safety ranged from 1
to 10 mm and differed between manufacturers.
The Leyland right-sided DLT has a greater margin
of safety than the Rusch (or Mallinkrodt) right-
sided DLT because it has a long (21 mm) right
upper lobe ventilation slot. However, it should be
remembered that the unique slanted doughnut
shape of the Mallinkrodt right endobronchial cuff
of the right-sided DLT allows the right upper lobe
ventilation slot to ride off the right upper lobe
orifice, thereby increasing (from 1 mm) to an un-
known extent the margin of safety in positioning
362 Separation of the Two Lungs (Double-Lumen Tube and Bronchial Blocker intubation)
Left Double Lumen Tube Position and
Head Flexion and Extension
Figure 929 Head flexion moves an endotracheal tube inward, and head extension moves an endotracheal tube outward. A,
Correct position of a left-sided double-lumen tube along with average values (in mm) for left main-stem bronchus, right to left
lumen tip, and left lumen tip to left upper lobe lengths; the latter length is the margin of safety. When the left cuff is just below the
tracheal carina, the margin of safety is 25 mm. B, Extreme head flexion can cause left upper lobe obstruction. C, Extreme head
extension can cause left main-stem bronchial decannulation. Data from Saito, Dohi, and Naito.
62
Figure 9-30 Mean tracheobronchial tree diameter (in cm)
Note that the diameter of the right upper lobe bronchial orifice
is 1 cm. (From Merendino KA, Keriluk LB: Human measure-
ments involved in tracheobronchial resection and reconstruc-
tion procedures. Surgery 35:590-597, 1954. Used with permis-
sion.)
this particular right-sided tube. The margin of
safety in positioning right-sided DLTs may be
negative if the right main-stem bronchus is less
than 10 mm (1:6 normal patients)
68
or if the right
Table 9-6 AVERAGE MARGIN OF SAFETY IN POSITIONING A MALLINKRODT RIGHT-SIDED
DOUBLE-LUMEN TUBE IS THE LENGTH OF THE RIGHT MAIN-STEM BRONCHUS MINUS
THE WIDTH OF THE RIGHT ENDOBRONCHIAL CUFF, AND THE MARGIN OF SAFETY IN
POSITIONING A RUSCH RIGHT-SIDED DOUBLE-LUMEN TUBE IS THE LENGTH OF THE
RIGHT UPPER LOBE VENTILATION SLOT MINUS THE DIAMETER OF THE RIGHT
UPPER LOBE BRONCHIAL ORIFICE*
upper lobe takes off from the trachea (1:250 in
normal patients
68
and 1:50 patients with congenital
heart disease
69
).
( 4 ) CLINICAL EXAMPLES/IMPLICATIONS OF EX-
CEEDING MARGIN OF SAFETY FOR RIGHT-SIDED
DOUBLE-LUMEN TUBES. This analysis of right-
sided DLT margin of safety has five clinical impli-
cations. First, with a right main-stem bronchus less
than 10 mm long or when the right upper lobe has
a tracheal takeoff, it will be impossible to position
a right-sided DLT without obstructing the right
upper lobe. Indeed, if the right upper lobe has a
low tracheal origin and a left-sided DLT is inserted
too deeply, the tracheal cuff may obstruct the right
upper lobe and the left lumen/cuff may obstruct
the left upper lobe (Fig. 9-31).
70
Second, a clinical
prospective study confirms that the Leyland right-
sided DLT has a greater positioning margin of
safety than the Mallinkrodt tube.
53
In 18 of 20
patients intubated with a Leyland right-sided DLT,
compared with only one of nine patients intubated
with a Mallinkrodt right-sided DLT, the right up-
per lobe was bronchoscopically patent when the
patient was supine. Third, the small margin of
safety for right-sided DLTs, along with the large
variation in the length of the right main-stem bron-
chus, indicates that whenever all other considera-
tions are equal, a left-sided DLT is preferable to a
right-sided DLT. Fourth, tolerance of head move-
ment with a right-sided DLT will be obviously
much less than for a left-sided DLT. Fifth, if a
right-sided DLT has to be used, the clear plastic
disposable ones may be best because of the slanted
doughnut shape of the right endobronchial cuff
Figure 9-31 The left double-lumen tube was advanced
down the trachea until moderate resistance to further passage
was encountered. With the tube in this position while ventilat-
ing the patient through both lumens of the double-lumen tube,
the left upper lobe bronchus was obstructed by the inflated
bronchial cuff while the right upper lobe bronchus, which orig-
inated in the trachea, was simultaneously obstructed by the
inflated tracheal cuff. (From Brodsky JB, Mark JBD: Bilateral
upper lobe obstruction from a single double-lumen tube. Anes-
thesiology 74:1163-1164, 1991. Used with permission.)
(Mallinkrodt), which decreases the likelihood of
right upper lobe obstruction by causing the right
upper lobe ventilation to be pushed away from the
lateral wall of the main-stem bronchus by the en-
dobronchial cuff (see Fig. 9-9). Finally, a larger,
rather than smaller, right-sided DLT should be
used. As tube size increases, length of Mallinkrodt
right cuff remains constant, and length of Rusch
right upper lobe ventilation slot increases; thus, the
margin of safety either remains constant (Mallin-
krodt right-sided tubes) or increases (Rusch right-
sided tubes) with increasing tube size. Although
the margin of safety remains constant or increases
with increasing DLT size, airway resistance and
difficulty in secretion removal decrease. In view of
the positive correlation between patient height and
main-stem bronchial length, and with the provision
that an intubation can be atraumatically performed,
the use of 39 and 41 French tubes in patients who
are 68 inches (170 cm) or taller, and 35 to 39
French tubes in patients who are shorter appears
indicated.
In summary, the analyses of margin of safety in
positioning left- and right-sided DLTs indicate that
if a question arises concerning DLT position
(using conventional unilateral clamping and aus-
cultation methods), blind attempts to adjust the
position of the DLT have a very good chance of
being unsuccessful and/or uncertain. Instead, the
DLT position should be checked by FOB, and for
a left-sided DLT the endobronchial cuff should be
positioned just below the tracheal carina (and,
therefore, left upper lobe obstruction will not be
possible). When a DLT is used for clinical re-
search purposes, a fiberoptic bronchoscope must
be used to confirm the proper DLT position to
prevent otherwise unrecognizable upper lobe ob-
struction and to eliminate this possibility of gath-
ering uninterpretable and nonrepresentative data.
2. Relationship of Fiberoptic
Bronchoscope Size to Double-Lumen
Tube Size
The clear plastic disposable right- and left-sided
double-lumen endotracheal tubes are manufactured
in five sizes: 28, 35, 37, 39, and 41 French. A 5.6-
mm outside-diameter diagnostic fiberoptic bron-
choscope will not pass down the lumens of any
sized DLT. A 4.9-mm external diameter fiberoptic
bronchoscope passes easily through the lumens of
the 41 French tube, passes moderately easily with
lubrication through the 39 French tube, causes a
tight fit that needs a liberal amount of lubrication
and a strong pushing force to pass through the 37
French tube, and does not pass through the lumen
of the 35 French tube. A silicon-based fluid (such
as that made by the American Cystoscope Co.) is
the best lubricant for a fiberoptic bronchoscope
because it does not dry out or crust and does not
interfere with the view even if it coats the tip of
the bronchoscope. Fortunately, from the point of
view of using a 4.9-mm outside-diameter fiberop-
tic bronchoscope, a 37 French tube or larger can
be used in almost all adult females and a 39 French
tube or larger can be used in almost all adult
males. A 3.6- to 4.2-mm outside-diameter (pediat-
ric) fiberoptic bronchoscope passes easily through
the lumens of all sized double-lumen endotracheal
tubes and, because the bronchoscope has an in-
creased amount of space, the maneuverability of
the tip of the bronchoscope is greatly increased.
Therefore, the 3.6- to 4.2-mm outside-diameter
bronchoscope is obviously the bronchoscope of
choice for DLTs. Table 9-7 summarizes these fi-
beroptic bronchoscope DLT relationships. Several
companies (Olympus, Machida, Pentax) presently
Table 9-7 RELATIONSHIP OF FIBEROPTIC BRONCHOSCOPE SIZE TO DOUBLE-LUMEN TUBE
SIZE
Fiberoptic Bronchoscope
Size Outside Diameter Double-Lumen Tube Size Fit of Fiberoptic Bronchoscope Inside
(mm) (French) Double-Lumen Tube
5.6 All sizes Does not fit
4.9 41 Easy passage
39 Moderately easy passage
37 Tight fit, needs lubricant,* hard push
55 Does not fit
3.6-4.2 All sizes Easy passage
*Lubricant recommended is a silicon-based fluid made by the American Cystoscope Company.
manufacture 4.9- and 3.6- to 4.2-mm outside-di-
ameter fiberoptic bronchoscopes that are of ade-
quate length and have a suction channel.
E. Use of Chest X-Ray to Determine
The chest roentgenogram can be used to deter-
mine DLT position. The usefulness of a chest
roentgenogram may be greater than conventional
unilateral auscultation and clamping in some pa-
tients, but it is always less precise than FOB. To
use the chest roentgenogram, the DLT must have
radiopaque markers at the end of the right and left
lumens. The key to discerning DLT position on
the chest roentgenogram is seeing where the
marker at the end of the tracheal lumen is in rela-
tion to the tracheal carina and whether the endo-
bronchial lumen located in the correct main-stem
bronchus. The end of the tracheal lumen marker
must be above the tracheal carina; however, this
does not guarantee correct position because this
technique may not reveal a subtle obstruction of
an upper lobe (see Fig. 9-25 and Margin of Safety
in Positioning Double-Lumen Tubes). If the tra-
cheal carina cannot be seen (which is often the
case with a portable anterior-posterior film), then
the chest roentgenogram method of determining
DLT position is not usable. Furthermore, the chest
roentgenographic method is time consuming (for
film transport, film development), costly, and awk-
ward to perform and may dislodge the tube (the
cassettes are often difficult to place under the op-
erating room table and may require movement of
the patient). Instillation of 1 ml of a radiopaque
marker into the endobronchial cuff (diatrizoate
meglumine and diatrizoate sodium solution [Re-
nografin 60] is water soluble and is easy to inject
but can be irritating to the bronchial mucosa if it
were to leak from a cuff) could enhance the utility
of using the chest X-ray to determine DLT posi-
tion.
F. Other Methods to Determine Double-
Lumen Tube Position
Three other methods may help to determine the
position of a DLT. First, comparison of capnogra-
phy (waveform and PETC0
2
value) from each lu-
men may reveal a marked discrepancy. For exam-
ple, and all other conditions being equal, one lung
may be very poorly ventilated in relation to the
other lung (high PETC0
2
) , indicating obstruction
to that lung; one lung may be extremely overven-
tilated in relation to the other lung (low PETC0
2
) ,
indicating, perhaps, just ventilation of a lobe of
that lung; or the capnogram from a given lung may
have a much steeper slope to the alveolar plateau,
indicating exhalatory obstruction.
72
73
Second, con-
tinuous spirometric data (Datex Capnomac Ul-
tima) from both lungs and each individual lung,
such as pressure-volume or flow-volume loops.
may be displayed and compared to a control loop
that is stored in memory.
74
Third, the surgeon may
be able to palpate the position of the DLT from
within the chest and may be able to redirect or
assist in changing the position of the DLT (by
deflecting the DLT away from the wrong lung.
etc.).
75
G. Securing the Double-Lumen Tube
After the correct placement of a double-lumen
endotracheal tube, it is customary to secure firmlv
the DLT at the level of the lips using adhesive or
umbilical tape. A simple and practical method for
securing the DLT is to use a simple single tie
around the tube at the level of the lips, followed
by a bow-tie knot at the bifurcation of the DLT
connector (Fig. 9-32).
76
Using this technique, sta-
bilization of the tube is not dependent on tight
knots around the tube itself but is instead accom-
plished by the bow-tie knot around the connector
bifurcation, preventing the tube from pulling out.
Fi gwe 932 TYi. dwtate-Vavwtn votat \% %ab\lV/fcd w\iK & t\.
around it at the ievei of hps and a bow tie at the bifurcation of
the lumens of the double-lumen tube. (From Cohen E, Koorn
R: An easy way to safely tie a double-lumen tube. J Cardi-
othorac Anesth 5:194-195, 1991. Used by permission.)
If readjustments need to be made, the DLT can be
freed by simply pulling on either end of the bow-
tie knot. The method is simple, avoids outward
displacement, and most importantly allows the
tube to be freed (untied) in a single maneuver for
proper positioning and manipulation. The time-
consuming struggle of untying multiple knots in a
contaminated, wet, and slippery environment is
avoided.
H. Quantitative Determination of Cuff-
Seal Pressure Hold
The use of FOB to determine DLT position does
not provide evidence or a guarantee that the two
lungs are functionally separated (i.e., against a
fluid and/or air pressure gradient). There are times,
such as during the performance of unilateral pul-
monary lavage, when the anesthesiologist must be
absolutely certain that functional separation has
been achieved. Complete separation of the two
lungs by the left endobronchial cuff can be dem-
onstrated in a left-sided tube by clamping the con-
necting tube to the right lung proximal to the right
suction port and attaching a small tube (i.e., intra-
venous extension tubing) to the open right suction
port (by appropriate adaptors) (Fig. 9-33). The
free end of this tube is submerged in a beaker of
water. When the left lung is statically inflated to
any pressure considered necessary, and the left
endobronchial cuff is not sealed, air will enter the
left lung as well as escape out from around the
unsealed left cuff, up the right lumen to the small
connecting tube, and bubble through the beaker of
water (Fig. 9335). If the left endobronchial cuff
is sealed, no bubbles should be observed passing
through the beaker of water (Fig. 9-33).
Following demonstration of functional lung sep-
aration, the right connecting tube is undamped,
the right suction port is closed, and ventilation to
both lungs is resumed. To test for lung separation
with the pressure gradient across the endobron-
chiJ baboon reversed, he tef airway connecting
tube is clamped proximai to the feft suction port,
the left suction port is opened to the beaker of
water via the small tube, the right lung is statically
inflated to any desired pressure, and the absence
or presence of air bubbles in the beaker of water
Yi *?&<&&. ft. &&&& fee ran&atserad tft&c, <2<r<sa
though the left endobronchial cuff may be ade-
quately sealed, it is possible that during these ma-
neuvers compression of the nonvemiiated Jung by
the ventilated lung may initially cause some small
amount of bubbling in the beaker, which will cease
with repeated inflation of the ventilated lung (no
bubbles should be seen following several
inflations).
61
62
The absence of airflow from the
nonventilated lung suction port is a very simple
but sensitive indicator of functional separation of
the two lungs. It is possible that a new capno-
graphic method for determining the seal of a bron-
chial blocker cuff (see section IV.A.5.) will prove
also to be useful for determining endobronchial
cuff seal.
I. Complications of Double-Lumen
Endotracheal Tubes
In addition to an impediment to arterial oxygen-
ation that is inherent in the use of double-lumen
endotracheal tubes for one-lung anesthesia, the
tubes themselves are very occasionally the cause
of other serious complications (Table 9-8). Many
of the complications reported in the literature in-
Air Bubble Method for Detection of Cuff Seal/Leak
Cuff Seal Cuff Leak
Figure 9-33 This schematic diagram shows the air bubble detection method for checking adequacy of the seal of the left
endobronchial cuff of a left-sided double-lumen tube. When the left lung is selectively ventilated or exposed to any desired
distending pressure and the left cuff is adequately sealed, no air will escape around the left cuff and out the open right suction port;
thus, no bubbles will be observed passing through the beaker of water (A). When the left lung is ventilated or exposed to any
desired distending pressure and the left endobronchial cuff is not adequately sealed, air will escape around the left cuff and out the
open right suction port; thus, air bubbles will be observed passing through the beaker of water (B). (IV = intravenous.)
volved the use of the Carlens tube, which reflects,
in part, the long history of the clinical use of this
tube. However, as experience with the new dispos-
able polyvinylchloride DLTs increases, so do the
number and type of complications associated with
this tube, especially tracheobronchial tree disrup-
tion.
In a group of 200 patients in whom the Carlens
tube was used, a 1.5 per cent incidence of trau-
matic laryngitis occurred, which was probably due
to malposition of the carinal hook at the time of
intubation.
15
Incorrect intraoperative positioning of
the tube was encountered in at least six patients in
this pre-FOB series and was believed to be respon-
sible for one intraoperative death. Another death
that occurred with the use of a Carlens tube during
Table 9-8 COMPLICATIONS OF DOUBLE-
LUMEN TUBES
1. Malpositioning
2. Tracheobronchial tree disruption
3. Traumatic laryngitis
4. Suturing of double-lumen tube to intrathoracic structure
pneumonectomy was caused by the inadvertent su-
turing of a pulmonary vessel to the tube.
77
The
possibility of a suture through the endotracheal
tube should be considered whenever excessive
resistance to extubation is encountered; the consid-
eration of this complication may warrant re-explo-
ration of the chest. The possibility of this compli-
cation occurring during pneumonectomy should be
minimized if an opposite-sided double-lumen en-
dotracheal tube is used or if the double-lumen en-
dotracheal tube is withdrawn into the trachea just
before the bronchus is clamped.
The most frequent serious complication is dis-
ruption of the tracheobronchial tree, which most
frequently occurs in the posterior membranous
wall. The vast majority of significant airway inju-
ries usually presents at operation with subcuta-
neous emphysema, pneumomediastinum and pneu-
mothorax, and, perhaps, cardiovascular instability
and hemoptysis, or they present shortly after the
trachea is extubated with the additional symptoms
of dyspnea, tachycardia, persistent cough, and he-
moptysis. There are reports of patients presenting
up to 24 hours after their tracheal injury.
In a series of approximately 2700 thoracic pro-
cedures in which a red rubber Carlens tube was
used, five cases of traumatic tracheobronchial rup-
ture were discovered.
78
In three cases the rupture
was noted intraoperatively, and in the other two it
was discovered early in the postoperative period.
All patients underwent successful direct repair of
the injuries. The authors suggested that factors
leading to this injury included the use of inappro-
priately sized DLTs, tube malpositioning, and
rapid and excessive inflation of tube cuffs (Table
9-9). These reports emphasize that postoperative
bronchoscopy should be performed to rule out the
diagnosis of tracheobronchial rupture when an
unexplained serious pneumothorax or pneumome-
diastinum persists after thoracotomy.
Bronchial rupture has also occurred with the use
of the red rubber Robertshaw double-lumen endo-
tracheal tube. An intraoperative diagnosis of bron-
chial rupture was made by the detection of medias-
tinal bubbles and subsequent direct inspection of
the left bronchus; intraoperative repair resulted in
an uneventful recovery.
79
The authors suggested
that this complication can be avoided by proper
selection of tube size, deflation of the endobron-
chial cuff prior to turning the patient to the lateral
decubitus position, and checking the integrity of
the previously intubated bronchus when testing the
bronchial stump for leaks (Table 9-9). Obviously,
great care must be taken when moving the patient
from the supine to the lateral decubitus position to
prevent tube movement, which can cause not only
tube malposition but also tracheobronchial tree
damage (Table 9-9).
Bronchial rupture has also occurred with the use
of the red rubber White DLT tube.
80
In this case
of right main-stem bronchial rupture, nitrous oxide
diffusion into the right endobronchial cuff, causing
Table 9-9 ENDOBRONCHIAL CUFF
CONSIDERATIONS TO MINIMIZE
TRACHEOBRONCHIAL WALL
DAMAGE (DISRUPTION)
1. Be particularly cautious in patients with bronchial wall
abnormalities.
2. Pick an appropriately sized tube.
3. Be certain tube is not malpositioned.* Use fiberoptic
bronchoscopy to confirm the position of the double-lumen
tube (especially if N
2
0 is introduced into the inspired
gases).
4. Avoid overinfiation of endobronchial cuff.*
5. Deflate endobronchial cuff during turning.
6. Inflate endobronchial cuff slowly.
7. Inflate endobronchial cuff with inspired gases or
periodically palpate pilot tube cuff and when appropriate,
remove cuff volume.
8. Do not allow tube to move during turning.*
9. Deflate endobronchial cuff during turning if possible.
*Most important consideration.
excessive right endobronchial cuff volumes, was
felt to be the cause. The authors, therefore, rec-
ommended that the cuffs be inflated with a sample
of the inspired mixture of gases rather than room
air or that cuff pressures be monitored so that
variation in cuff pressure can be observed and
corrected (Table 9-9). As a simple method of
monitoring cuff pressures and dealing with exces-
sive cuff volumes and pressures, the tension in the
pilot balloons to the cuffs can be periodically pal-
pated and gas removed from the cuffs if the ten-
sion appears to be increasing (same standard of
practice as with single-lumen tubes) (Table 9-9).
The authors also noted that bronchial rupture is
more likely to occur in patients with congenital
abnormalities of the bronchus; weakness of the
bronchial wall caused by infiltration of tumor, by
infection, or by poor quality tissues (sepsis, alco-
holism, drug abuse); and distortion of the bron-
chial tree by enlarged mediastinal lymph glands or
by extrabronchial tumors (Table 9-9). In addition,
the authors felt that the bevel of the tube or carinal
hook can potentially dissect underneath the mu-
cosa. Finally, the authors correctly warn of the
dangers of advancing the stylet beyond the vocal
cords and of using force to advance the tube fur-
ther whenever resistance is encountered.
A common thought in the preceding reports is
that excessive air volume and pressure in the bron-
chial balloon may be a major factor in the genesis
of tracheobronchial tree tears following DLT in-
sertion. Consequently, this complication was the
logical inspiration for the development of clear
plastic, tissue-implantable DLTs with high-vol-
ume-low-pressure cuffs (as it was the earlier in-
spiration for single-lumen tubes). The mean (
standard deviation) intracuff volume and pressure
for the Sheridan, Mallinckrodt, and Rusch left-
sided DLTs during clinical one-lung ventilation (n
= 48 patients) at peak inspiratory pressures of 35
to 40 cm H
2
0 are 2.8 1.3 ml and 27.9 17.7
cm H
2
0, 1.7 0.9 ml and 17.6 8.5 cm H
2
0,
and 1.5 0.9 ml and 14.1 8.6 cm H
2
0, respec-
tively (see Fig. 9-34).
81
These pressures are less
than levels that are associated with mucosal ische-
mia.
82
Thus, when the tracheobronchial tree is in-
spected with a fiberoptic bronchoscope after DLT
insertion, much less mucosal damage is caused by
the tissue-implantable, low-pressure cuffed tubes
than with the red rubber, high-pressure cuffed
tubes.
12
However, in spite of these expected find-
ings, there have been at least four reports of tra-
cheobronchial tree disruption after use of the tis-
sue-implantable, low-pressure cuffed DLTs,
83-86
and the precautions listed in Table 9-9 must be
considered even with the modern DLTs. One rea-
son why the cuffs of the polyvinylchloride DLTs
cause tracheobronchial tree damage is that the
Figure 934 Mean standard deviation intracuff
volume and pressure during bronchial seal for one-lung
ventilation in 48 patients for variously manufactured
double-lumen tubes. (From Slinger P, Chripko D: A
clinical comparison of bronchial cuff pressure in dis-
posable left double-lumen tubes. Anesthesiology
77:A1233, 1992. Used with permission.)
low-pressure cuff assumes the high-pressure char-
acteristics of red rubber cuffed tubes with volumes
greater than 3 ml.
10,86
Of course, as with all medi-
cal instruments, manufacturing defects may always
be unexpectedly encountered.
87
There are two other possible relevant predispos-
ing conditions for tracheobronchial tree damage.
First, esophageal surgery may cause a specific haz-
ard of tracheobronchial tree damage by a DLT in
that dissection of the upper and cervical esophagus
away from the posterior membranous trachea,
when it contains an inflated cuff, increases the risk
of tracheal tears. This increased risk may be secon-
dary to undermining the support and vascular sup-
ply of the membranous trachea during dissection.
88
Second, chronic obstructive airway disease has
been suggested as a possible risk factor in tracheal
trauma. Emphysema causes the tracheal rings to
open and the trachea itself to enlarge. This results
in an increase in the size of the membranous por-
tion of the tracheal wall. This is inevitably the
portion damaged at the time of or during intuba-
tion. The increase in surface area, along with atten-
uation of this already vulnerable area, could ex-
plain its increased propensity for damage.
J. Relative Contraindications to the Use
of Double-Lumen Endotracheal Tubes
There are several situations in which lung sepa-
ration by a DLT may be relatively contraindicated
because insertion of the DLT is either difficult or
dangerous (Table 9-10). These situations include
patients who have a full stomach and in whom
DLT intubation might be considered time consum-
ing (risk of aspiration); patients who have a lesion
(airway stricture,
89
9<)
endoluminal tumor) that is
present somewhere along the pathway of the DLT
that prevents proper positioning and/or that could
be traumatized; small patients in whom a 35
French tube is too large to fit comfortably through
the larynx and in whom a 28 French tube is con-
sidered too small; patients whose upper airway
anatomy either precludes safe insertion of the tube
or indicates that a required intraoperative DLT to
single-lumen tube change may be very difficult
(large tongue, recessed jaw, prominent teeth, bull
neck, restricted head and neck motion, anterior
larynx); extremely critically ill patients who have
a single-lumen tube already in place and who will
not tolerate being taken off mechanical ventilation
Table 9-10 RELATIVE CONTRAINDICATIONS
TO USE OF DOUBLE-LUMEN
TUBE
1. Presence of lesion along double-lumen tube
pathway
2. Very distorted tracheobronchial tree preventing
cannulation of a main-stem bronchus
3. Requirement for a difficult intraoperative change
from a double-lumen tube to a single-lumen tube
4. Difficult/impossible conventional direct vision
intubation
5. Extremely critically ill patients with single-lumen
tube in situ who cannot tolerate even a short period
off mechanical ventilation
6. Full stomach/high risk of aspiration
7. Some combination of above
and PEEP (even for the short period of 1 min);
and patients having some combination of all these
problems. Under these circumstances, it is still
possible to separate the lungs safely and ade-
quately by using a single-lumen tube and fiberop-
tic bronchoscopic placement of a bronchial blocker
or by fiberoptic bronchoscopic placement of a sin-
gle-lumen tube in a main-stem bronchus.
IV. BRONCHIAL BLOCKERS (WITH
SINGLE-LUMEN ENDOTRACHEAL
TUBES)
Lung separation can be effectively achieved
with the use of a single-lumen tube and a fiberop-
tically placed bronchial blocker (Figs. 9-35 to 9-
44). This is often necessary in children because
double-lumen endotracheal tubes are too large to
be used in these patients. The smallest DLT avail-
able is a size 28 French and may be potentially
used in patients in the range of 10 to 14 years and
weighing 30 to 45 kg. The bronchial blocker most
widely used for adults is the movable bronchial
blocker that is contained in, and is an integral part
of, the Univent single-lumen tube system (Fuji
Systems Corporation, Tokyo, Japan) (see Fig. 9-
35).
9I
~
97
j
n e
physiology of one-lung ventilation
produced by bronchial blockade is identical to that
produced by clamping one of the lumens of a
DLT.
A. Univent Bronchial Blocker Tube
1. Description of Univent Bronchial
Blocker Tube
The Univent bronchial blocker tube is a poly-
meric silicone single-lumen tube (6.0-9.0 mm in-
ternal diameter, 31-37 French) of normal configu-
ration that has a separate small lumen along the
internal anterior concave wall of the tube (i.e., at
the 12 o'clock position when the tube is held with
the natural concavity facing anteriorly) (see Fig.
9-35). The small lumen on the anterior concave
side of the tube, in turn, contains a small hollow
lumen catheter (2 mm internal diameter, approxi-
mately 17 gauge) that has a cuff at the end of it;
this secondary movable (by 8 cm beyond the tip
of the single-lumen tube)/retractable, small-lumen
cuffed catheter serves as a bronchial blocker when
the cuff is inflated (and, of course, allows two-
lung ventilation when the cuff is deflated). The
extra bronchial blocker channel adds significantly
to the anteroposterior outside diameter so that, for
a given inside diameter, the outside diameter of
the Univent tube is greater than an equivalent in-
ternal diameter single-lumen tube, and the antero-
posterior diameter is greater than the lateral diam-
eter (Table 9-11). The cuff of the bronchial
blocker has a natural volume of 2 ml, which cre-
ates a sphere or ellipsis shape with a midcuff di-
ameter of 5 mm; intracuff volume greater than 2
ml converts the cuff from a low-pressure cuff to
high-pressure cuff (see Section 3). However, at
present, the natural volume of the cuff is being
increased to 3 ml (Fujii Corp., personal communi-
cation).
Table 9-11 COMPARATIVE TUBE SIZES'
*Based on data from MacGillvray
93
and Slinger.
103
The AP diameter is greater than the lateral diameter because of the presence of bronchial blocker lumen.
Abbreviations: DLT = double-lumen tube, Bronchocath; SLT = single-lumen tube, Shiley; AP = anteroposterior; ID =
internal diameter; OD = outside diameter; FG = French gauge (OD in mm X 3).
UNIVENT BRONCHIAL BLOCKER TUBE
Figure 9-35 The Univent single-lumen
tube of bronchial blocker (BB) system.
2. Insertion of Univent Tube and
Positioning of Bronchial Blocker
The tube is inserted in the following manner.
First, the single-lumen tube along with the bron-
chial blocker (in the fully retracted position) is
inserted as a unit into the trachea (see Fig. 9-36,
panel A). The cuff on the main endotracheal tube
lumen is inflated, and the patient is ventilated and
oxygenated (see Fig. 9-36, panel A). A fiberoptic
bronchoscope is inserted through a self-sealing
diaphragm in the elbow connector to the single-
lumen tube while ventilation is maintained around
the fiberoptic bronchoscope (but within the single-
lumen tube) (see Figs. 9-36, panel and 9-37).
The right and left main-stem bronchi are identified
(by noting the relationship of the main-stem bron-
chi to the posterior membrane and the anterior
cartilaginous rings [Fig. 9-36, panel BJ), and the
tube of the bronchial blocker is located (by mov-
ing the bronchial blocker in and out just beyond
the end of its own and the main lumen of the
Univent tube [see Figs. 9-36, panel C and 9-37]).
The bronchial blocker cuff is colored blue and
is easy to see. It will be seen that the bronchial
blocker usually (almost always) enters the right
main-stem bronchus if it is simply pushed in (and
the main single-lumen tube is not turned). If
the left main-stem bronchus is to be blocked, then
the main single-lumen tube is turned 90 degrees to
the left (counterclockwise) so that the concavity of
the tube is facing toward the left side (Fig. 9-36.
panel C) (and vice versa for the right side, if nec-
essary). The bronchial blocker can also be rotated
at its distal end a slight amount (obtain 1-3 mm of
laterality) by twirling the proximal end in the fin-
gers. The bronchial blocker is then advanced into
the main-stem bronchus under direct vision (see
Fig. 9-36, panel D and 9-37). Attempting to ad-
vance the bronchial blocker blindly into the appro-
Insertion and Positioning of
Univent Bronchial Blocker (BB) System
Figure 9-36 The sequential steps of the fiberoptic-aided method of inserting and positioning the Univent bronchial blocker (BB)
in the left main-stem bronchus are illustrated. One- and two-lung ventilation is achieved by simply inflating and deflating,
respectively, the bronchial blocker balloon. (FOB = fiberoptic bronchoscope.)
Figure 9-37 Univent tube with the endobronchial blocker extended and inflated, showing easy passage of the flexible broncho-
scope. (From Karwande SV: A new tube for single lung ventilation. Chest 92:761-763, 1983. Used with permission.)
priate main-stem bronchus (particularly the left)
will be unsuccessful 87 per cent of the time, and
repeated attempts may cause excoriation of the
tracheal mucosa.
93
In fact, blindly pushing the
somewhat stiff bronchial blocker may result in
perforation of the tracheobronchial tree and ten-
sion pneumothorax.
94
The balloon is inflated only
to the point at which the cephalad surface of the
balloon is just below the tracheal carina (see Fig.
9-36, panel E) (so that the upper lobe of the
blocked lung may also distend if CPAP is applied
to the blocked lung; see later discussion) and the
FOB is then withdrawn (see Fig. 9-36, panel F).
3. Advantages/Noteworthy Positive
Attributes of Univent Bronchial Blocker
Tube System
The Univent bronchial blocker tube has six im-
portant attributes that require special mention (Ta-
ble 9-12). First, and foremost, the degree of diffi-
culty in inserting the Univent tube is equivalent to
that of a standard single-lumen tube and therefore,
in many instances, will be an easier and quicker
way to separate the lungs to obtain simple one-
lung ventilation (compared with a double-lumen
tube).
95 97
Thus, for example, the Univent tube
may be preferable when difficult intubation is an-
ticipated and in an anticoagulated patient.
Second, the patient can be continuously venti-
lated while the bronchial blocker is being placed
into a main-stem bronchus, and the bronchial
blocker can be placed into a main-stem bronchus
just as easily in the lateral decubitus position as in
the supine position.
Third, and provided the postanesthesia care unit
and the ICU personnel are instructed in the design
Table 9-12 ADVANTAGE/NOTEWORTHY
POSITIVE ATTRIBUTES OF THE
UNIVENT BRONCHIAL BLOCKER
TUBE (RELATIVE TO A DOUBLE-
LUMEN TUBE AND OTHER
BRONCHIAL BLOCKERS)
1. Easier to insert and properly position
2. Can properly position during continuous ventilation
and in the lateral decubitus position
3. Do not need to change the tube for postoperative
mechanical ventilation
4. Do not need to change the tube intraoperatively
when turning from the supine to the prone position
5. Selective blockade of some lobes of each lung
6. Can apply nonventilated, operative lung continuous
positive airway pressure
and function of the Univent tube (particularly the
ventilatory consequence of inflating the bronchial
blocker cuff when the bronchial blocker cuff is
just distal to the main lumen; i.e., the main lumen
will be obstructed),
98
the Univent tube may be left
in situ for postoperative mechanical ventilation;
therefore, the risk of a potentially difficult tube
change (e.g., from a DLT to a single-lumen tube)
is avoided.
Fourth, and similarly, the Univent tube may be
left in situ if a patient is turned from the supine to
the prone position midway through a surgical pro-
cedure (common occurrence with surgery on the
thoracic spine).
Fifth, the unique characteristic of a movable
endobronchial blocker permits the Univent endo-
tracheal tube to create selective, partial (a lobe), or
total collapse of the targeted lung." The capability
of selectively blocking lung segments is extremely
important in cases of isolated pulmonary hemor-
rhage. Partial versus total one-lung ventilation may
allow for an improvement in P
a
0
2
in cases of intra-
operative hypoxemia during thoracic operations.
Finally, although not a distinct advantage over
a DLT, it should be noted that it is possible to
apply CPAP to the nonventilated operative lung
(see chapter 11) through the lumen of the bron-
chial blocker;
100
therefore, the Univent tube pro-
vides the same best solution to hypoxemia during
one-lung ventilation as does a DLT (see chapter
11). In fact, Figure 9-38 shows that, except for
independent unilateral intermittent positive-pres-
sure ventilation and suctioning, all selective differ-
ential lung functions that are possible with a DLT
are possible with a Univent bronchial blocker tube.
4. Potential Limitations of the Univent
Bronchial Blocker Tube System and
Solutions
There are several distinct limitations to the Uni-
vent bronchial blocker tube system, but fortunately
all have a relatively simple remedy (Table 9-13).
First, the small lumen of the bronchial blocker
results in slow inflation of the lung if gases are
just insufflated (e.g., at a flow rate of 10 L/min, a
lung will require at least 20 sec to reach total lung
capacity) or pushed in by conventional positive
pressure. The operative lung may be made to ex-
pand rapidly if the bronchial blocker cuff is de-
flated (the operative lung will expand with one
positive-pressure breath from the main single lu-
men) or one very short (e.g., < 0.5 sec) wall
oxygen powered 20 to 30 psi jet ventilation (re-
duced from 50 psi) is administered; however, con-
nection of the bronchial blocker lumen to a jet
ventilator is potentially dangerous (i.e., can cause
Table 9-13
Limitation
LIMITATIONS TO THE USE OF
THE UNIVENT BRONCHIAL
BLOCKER TUBE AND SOLUTIONS
Solution
Slow inflation time
Slow deflation time
Blockage of bronchial
blocker lumen by
blood pus
High-pressure cuff
Intraoperative leak in
bronchial blocker
cuff
(l) Deflate bronchial blocker cuff
and administer a positive-
pressure breath through the main
single lumen; (2) carefully
administer one short high-
pressure (20-30 psi) jet
ventilation.
(I) Deflate bronchial blocker cuff,
and compress and evacuate the
lung through the main single
lumen; (2) apply suction to
bronchial blocker lumen.
Suction, use stylet, and then suction.
Use just-seal volume of air.
Make sure bronchial blocker cuff is
subcarinal, increase inflation
volume, and rearrange surgical
field.
barotrauma) because the lung can expand ex-
tremely rapidly, and it is of paramount importance
that the anesthesiologist directly observe the lung
and that the ventilation be very short or the psi
limited to 20 to 30 psi by an additional in-line
regulator.
Second, and also because the bronchial blocker
lumen is small, the lung will deflate very slowly
when blocked. This is easily remedied by deflating
the bronchial blocker cuff (which re-establishes
continuity between the operative lung and the
main single lumen), disconnecting the patient from
the ventilator and leaving the endotracheal tube
open to air while the surgeon gently compresses
the lung to evacuate air from the operative lung
through the main single lumen. After the lung is
thus fully collapsed, the blocker balloon is inflated
and ventilation resumed.
95
96
Alternatively, the lu-
men of the bronchial blocker may be connected to
the suction apparatus while the cuff is inflated; a
normal amount of wall suction greatly facilitates
lung collapse.
Third, because the bronchial blocker lumen is
small, the lumen is relatively easily blocked by
blood and/or pus. High suction will usually clear
the lumen of these materials, and total blockage by
inspissated secretions can be broken up by a wire
stylet.
Fourth, the Univent bronchial blocker behaves
as a high-pressure cuff when intracuff volume is
greater than 2 ml (the resting volume of the cuff)
and may be expected to have an intracuff pressure
between 150 and 250 mm Hg and a transmural
Independent
Collapse
IPPV
Sigh
CPAP
HFV
Suction
PEEP
Operative
Lung
Independent
Collapse
IPPV
Sigh
CPAP
HFV
Suction
PEEP
Nonoperative
Lung
Bilateral
IPPV
Sigh
PEEP
Operative
Lung
Nonoperative
Lung
Independent
Collapse
Sigh
CPAP
HFV

Operative
Lung
Independent
IPPV
Sigh
CPAP
HFV
Suction
PEEP
Nonoperative
Lung
Double-Lumen
Tube
Univent BB
Deflated
Univent BB
Inflated
Figure 9-38 Except for independent unilateral intermittent positive-pressure ventilation (IPPV) and suctioning, all selective differential lung functions possible
with a double-lumen tube are possible with a Univent BB tube. (CPAP = continuous positive airway pressure; HFV = high-frequency ventilation; PEEP =
positive end-expiratory ventilation.)
pressure (intracuff pressure within the airway mi-
nus intracuff pressure outside of the airway [free
in the room]) between 50 and 60 mm Hg when
intracuff volumes of 4 to 6 ml are used to seal
airways of 12 to 18 mm against the usual proximal
airway pressures
101. 102
(Thus, the order of usual
bronchial cuff pressures are left-sided polyvinyl-
chloride DLT less than right-sided polyvinylchlor-
ide DLT less than Univent Bronchial Blocker cuff
less than red rubber DLT.)
103
These findings un-
derscore the need to inflate the bronchial blocker
cuff with a just-seal volume of air (see next section
for three alternative methods to obtain a just-seal
volume).
Fifth, the Univent bronchial blocker has been
reported to have a minor leak during surgery on
occasion (25 per cent in one series),
93
but this is
not understandable in view of experiments that
show that the Univent bronchial blocker cuff seals
within normal-size main-stem bronchi against
proximal airway pressures as great as 100 cm H
2
0
with inflation volumes that are within the manu-
facturer's recommendation.
102
Consequently, if an
intracuff volume of less than 6 to 7 ml has been
used, the bronchial blocker cuff is completely sub-
carinal (determined by FOB) and intact, and an
intraoperative leak is present, then the intracuff
volume should be increased. If an adequate cuff
inflation volume has been used (the bronchial
blocker can be seen [fiberoptically] to fill the
main-stem bronchus in question), the bronchial
blocker cuff is completely subcarinal (determined
fiberoptically) and intact, and an intraoperative
leak develops, then the relationship between the
main-stem bronchus and the bronchial blocker cuff
may no longer be a simple matter of a sphere or
ellipsis shape being inflated within a cylinder. Un-
der these circumstances, the surgeon may need to
rearrange the surgical field so that the main-stem
bronchus and bronchial blocker cuff are less dis-
torted. Finally, the addition of the lumen for the
bronchial blocker results in an endotracheal tube
that has a large outside anterior-posterior diameter
relative to its inside diameter.
5. Methods to Obtain a Just-Seal
Volume in the Bronchial Blocker Cuff
There are three methods to obtain a just-seal
volume of air in the bronchial blocker cuff. The
first method is the same as that already described
for obtaining a just-seal volume of air in the en-
dobronchial cuff of a DLT. It consists of pressur-
izing the main single lumen until air ceases to
escape from the bronchial blocker lumen (detected
by connecting the bronchial blocker lumen to a
catheter that is submerged beneath the surface of a
beaker of water; when air bubbles cease to come
out, the bronchial blocker cuff has sealed) (Fig. 9-
39; compare with Fig. 9-33).
Confirmation
of Univent
Bronchial
Blocker (BB)
Cuff Seal
1. Pressurize tracheal lumen
2. Connect BB lumen to
underwater seal
a. No bubbles = seal
b. Bubbles = no seal
Figure 9-39 The positive-pressure ventilation-bubble under water method for determining just-seal volume of bronchial blocker
cuff is the same as for an endobronchial cuff on a double-lumen tube (Fig. 9-33).
Figure 940 When the bronchial cuff is deflated, negative
pressure applied at the proximal end of the lumen of the
blocker is freely transmitted to the reservoir bag of the breath-
ing system. (FGF = fresh gas flow.) (From Hannallah M: The
Univent tube: Bronchial cuff inflation. Anesthesiology
The second method is opposite in concept and
uses the negative pressure of the suction
apparatus.
35
36
After correct placement of the tube
in the trachea and the blocker in the bronchus, the
tracheal cuff is inflated in the usual manner, and
the patient's lungs are ventilated with 100 per cent
oxygen. Ventilation then is discontinued, and a
fresh gas flow of 5 to 6 L/min is delivered to the
system. The reservoir bag will then fully distend.
Using appropriate connections, negative pressure
is applied to the proximal end of the lumen of the
blocker; with the bronchial cuff deflated, the distal
end of the lumen of the blocker will be freely
connected with trachea and the circle system, the
volume of oxygen suctioned out of the system will
be more than the volume delivered to it, and the
reservoir bag will then begin to deflate (Fig. 9-
40). At this point, the bronchial cuff is inflated
slowly with 1-ml increments of air, and the breath-
ing bag is watched carefully until it ceases to de-
flate. This will indicate that there is no longer any
communication between the tip of the blocker,
where the negative pressure is being applied, and
the rest of the breathing system. Complete sealing
of the bronchus will have been accomplished at
this point. This is confirmed by auscultation of the
breath sounds initially and later by observation of
the lung's actual collapse, with continued suction,
when the chest is opened. Comparison of the first
two methods in the same patient show that they
yield nearly identical results.
35
36
The third method appears very promising and
uses capnography.
37
End-tidal C0
2
analyzers draw
gas samples from the anesthesia breathing circuit
via tubing terminating in a standard Luer lock
male connector that inserts into a female port in
the breathing circuit. The male connector also at-
taches to the female port at the proximal end of
the Univent's bronchial blocker. The tracing from
a gas analyzer, connected to the blocker with its
cuff deflated, shows a typical respiratory wave-
form. As the blockers cuff is steadily inflated, a
point is reached at which the respiratory waveform
abruptly ceases, and a straight line is seen, indicat-
ing that lung isolation has occurred (Fig. 9-41).
C0
2
concentration remains near its end-tidal value
until the blocked lung has collapsed and then rap-
idly decreases. The strengths of this method in-
clude simplicity, repeatability, and ability to venti-
late the unblocked lung continuously throughout
the procedure.
6. Firm Clinical Indications for Use of
Univent Bronchial Blocker System
There are several clinical situations in which the
use of the Univent bronchial blocker tube is rela-
tively indicated. First, whenever it is anticipated
that postoperative ventilation will be necessary
(e.g., poor pulmonary function preoperatively, an-
ticipated lung damage or massive fluid/blood in-
fusion intraoperatively, anticipated very long
< Q
Q
Fe: 02=94 N2= 0 N20= 0~ISO=.25 C02=35
Figure 941 Capnogram tracing showing normal respiratory waveform changing to a straight line as bronchial seal occurs.
(From Essig K, Freeman JA: Alternative bronchial cuff inflation technique for the Univent tube. Anesthesiology 76:478^479,
case), use of the Univent bronchial blocker tube
for lung separation may avoid a risky postopera-
tive DLT to single-lumen tube change.
Second, and similarly, use of the Univent bron-
chial blocker tube will avoid a potentially danger-
ous DLT to single-lumen tube change in cases of
surgery on the thoracic spine in which a supine or
lateral decubitus position thoracotomy is followed
by surgery in the prone position.
Third, a very severely distorted airway may pre-
vent successful placement of a DLT, whereas such
distortion may have much less of an effect on the
proper placement of the Univent tube. Finally, but
least predictable/compelling, are situations in
which both lungs may need to be blocked (e.g.,
bilateral operations, indecisive surgeons).
B. Bronchial Blockers That Are
Independent of a Single-Lumen Tube
Bronchial blockers may also be passed indepen-
dently of the single-lumen tube. Independent bron-
chial blockers that are balloon-tipped luminal cath-
eters (e.g., Fogarty embolectomy, MaGill, Foley,
and pulmonary artery catheters) have the advan-
tage of allowing suctioning and injection of oxy-
gen down the central lumen (as with the Univent
bronchial blocker tube). However, most indepen-
dent bronchial blockers have the disadvantage of
sometimes requiring rigid bronchoscopy for place-
ment, and, because they have high-pressure, spher-
ically inflating balloons, they tend to back out of
the bronchus into the trachea (see the following
discussion).
The independent bronchial blocker most often
used for adults is a Fogarty occlusion (embolec-
tomy catheter with a 3-ml balloon).
104
The Fogarty
embolectomy catheter comes with a stylet in place
so that it is possible to place a curvature at the
distal tip to facilitate entry into the larynx and
either main-stem bronchus (by twirling the proxi-
mal end). If no endotracheal tube is in place, the
operator exposes the larynx and places a single-
lumen tube with a high-volume cuff in the trachea.
The Fogarty catheter is then placed either
inside
105
l06
or alongside the single-lumen tube
(Figs. 9-42 and 9-43). Placement of the Fogarty
catheter inside the single-lumen tube can be
greatly facilitated by use of two elbow connectors
with self-sealing diaphragms that are connected in
series with the anesthesia circuit attached to the
proximal end of the connectors (Fig. 9-44).
I05
The
distal end of the elbow connectors is connected to
the patient's single-lumen endotracheal tube. The
Fogarty catheter can be easily introduced through
the diaphragm of one of the elbow connectors
while the other diaphragm allows insertion of a
fiberoptic bronchoscope used to verify correct
placement of the catheter in the main-stem bron-
chus. In either case (bronchial blocker inside or
outside the single-lumen tube), a fiberoptic bron-
choscope is passed down to the end of the single-
lumen tube through a self-sealing diaphragm in the
elbow connector (which permits continued posi-
tive-pressure ventilation around the fiberoptic
bronchoscope), and the Fogarty catheter is visual-
ized below the tip of the single-lumen tube. The
proximal end of the bronchial blocker is then
twirled in the finger tips until the distal tip locates
in the desired main-stem bronchus. The catheter
balloon is then inflated under direct visualization,
and the fiberoptic bronchoscope is withdrawn
through the self-sealing diaphragm. The self-
sealing diaphragm in the elbow connector contain-
ing the bronchial blocker should be made airtight.
The bronchial blocker within the single-lumen
Text continued on page 383
A. Lung Separation With Single Lumen Tube
and Left Lung Bronchial Blocker
Inside of Single Lumen Tube
Figure 9-42 See legend on following page
B. Lung Separation With Single Lumen Tube
and Right Lung Bronchial Blocker
Inside of Single Lumen Tube
LL RL LL RL
LL RL LL RL LL RL
Figure 942 This figure shows how to separate the two lungs with a single-lumen tube, fiberoptic bronchoscope, and a left lung
(A) and a right lung (B) bronchial blocker. The sequence of events is as follows: A single-lumen tube is inserted, and the patient is
ventilated (upper left diagram, A and B). A bronchial blocker is passed inside the indwelling endotracheal tube (upper right diagram,
A and B). A fiberoptic bronchoscope is passed through a self-sealing diaphragm in the elbow connector to the endotracheal tube
and is used to place the bronchial blocker into the appropriate main-stem bronchus under direct vision (lower left diagram, A and
B). The balloon on the bronchial blocker is also inflated under direct vision and is positioned just below the tracheal carina (lower
middle diagram, A and B). During the lower panel sequence (insertion and use of fiberoptic bronchoscope), the self-sealing
diaphragm allows the patient to continue to be ventilated with positive-pressure ventilation (around the fiberoptic bronchoscope but
within the lumens of the endotracheal tube). (LL = left lung; RL = right lung.)
A. Lung Separation With Single Lumen Tube
and Left Lung Bronchial Blocker
Outside of Single Lumen Tube
LL RL LL RL
LL RL LL RL LL RL
Figure 9-43 See legend on following page
B. Lung Separation With Single Lumen Tube
and Right Lung Bronchial Blocker
Outside of Single Lumen Tube
LL RL LL RL LL RL
Figure 9-43 This figure shows how to separate the two lungs with a single-lumen tube, fiberoptic bronchoscope, and a left lung
(A) and a right lung (B) bronchial blocker. The sequence of events is as follows: A single-lumen tube is inserted, and the patient is
ventilated (upper left diagram, A and B). A bronchial blocker is passed alongside the indwelling endotracheal tube (upper right
diagram, A and B). A fiberoptic bronchoscope is passed through a self-sealing diaphragm in the elbow connector to the endotracheal
tube and is used to place the bronchial blocker into the appropriate main-stem bronchus under direct vision (lower left diagram, A
and B). The balloon on the bronchial blocker is also inflated under direct vision and is positioned just below the tracheal carina
(lower middle diagram, A and B). During the lower panel sequence (insertion and use of fiberoptic bronchoscope), the self-sealing
diaphragm allows the patient to continue to be ventilated with positive-pressure ventilation (around the fiberoptic bronchoscope but
within the lumens of the endotracheal tube). (LL = left lung; RL = right lung.)
Figure 9 -44 Two elbow connec-
tors (A) are shown interposed be-
tween the anesthesia circuit and the
endotracheal tube with a Fogarty
catheter (B) inserted through one
self-sealing diaphragm and a fiber-
optic bronchoscope (C) inserted
through the other self-sealing dia-
phragm. (From Larson CE, Gaisor
TA: A device for endobronchial
blocker placement during one-lung
anesthesia. Anesth Analg 71:311-
tube technique may be used with a tracheos-
tomy.
106
A simplified coaxial double-tube system (see
section III..), but in which the inner tube is used
primarily as a bronchial blocker, has been re-
ported.
107
Conventional one-lung ventilation is
conducted through the outer tracheal tube. How-
ever, and unfortunately, the bronchial blocker in
this system is best situated with a rigid broncho-
scope.
For bronchial blockade in very small children
(10 kg or less), a Fogarty embolectomy catheter
with a balloon capacity of 0.5 ml or a Swan-Ganz
catheter (1-ml balloon) should be used.
108
Of
course, these catheters have to be positioned under
direct vision; a fiberoptic bronchoscope method, as
depicted in Figure 9-43, is perfectly acceptable,
except the fiberoptic bronchoscope outside diame-
ter must be approximately 2 mm to fit inside the
endotracheal tube. Otherwise, the bronchial
blocker must be situated with a rigid broncho-
scope. Pediatric patients of intermediate size will
require intermediately sized occlusion catheters
and a judgment on the mode of placement (i.e.,
rigid vs. fiberoptic bronchoscope) (see chapter 18).
Disadvantages of the independent bronchial
blockers compared with double-lumen endotra-
cheal tube lung separation include the inability to
extensively suction and/or to ventilate the lung
distal to the blocker, increased placement time (as
well as compared with the Univent tube), and the
definite need for a fiberoptic or rigid broncho-
scope. In addition, if a main-stem bronchial
blocker backs out into the trachea, the seal be-
tween the two lungs will be lost, and two cata-
strophic complications may occur. First, if the
bronchial blocker was being used to seal off a fluid
(blood or pus) in one lung, then both lungs may
become contaminated with the fluid. Second, the
trachea will be at least partially obstructed by the
blocker, and ventilation will be greatly impaired.
Therefore, bronchial blockage requires that the
anesthesiologist continuously and intensively
monitor the compliance and breath sounds of the
ventilated lung.
V. ENDOBRONCHIAL INTUBATION
WITH SINGLE-LUMEN TUBES
A single-lumen tube may be electively placed in
a main-stem bronchus, thereby blocking off the
contralateral lung and permitting ventilation of just
the ipsilateral lung. If the single-lumen tube has to
be pushed in blindly, it will almost always enter
(near 100 per cent of the time) the right main-stem
bronchus when the concavity of the single-lumen
tube points anteriorly (i.e., the way it is usually
held) and the head is in a normal straight-ahead
midposition. The single-lumen tube will enter the
left main-stem bronchus 92 per cent of the time
when the concavity of the single-lumen tube is
facing posteriorly (180-degree rotation from the
way it is usually held) and the head is turned to
the right.
109
The single-lumen tube enters the right
and left main-stem bronchi when the concavity of
the tube is anterior or posterior facing, respec-
tively, because the bevel is left and right facing,
respectively (i.e., left-facing bevel enters the right
main-stem bronchus and a right-facing bevel en-
ters the left main-stem bronchus) (see Pediatric
One-lung Ventilation in chapter 18)." Unfortu-
nately, blindly placing a single-lumen tube into a
main-stem bronchus incurs a high risk of blocking
off the upper lobe of the intubated lung and caus-
ing hypoxemia."
1,112
However, if the single-lumen
tube can be placed proximal to either the right or
left upper lobe either with bronchoscopic vision or
by surgical manipulation, then the resulting one-
lung ventilation gas exchange will be the same as
expected for one-lung ventilation with a DLT
or with an appropriately placed bronchial
blocker.
111
112
In adults presenting with hemoptysis, endobron-
chial intubation with a single-lumen tube is often
the easiest, quickest way of effectively separating
the two lungs, especially if the left lung is bleed-
ing. If the left lung is bleeding, one can simply
take an uncut single-lumen endotracheal tube and
advance it inward until moderate resistance is felt
(Fig. 9-45). In the vast majority of patients, the
single-lumen tube will locate in the right main-
Single Lumen Tube: Lung Bleeding
Figure 9-45 This figure shows how to separate the two lungs with a single-lumen tube in the presence of massive lung bleeding.
When the left lung is bleeding (A), an uncut single-lumen tube may simply be inserted its full length, and, in the vast majority of
cases, it will enter the right main-stem bronchus, thereby effectively sealing off the right lung from the left lung. However, one can
expect that the cuff of the single-lumen tube will obstruct the right upper lobe. When the right lung is bleeding (B), a fiberoptic
bronchoscope, which is jacketed on its proximal end with an endotracheal tube, can be passed through a self-sealing diaphragm in
the elbow connector to the endotracheal tube (which allows continued positive-pressure breathing) and, if a moment's view of the
tracheal carina can be obtained, passed into the left main-stem bronchus. Using the fiberoptic bronchoscope as a stylet, the
endotracheal tube can be passed over the fiberoptic bronchoscope into the left main-stem bronchus. The fiberoptic bronchoscope is
then withdrawn. (LL = left lung; RL = right lung; FOB = fiberoptic bronchoscope.)
Figure 9-46 The Wilson tube
constructed by attaching an exten-
sion tube onto a square-ended, flexi-
ble armored tube. (From Newton JR,
Gulls HC, Mathisen DJ: Main bron-
chial sleeve resection with pulmo-
nary conservation. Ann Thorac Surg
52:1272-1280, 1991. Used with per-
mission.)
stem bronchus, thereby blocking off the bleeding
left lung and allowing for selective ventilation of
just the right lung. Under these circumstances, it is
very possible that the right upper lobe bronchus
will be blocked off as well, resulting in ventilation
of just the right middle and lower lobes. Ventila-
tion of just a soiled right lung or ventilation of
just the right middle and lower lobes (even if
they are unsoiled) incurs the risk of serious hy-
poxemia because of the very large transpulmo-
nary shunt that is necessarily created by the sin-
gle lung (and perhaps bilobar) endobronchial
intubation/ventilation.
If the right lung is bleeding, a fiberoptic bron-
choscope can be passed through a self-sealing dia-
phragm in the single-lumen tube elbow connector
and directed into the left main-stem bronchus. Per-
sistent large, soft catheter suctioning of the carinal
area through the single-lumen tube before use of
the fiberoptic bronchoscope and suctioning
through the fiberoptic bronchoscope (through the
single-lumen tube) may be required to visualize
the tracheal carina (Fig. 9-45B). The single-lumen
tube can then be passed over the fiberoptic bron-
choscope into the left main-stem bronchus, thereby
sealing off the bleeding right lung and allowing
for selective ventilation of the left lung. Passing
the fiberoptic bronchoscope through a self-sealing
diaphragm allows for the continuance of positive-
pressure ventilation and PEEP around the bron-
choscope. However, it should be realized that vis-
ualization of the carina may not be possible when
the bleeding is copious and that the only hope for
the patient may lie in rapid thoracotomy and con-
trol of bleeding from within the chest. In addition,
under these very adverse conditions, conventional
passage of a double-lumen tube may more rapidly
and effectively separate the two lungs than trying
to visualize anatomy with a fiberoptic broncho-
scope. Alternatively, a right main-stem endobron-
chial tube, which has tracheal and right endobron-
chial cuffs and a right upper lobe ventilation slot,
can be passed, but the endobronchial tube has lim-
itations that the DLT does not (see discussion of
this tube immediately following and Figs. 9-46
through 9-48).
Figure 9-47 Extended endotra-
cheal tube compared with an ordi-
nary endotracheal tube. The distal
section of the extended tube is cho-
sen in a size appropriate for the pa-
tient. The size of the proximal sec-
tion is selected for a snug fit over the
distal section. (From Bragg CL, Vuk-
elich GR: Endotracheal tube exten-
sion for endobronchial intubation.
Anesth Analg 69:548-549,
1989.
Gordon Green Tube
Figure 9-48 Schematic diagram of the placement of a Gordon
Green endobronchial tube at the carina.
For adults, a variety of reusable endobronchial
tubes have been developed for insertion into a
main-stem bronchus to provide one-lung anesthe-
sia during thoracic surgery (see Table 1-1). Some
of these can be placed blindly, while others require
placement via surgical manipulation from within
the surgical field or an intubating bronchoscope.
These tubes have the advantage of having a large
diameter and therefore a low airway resistance.
Major disadvantages of some of these tubes in-
clude the inability to suction the operative site,
difficulty in positioning the bronchial cuff, the
possibility of the thin-walled tube kinking in the
posterior pharynx, and the very considerable haz-
ard of inadequately ventilating the right upper lobe
after right endobronchial intubation. Two newly
described endobronchial tubes consist of single-
lumen tubes with an extension (which is necessary
because standard single-lumen tubes are of insuf-
ficient length to extend from a main-stem bronchus
to the mouth). The Wilson tube (see Fig. 9-46)
113
is a long, flexible, single-lumen tube constructed
by attaching an extension onto a flexible armored
tube. There is also an extender on the balloon cuff
inflation port. The end is square, not beveled, for
better fit into the main bronchi, and the balloon
cuff is smaller to minimize herniation. The tube is
advanced into the opposite bronchus before thora-
cotomy. Tube position is verified by examination
or by use of a flexible pediatric bronchoscope.
Alternatively, the length of a single-lumen tube
may be extended by cutting a 9.0-mm endotra-
cheal tube 12 cm from its proximal end and plac-
ing it over a 6.5-mm endotracheal tube from which
the 15-mm adaptor has been removed. The tubes
fit snugly together with an airtight seal; however,
for added security, the tubes should be sutured
together with use of 2.0 silk (Fig. 9-47).
m
The
Gordon Green tube, having both a tracheal cuff
and an endobronchial cuff, offers several advan-
tages if a right-sided endobronchial tube is needed
(see Fig. 9-48)."
5
A carinal hook facilitates blind
positioning. Inflation of the bronchial cuff isolates
the left lung; when the right endobronchial balloon
is not inflated, the entire lung can be ventilated by
backflow around the distal balloon. Finally, the
endobronchial cuff is slotted to provide ventilation
to the right upper lobe during one-lung ventilation.
In spite of these advances in tube design, endo-
bronchial tubes are generally less satisfactory than
double-lumen endotracheal tubes and are used in-
frequently today.
22
In children, the simplest method for achieving
lung separation is to pass a standard single-lumen
tube into a main-stem bronchus. Right main-stem
bronchial intubation is easily achieved blindly; left
main-stem bronchial intubation may require FOB
(see preceding discussion), fluoroscopy, or guid-
ance of the tube by the surgeon from within in the
chest (see chapter 18). The best method of lung
separation in children, however, is with a bron-
chial blocker.
In summary, double-lumen endotracheal tubes
are the method of choice for separating the lungs
in most adult patients. If there is any question, the
precise location of a DLT can be determined by
FOB at any time. There are a number of situations
in which insertion of a DLT may be difficult
and/or dangerous, and under these circumstances
consideration should be given to separating the
lungs with a single-lumen tube alone or in combi-
nation with a bronchial blocker (e.g., the Univent
tube). However, when using a single-lumen tube
in a main-stem bronchus or when using a bron-
chial blocker, the ability to suction the operative
site and control ventilation is limited. In addition,
the placement of the single-lumen tube into one or
the other main-stem bronchus and the proper
placement of a bronchial blocker require FOB.
Therefore, no matter which method of separating
the lungs is chosen, there is a real need for the
immediate availability of a small-diameter fiber-
optic bronchoscope (for checking the position of
the DLT, placing a single-lumen tube in the left
main-stem bronchus, and placing a bronchial
blocker) that has a suction port (to clear secretions
and blood from the airway).
CHAPTER 10
Routine Surgical
Considerations that have
Anesthetic Implications
1. Introduction
II. Positioning the Patient
A. Posterior Lateral and Limited Lateral
Thoracotomy
B. Anterior Thoracotomy
C. Posterior Thoracotomy
D. Median Sternotomy
E. The Minithoracotomies
III. Thoracic Incisions
Thoracotomy
E.
F.
Minithoracotomy
Drainage of the Pleural Space
IV. Common Major Elective Thoracic
Operations
A.
B.
Pulmonary Resections
1. Pneumonectomy
2. Lobectomy
3. Segmentectomy
4. Wedge and Limited Resections
5. Laser-Assisted Parenchyma-
Sparing Pulmonary Resection
Surgery of the Thoracic Esophagus
and Aorta
390
I. INTRODUCTION
There are a number of common surgical consid-
erations that have important routine management
implications for the anesthesiologist (Table 10-1).
First, the anesthesiologist should be aware of the
planned patient position and how to position the
patient correctly. Knowledge of the position of the
patient may influence or determine the timing of
such anesthetic procedures as epidural catheter
placement, location of intravenous and monitoring
lines, and whether anesthesia is induced on the
transporting gurney or on the operating room ta-
ble. In addition, the anesthesiologist is responsible,
at least in part, for positioning the patient and
avoiding complications related to patient position.
Second, the anesthesiologist should be familiar
with the various thoracic incisions. The time of
making the incision is important to the anesthe-
siologist, for this is when he must closely observe
the patient and operating field (watch for signs of
inadequate anesthesia, the color of the shed blood,
and the opening of the pleura [which entails stop-
ping breathing and initiating one-lung ventila-
tion]). The anesthesiologist must also closely ob-
serve the closure of the pleura (sigh patient to
reverse atelectasis, expel air and fluid from the
pleural space) and the skin incision so that the
administration or discontinuance of drugs (reversal
of paralysis, discontinuance of inhalational drugs,
administration of intravenous narcotics) can be
Table 10-1 SURGICAL EVENTS THAT
REQUIRE THE
ANESTHESIOLOGIST'S
ATTENTION
1. Positioning
2. Skin incision
a. Color of shed blood
b. Signs of inadequate anesthesia
3. Pleural incision
a. Freedom of movement of lung beneath pleura
b. Cessation of breathing
c. Initiation of one-lung ventilation
4. Pulmonary resection
a. Prevent any patient movement while vessels are being
secured
b. Note whether pulmonary veins are taken before pulmo-
nary arteries (causes entrapment of much more blood
within specimen)
c. Observe surgical field for blood loss and compression of
vital structures
d. Test integrity of bronchial stump with sigh
5. Pleural closure
a. Sigh to reverse atelectasis
b. Sigh to expel air and fluid from pleural space while last
stitch is being placed
6. Skin closure
a. Time administration or discontinuance of anesthetic
drugs
made with maximum precision. Third, the anesthe-
siologist should have some idea of the sequence of
events that occur during the surgical procedure so
that autonomic reflexes, arrhythmias, blood loss,
and changes in gas exchange may be more readily
anticipated and responded to. Consequently, this
chapter considers, in sequence, positioning of the
patient, the various thoracic incisions, and briefly
the common major thoracic operations.
II. POSITIONING THE PATIENT
Following the induction of anesthesia and dou-
ble-lumen tube insertion, the patient is ready to be
positioned for the surgical procedure. Although it
is not the primary responsibility of the anesthesiol-
ogist to position the patient, in practice this duty
falls at least in part on the anesthesiologist. In
addition, the anesthesiologist is liable for compli-
cations caused by patient position and, therefore.
the anesthesiologist should actively contribute to
preventing these complications. Consequently, this
section is written with these duties in mind and
describes the correct positioning of the patient for
the various thoracic incisions (see next section).
Thoracotomy
A standard posterior lateral thoracotomy is the
incision of choice for the majority of intrathoracic
operations and requires that the patient be placed
in the lateral decubitus position (Fig. 10-1). A
folded gauze pad or towel is placed just caudad to
the axilla to prevent compression of the axillary
structures against the rib cage. The hands and arms
are first outstretched in the position of an athlete
performing the hammer throw and are supported
either on special, curved, padded high/low arm
boards or on pads placed below the lower elbow
and between the two elbows. This brings the ver-
tebral border of the scapula forward, thus permit-
ting the incision to be extended superiorly as high
as necessary. The thigh closest to the table is
flexed slightly and the knee bent; the thigh and leg
in the superior position are kept straight. This pose
maintains the hips and trunk in a vertical plane. A
pillow is placed between the thighs, knees, and
legs to avoid pressure on the dependent extremity
by the sheer weight of the nondependent extrem-
ity. Some surgeons prefer a small pillow roll or
foam-rubber pad to fill the contour of the waistline
above the operating table and perhaps to extend
the uppermost intercostal spaces. Small blanket
rolls are placed anteriorly and posteriorly, snugly
tied to each other, and pulled against the chest wall
Lateral Decubitus Position for
Posterior Lateral Thoracotomy
Flex
Thigh
Pillow Between
Thighs, Knees,
Legs
Bend
Knee
Inflatable Plastic
Bead Bag or Small
Rolls Snugly Against
Anterior and Posterior
Chest
Small Roll Just
Caudal to Dependent
Axilla
Big Pads or Pillow
Underneath and
Between Bent Elbows
Inset:
Alternative
Method of
Arm Fixation
Utilizing Arm
Boards
Superior Leg
Nearly Straight
Hip and Trunk
in Vertical
Plane
Towel to Fill
Waist-Line
Contour
Elevate Arm Superiorly
to Bring Vertebral
Border of Scapula
Forward and Release
Tension on Nondependent
Axilla
Figure 10-1 Lateral decubitus position for the performance of a posterior lateral thoracotomy.
front and back. Alternatively, an inflatable plastic
bead bag may be used. A 3- or 4-inch strap of
adhesive tape is placed across the hip and attached
to the table to steady the patient in position. Such
fixation places no pressure on the skin of the pa-
tient and permits tilting of the table as required
(see Fig. 10-1).
With the patient in the supine position the hemi-
thorax to be operated upon is elevated 30 degrees
on a folded blanket, sponge-rubber pad, or the
newer pad filled with plastic beads (Fig. 10-2).
The patient's ipsilateral arm may be elevated and
carefully fixed to a cross brace, without causing
pressure on the neurovascular structures of the arm
or excessive stretch on the brachial plexus. In
other instances, the ipsilateral elbow can be bent
slightly, and the hand can be placed palm down
under the hip, to hold the arm away from the
lateral chest wall, again avoiding pressure upon, or
angulation of, blood vessels and nerves.
The standard posterior thoracotomy, made with
the patient in a prone position, is not used often
today (except for removal of dumbbell-shaped
neurogenic tumors of the thorax) (see Fig. 10-7).
The patient lies prone on the table, and a foam-
rubber or blanket pad is placed under the side
opposite that to be operated on, elevating it
slightly (about 10 degrees) (see Fig. 10-6). Sup-
ports may be placed underneath the upper chest
and pelvis to allow free diaphragmatic movement.
The arm is moved cephalad so that the scapula is
drawn away from the site of surgery. Many years
ago, in 1949, Overholt and Langer had a table
especially designed for this position to permit the
surgeon to sit while performing the operation. For
a number of reasons, not the least of which are
pressure injuries to skin, brachial plexus, and axil-
lary vessels, this position is not presently used.
The patient is supine with arms by the sides (see
Fig. 10-8A). A towel roll is placed underneath and
across the shoulders, and the head is slightly ex-
tended on the neck and placed in a sponge dough-
nut for stability. The shoulder roll and head exten-
sion provide access to the supersternal notch,
where the incision is begun.
E. The Minithoracotomies
There are numerous surgical diagnostic proce-
dures that require small incisions. The scalene
Supine Position for Anterior Thoracotomy
Elevate Ipsilateral Arm
Without Stretching Axilla Elevate Operative
Hemithorax 30 on
Blanket on Pad
Figure 10-2 Supine position with elevated arm for performance of an anterior thoracotomy. The elevated arm can be stabilized
on an ether screen.
394 Routine Surgical Considerations that have Anesthetic Implications
node biopsy is made with the patient in the supine
position with the head turned slightly to the oppo-
site side of the incision. Mediastinoscopy is per-
formed in the supine position with the patient's
head in moderate extension (incision is made over
the suprasternal notch). Mediastinotomy is made
in the supine position (incision is over the second
costal cartilage). Other diagnostic procedures such
as pleural biopsy, thoracentesis, needle biopsy of
the pleura, and thoracoscopy are most often done
under local anesthesia. If these procedures are per-
formed under general anesthesia, they are done in
the lateral decubitus position, with the affected
side up to facilitate the examination.
In a patient with a diffuse pulmonary infiltrate,
open-lung biopsy is the procedure that most likely
will establish a definitive diagnosis. Despite a thor-
ough history and physical examination, collection
of sputum for cytologic studies and cultures, and
examination with a flexible fiberoptic endoscope
(including visualization, brushing, lavage, and
possibly biopsy of pulmonary parenchyma via
either the transbronchial and/or percutaneous
routes), approximately one third of patients with
diffuse infiltrates will come to open-lung biopsy
because of negative results from these less direct
methods. These pulmonary examinations may fail
because of a contraindication to transbronchial or
percutaneous biopsy or because of rapidly progres-
sive clinical deterioration for which an immediate
diagnosis is needed.
1
The decision to perform an
open-lung biopsy is based on the likelihood that
pathologic examination of the tissue obtained will
yield specific information about the cause of the
disease process and that this information can be
used to alter the treatment being received by the
patient. For an open-lung biopsy, the incision may
be made anywhere in the thorax, as determined
preoperatively by the roentgenographic distribu-
tion of the disease process.
1
In most patients with
diffuse lung disease, an anterior incision provides
excellent exposure with minimal postoperative dis-
comfort; thus, the patient is most often in the su-
pine position, and a short submammary skin inci-
sion is utilized. An attempt is made to obtain the
biopsy specimen from the "transition" zone be-
tween diseased and normal-appearing lung tissue.
A wedge of tissue is taken, and hemostasis and
pneumostasis are achieved with either a running
absorbable suture or stapling device. Routine
drainage of the pleural space is done usually with
a single chest tube. The "minithoracotomy" (see
later discussion) usually utilizes a small transaxil-
lary approach. The procedure is performed with
the patient in the lateral decubitus position, with
the arm extended upward over the patient's head.
Otherwise, the considerations are the same as for
the open lung biopsy.
III. THORACIC INCISIONS
2
The anesthesiologist must pay close attention to
the operating field during the initial and final
phases of thoracotomy (Table 10-1). The attention
at the initial incision is directed to the color of the
shed blood, signs of sympathetic nervous system
stimulation, and the opening of the pleura (observe
how freely the lung moves below the pleura, stop
ventilating the patient to avoid having the surgeon
incise the lungs beneath the pleura, and time the
initiation of one-lung ventilation). Attention to the
closure of the incision is important to expand the
lungs fully but slowly (to avoid re-expansion pul-
monary edema),
3
under direct observation just be-
fore closure (reverse atelectasis), and as the last
stitch closing the pleura is tied (expel all possible
air and fluid from the pleural space) and to time
the administration or discontinuance of the various
anesthetic drugs, depending on whether a smooth
transition to postoperative mechanical ventilation
in the recovery room or intensive care unit is de-
sired as opposed to extubation in the immediate
postoperative period. The usual incisions provid-
ing access to the thorax are posterior lateral, ante-
rior, and posterior thoracotomy and median ster-
notomy. If the patient is to undergo postoperative
irradiation, the surgical incision may be made so
as to keep the incision outside the irradiated field
so as to ensure good healing.
Thoracotomy
A standard posterior lateral thoracotomy is the
incision of choice for the majority of intrathoracic
operations. This incision provides good access to
all areas of the lung, lung hilum, and most of the
mediastinum.
With the patient in the lateral decubitus position
(see Positioning the Patient), the skin incision is
begun at the anterior axillary line at the level at
which the surgeon wishes to enter the chest, usu-
ally the fifth intercostal space, and runs posteriorly
several fingerbreadths below the angle of the scap-
ula, and then turns superiorly to run midway be-
tween the vertebral border of the scapula and the
spinous processes of the vertebrae (Fig. 10-3).
In the non-muscle-sparing approach, dissection is
then made to expose the rib cage by dividing the
serratus anterior, latissimus dorsi, trapezius, and,
perhaps, the rhomboid muscles (Fig. 10-35) and
in the muscle-sparing approach by developing ap-
propriate skin and muscle flaps (see Figs. 10-4
and 10-5). Entry into the pleural space can then
be gained by either resecting a rib or using an
intercostal space. If the rib is to be resected, the
periosteum is incised and then elevated from the
bone. The rib is divided with a rib shears at the
costotransverse junction posteriorly and at the an-
terior axillary line. It is necessary to ligate and to
divide the intercostal bundles if the rib is to be
divided, to obviate tearing these vessels when the
ribs are subsequently spread. If the intercostal in-
cision is used, it is extended down to the pleura,
while injury to the intercostal vessels and nerve is
avoided. Entry into the pleural space is gained by
making an incision through the bed of the resected
rib or through the exposed pleura in the intercostal
space. In either instance, positive pressure should
be removed from the airway system to permit the
lung to collapse away (inward) from the thoracic
wall. If adhesions are present between the visceral
and parietal pleura (i.e., the lung does not move
"freely" in the pleural space), these must be di-
vided, and if vascular, they must be ligated.
Closure of the posterior lateral thoracotomy is
made by first placing several sutures pericostally
above and below the ribs adjacent to the intercostal
incision (Fig. 10-3C). The intercostal nerve, ar-
tery, and vein can be dissected free from the costal
groove or protected by inclusion of a thick layer
of the intercostal muscle. There may be less pain
in the postoperative period if the intercostal nerve
is excluded from the pericostal suture. No attempt
is made to approximate the parietal pleura as a
separate layer, and the parietal pleura is included
in the suture of the intercostal muscles. These
muscles are approximated with either continuous
or interrupted sutures. All layers of the divided
extracostal muscles should be reapproximated an-
atomically.
The main disadvantages of the posterior lateral
thoracotomy are due to the division of the major
Posterior Lateral Thoracotomy
Latissimus dorsi
muscle
B. Incision
Serratus
anterior
muscle
C. Closure
Trapezius
muscle
Pericostal suture
Figure 103 Posterior lateral thoracotomy. A, Outline of incision proceeding anteriorly to posteriorly following the course of the
fifth or the sixth rib and then around the angle of the scapula to the head superiorly between the scapula and vertebral column. B,
The extracostal muscles have been divided or retracted, and the incision through the periosteum of the fifth rib is shown. C. Closure
of the thoracotomy wound, with pericostal sutures in place.
chest wall muscles (latissimus dorsi, trapezius,
rhomboid major, and serratus anterior) and include
increased blood and operating time, severe post-
thoracotomy pain, ineffective coughing and poor
performance of chest physiotherapy exercises, lim-
ited ipsilateral shoulder mobility, and delayed am-
bulation. These limitations/deficits obviously in-
crease postoperative morbidity. Several authors
described the technique of muscle sparing thora-
cotomy and found a remarkable decrease in the
complications that occur with the standard thora-
cotomy.
4-8
However, others found only slight
benefits.
9
10
The limited lateral thoracotomy has been used
as a standard incision for many pulmonary opera-
tions including lobectomy, pneumonectomy,
wedge resection, resections of blebs or bullae, and
decortication.
4-10
In addition, the incision has been
used for automatic implantable cardioverter defib-
rillation insertion." The use of the lateral thoracot-
omy approach to the vertebral column is new and
can be used to treat fractures, spinal deformities,
and destructive lesions secondary to tumor or
infection.
12
13
Midthoracic lesions can be ap-
proached through a lateral thoracotomy, and lower
thoracic and lumbar lesions can then be ap-
proached with a thoracoabdominal incision ex-
tending into the flank and entering the retroperito-
neal space. The choice of laterality is dependent
on any concomitant scoliosis that may place the
apex of the curve closer to one side or the other.
13
For the limited lateral thoracotomy, the skin in-
cision extends from the submammary fold at the
anterior axillary line to a point just below the scap-
ular tip (see Fig. 10-4, left panel). Extensive ceph-
alad and caudad subcutaneous flaps are developed
to free the latissimus dorsi and serratus anterior
muscles. Coagulation of small blood vessels be-
tween the subcutaneous tissue and muscle is essen-
tial to prevent wound hematoma. The anterior bor-
der of the latissimus is freed and retracted
posteriorly, exposing the poster edge of the serra-
tus, which is then retracted anteriorly (see Fig. 10-
4, right panel). Two retractors are placed at right
angles to each other and are opened incrementally.
The chest is usually entered through the fifth or
sixth intercostal space. Rib resection is not re-
quired. As can be seen from Figure lO^l, the
limited lateral thoracotomy provides less exposure
than the standard posterior lateral thoracotomy,
and one-lung ventilation (collapse of the operative
lung) is extremely useful. No muscle closure is
Figure 10-4 Left panel, The stippled area of the inset represents the generous subcutaneous dissection that allows adequate
mobilization of the latissimns dorsi and serratus anterior musculature. Right panel, After the chest had been entered, one rib
retractor was positioned to retract the ribs, and a second retractor separated the serratus anterior and latissimus dorsi muscles.
(Based on Hazelrigg SR, Boley TM, Schmaltz RA, Johnson JA, Curtis JJ: The effect of muscle-sparing versus standard posterolat-
eral thoracotomy on pulmonary function, muscle strength, and postoperative pain. J Thorac Cardiovasc Surg 101:394-401, 1991.
Anterior Thoracotomy
Figure 105 Anterior thoracotomy. A, Outline of skin incision. B, Pectoralis major muscle separated and retracted to expose the
anterior chest wall. C, Placement of sutures pericostally above and below the adjacent ribs to close the intercostal incision as well
as to approximate the divided costal cartilage.
required because the serratus and latissimus are
allowed to resume their normal positions. A drain
is positioned to evacuate the subcutaneous flap.
The limited lateral thoracotomy has not, how-
ever, become widely accepted, probably because
of the required extensive subcutaneous detachment
of the latissimus dorsi (down to the iliac crest) (see
Fig. 10-4, left panel), the use of two overlapping
rib retractors that obscure the surgical field (see
Fig. 104, right panel), the impossibility of using
it in muscular patients, and the belief that much of
the postoperative pain is due to rib frac-
ture/dislocations and costochondral separations.
14
The muscle-sparing incision requires more time to
open because of the dissection of the skin and
muscle flaps, but the time to close is much less
because time is saved in suturing the latissimus
dorsi and serratus anterior muscles.
The left thoracoabdominal incision is used for
esophagectomy and exposure of the stomach, dia-
phragm, spleen, and thoracoabdominal aortae.
15
The incision involves extension of a standard left
posterolateral thoracotomy incision anteriorly to-
ward a point midway between the umbilicus and
the xiphoid process. The left arm and the neck are
included in the operative field if a cervical esoph-
ageal anastomosis is anticipated.
A standard anterior thoracotomy provides speed
of entry into the chest and minimal disturbance of
respiratory and circulatory functions, and closure
is relatively easy. It provides adequate exposure to
the anterior mediastinum, to the anterior portions
of the upper lobes, and to areas necessary for some
cardiac procedures. Lower lobe segments are dif-
ficult to expose through this incision, especially on
the left, where the heart would have to be re-
tracted, leading to decreased cardiac output and
arrhythmias.
The incision extends from the lateral edge of the
sternum in a curvilinear manner to the midaxillary
line (see Fig. 10-5). In men, the incision parallels
the intercostal space to be entered, and in women,
it is made along the inframammary fold. The pec-
toralis major and minor are then divided and re-
tracted (Fig. \0-5B). The intercostal muscles are
divided down to the parietal pleura, so that the
surgeon can determine whether a "free" pleural
space is present by observing lung motion with
respiration. If this space is present, the pleura can
be incised safely and the lung permitted to retract
from the chest wall. Frequently, when performing
an anterior thoracotomy, it is unnecessary to resect
or to divide a rib to gain adequate exposure, al-
though one or two costal cartilages may be divided
for greater access to the pleural space.
The incision through the pleura and intercostal
muscles is carried posteriorly toward the axilla,
and the fibers of the serratus are split back to the
long thoracic nerve, which is preserved. The latis-
simus dorsi muscle is usually not divided but can
be retracted posteriorly, if necessary, to obtain
greater exposure.
Closure of the anterior thoracotomy is essen-
tially the same as that described for the posterior
lateral incision (Fig. 10-5C). The pectoralis major
and minor muscles must be anatomically reat-
tached.
The standard posterior thoracotomy made with
the patient in a prone position is not often used
today. The advantages claimed for this incision in
the past included better ventilation of the lung not
under operation and reduced possibility of aspira-
tion of infective secretions into the unoperated
side. However, the risk of injury to the skin, bra-
chial plexus, and axillary vessels is high. In addi-
tion, without use of a double-lumen tube, the pre-
vention of aspiration is not certain. The patient is
difficult to position, and exposure is poor for an-
teriorly located lesions and is especially poor for
exposure of the anterior hilar and mediastinal
structures.
The skin incision for the posterior thoracotomy
extends from the level of the spine of the scapula
superiorly, along the midline between the spinous
processes of the vertebrae and the vertebral border
of the scapula, and swings laterally around the
angle of the scapula to about the midaxillary line
or slightly posterior to that point (Fig. 10-6A). The
muscles are incised as previously described for a
standard posterior lateral thoracotomy (Fig. 10-
6B). The closure of the incision is similar to that
used in the posterior lateral approach (Fig. 10-6C).
The posterior approach is essential for excision
of neurogenic tumors of the thorax.
16-18
When ex-
tension into the spinal canal has been demon-
strated, a one-stage, two-team approach is in-
dicated first to remove the intraspinal canal
extension by the appropriate hemilaminectomy
and then to remove the intrathoracic portion of
the growth. As a rule, this may be accomplished
through a single incision (Fig. 10-7). The patient
is placed in the prone position and the incision
begun in the midline above the level of the highest
extension of the intraspinal portion of the tumor
and carried inferiorly to below the lower level of
the tumor and then swung ipsilaterally to the mid-
axillary line. The vertebral column is exposed and
ipsilateral hemilaminectomy of one or more of the
vertebrae is carried out to enable mobilization of
the intraspinal canal portion of the tumor. The
pleural space is opened posteriorly with or without
a partial rib section to expose the intrathoracic
portion of the tumor. The intraspinal extension of
the tumor may then be excised or, when possible,
extracted through the enlarged intervertebral fora-
men and removed in toto with the intrathoracic
portion of the tumor.
Longitudinal median sternotomy is the proce-
dure of choice for exposure of the anterior medias-
tinum, great vessels, and the heart. By and large,
other incisions, such as transverse sternotomy with
bilateral thoracotomy and partial sternotomy with
partial resection of the clavicle or disarticulation at
the sternal clavicular joint, have been abandoned.
Recently, median sternotomy has gained popular-
ity as an approach to the excision of bilateral pul-
monary processes such as excision of bilateral
multiple metastatic pulmonary nodules and bilat-
eral pulmonary bullous disease,
19-21
22
and simul-
taneous resection of synchronous bilateral bron-
chogenic carcinoma.
23
Because the median
sternotomy approach has been thought to be asso-
ciated with less postoperative pain and respiratory
dysfunction, it has even been recommended for
routine pulmonary resections.
23
24
Although pul-
monary resections may be accomplished by this
route, the anatomic disadvantages are similar to
those of the standard anterior thoracotomy noted
previously (left lower lobectomy is especially dif-
ficult) and usually require one-lung ventilation
(collapse of the operative lung).
The skin incision is made from just below the
top edge of the manubrium at the supersternal
notch to about 2 to 3 inches below the xiphoid
process (Fig. 10-&4). The incision is not carried
higher, since if a tracheostomy is necessary, it may
Routine Surgical Considerations that have Anesthetic Implications
399
Prone Posterior Position Thoracotomy
Elevate Nonoperative
Hemithorax 10
Raise Arm
Bend Elbow
Latissimus
dorsi muscle
B. Incision
Serratus anterior muscle
Scapula
Scapula
Figure 10-6 Posterior thoracotomy. A, Outline of skin incision. B, Latissimus dorsi and serratus anterior muscles are divided,
and the scapula and upper margins of the muscles are retracted cephalad. C, Reapproximation of muscles by sutures.
Median Sternotomy Supine Position
Chin Elevated
or Neck Slightly
Extended
Towel Under
Shoulders
Arms at
Side
Figure 10-8 Median sternotomy. A. Incision from just below the suprasternal notch to below the top of the xiphoid. B. Blum
dissection of the posterior aspect of the sternum at the suprasternal notch. C, Division of the sternum from below upward with the
use of an oscillating saw. D, Closure of the divided sternum with wire sutures.
Should untoward circumstances arise (e.g., marked
hemorrhage), the incision can be extended.
For young patients with recurrent pneumo-
thorax, management by pleurectomy and/or exci-
sion or ligation of bullae via a triangle of auscul-
tation thoracotomy, without transsecting chest wall
muscles, has been advocated.
27
This approach
through a limited thoracotomy is possible because
of the flexibility of the posterior vertebral joint in
young people. Indeed, in infants this approach may
be used for the repair of esophageal atresia.
28
If
needed, conversion to a full thoracotomy is simple.
F. Drainage of the Pleural Space
After thoracotomy, the pleural space should be
drained, especially when there has been resection
of lung tissue or a portion of the chest wall. For
tube placement in a posteroinferior location, a skin
incision 2 cm long is made just anterior to the
posterior axillary line in about the eighth intercos-
tal space. A heavy duty 8-inch hemostat or forceps
is inserted into the skin incision at an angle and is
pushed through the intercostal muscles so that the
forceps enters the pleural space cephalad to the
skin incision. The forceps then pulls the distal end
of the chest drain out of the chest cavity; this
prevents contamination of the proximal end of the
drain (the indwelling end) by contact with the skin.
By palpating the level of the dome of the dia-
phragm, the proximal end of the tube is positioned
so that the proximal tip of the tube is at the dome
and the side opening of the tube is in the sulcus.
For anterior tube placement the tube is inserted
just lateral to the border of the pectoralis major
muscle. A skin incision is made at the level of the
third intercostal space, and with cephalad direc-
tion, the tube usually enters the chest through the
second intercostal space. The tip of the catheter is
fixed to the anterior parietal pleura by a slip knot
to keep the tip from becoming angulated medially
and lying adjacent to the great vessels of the me-
diastinum, where, by constant motion, penetration
of the wall of the vessel could conceivably occur.
If such movement of the tip of the tube is noted
on postoperative roentgenograms of the chest, the
tube should be withdrawn promptly. The loosely
tied slip-knot suture permits easy removal of the
tube. A firm stitch is used to anchor any chest
drain to the skin, and a separate loose mattress
suture is placed around the drain so that this can
be cut and tied to seal the wound when the drain
is withdrawn postoperatively. See chapter 13 for a
complete discussion of the design and function of
a modern chest tube drainage system and potential
chest tube drainage system complications.
IV. COMMON MAJOR ELECTIVE
THORACIC OPERATIONS
In a typical modern thoracic surgery service,
pulmonary resections are the most frequent major
cases done, with esophageal and thoracic aortic
surgery making up most of the remainder. Opera-
tions such as thoracoplasty, drainage of empyema,
thoracoscopy, and phrenic nerve crush/evulsion
are infrequently performed or abandoned today.
Consequently, this section describes pulmonary re-
sections (pneumonectomy, lobectomy, segmente -
tomy) in moderate detail; the conduct of surgery
for esophageal and thoracic aortic operations is
described in detail in chapters 16 and 17, respec-
tively.
A. Pulmonary Resections
29
1. Pneumonectomy
Following one of the standard thoracic incisions
(see preceding section), entrance is made into the
pleural space. The pleura is then incised as it re-
flects upon the hilus anteriorly, superiorly, and
posteriorly. On the right, the azygos vein may be
isolated and divided to give greater access to the
hilus superiorly. Generally, the pulmonary artery
is the first structure to be isolated and divided
(rather than the pulmonary vein, which would lead
to increased blood loss [retained within the speci-
men]). After control of the pulmonary artery, the
superior and then the inferior pulmonary veins are
dissected, isolated, and divided, and the cut ends
are secured. All the remaining tissue reflections
are divided, and the main-stem bronchus is freed
up to its junction with the trachea. A bronchial
clamp is placed just distal to this junction and the
bronchus divided proximal to the clamp. The bron-
chial stump is then closed with a stapling device
(or tissue adhesive)/
0
and the anesthesiologist may
be asked to test the integrity of the bronchial clo-
sure with a sustained positive-pressure breath (the
surgeon may also immerse the bronchial stump in
saline to see whether gas bubbles through the sa-
line). Not infrequently, one or two bronchial arte-
ries will need to be ligated after the bronchus has
been divided. It is important that bleeding from
the bronchial vessels be controlled, since these
vessels may cause significant postoperative blood
loss. After closure of its proximal end, the bron-
chial stump is covered with adjacent tissue, such
as a pleural flap, the azygos vein, a pedicle graft
of pericardial fat, or adjacent pericardium. This is
done to provide the stump with a viable tissue
cover to help prevent the possible development of
a leak from the stump, which normally heals by
secondary intention.
Tracheal-sleeve pneumonectomy has been used
to excise high line carcinomas of either main-stem
bronchus. Resection of the tracheal carina with the
ipsilateral lung is carried out, and the contralateral
bronchus is then anastamosed to the distal end of
the trachea. When extensive pleural disease coex-
ists with parenchymal disease that requires a pneu-
monectomy, a pleuropneumonectomy may be per-
formed. Essentially, a plane of cleavage is
developed between the endothoracic fascia and the
parietal pleura, and the pleura and lung are freed
as one from the chest wall, diaphragm, and me-
diastinal structures down to the hilus. From this
stage on, the vessels and bronchus are managed as
in a standard pneumonectomy.
2. Lobectomy
After the pleural space is entered (see Thoracic
Incisions), the pleura is incised as it reflects upon
the hilar structures. The oblique (major) fissure is
then opened to expose the interlobar portion of the
pulmonary artery. The arterial branches of the lobe
to be resected are then divided, double ligated,
transfixed, and divided. After control of the arte-
ries, the veins draining the lobe are identified and
managed in a similar fashion. The lobar bronchus
is then isolated and clamped distally to the pro-
posed line of division. After division of the bron-
chus, the proximal end of the stump is closed with
a stapling device. Coverage of the stump with ad-
jacent tissue is practiced by most thoracic sur-
geons, although the remaining lung tissue gener-
ally will cover the stump effectively when the
remaining lung is re-expanded. Removal of the
lobe is then completed by division of any remain-
ing connections to the other lobe or lobes. The
integrity of the bronchial suture line is tested by
pouring saline into the pleural cavity to cover the
stump while sustained positive pressure of 30 to
40 cm H
2
0 is maintained by manual compression
of the rebreathing bag. Some gas usually escapes
during this maneuver, but it is generally from the
raw, now separated, surface of the remaining lung
and not from the bronchial stump.
A radical lobectomy may be utilized in certain
instances for treatment of carcinoma of the lung.
In this procedure the lymphatic nodes in the me-
diastinum are removed en bloc with the lobe. This
is most easily and satisfactorily performed on the
right upper lobe, although it may be done almost
as well on the left upper lobe. Both right and left
lower radical lobectomies are more difficult to per-
form, and it is doubtful whether an adequate radi-
cal lymphatic dissection is truly possible in these
anatomic locations. A lobectomy may also include
a sleeve resection of a portion of the main-stem
bronchus. This is indicated for the preservation of
lung tissue either when the pathologic lesion to be
removed is benign or when the patient has insuffi-
cient pulmonary reserve to tolerate a pneumonec-
tomy. The initial steps of the operative procedure
are carried out as in a standard lobectomy. The
arterial and venous branches to the lobe are iso-
lated, ligated, and divided. Next, the lobar bron-
chus and adjacent main-stem bronchus are dis-
sected free from the remaining vascular structures.
A clamp is placed proximally on the main-stem
bronchus, and the bronchus is then divided just
distal to the clamp. The main-stem bronchus is
divided a second time distal to the lobar orifice,
and this segment of the bronchus is removed along
with the resected lobe. The cut ends of the bron-
chus are then tailored, as necessary, to fit each
other and are reapproximated.
3. Segmentectomy
The initial steps in a segmentectomy (removal
of a discrete bronchopulmonary segment) are car-
ried out in the same manner as that in a lobectomy.
After exposing the hilar structures and opening the
major fissure, the pulmonary artery is identified,
and the arterial branch of the segment or segments
to be removed are identified, isolated, secured, and
divided. After control of the arterial supply, the
segmental veins and the bronchus are divided, al-
though at times the order of these procedures may
be reversed. Before dividing the bronchus, differ-
ential deflation and inflation of the segment to be
removed should be done by clamping and un-
clamping to help delineate the intersegmental
planes. It is to be remembered that filling of the
deflated segment may occur from the adjacent seg-
ments by means of collateral ventilation. Once the
segmental planes are identified, the bronchus is
divided and the proximal stump is closed with a
stapling device. Additional coverage of the stump
is not usually carried out. Separation of the dis-
eased segment is accomplished by blunt dissec-
tion. Vascular and small airway connections be-
tween the adjacent segments must be clamped,
divided, and ligated.
4. Wedge and Limited Resections
Wedge resection consists of excision of a dis-
eased portion of the lung (usually small) irrespec-
tive of bronchopulmonary segmentology. Entry is
made into the pleural space, and the area of lung
to be resected is identified. Stapling devices are
applied so that the specimen will consist of a mar-
gin of surrounding normal lung parenchyma as
well as the lesion. The specimen is removed by
cutting along the staples. These procedures are
usually short.
In a limited or local resection (small discrete
tumors that can be enucleated), the visceral pleura
is incised over the mass, and the mass is removed
by blunt and sharp dissection, with the raw sur-
face, blood vessels, and minor bronchial structures
being individually clamped and ligated or stapled.
The visceral pleura is reapproximated with suture.
These procedures are usually also short.
5. Laser-Assisted Parenchyma-Sparing
Pulmonary Resection
Several studies described the use of a neodym-
ium:yttrium aluminum garnet (Nd:YAG) laser to
resect pulmonary lesions.
22
31
32
The power setting
used for all resections was 30 to 60 W continuous-
wave energy. All personnel were required to wear
protective glasses specifically for 1060 nm. All
windows into the operating room were shaded, and
access to the room was limited while the laser was
in use. The surgeon controlled the energy delivery
to the tissue with a foot switch. A thermal printer
attached to the laser recorded the number of pulses
and the output so that total energy delivery of the
procedure could be documented.
The anesthetic was delivered through a double-
lumen endotracheal tube in most patients. The le-
sion was located usually through a muscle-sparing
thoracotomy, but sometimes through a standard
thoracotomy or median sternotomy (bilateral me-
tastasis), and delivered into the field. At this point,
the lung was deflated. An incision in the visceral
pleura was made. For superficial lesions, an encir-
cling incision was used first. For deep-seated le-
sions, the lung was incised down to the lesion
before a circumscribing incision was made. With
gentle traction and countertraction between the le-
sion and the surrounding lung tissue, a plane was
developed evenly around the lesion. Suction was
always placed in close proximity to the laser to
remove the smoke plume. Periodic palpation dur-
ing the procedure ensured complete excision of the
tumor (and no excision of normal lung) and al-
lowed the surgeon to adjust the depth of the laser
excision. When the lesion had been removed, a
coagulated crater remained. Intraoperative blood
loss has been minimal. The pleura and wound mar-
gins may be reapproximated or the crater left open
and the lung re-expanded. A single chest tube has
been placed and the thoracotomy incision closed
in the usual fashion.
Postoperative air leak has not been a problem,
and early postoperative pulmonary function equals
preoperative pulmonary function. During the early
postoperative period, chest roentgenograms fre-
quently will show a "hal o" at the site of laser
excision. This area eventually collapses and dis-
appears, a process that may take 6 months to 1
year. Such a scar may be difficult to evaluate in
follow-up of malignant lesions, especially on CT
scan. Theoretically, since the Nd:YAG laser de-
stroys up to 4 mm beyond the cut edge, microme-
tastasis near the primary lesion should be de-
stroyed, and this scar should be of no great
concern.
B. Surgery of the Thoracic Esophagus
and Aorta
These surgical procedures are extensively de-
tailed in chapters 16 and 17; this section is in-
tended as a brief overview of surgical inci-
sions/approach. Resectional surgery of the
midthoracic esophagus is technically very compli-
cated (immediate mortality is highapproxi-
mately 20 per cent) and may involve an additional
proximal cervical incision (to approach the upper
third of the esophagus) and a distal abdominal
incision (to approach the lower third of the esoph-
agus). The midthoracic esophagus can be ap-
proached through either a left-sided or right-sided
thoracotomy (because the midthoracic portion of
the esophagus, except for its most distal part, is
primarily on the right side of the mediastinum).
Advantages of a left thoracotomy include the abil-
ity to split open the left diaphragm, mobilize the
stomach, and bring it up into the chest without
having to change the patient's position. The pa-
tient should be placed in true lateral thoracotomy
position for a left thoracoabdominal esophagec-
tomy, rather than elevating the chest only 45 de-
grees because the full lateral position allows the
anastomosis to be performed more easily with the
heart lying in a more medial position.
33
A right
thoracotomy has the advantage of allowing the
surgeon to perform a by-pass between the esopha-
gus and the fundus of the stomach with complete
relief from dysphagia in the presence of an unre-
sectable lesion. Also, if the azygos vein is infil-
trated by tumor, it is safer to dissect it from the
right side. The disadvantages of a right thoracot-
omy are that the position of the patient must be
changed after the abdominal part of the operation
(laparotomy followed by right thoracotomy is
called the Ivor Lewis technique) is completed. In
addition, after an extensive procedure in the abdo-
men and an opening of the chest, the lesion may
be found to be unresectable.
Esophageal surgery involves a number of other
controversies; they include questions of palliation
versus cure and degree of blood loss with transhia-
tal esophagectomy (pull-through esophagectomy
without thoracotomy) followed by cervical esoph-
agogastric anastomosis versus transthoracic esoph-
agectomy,
34
delayed versus immediate reconstruc-
tion, and the use of colon or stomach to restore
continuity. The multiple surgical options available
for treatment of esophageal strictures include mul-
tiple dilations, antireflux procedures, esophago-
plasty, and resection and replacement with inter-
posed segments of the gastrointestinal tract (colon,
stomach).
35
When patients can be dilated (most
patients), a subsequent standard transthoracic anti-
reflux procedure combined with intraoperative di-
lation provides good results (eliminates esophagi-
tis and results in comfortable swallowing). If the
stricture cannot be dilated, then resection and re-
placement with healthy tissue are performed.
35
36
Similarly, repair of thoracic aortic aneurysms
requires a high degree of technical sophistication
and also usually involves proximal and distal per-
fusion connections and considerations (see chapter
16 for an extensive discussion). Since the aortic
arch lies over the esophagus on the left, a left fifth
intercostal space thoracotomy is usually used for
thoracic aortic surgery. The mediastinal pleura is
usually opened over its entire length to allow for
mobilization of the entire thoracic aorta. Signifi-
cant controversy exists regarding the best method
for by-passing the thoracic aorta (see chapter 17).
Further description of the ensuing steps of proxi-
mal and distal control of the esophagus and aorta
is beyond the scope of this chapter.
REFERENCES
1. Walker WA, Cole HF, Khandekar A, Mahfood SS, Watson
DC: Does open lung biopsy affect treatment in patients
with diffuse pulmonary infiltrates? J Thorac Cardiovasc
Surg 97:534-540, 1989.
2. Lees WM: Thoracic incisions. In Shields TW (ed): General
Thoracic Surgery. 2nd ed. Philadelphia, Lea & Febiger,
1983, pp 305-314.
3. Mahfood S, Hix WR, Aaron BL, Blaes P, Watson DC:
Reexpansion pulmonary edema. Ann Thorac Surg 45:340-
345, 1988.
4. Ashour M: Modified muscle sparing posterolateral thora-
cotomy. Thorax 45:935-938, 1990.
5. Bethencourt DM, Holmes EC: Muscle-sparing posterolat-
eral thoracotomy. Ann Thorac Surg 45:337, 1988.
6. Hazelrigg SR, Boley TM, Schmaltz RA, Johnson JA, Cur-
tis JJ: The effect of muscle-sparing versus standard pos-
terolateral thoracotomy on pulmonary function, muscle
strength, and postoperative pain. J Thorac Cardiovasc Surg
101:394-401, 1991.
7. Karwande SV, Pruitt JC: A muscle-saving posterolateral
thoracotomy incision. Chest 96:1426-1427, 1989.
8. Mitchell RL: The lateral limited thoracotomy incision:
Standard for pulmonary operations. J Thorac Cardiovasc
Surg 99:590-596, 1990.
9. Poon RB: Comparison of later pulmonary function after
posterolateral and muscle-sparing thoracotomy. Ann
Thorac Surg 53:675-679, 1992.
10. Lemmer JH, Gomez MN, Symreng T, Ross AF, Rossi NP:
Limited lateral thoracotomy: Improved postoperative func-
tion. Arch Surg 125:673-677, 1990.
11. Slater AD, Singer I, Stavens C, Springer WJ, Gray LA:
Lateral thoracotomy for the automatic implantable defibril-
lator. Arch Surg 126:778-781, 1991.
12. McElvein RB, Nasca RJ, Dunham WK, Zorn GL Jr: Trans-
thoracic exposure for anterior spinal surgery. Ann Thorac
Surg 45:278-283, 1988.
13. Westfall SH, Behooz AA, Merenda JT, Naunheim KS,
Conners RH, Kaminski DL, Weber TR: Exposure of the
anterior spine: Technique, complications and results in 85
patients. Am J Surg 154:700-704, 1987.
14. Heitmiller RF: Thoracic incisions. Ann Thorac Surg
46:601, 1988.
15. Heitmiller RF: The left thoracoabdominal incision. Ann
Thorac Surg 46:250-253, 1988.
16. Shields TW, Reynolds M: Neurogenic tumors of the
thorax. Surg Clin North Am 68:645-668, 1988.
17. Grillo HC, Ojemann RG, Scannell JG, Zervas NT: Com-
bined approach to "dumbbell" intrathoracic and intraspi-
nal neurogenic tumors. Ann Thorac Surg 36:402-407,
1983.
18. Ricci C, Rendina EA, Venuta F, Pescarmona EO, Gagliardi
F: Diagnostic imaging and surgical treatment of dumbbell
tumors of the mediastinum. Ann Thorac Surg 50:586-589,
1990.
19. Takita H, Merrin C, Didolkar MS, Douglass HO, Edgerton
F: The surgical management of multiple lung metastasis.
Ann Thorac Surg 24:359-364, 1977.
20. Cooper JD, Nelam JM, Pearson FG: Extended indications
for median sternotomy in patients requiring pulmonary re-
section. Ann Thorac Surg 26:413^118, 1978.
21. Roth JA, Pass HI, Wesley MN, White D, Putnam JB, Seipp
C: Comparison of median sternotomy and thoracotomy for
resection of pulmonary metastasis in patients with adult
soft-tissue sarcomas. Ann Thorac Surg 42:134-138, 1986.
22. Kodama K, Doi O, Higashiyama M, Tatsuta M, Iwanaga
: Surgical management of lung metastases: Usefulness of
resection with the neodymium:yttrium-aluminum-garnet
laser with median sternotomy. J Thorac Cardiovasc Surg
101:901-908,1991.
23. Dalton M, Coltharp WH: Simultaneous resection of syn-
chronous bilateral bronchogenic carcinoma using median
sternotomy. J Thorac Cardiovasc Surg 94:916-918, 1987.
24. Urschel HC Jr: In discussion: Cooper JD, Nelam JM, Pear-
son FG. Extended indications for median sternotomy in
patients requiring pulmonary resection. Ann Thorac Surg
26:419, 1978.
25. Daly BDT, Faling LJ, Diehl JT, Bankoff MS, Gale ME:
Computed tomographyGuided minithoracotomy for the
resection for the small peripheral pulmonary nodules. Ann
Thorac Surg 51:465-469, 1991.
26. Massimiano P, Ponn RB, Toole AL: Transaxillary thora-
cotomy revisited. Ann Thorac Surg 45:559-560, 1988.
27. Nazarian J, Down G, Lau OJ: Pleurectomy through the
triangle of auscultation for treatment of recurrent pneumo-
thorax in younger patients. Arch Surg 123:113-114, 1988.
28. Goh DW, Brereton RJ: Triangle of auscultation thoracot-
omy for esophageal atresia. J Thorac Cardiovasc Surg
103:14-16, 1992.
29. Shields TW: Pulmonary resections. In Shields TW (ed):
General Thoracic Surgery. 2nd ed. Philadelphia, Lea &
Febiger, 1983, pp 315-330.
30. Eng J, Sabanathan S: Tissue adhesive in bronchial closure.
Ann Thorac Surg 48:683-685, 1989.
31. LoCicero J, Frederiksen JW, Hartz RS, Michaelis LL:
Laser-assisted parenchyma-sparing pulmonary resection. J
32. Landreneau RJ, Hazelrigg SR, Johnson JA, Boley TM,
Nawarawong W, Curtis JJ: Neodymiuny.yttrium-aluminum
garnet laser-assisted pulmonary resections. Ann Thorac
Surg 51:973-978, 1991.
33. Mathisen DJ, Grillo HC, Wilkins EW Jr, Moncure AC.
Hilgenberg AD: Transthoracic esophagectomy: A safe ap-
proach to carcinoma of the esophagus. Ann Thorac Surg
45:137-143, 1988.
34. Orringer MB: Transthoracic versus transhiatal esophagec-
tomy: What difference does it make? Ann Thorac Surg
44:116-118, 1987.
35. Little AG, Naunheim KG, Ferguson MK, Skinner DB:
Surgical management of esophageal strictures. Ann Thorac
Surg 45:144-147, 1988.
36. Stirling MC, Orringer MB: The combined Collis-Nisson
operation for esophageal reflux strictures. Ann Thorac
Surg 45:148-157, 1988.
CHAPTER 11
Conventional and Differential
Lung Management of
I. Introduction
II. One-Lung Ventilation Situation
A. Nondependent, Nonventilated Lung
B. Dependent, Ventilated Lung
III. Conventional Management of One-
Lung Ventilation
A. Inspired Oxygen Concentration
B. Tidal Volume
C. Respiratory Rate
D. Dependent-Lung PEEP
IV. Differential Lung Management of
A. Intermittent Inflation of the
Nondependent, Operative Lung
B. Selective Dependent-Lung PEEP
C. Selective Nondependent-Lung
CPAP
D. Differential Lung PEEP and CPAP
E. Selective Nondependent-Lung High-
Frequency Ventilation
F. Re-expansion of the Lung and
Return to Two-Lung Ventilation at
the End of the Case
V. Recommended Combined
Conventional and Differential Lung
Management of One-Lung Ventilation
406
I. INTRODUCTION
Both the patient's lungs are ventilated with in-
termittent positive-pressure ventilation during the
induction of anesthesia, before and after insertion
of the double-lumen endotracheal tube, during po-
sitioning of the patient (in the lateral decubitus
position in the vast majority of cases), and during
the chest wall incision. However, once the pleura
has been incised, it is usually useful and helpful to
the surgeon to collapse the lung that is being op-
erated on (the nondependent lung) to facilitate sur-
gical exposure.
Before describing the various one-lung ventila-
tion management techniques, it is important to re-
view briefly the one-lung ventilation situation to
understand how the various one-lung management
techniques might influence the distribution of
blood flow between the two lungs during one-lung
ventilation. Conventional ventilatory management
of the one-lung ventilation situation is described
firstconventional management of one-lung ven-
tilation provides satisfactory gas exchange in the
majority of cases. In a few situations conventional
management of one-lung ventilation does not pre-
vent severe hypoxemia, and in these few cases the
addition of differential lung ventilation to conven-
tional mechanical ventilation almost always re-
solves the problem. Accordingly, the chapter ends
with a combined conventional and differential
lung management plan for providing adequate ar-
terial oxygenation during one-lung ventilation.
II. ONE-LUNG VENTILATION
SITUATION
Anesthesia for thoracic surgery is most com-
monly performed with the patient in the lateral
decubitus position, with the nondependent hemi-
thorax comprising the operative field. When one-
lung ventilation is employed, the nondependent
lung is the nonventilated and collapsed (atelec-
tatic) lung, and the dependent lung is the ventilated
lung. In this section (which essentially summarizes
part of chapter 4), it is important to remember that,
with a constant cardiac output, whatever increases
blood flow to one lung usually decreases blood
flow to the other, and vice versa.
A. Nondependent, Nonventilated Lung
The nondependent, nonventilated lung has a
ventilation-perfusion ratio of zero. Consequently,
one-lung ventilation creates an obligatory right-to-
left transpulmonary shunt through the nonventi-
lated, nondependent lung that is not present during
two-lung ventilation (see Fig. 4-7). Thus, one-
lung ventilation results in a much larger alveolar-
to-arterial oxygen tension difference and lower ar-
terial oxygen tension (P
a
0
2
) than does two-lung
ventilation.
1
Fortunately, blood flow to the nonde-
pendent, nonventilated lung is usually reduced by
both passive mechanical and active vasoconstric-
tion mechanisms that are usually operant and.
thereby, prevent the shunt from increasing and the
P
a
0
2
from decreasing as much as might be ex-
pected on a two-lung ventilation blood flow distri-
bution basis (Fig. 11-1). The passive mechanical
factors that decrease blood flow to the nondepen-
dent lung consist of the force of gravity, surgical
interference with blood flow (retraction and com-
pression of vessels, tying off of vessels), and a low
dependent lung airway pressure. The active vaso-
constrictor mechanism decreasing blood flow to
the nondependent lung is nondependent-lung
hypoxic pulmonary vasoconstriction (HPV) (Fig.
11-1).
B. Dependent, Ventilated Lung
Unfortunately, the dependent lung may have a
reduced lung volume and may be poorly ventilated
for several reasons (Fig. 11-1) (see chapters 3 and
4). First, in the lateral decubitus position the ven-
tilated, dependent lung usually has a reduced lung
volume owing to the combined factors of induc-
tion of general anesthesia and circumferential
(and, perhaps, severe) compression by the weight
of the mediastinum from above, by the abdominal
contents pressing against the diaphragm from the
caudad side, and by suboptimal positioning effects
(rolls, packs, shoulder supports) pushing in from
the operating room table. Reduced lung volume
usually results in low ventilation-perfusion and
atelectatic regions (see Fig. 3-27). Second, absorp-
tion atelectasis can also occur in regions of the
dependent lung that have low ventilation-perfusion
ratios when they are exposed to high inspired ox-
ygen concentration.
2
3
Third, difficulty in secretion
removal may also cause the development of poorly
ventilated and atelectatic areas in the dependent
lung. Finally, maintaining the lateral decubitus po-
sition for prolonged periods of time may cause
fluid to transudate into the dependent lung (which
may be vertically below the left atrium) and cause
further decreased lung volume and increased air-
way closure in the dependent lung.
4
A decrease in
lung volume and an increase in airway closure in
the dependent lung additively creates areas that
have low ventilation-perfusion ratios or atelec-
tasis.
2
The development of low ventilation-perfu-
sion and atelectatic areas in the dependent lung
can elicit dependent-lung HPV, resulting in in-
408 Conventional and Differential Lung Management of One-Lung Ventilation
One Lung Ventilation: Determinants
of Blood Flow Distribution
Figure 11-1 This schematic diagram shows the major determinants of blood flow distribution during one-lung ventilation. Blood
flow to the nonventilated, nondependent lung is reduced by the force of gravity, surgical interference (compression, tying off of
vessels), hypoxic pulmonary vasoconstriction, and/or lung disease (vascular obliteration, thrombosis). Blood flow to the ventilated,
dependent lung is increased by gravity; however, dependent lung blood flow and vascular resistance may be altered in either
direction depending on the method of ventilation (amount of airway pressure, F,0
2
, and the amount of dependent lung disease
and/or hypoxic pulmonary vasoconstriction). Factors that may increase the amount of dependent lung disease intraoperatively are
listed on the extreme right of the figure (see also chapter 3).
creased dependent-lung pulmonary vascular resis-
tance that will divert blood flow to the nondepen-
dent lung, thereby increasing the shunt across the
lungs.
In view of these factors, which can affect the
amount of nondependent-lung blood flow, the
method used to ventilate the dependent lung is
especially important for several reasons (Fig.
11-2). First, if the dependent-lung method of ven-
tilation involves an excessive amount of airway
pressure, due to either use of high positive end-
expiratory pressure (PEEP) levels or very large
tidal volumes, then dependent-lung vascular resis-
tance may be increased, which would increase
nondependent-lung blood flow. The deleterious ef-
fects of increased nondependent-lung blood flow
(shunt) may outweigh the beneficial effects of the
opening of atelectatic and low ventilation-perfu-
sion areas in the dependent lung. Second, if the
dependent lung is hyperventilated (which may re-
quire an excessive amount of airway pressure), the
resultant hypocapnia may inhibit HPV (see chapter
4). Third, if the dependent lung is ventilated with
too small of a tidal volume, dependent-lung atelec-
tasis may develop. Fourth, a high inspired oxygen
concentration to the dependent lung may vasodi-
late the dependentjung, enhancing nondependent-
lung HPV.
5
6
On the other hand, a high inspired
oxygen concentration to the low ventilation-perfu-
sion-dependent lung may promote dependent-lung
absorption atelectasis.
3
7
III. CONVENTIONAL MANAGEMENT
OF ONE-LUNG VENTILATION
The proper initial conventional management of
one-lung ventilation is logically based on the pre-
ceding determinants of blood flow distribution
during one-lung ventilation. In view of the fact
that one-lung ventilation incurs a high risk of caus-
ing systemic hypoxemia, it is extremely important
that dependent-lung ventilation, as it affects these
determinants, be optimally managed. This section
considers the usual management of one-lung ven-
tilation in terms of the most appropriate inspired
oxygen concentration, tidal volume, respiratory
rate, and dependent-lung PEEP level that should
be used (Table 11-1).
A. Inspired Oxygen Concentration
Although the theoretical possibilities of absorp-
tion atelectasis and oxygen toxicity exist, the ben-
efits of ventilating the dependent lung with 100
per cent oxygen far exceed the risks (Fig. 11-3)
(except, perhaps, in patients receiving bleomycin
preoperatively who may be sensitized to oxygen-
induced lung injury; see later discussion). A high
F,0
2
in the single ventilated lung may critically
increase the P
a
0
2
from arrhythmogenic and life-
threatening levels to safer levels. In addition, a
high F,0
2
in the dependent lung will cause vaso-
Method of Ventilation of the Dependent Lung
During One-Lung Ventilation Can Determine Amount of
Nondependent Lung Blood Flow
Figure 11-2 The method of ventilation of the dependent lung during one-lung ventilation can be a major determinant of the
amount of nondependent lung blood flow.
The Good Effects of a High Fj0
2
to Ventilate Dependent Lung Outweigh the Bad Effects
High FT0
2
Dependent Lung
Figure 11-3 The good effects of a high F, 0
:
to the ventilated dependent lung outweigh the bad effects. The good effects consist
of vasodilation in the dependent lung, which increases dependent lung blood flow (and decreases nondependent lung blood flow)
and may critically increase the P
a
O
:
. The bad effects are theoretical and consist of dependent lung absorption atelectasis and oxygen
toxicity.
Table 11-1 INITIAL CONVENTIONAL
VENTILATORY MANAGEMENT OF
ONE-LUNG ANESTHESIA
1. Maintain two-lung ventilation as long as possible.
2. Begin one-lung ventilation with tidal volume of 10 ml/kg.
3. Adjust the respiratory rate so that P
a
C0
2
= 40 mm Hg.
4. UseF,0
2
= 1.0.
5. Utilize frequent or continuous arterial P0
2
and Pco
2
monitoring.
dilation, thereby increasing the dependent-lung ca-
pability of accepting blood flow redistribution due
to nondependent-lung HPV.
6-8
Direct chemical
100 per cent oxygen toxicity will not occur during
the time frame of the operative period,
9
and ab-
sorption atelectasis in the dependent lung
3
is un-
likely to occur in view of the remaining one-lung
ventilation management characteristics (moder-
ately large tidal volumes with intermittent posi-
tive-pressure, low-level PEEP; see later discus-
sion). Although not previously studied during
one-lung ventilation, use of a dependent-lung F,0
2
of 0.8 to 0.9 may be ideal in view of the fact that
an F,N
2
of 0.1 to 0.2 greatly reduces the possibility
of absorption atelectasis (by allowing some nitro-
gen to splint open the low V/Q regions),
3
whereas
a reduction in F,0
2
of 0.1 to 0.2 (from 1.0) will
probably cause only a small decrease in P
a
0
2
(ob-
serve the flatness of the high isoshunt lines in Fig.
3-33).
It may desirable/indicated to decrease the F,0
2
to bleomycin-treated patients. Bleomycin is known
to induce interstitial lung disease and may sensi-
tize the lung to oxygen-induced lung injury. How-
ever, the relationship between increased, but nor-
mally nontoxic, concentrations of oxygen in
bleomycin-treated patients and in patients with
bleomycin-induced pulmonary fibrosis is contro-
versial. Although human and animal data suggest
that the use of elevated oxygen concentrations
(F,0
2
> 0.3) in bleomycin-treated patients in-
creases the risk of pulmonary damage,
10
~
17
the data
from humans fail to establish a cause and effect
relationship, and the animal data contain experi-
mental design aberrations such as direct adminis-
tration of bleomycin intratracheally (as opposed to
intravenous administration, as is done clinically).
The fact that bleomycin's antitumor activity is de-
pendent on the production of superoxide- and hy-
droxyl-free radicals, which can selectively damage
deoxyribonucleic acid (DNA),
18
provides a theo-
retical cellular basis for the contention that ele-
vated F,0
2
and^bleomycin may interact in a posi-
tive way to produce pulmonary damage. However,
if a bleomycin-treated patient, or any other patient
with pulmonary disease, were ventilated with F,0
2
less than 0.3 during one-lung ventilation, the risk
of hypoxemia would indeed be great. The fact that
there are both human
19
and animal
20
2I
data that
equally strongly suggest that increased oxygen
concentrations (F,0
2
> 0.3) in bleomycin-treated
patients do not cause acute pulmonary damage
blunts the need for reduced F,0
2
during one-lung
ventilation.
Similarly, concern has been registered regarding
mitomycin, which has also been linked to preop-
erative lung damage and postoperative interstitial
pneumonitis (because it forms superoxide radicals
in the presence of elevated oxygen tensions).
However, the only study to examine the relation-
ship between anesthetic F,0
2
and interstitial pneu-
monitis concluded that elevated F,0
2
may not
contribute to interstitial pneumonitis in mitomy-
cin-treated patients.
22
At present, the matter regarding preoperative
bleomycin and mitomycin administration and in-
traoperative F,0
2
is considered unsettled. Never-
theless, it is reasonable to minimize the F,0
2
(<
0.3) to both the nonoperative and operative lungs
as long as it is documented that it is safe to do so
by measuring S
p
0
2
and P
a
0
2
. If the anesthesiolo-
gist is concerned about and wishes to limit the
F,0
2
in either bleomycin- or mitomycin-treated pa-
tients, then it is possible to carefully increase or
decrease F,0
2
to both lungs independently in an
attempt to have the lowest F,0
2
to both lungs (even
when continuous positive airway pressure [CPAP]
is administered to the nonventilated lung) by using
the system shown in Figure 1 l^-.
23
B. Tidal Volume
The dependent lung should be ventilated with a
tidal volume of 10 ml/kg. Use of a tidal volume
much less than 10 ml/kg might promote depen-
dent-lung atelectasis.
24
Use of a tidal volume much
greater than 10 ml/kg might excessively increase
dependent-lung airway pressure and vascular resis-
tance
25
and thereby increase nondependent-lung
blood flow (decrease nondependent-lung HPV).
26-28
A dependent-lung tidal volume of 10 ml/kg rep-
resents a volume that is in the middle of a range
of tidal volumes (8 to 15 ml/kg) that have been
found to not greatly affect arterial oxygenation
during one-lung ventilation. The dependent-lung
tidal volume was systematically changed from 8 to
15 ml/kg during one-lung ventilation, and blood
gases and transpulmonary shunt were measured at
the following times: sample 1, lateral decubitus
position with the chest closed and two-lung tidal
volume of 15 mg/kg; sample 2, chest open with
the same two-lung ventilation; sample 3, 10 min
after collapse of the nondependent lung and with
dependent-lung tidal volume of 15 ml/kg; sample
Figure 11-4 Schematic of the variable F,0
;
, operative lung, continuous positive airway pressure (CPAP) system used to limit
F,0
2
to both lungs independently. Oxygen and room-air tanks are connected by a Y-piece leading to an F,0
:
analyzer; the F,0
:
analyzer is connected in series to a CPAP device consisting of a reservoir bag with pressure-relief valve and a pressure manometer.
(From Hughes S, Benumof JL: Use of operative lung CPAP to decrease F, 0
:
during one-lung ventilation in a bleomycin treated
patient. Anesth Anag 71:92-95, 1990. Used with permission.)
4, 10 min after tidal volume to the dependent lung
was reduced from 15 to 8 ml/kg; sample 5, 10 min
after dependent-lung tidal volume was increased
from 8 to 15 mg/kg; sample 6, 10 min after occlu-
sion of the pulmonary artery to the upper lung.
29
This study reported a consistent decrease in P.,0
2
and increase in shunt during one-lung anesthesia
(sample 3 compared with sample 2). Changes in
P
a
0
2
with alterations in the tidal volume (sample 4
compared with samples 3 and 5) in individual pa-
tients were variable and unpredictable in both de-
gree and direction (although the mean value for
the group did not change). Thus, it appears that
changing the tidal volume from 15 to 8 ml/kg
during one-lung ventilation has an unpredictable
but usually not a great impact on arterial oxygen-
ation. However, increases in tidal volume or use
of a large tidal volume may be associated with an
increased alveolar dead space (zone I, dead space
ventilation) in the ventilated lung.
30
C. Respiratory Rate
The respiratory rate should be set so that the
P
a
C0
2
remains at 40 mm Hg. Because a depen-
dent-lung tidal volume of 10 ml/kg represents a 20
per cent decrease from the usual two-lung tidal
volume of 12 ml/kg, the respiratory rate usually
needs to be increased by 20 to 30 per cent to
maintain carbon dioxide hemostasis (10 ml/kg to
one lung may make C0
2
elimination less efficient
because of increased alveolar dead space).
30
The
tradeoff between decreased tidal volume and in-
creased respiratory rate is usually a constant min-
ute ventilation; although ventilation and perfusion
are considerably mismatched during one-lung ven-
tilation, an unchanged minute ventilation during
one-lung ventilation (compared with two-lung
ventilation) can continue to eliminate a normal
amount of carbon dioxide because of the high dif-
fusibility of carbon dioxide.
25
3I
~
33
In addition, an
increase in respiratory rate does not increase dead
space breathing
30
but may be expected to increase
peak inspiratory pressure (compliance is frequency
dependent). In fact, Figure 11-5 shows that low-
ering the minute ventilation by approximately one
half (tidal volume reduced from 15 to 8 ml/kg
while respiratory rate is constant) has little effect
on P
a
C0
2
(see chapter 4 for explanation). Hypo-
capnia should be avoided because use of the air-
way pressure in the dependent lung necessary to
produce systemic hypocapnia may excessively in-
crease dependent-lung vascular resistance. Fur-
thermore, hypocapnia may directly inhibit HPV in
the nondependent lung.
34
35
SAMPLE
Figure 11-5 The effects of changing tidal volume on arterial blood gas values, peak airway pressure, pulmonary artery pressure,
and shunt during one-lung ventilation. P
a
0
2
and percentage shunt are not significantly affected by changing the tidal volume from
15 to 8 ml/kg or vice versa. (From Flacke JW, Thompson DS, Reed RC: Influence of tidal volume and pulmonary artery occlusion
on arterial oxygenation during endobronchial anesthesia. South Med J 69:619, 1976. Redrawn and reprinted by permission from the
Southern Medical Journal.)
D. Dependent-Lung PEEP
No, or just a very low level of, dependent-lung
PEEP (less than 5 cm H
2
0) should be used ini-
tially because of concern for unnecessarily increas-
ing dependent-lung vascular resistance (although it
is unlikely to occur when dependent-lung PEEP is
< 5 cm H
2
0,
36
the presence of intrinsic PEEP
during one-lung ventilation may make the total
PEEP excessive
37
) (see Selective Dependent-Lung
PEEP).
In summary, at the commencement of one-lung
ventilation, 100 per cent oxygen (although it may
be argued on theoretical grounds that 80 to 90 per
cent oxygen decreases the risk of absorption ate-
lectasis and only minimally, or not at all, increases
the risk of hypoxemia), a tidal volume of 10 ml/kg,
and a 20 per cent increase in respiratory rate are
used as initial ventilation settings (see Table
11-1). Ventilation and arterial oxygenation are
monitored by use of frequent arterial blood gases,
end-tidal carbon dioxide concentration, and pulse
oximetry or transcutaneous tensions. If there is a
problem with either ventilation or arterial oxygen-
ation, then one or more of the differential lung
management techniques are used.
IV. DIFFERENTIAL LUNG
MANAGEMENT OF ONE-LUNG
VENTILATION
A. Intermittent Inflation of the
Nondependent, Operative Lung
Intermittent inflation with oxygen of the col-
lapsed lung during one-lung ventilation may be
expected to increase P
a
0
2
for a variable period of
time. In a group of thoracic surgery patients
undergoing one-lung ventilation with F,N
2
0 = 0.5
and F,0
2
= 0.5, the collapsed lung was manually
inflated with a breath every 5 min with 2 L of
oxygen (for practical useful independent-lung ven-
tilation systems, see Figs. 11-4, 11-12, 11-13),
and the lung was then allowed to collapse again;
P
a
0
2
increased more than 4 kPa (28 mm Hg) after
each inflation (Fig. 11-6).
38
It can be seen from
Figure 11-6 that the beneficial effect of each infla-
tion persisted to a large extent to the next breath
even if at a gradually decreasing level. The control
group (no intermittent inflation) had a mean P.,0
2
of 19.2 (range = 11.2-30.2) kPa before one-lung
ventilation and 10.2 (8.2-16.0) kPa after 9 min of
one-lung ventilation without further reduction in
P
a
0
2
for another 10 min. Although P
a
0
2
decreased
between inflations, it never reached the level ob-
served in controls during 19 min of one-lung ven-
tilation.
There are two major drawbacks to the use of
intermittent inflation. First, it may not be compati-
ble with the performance of surgery. Second, be-
cause it may disturb the surgeon, the timing of
inflation must be agreed on by both parties. In
view of the systems recommended for CPAP of
the nondependent, nonventilated lung (see Figs.
11-4, 11-12, and 11-13), use of intermittent rein-
flation of the nondependent, nonventilated lung is
entirely compatible with nondependent-, nonven-
tilated-lung CPAP.
B. Selective Dependent-Lung PEEP
Since the ventilated dependent lung often has a
decreased lung volume during one-lung ventilation
(Figs. 4-6 and 11-1), it is not surprising that sev-
eral attempts have been made to improve oxygen-
ation by treating the ventilated lung with PEEP.
:fv
28. 39-42 yh
e m0
st accepted mechanism by which
PEEP is thought to be of benefit is that PEEP
causes an increase in lung volume at end-expira-
tion (by definition, the functional residual capac-
ity, or FRC) (Fig. 3-28). The increase in FRC
contributes to the prevention of airway and alveo-
lar closure at end-expiration and to the recruitment
of airways and alveoli during inspiration. The in-
creases in lung volume and airway and alveolar
openings result in increases in lung compliance,
ventilation, and the ventilation-perfusion ratio of
the single ventilated lung (Fig. ll-T).
43
44
Conse-
quently, it may be expected that the application of
PEEP to the compressed dependent lung would
improve dependent-lung volume and ventilation-
perfusion relationships.
An accepted risk of PEEP is that the PEEP-
induced increase in lung volume can cause com-
pression of the small intra-alveolar vessels. If the
PEEP-induced intra-alveolar vessel compression is
Figure 11-6 Arterial oxygen ten-
sion (P
a
0
2
. median and range) in the
control group (O-O) and inflation
group ( ) . Arrows indicate in-
flations. Between-groups differences:
*p < .05; **p < .02; ***p < .01.
(From Malmquist G: Maintenance of
oxygenation during one-lung venti-
lation: Effect of intermittent reinfla-
tion of the collapsed lung with oxy-
gen. Anesth Analg 68:763-766,
One Lung Ventilation: Dependent Lung PEEP
Nondependent.Lung;
Nonventilated
Dependent Lung;
Ventilated
t Shunt
Compression of
Small Intra-Alveolar
Vessels
Figure 11-7 Selective positive end-expiratory pressure (PEEP) to the ventilated-dependent lung can increase dependent lung
ventilation-perfusion ratios ( V
A
/Q). However, dependent lung PEEP can also cause compression of the small intra-alveolar
vessels in the dependent lung, causing blood flow diversion to the nonventilated, nondependent lung, thereby increasing the shunt
through the nonventilated nondependent lung. Therefore, the overall arterial oxygenation effect of dependent-lung PEEP will be a
trade-off between the good effect of an increase in dependent lung V
A
/Q and the bad effect of increased nonventilated lung blood
flow.
geographically widespread, then total pulmonary
vascular resistance increases and cardiac output
decreases. If the intra-alveolar vessel compression
is limited to the ventilated lung, then the pulmo-
nary vascular resistance of the lung just being ven-
tilated and receiving PEEP will increase, which
will cause diversion of blood flow away from the
ventilated lung to the nonventilated lung (see Fig.
11-7), the shunt will increase, and P
a
0
2
will de-
crease. With one-lung PEEP of 5 to 15 cm H
2
0,
pulmonary vascular resistance for the whole lung
does not greatly increase, and the cardiac output
does not decrease.
26
28
However, if the dependent-
lung PEEP is 20 cm H
2
0, then cardiac output and
S
v
0
2
may decrease.
45
The fact that increases in
both PEEP and tidal volume in the dependent,
ventilated lung have an additive effect in decreas-
ing P
a
0
2
during one-lung ventilation greatly sup-
ports the one-ventilated lung volume versus vas-
cular resistance hypothesis.
26
Consistent with these
observations is the fact that the application of 10
cm H
2
0 PEEP to the dependent lung during one-
lung ventilation decreases P
a
0
2
in some patients
but has had no significant systemic hemodynamic
effect in any patient.
45
"*
7
In summary, the effect of dependent-lung PEEP
on arterial oxygenation is a tradeoff between the
positive effect of increasing dependent-lung FRC
and ventilation-perfusion ratio and the negative ef-
fect of increasing dependent-lung vascular resis-
tance and shunting blood flow to the nonventilated
lung (and perhaps a decrease in cardiac output
with very high PEEP). Thus, the various one-lung
ventilation PEEP studies have contained patients
who have had an increase,
27,39
48
49
no change,
27
40

41,48.49
o r a
d
e C r e a S e
27. 39, 42, 48, 49 j
R
oxygenati on.
In fact, the only study to perform a dose-(depen-
dent-lung PEEP, 5-20 cm H
2
0) response (P
a
0
2
,
cardiac output) curve found a flat to slightly bi-
phasic P
a
0
2
/PEEP relationship (P
a
0
2
increased
slightly at 10 cm H
2
0 PEEP and 15 cm H
2
0
PEEP), whereas cardiac output was decreased sig-
nificantly at 15 cm H
2
0 PEEP and at 20 cm H
2
0
PEEP (Table 11-2).
45
The authors concluded that
there is no evidence that dependent-lung PEEP 5
to 20 cm H
2
0 is beneficial in improving P
a
0
2
during one-lung ventilation and cannot be recom-
mended for treating hypoxemia. Furthermore,
high-dose dependent-lung PEEP (15-20 cm H
2
0)
compromises cardiac output and mixed venous ox-
ygenation because of a reduction in oxygen deliv-
ery.
Table 11-2
Stages
P
a
0
2
(mm Hg)
CcT(L/min)
S
v
0
2
(%)
EFFECT OF INCREMENTAL DEPENDENT-LUNG PEEP*
OLV
147 71
4.7 1
80 5
PEEP 5
147 87
5.0 1.2
78 7
PEEP 10
160 66
4.6 0.9
80 6
PEEP 15
162 68
4.1 l
75 7t
PEEP 20
111 48
3.8 0.9t
68 8t
OLV
143 67
4.3 1
76 8
*From Cohen E, Salter O, Ali J: Effect of incremental PEEP on P
a
0
2
and S
v
0
2
during OLV. Anesth Analg 72:543, 1991. Used
with permission,
tp <.05.
<.001.
Values are expressed as Mean SD.
Abbreviations: PEEP = positive end-expiratory pressure; OLV = one-lung ventilation.
It may be expected that in patients with a very
diseased dependent lung (low lung volume and
low ventilation-perfusion ratio), the positive effect
of selective dependent-lung PEEP (increased lung
volume and increased ventilation-perfusion ratio)
might outweigh the negative effects of selective
dependent-lung PEEP (shunting of blood flow to
the nonventilated, nondependent lung), whereas in
patients with a normal dependent lung the negative
effects of dependent-lung PEEP would outweigh
the benefits. Indeed, in one study in which 10 cm
H
2
0 PEEP was selectively applied to the depen-
dent lung, P
a
0
2
increased in those patients with
P
a
0
2
less than 80 mm Hg (F,0
2
= 0.5), whereas
P
a
0
2
decreased or remained constant in patients
with P
a
0
2
higher than 80 mm Hg (F,0
2
= 0.5).
4H
Presumably, in the patients with P
a
0
2
lower than
80 mm Hg (F,0
2
= 0.5), the dependent lung had
a low FRC (low ventilation-perfusion ratio and
atelectatic regions); therefore, the positive effect of
increased dependent-lung volume predominated
over the negative effect of shunting blood flow to
the nonventilated lung. Conversely, the patients
with the higher P
a
0
2
presumably had a dependent
lung with an adequate FRC and ventilation-perfu-
sion ratio, and the negative effect of shunting
blood flow to the nonventilated lung predominated
over the positive effect of increased dependent-
lung volume.
Although only one of the previously listed stud-
ies described a dose-(ventilated-lung PEEP) re-
sponse (P
a
0
2
, Q
s
/Q cardiac output value) relation-
ship,
45
it seems reasonable to state on the basis of
all the results that the therapeutic margin of using
PEEP to just the ventilated lung in a given patient
to increase P
a
0
2
during one-lung ventilation is
quite narrow. Furthermore, it should be remem-
bered that intrinsic PEEP has been documented in
ventilated patients with respiratory failure who
have a persistent end-expiratory flow
50
(the contin-
ued end-expiratory flow generates a PEEP).
End-expiratory flow has been noted during one-
lung ventilation for thoracic surgery
51
and can re-
sult in significant intrinsic PEEP (range - 4-8 cm
H
2
0 when tidal volume is 10 ml/kg, respiratory
rate is 10 breaths/min, inspiration to expiration
ratio [I:E] = 1:3) during one-lung ventilation (Fig.
11-8).
17
When extrinsic PEEP is added during
one-lung ventilation, the resultant total PEEP is
increased (see Fig. 11-8). However, the increase
is less than the value of the extrinsic PEEP itself
(see Fig. 11-8). The presence of intrinsic PEEP
Figure 11-8 Mean total positive
end-expiratory pressure (PEEP) (
standard deviation) versus expiratory
time. (OLV = one-lung ventilation;
2LV = two-lung ventilation.) (From
Slinger PD, Hickey DR, Lenis SG,
Gottfried SB: Intrinsic PEEP during
one-lung ventilation. Anesth Analg
68:S269, 1989. Used with permis-
sion.)
should be considered whenever the use of extrinsic
PEEP is considered during one-lung ventilation.
PEEP to just the dependent, ventilated lung may
be delivered by the same anesthesia machine ap-
paratus that is ordinarily used to deliver PEEP to
the whole lung. Other studies showed that high
tidal volumes,
52
variations in the I:E ratio,
40
and
intermittent manual hyperventilation of the lower
lung are not beneficial in increasing P
a
0
2
during
40
C. Selective Nondependent-Lung CPAP
Positive pressure can be selectively and stati-
cally applied to just the nonventilated, nondepen-
dent lung. Since under these conditions the non-
ventilated lung is only slightly but constantly,
distended by oxygen, an appropriate term for this
ventilatory arrangement is nonventilated-lung
CPAP. Two reportsone in humans
42
and one in
dogs
53
were the first to show that the application
of CPAP (without tidal ventilation) to only or just
the nonventilated lung significantly increased ox-
ygenation. The latter study was performed with the
dogs in the lateral decubitus position and showed
that low levels of CPAP (5 to 10 cm H
2
0) to the
nonventilated, nondependent lung greatly in-
creased P
a
0
2
and decreased shunt, while blood
flow to the nonventilated lung remained un-
changed; presumably, this level of CPAP does not
compress the small intra-alveolar vessels in the
nondependent lung. Thus, it is not at all surprising
to find that the institution of 10 cm H
2
0 nondepen-
dent-lung CPAP in patients has had no significant
hemodynamic effect.
46
47
In summary, low levels
of CPAP simply maintain the patency of nonde-
pendent-lung airways, allowing some oxygen dis-
tention of the gas-exchanging alveolar space in the
nondependent lung (Fig. 11-9) without signifi-
cantly affecting the pulmonary vasculature. In all
clinical studies,
42
4647
54
the application of 5 to 10
cm H
2
0 CPAP has not interfered with the per-
formance of surgery and may, in fact, facilitate
intralobar dissection; this is not surprising in view
of the fact that the initial compliance of collapsed
lung is only 10 to 15 ml/cm H
2
0, and 5 to 10 cm
H
2
0 CPAP should only create a slightly distended
lung that occupies a volume of 50 to 75 to 100 to
150 ml, which is hardly or not at all noticed by
the surgeon.
42
55
56
On the other hand, in the canine study,
53
15 cm
One Lung Ventilation: Nondependent Lung CPAP
Nondependent Lung;
Nonventilated
Dependent Lung;
Ventilated
Figure 11- 9 Selective continuous positive airway pressure (CPAP) to the nonventilated, nondependent lung (static distension
without tidal movement) 'allows this lung to participate in oxygen uptake and markedly decreases the shunt through the nonventi-
lated, nondependent lung. Even if the nonventilated, nondependent lung CPAP causes blood flow diversion to the ventilated
dependent lung, the diverted flow still can participate in oxygen uptake and C0
2
elimination in the ventilated dependent lung.
Usually 5 to 10 cm H
2
0 of nondependent lung CPAP is all that is clinically needed, and this amount of CPAP does not cause any
significant surgical interference.
Conventional and Differential Lung Management of ne-Lung Ventilation
Figure 11-10 Histogram showing P0,, (A-a)DO
:
and Q
s
/Q, during two-lung ventilation, and ventilation of only the dependent
lung with and without up-lung continuous positive airway pressure (CPAP). (ZEEP-zero end expiratory pressure; DEFL-deflation.)
Note the approximation of P
a
0
2
, (A-a)DO, and Q/Q, during one-lung ventilation with CPAP to values obtained with two-lung
ventilation. (Reprinted with permission Capan LM, Turndorf H, Chandrakant P, et al: Optimization of arterial oxygenation during
one-lung anesthesia. Anesth Analg 59:847-851, 1980.)
H
:
0 of nondependent-lung CPAP caused changes
in P
a
0
2
and shunt similar to those of 5 to 10 cm
H
2
0 nondependent-lung CPAP, while blood flow
to the nonventilated, nondependent lung decreased
significantly. Therefore, high levels of nonventi-
lated lung CPAP act by permitting oxygen uptake
in the nonventilated lung as well as by causing
blood flow diversion to the ventilated lung, where
both oxygen and carbon dioxide exchange can take
place (see Fig. 11-9). Since low levels of nonven-
tilated-lung CPAP are as efficacious as high levels
of nonventilated-lung CPAP and have less surgical
interference and fewer hemodynamic implications,
it is logical to first use low levels of nonventilated
CPAP.
In all patients in all clinical studies to date, 5 to
10 cm H
2
0 of nondependent-lung CPAP signifi-
cantly increased P
a
0
2
during one-lung ventila-
tion.
42 4ft 47 49 54
"
56
" In most of these studies, P
a
O
:
during two-lung ventilation was not significantly
different from P
a
0
2
during one-lung ventilation
plus 10 cm H
2
0 of nondependent, nonventilated
lung CPAP; Figures 11-10 and 11-11 show re-
sults from two of these studies (one with a double-
lumen tube
42
and one with a bronchial blocker,
55
respectively) that are typical and representative of
all the studies.
Obviously, the results from these two human
studies, which produced one-lung ventilation and
operative-lung CPAP in different ways, are nearly
identical. It should be concluded that the single
most efficacious maneuver to increase P
a
O
:
during
one-lung ventilation is to apply 5 to 10 cm H
:
0
CPAP to the nondependent lung. In my experi-
ence, low levels of nonventilated CPAP have cor-
rected severe hypoxemia (P
a
0
2
< 50 mm Hg)
greater than 90 per cent of the time, provided the
double-lumen tube was correctly positioned. How-
ever, the nondependent-lung CPAP must be ap-
plied during the deflation phase of a previous large
tidal volume so that the deflating lung can lock
into a CPAP level with uniform expansion and
avoid the need to overcome critical opening pres-
sures. Indeed, P
a
0
2
plus 5 cm H
2
0 nondependent-, nonventilated-lung
CPAP is significantly higher (by 100 mm Hg)
when the lung is allowed to deflate (from a pre-
vious tidal breath) to just 2 cm H
2
0 (before apply-
ing the steady-state 5 cm H
2
0 CPAP) compared
with deflation 0 cm H
2
0 (before applying the
steady-state 5 cm H
2
0 CPAP).
56
In both the human
42
and canine
51
studies, oxy-
gen insufflation at zero airway pressure did not
significantly improve P
a
0
2
and shunt, and this re-
sult was probably due to the inability of zero trans-
bronchial airway pressure to maintain airway pa-
tency and overcome critical opening pressures.
Although one study in patients has concluded that
Figure 11-11 P
a
0
2
values for
the 12 individual experimental se-
quences of two-lung ventilation (2-
LV), one-lung ventilation (1-LV),
and one-lung ventilation with 10 cm
H
2
0 continuous positive airway pres-
sure (CPAP) (1-LV + 10 CPAP).
The mean standard deviation P
a
O
:
values for each of the three experi-
mental sequence steps (2-LV, 1-LV,
1-LV + 10 CPAP) is offset to the
right of the individual data points.
P
a
0
2
during 1-LV was significantly
less (p < .001) than during both 2-
LV and 1-LV + 10 CPAP, but P
a
O
:
during 2-LV was not significantly
different from 1-LV + 10 CPAP.
(From Benumof JL, Gaughan S,
Ozaki GT: Operative lung constant
positive airway pressure with the
Univent bronchial blocker tube.
Anesth Analg 74:406-410, 1992.
Nondependent Lung CPAP Systems Has
Three Essential Components
ZEEP or PEEP
from Anesthesia
Circle System
Figure 11-12 The three essential components of a nondependent lung continuous positive airway pressure (CPAP) system are
(1) an oxygen source, (2) a pressure relief valve, and (3) a pressure manometer to measure the CPAP. The CPAP is created by the
free flow of oxygen into the lung versus the restricted outflow of oxygen from the lung by the pressure relief valve. (ZEEP = zero
end-expiratory pressure; PEEP = positive end-expiratory pressure.)
insufflation of 0
2
at zero airway pressure does
increase P
a
0
2
, the study is difficult to interpret
because the patients did not serve as their own
control.
58
Several selective nondependent-lung CPAP sys-
tems that are easy to assemble have been
described.
41, 59_61
All these nondependent-lung
CPAP systems have three features in common
(Fig. 11-12). First, there must be a source of oxy-
gen to flow into the nonventilated lung.
Second, there must be some sort of restrictive
mechanism (hand-screw valve, pop-off valve,
weight-loaded valve) to prevent the egress of oxy-
gen out of the nonventilated lung so that the non-
ventilated lung may become distended. Thus, a
free-flowing pressurized source of oxygen flows
into a lung, but the escape of the oxygen is re-
stricted; the unrestricted flow in and the restricted
flow out create a constant distending pressure.
Third, the distending pressure must be measured
by a manometer. In practice, it is simplest to keep
the restrictive mechanism constant and adjust the
distending pressure with a relatively fine sensitiv-
ity by changing the oxygen flow rate. If the non-
dependent-lung CPAP system has a reservoir bag
included, the nondependent lung may also be ven-
tilated with intermittent positive pressure when-
ever desired.
The exact arrangement of the oxygen source,
the restrictive mechanism, and the manometer is
not important, and many homemade systems have
been proposed (all of which may be rendered ob-
solete in the next few years by the availability of
commercial systems; see later discussion). For ex-
ample, Figure 11-13 shows that these three es-
sential nondependent-lung CPAP components may
be arranged, proceeding proximally to distally
(toward the double-lumen tube), as anesthesia res-
ervoir bag, restrictive mechanism, pressure ma-
nometer, and oxygen source,
59
62
whereas Figure
11-13Z? shows a similar type of nondependent-
lung CPAP system but with the place of the anes-
thesia reservoir bag and restrictive mechanism re-
versed (unpublished personal experience).
The arrangement shown in Figure 11 4
23
con-
sists of reservoir bag, oxygen source, restrictive
Nondependent Lung CPAP Systems With Reservoir Bags
Figure 11-13 This schematic diagram shows two (A and B) nondependent lung continuous positive airway pressure (CPAP)
systems with reservoir bags. Both of these systems contain an oxygen source, some type of pressure relief valve, and a pressure
manometer to measure the CPAP. The presence of a reservoir bag allows for intermittent positive-pressure breathing and sighing.
if desired. (A is from Aalto-Setala et al.
4
', and is the system used by the author.)
Nondependent Lung CPAP Systems Without Reservoir Bags
Nondependent
v
Lung
Figure 11-14 This schematic diagram shows two (A and B) nondependent lung continuous positive airway pressure (CPAP)
systems without reservoir bags. Both contain an oxygen source and a pressure relief valve, but the upper panel has a pressure
manometer to measure the CPAP, whereas the lower panel does not. (A is from Baraka et al.
57
and Hannenberg et al.,
60
and is
from Brown and Davis.
61
)
Figure 11-15 The schematic shows the operative lung continuous positive airway pressure (CPAP) system for the Univent
bronchial blocker (BB) used in this study. The CPAP was measured by a pressure gauge. (From Benumof JL, Gaughan S, Ozaki
GT: Operative lung constant positive airway pressure with the Univent bronchial blocker tube. Anesth Analg 74:406-410, 1992.
mechanism, and pressure manometer.
21
The pres-
ence of an anesthesia reservoir bag in the nonde-
pendent-lung CPAP system allows for the capabil-
ity of delivering an independent tidal breath or
sigh to the nondependent lung.
63
64
Figure 11-14
shows two nondependent-lung CPAP systems that
do not include an anesthesia reservoir bag. In the
first nonreservoir bag CPAP system (Fig. 11-144),
the arrangement of the necessary three components
is oxygen source, pressure manometer, and the
restrictive mechanism,
57
"*
65
whereas in the se-
cond nonreservoir bag CPAP system (Fig. 11-
145), the restrictive mechanism is most distal and
the oxygen source most proximal.
61
In most of
these systems using these readily available restric-
tive valves, an oxygen flow rate of 5 to 10 L/min
creates a nondependent-lung CPAP of 5 to 10 cm
H
2
0. Nonventilated-lung CPAP may also be ap-
plied when using a Univent bronchial blocker; Fig-
ure 11-15 shows the system used, and with this
system, oxygen flow rates of 1.6 to 3.3 L/min
produced 5 to 20 cm H
2
0 CPAP.
55
The availability of a commercial nondependent,
nonventilated lung CPAP device (Mallinckrodt
Inc.) may replace the need for all these homemade
devices. The Mallinckrodt device (Fig. 11-16) is
similar in concept to the system shown in Figure
11-13 except that the restrictive mechanism is a
series of variously sized holes; the larger the hole,
the lower the CPAP, and the smaller the hole, the
higher the CPAP. The level of CPAP can be se-
lected by simply turning the calibrated dial (cali-
brated for an oxygen flow of 5 L/min), which
uncovers the appropriately sized decompression
(pop-off) hole. Because the dial is calibrated (for
an oxygen flow of 5 L/min) and marked with the
CPAP level selected, a pressure manometer is un-
necessary. The rest of the CPAP system consists
of the other components shown in Figure 11-13/4
(oxygen tubing, reservoir bag). This small, light-
weight CPAP device can be included within the
double-lumen tube package (the double-lumen
tube is available with or without the CPAP de-
vice), is extremely easy to use, is preassembled,
cost-effective, sterile, reliable, and disposable. Per-
haps the most important attribute is that a preas-
sembled device is obviously more readily available
than devices that require a search for spare parts,
and the commercial process has a built-in quality
assurance that cannot be duplicated by individual
remedies.
The inflow of fresh gas into a statically dis-
tended airway/alveolar air space will wash out
some carbon dioxide from the lung. To date, it is
not known how much carbon dioxide is removed
by causing nondependent-lung CPAP with 5 to 10
L/min of oxygen. Obviously, the higher the oxy-
gen flow rate, the greater the washout of carbon
dioxide, and as the oxygen flow increases, an ap-
proach to continuous high-flow apneic ventilation
(see chapter 12), in which adequate gas exchange
can be maintained without any tidal exchange, will
be made.
D. Differential Lung PEEP and CPAP
In theory, and from the preceding considera-
tions, it appears that the ideal way to improve
oxygenation during one-lung ventilation is to ap-
ply differential lung CPAP/PEEP (Fig. 11-17; see
section V) for the step-by-step approach. In this
situation, the ventilated (dependent) lung is given
PEEP in the usual conventional manner in an ef-
fort to improve ventilated lung volume and venti-
lation-perfusion relationships. Simultaneously, the
nonventilated (nondependent) lung receives CPAP
in an attempt to improve oxygenation of the blood
perfusing this lung. Therefore, with differential
lung PEEP or PEEP/CPAP, it does not matter
where the blood flow goes nearly as much as dur-
ing simple one-lung ventilation, since wherever it
goes (to either ventilated or nonventilated lung), it
has at least some chance to participate in gas ex-
change with alveoli that are expanded with oxy-
gen. In indirect support of this contention, arterial
oxygenation has been increased significantly in pa-
tients during thoracotomy in the lateral decubitus
position (utilizing two-lung ventilation) when
PEEP has been added to the ventilated, dependent
lung, while the nondependent lung was also able
to participate in gas exchange by virtue of
being ventilated at zero end-expiratory pressure
(ZEEP).
44
In direct support of this contention, in patients
undergoing thoracotomy and one-lung ventilation,
arterial oxygenation was unchanged by the appli-
cation of 10 cm H
2
0 dependent-lung PEEP alone
(consistent with an equal positive/negative effect
tradeoff), was significantly improved by 10 cm
H
2
0 nondependent-lung CPAP alone, and was fur-
ther and even more significantly increased by use
of 10 cm H
2
0 nondependent-lung CPAP and 10
cm H
2
0 dependent-lung PEEP together (differen-
tial lung ventilation).
46
47
49
The use of 10 cm H
;
0
nondependent-lung CPAP together with 10 cm
H
2
0 dependent-lung PEEP in patients caused
only small, clinically moderate hemodynamic ef-
fects.
46
47
49
There are now multiple reports of significant
increases in oxygenation obtained with the appli-
cation of differential lung ventilation and posi-
tive end-expiratory pressure (PEEP/PEEP, PEEP/
CPAP, or CPAP/CPAP) through double-lumen en-
dotracheal tubes to patients in the intensive care
unit with acute respiratory failure due to predomi-
B R O N C H O - C A T H
E N D O B R O N C H I A L T U B E
P R E S S U R E R E G U L A T I N G O R I F A C E
A D J U S T A B L E S E T T INGS
1cm - 10cm H 2
O P T I - P O R T
R I G H T A N G L E D O U B L E S W I V E L
C O N N E C T O R S
1 L I T ER A N E S T H E S I A B AG

Figure 11-16 The Mallinckrodt Incorporated continuous positive airway pressure (CPAP) device. A, Photograph of the entire
CPAP system connected to a double-lumen tube. B, Schematic of the entire CPAP system. C, Close-up of the CPAP-generating
restrictive mechanism.
One Lung Ventilation: Differential Lung CPAP/ PEEP
Nondependent Lung;
Nonventilated
Dependent Lung;
Ventilated
Figure 11-17 Differential lung continuous positive airway pressure (CPAP)/positive end-expiratory pressure (PEEP) during one-
lung ventilation allows for all the lung to participate in oxygen uptake and markedly reduces the shunt during one-lung ventilation.
The situation depicted here is a combination of the situations shown in Figures 11-7 and 11-9 (see the legends of those figures for
further explanation). With differential lung CPAP/PEEP, the distribution of blood flow is not critically important because all the
lung can participate in oxygen uptake.
nantly unilateral lung disease.
66-79
In all cases, con-
ventional two-lung therapy (mechanical ventila-
tion, PEEP, CPAP) was administered via a
standard single-lumen tube and either failed to im-
prove or actually decreased oxygenation. In most
cases, the amount of PEEP initially administered
to each lung was inversely proportional to the
compliance of each lung; ideally, this PEEP ar-
rangement should result in equal FRC in each
lung. In some cases, the amount of PEEP that each
lung received was later readjusted and titrated in
an effort to find a differential lung PEEP pressure
combination that resulted in the lowest right-to-left
transpulmonary shunt. It has not been necessary to
ventilate the two lungs synchronously, and good
success has been obtained with asynchronous in-
dependent lung ventilation.
78
-
79
Apparently, the
mediastinum does not move enough when the
lungs expand at different times to cause significant
hemodynamic effects. In addition, the compliance
of each lung may be increased by asynchronous
independent lung ventilation, since each lung does
not have to compete with the other lung for space
within the thorax. Special equipment has been de-
veloped to facilitate the application of differential
lung PEEP and tidal ventilation.
42
69
75
80
The use
of a double-lumen tube adapter that permits selec-
tion of, access to, and blockage of each lung inde-
pendently
80
is particularly helpful in administering
differential lung ventilation.
E. Selective Nondependent-Lung High-
Frequency Ventilation
The studies of selective nondependent-lung
CPAP indicate that delivery of any amount of ox-
ygen to the nondependent-lung alveolar space will
cause some oxygen uptake from the nondependent
lung and decrease the nondependent-lung right-to-
left shunt. Consequently, it is not surprising that
selective high-frequency ventilation of the nonde-
pendent lung (see chapter 12 for extensive discus-
sion of this ventilation modality), while the de-
pendent lung is ventilated with conventional
intermittent positive-pressure breathing, increases
P
a
0
2
compared with simple collapse of the nonde-
pendent lung and conventional mechanical venti-
lation of the dependent lung (Fig. 11-18).
81-84
However, because the same increase in arterial
oxygenation may be obtained by selective nonde-
pendent-lung CPAP with much simpler equipment
Nondependent Lung High Frequency Ventilation
Nondependent Lung
Dependent Lung
Figure 11-18 The nondependent lung can be ventilated with high-frequency ventilation (HFV) in order to improve arterial
oxygenation during ventilation of the dependent lung with intermittent positive-pressure breathing (IPPB).
(can either be assembled from equipment already
available in any anesthesia department or can be
purchased for an extra few dollars with the double-
lumen tube [Mallinckrodt, Inc.]; see Fig. 11-16)
compared with high-frequency ventilation appa-
ratus (typical cost is $8000), it is more logical to
use selective nondependent-lung CPAP than high-
frequency ventilation to improve arterial oxygena-
tion during "one-lung ventilation."
83
There are several other potentially important
considerations concerning the use of nondepen-
dent-lung high-frequency ventilation (both posi-
tive and negative; Table 11-3). On the negative
side, although high-frequency ventilation may be
superior to conventional mechanical ventilation for
peripheral lung work, it makes surgery around the
hilum or mediastinal structures difficult because of
vibration of the surgical field and changes in the
large-diameter airways.
85
In addition, high-fre-
quency ventilation can result in intrapulmonary air
trapping, especially in chronic obstructive pulmo-
nary disease and asthmatic patients, which may be
a mechanism for the improved oxygenation ["oc-
cult PEEP"]);
86
87
even though air trapping and
the size of the lung may be decreased by adjusting
respiratory parameters (driving pressure, rate. I:E
ratio, etc.), inability to reduce lung size will re-
quire abandonment of the technique in a few
patients.
83
85
On the positive side, more effective CO
:
Table 11-3 NONDEPENDENT-LUNG HIGH-FREQUENCY VENTILATION (WITH DEPENDENT-LUNG
CONVENTIONAL INTERMITTENT POSITIVE PRESSURE VENTILATION)
Positive Attributes Negative Attributes
1. More effective C0
2
exchange compared with simple
nondependent-lung atelectasis or CPAP
2. Equal oxygenation compared with nondependent-lung
CPAP
3. May allow decrease in tidal volume to dependent lung
4. Provides good operating conditions for peripheral lung work
5. Questionable increase in cardiac output compared with
nondependent-lung CPAP
1. Expensive, complicated equipment/procedure compared
with nondependent-lung CPAP
2. May result in intrapulmonary air trapping and therefore poor
surgical conditions for peripheral work
3. Poor operating conditions for surgery around the hilum.
large airways, and mediastinum
Abbreviations: CPAP = continuous positive airway pressure.
exchange is achieved with nondependent-lung
high-frequency ventilation than with nondepen-
dent-lung CPAP. However, C0
2
elimination is not
usually a significant problem during one-lung an-
esthesia provided that the dependent lung is venti-
lated with a minute volume that is equal to or
slightly greater than that used during two-lung
ventilation, the dependent lung is not extensively
diseased, and the lumen of the endobronchial tube
ventilating this lung is not narrowed by kinking,
twisting, or accumulation of secretions. Although
it can be argued that increased C0
2
elimination
from the nondependent lung with high-frequency
ventilation may allow reduction of the tidal vol-
ume of the dependent lung (which might improve
surgical conditions because of diminished medias-
tinal movement), reduced tidal volume of the de-
pendent lung may cause dependent-lung atelec-
tasis. Finally, high-frequency ventilation compared
with 10 cm H
2
0 CPAP may be associated with a
25 per cent higher cardiac output.
84
There are three relative indications for the com-
bination of nondependent-lung high-frequency
ventilation and dependent-lung intermittent posi-
tive-pressure breathing (see chapter 12).
72
First, if
the nondependent lung has a major bronchopleural
fistula, high-frequency ventilation may be indi-
cated because it may minimize airway leak.
88
Second, if prolonged surgery has to be per-
formed on a major nondependent-lung conducting
airway, high-frequency ventilation may be indi-
cated because it may permit a much smaller cath-
eter to pass through the operating field.
Third, when it is considered essential to recruit
some alveoli from the operative lung to participate
in ventilation (because the nonoperative lung has
very poor function), high-frequency ventilation of
the operative lung may be indicated.
89
F. Re-Expansion of the Lung and
Return to Two-Lung Ventilation at the
End of the Case
Unilateral pulmonary edema of the collapsed
lung after re-expansion of the collapsed lung at the
end of open thoracotomy has been described after
evacuation of a malignant pleural effusion,
90
re-
moval of a compressing mediastinal mass,
91
repair
of ruptured diaphragm,
92
esophagectomy,
93
pleu-
rodesis for pneumothorax,
94
as well as the many
more numerous closed-chest cases after evacuation
of pleural fluid and air
95
"
97
and withdrawal of an
endotracheal tube from the right main-stem bron-
chus (left lung became edematous)
95
98
(see chapter
18). All of the just described clinical literature
indicates that the previously collapsed lung should
be expanded slowly (e.g., over 10 sec) and that
several staged or staggered breaths should be used
to reverse atelectasis and reach unilateral total lung
capacity.
V. RECOMMENDED COMBINED
CONVENTIONAL AND
DIFFERENTIAL LUNG
MANAGEMENT OF ONE-LUNG
VENTILATION
Figure 11-19 summarizes the recommended
plan for obtaining satisfactory arterial oxygenation
during one-lung anesthesia. Two-lung ventilation
is maintained for as long as possible (usually until
the pleura is opened). When one-lung ventilation
is commenced, a tidal volume of 10 ml/kg is used,
and the respiratory rate is adjusted so that P
a
C0
2
is 40 mm Hg. A high inspired oxygen concentra-
tion (F,0
2
= 0.8-1.0) should be used, and arterial
blood gases should be monitored frequently.
If severe hypoxemia is present after this initial
conventional approach, three major causes of hy-
poxemianamely malposition of the double-lu-
men tube, poor hemodynamic status, and any leaks
or disconnects in the ventilating systemmust be
ruled out. If the double-lumen tube is correctly
positioned, there are no leaks or disconnects, and
the hemodynamic status is satisfactory, then sim-
ple tidal volume and respiratory rate adjustments
should be made.
29
For example, if the tidal venti-
lation is thought to be too high, it should be de-
creased, and if the tidal ventilation is thought to be
too low, it should be increased. If these simple
maneuvers do not quickly resolve the problem, the
studies of selective nondependent-lung CPAP,
42
46
47.49.55-57.59-61
a n d
differential lung PEEP
66
-
79
sug-
gest that the next treatment should be to apply 5
to 10 cm H
2
0 CPAP to the nondependent lung
(Fig. 11-20). Nondependent-lung CPAP should be
applied during the deflation phase of a large tidal
volume breath to overcome critical opening pres-
sures in the atelectatic lung. If oxygenation does
not improve with nondependent-lung CPAP
(which it does in the large majority of cases), 5 to
10 cm H
2
0 of PEEP to the ventilated, dependent
lung should then be applied. If dependent-lung
PEEP does not improve oxygenation, nondepen-
dent-lung CPAP should be increased to 10 to 15
cm H
2
0 while the dependent lung is maintained at
5 to 10 cm H
2
0 of PEEP. If arterial oxygenation
is still not satisfactory, then the nondependent-lung
CPAP level should be matched with an equal
amount of dependent-lung PEEP. In this way, a
differential lung CPAP/PEEP search for the maxi-
mum compliance and a minimum right-to-left
Overall One Lung Ventilation Plan
1. Maintain Two Lung Ventilation Until Pleura is Opened
2. Dependent
Lung
F A = 1.0
TV-8-10 ml/kg
RR = So That P
a
C0
2
= 40 mm Hg
PEEP = 0-5 mm Hg
3. If Severe
Hypoxemia
Occurs
(a) Check Position of Double-Lumen Tube
With Fiberoptic Bronchoscopy
(b) Check Hemodynamic Status
(c) Nondependent Lung CPAP
(d) Dependent Lung PEEP
(e) Two Lung Ventilation
(f) Clamp Pulmonary Artery ASAP
(For Pneumonectomy)
Figure 11-19 The figure shows an overall one-lung ventilation plan. (TV = tidal volume; RR = respiratory rate; PEEP
positive end-expiratory pressure; CPAP = continuous positive airway pressure; ASAP = as soon as possible.)
Nondependent Lung CPAP and Dependent Lung PEEP Search
Nondependent Lung
CPAP (leading)
Dependent Lung
PEEP
Figure 11-20 A search for optimal nondependent-lung continuous positive airway pressure (CPAP) and dependent-lung positive
end-expiratory pressure (PEEP) is performed by leading with nondependent-lung CPAP and following with dependent-lung PEEP.
Leading with nondependent-lung CPAP removes the deleterious arterial oxygenation blood flow diversion effects of dependent-
lung PEEP.
Ligation of Nondependent Lung Pulmonary Artery
During One-Lung Ventilation Restores a
Matching of Ventilation to Perfusion
With PA Ligation
Two-Lung
Ventilation
Figure 11-21 Ligation of the nondependent lung pulmonary artery during one-lung ventilation restores a matching of ventilation-
perfusion ratios; the nondependent lung is now neither ventilated nor perfused. Consequently, arterial oxygenation during one-lung
ventilation with pulmonary artery (PA) ligation is nearly the same as during two-lung ventilation.
transpulmonary shunt is done in an attempt to find
the optimal end-expiratory pressure for each lung
and the patient as a whole.
If severe hypoxemia is still present following
the application of differential lung CPAP/PEEP
(which would be extremely rare), it should be re-
membered that the nondependent lung may be in-
termittently ventilated with positive pressure with
oxygen. Finally, most of the ventilation-perfusion
imbalance is eliminated during a pneumonectomy
by tightening a ligature around the nonventilated
lung pulmonary artery as early as possible, which
directly eliminates all shunt flow through the non-
ventilated lung (see Fig. 11-19). Indeed, clamping
the pulmonary artery to a collapsed lung function-
ally resects the entire lung, and the P
a
0
2
is restored
to a level not significantly different from a two-
lung ventilation or postpneumonectomy one-lung
ventilation value (Fig. 11-21).
REFERENCES
1. Tarhan S, Lundborg RO: Carlens endobronchial catheter
versus regular endotracheal tube during thoracic surgery:
A comparison of blood gas tensions and pulmonary shunt-
ing. Can Anaesth Soc J 18:594-599, 1971.
Benumof JL: Respiratory physiology and respiratory func-
tion during anesthesia. In Miller R (ed): Anesthesia. 2nd
ed. New York, Churchill-Livingstone, 1986, chapter 40, pp
1371-1462.
Dantzker DR, Wagner PD, West JB: Instability of lung
units with low V/Q ratios during 0
2
breathing. J Appl
Physiol 38:886-895, 1975.
Ray JF III, Yost L, Moallem S, Sanodos GM, Villamena
P, Paredes RM, Clauss RH: Immobility, hypoxemia and
pulmonary arteriovenous shunting. Arch Surg 109:537-
541, 1974.
Benumof JL, Pirlo AF, Trousdale FR: Inhibition of hy-
poxic pulmonary vasoconstriction by decreased Pv0
2
: A
new indirect mechanism. J Appl Physiol 51:871-874,
1981.
Johansen I, Benumof JL: Flow distribution in abnormal
lung as a function of F,0
2
(abstract). Anesthesiology
51:369, 1979.
Lumb TD, Silvay G, Weinreich AI, Shiang W: A compar-
ison of the effects of continuous ketamine infusion and
halothane on oxygenation during one-lung anesthesia in
dogs. Can Anaesth Soc J 26:394-401, 1979.
Scanlon TS, Benumof JL, Wahrenbrock EA, Nelson WL:
Hypoxic pulmonary vasoconstriction and the ratio of hy-
poxic lung to perfused normoxic lung. Anesthesiology
49:177-181, 1978.
Winter PM, Smith G: The toxicity of oxygen. Anesthesiol-
ogy 37:210-241, 1972.
CHAPTER 12
High-Frequency and High-
Flow Apneic Ventilation During
Thoracic Surgery
1. Introduction
II. High-Frequency Ventilation
A. General Considerations
1. High-Frequency Positive-
Pressure Ventilation
2. High-Frequency Jet Ventilation
3. High-Frequency Oscillation
Ventilation
B. Use in Major Conducting Airway
Surgery
C. Use in Bronchopleural Fistula and
Tracheobronchial Tree Disruptions
D. Use in Minimizing Movement of the
Operative Field
III. Low- and High-Flow Apneic
Ventilation
A. Low-Flow Apneic Ventilation
B. Interrupted High-Flow Apneic
Ventilation
1. Low-Frequency Interrupted High
Flow
a. Use in Major Conducting
Airway Surgery
2. Continuous High-Flow Apneic
Ventilation
432
High-Frequency and High-Flow Apneic Ventilation During Thoracic Surgery 433
I. INTRODUCTION
Conventional positive-pressure breathing can
interfere with the performance of thoracic surgery
in four ways (Fig. 12-1A). First, conventional pos-
itive-pressure breathing requires large endotra-
cheal tubes (either single- or double-lumen tubes),
which can greatly interfere with the performance
of surgery on major conducting airways. Second,
the expansion and movement of the nondependent
lung by intermittent positive-pressure breathing
may greatly interfere with access to the surgical
field. Third, movement of the dependent lung due
to intermittent positive-pressure breathing will
cause the mediastinum, and therefore the floor of
the surgical field, to move up and down, which
may hamper the performance of surgery. Fourth,
the airway pressure in the ventilated dependent
lung, if too high, can compress the small intra-
alveolar vessels in the dependent lung, increase
dependent-lung pulmonary vascular resistance,
and shunt blood flow to the nonventilated nonde-
pendent lung.
In theory, high-frequency and high-flow apneic
ventilation do not have these disadvantages (Fig.
12-1 ). Both these forms of ventilation require
only small-bore catheters (tidal volumes are very
small and/or flow rates are very high) and thereby
can potentially facilitate the performance of sur-
gery on major conducting airways. Second, the
tidal movements with high-frequency and high-
flow apneic ventilation are very small or nonexis-
tent, respectively, thereby minimizing movements
of both the nondependent and dependent lungs. In
addition, both of these forms of ventilation utilize
low airway pressures so that it is possible to obtain
adequate gas exchange whenever airway resistance
is extremely low (as it might be in surgery on a
major conducting airway or with a large broncho-
pleural fistula) without a large air leak. Finally, the
lower airway pressure associated with both these
forms of ventilation
1
and release of vasodilator
prostaglandins with high-frequency ventilation
(HFV)
2
can, in theory, lower dependent-lung vas-
cular resistance and improve diversion of blood
flow away from an atelectatic nondependent lung.
This chapter reviews clinical experience with high-
frequency and high-flow apneic ventilation during
intrathoracic surgery. The use of HFV for extra-
thoracic procedures (i.e., bronchoscopy) is consid-
ered in chapter 14.
II. HIGH-FREQUENCY VENTILATION
Conventional intermittent positive-pressure ven-
tilation (IPPV) delivers relatively large tidal vol-
umes (10 to 15 ml/kg) at respiratory rates usually
less than 30 breaths per minute. With IPPV the
basic mechanism of gas transport is by mass
movement to the smaller airways and then more
distally by mass movement and molecular diffu-
sion.
3
In contrast to IPPV, HFV delivers very
small tidal volumes (< 2 ml/kg) at rates between
60 and 2400 breaths per minute. Under these cir-
cumstances, the proportion of the tidal volume
ventilating the dead space increases so that normal
alveolar ventilation can be maintained only by the
use of high respiratory rates and minute volumes.
With HFV, gas transport by mass movement is
still important
4
but may be combined with varying
degrees of other mechanisms of gas transport such
as enhanced molecular diffusion (Taylor disper-
sion),
5
high-velocity gas flow,
6
coaxial gas flow
(bidirectional flow with gases in the center of the
airway moving distally [0
2
] and gases on the edge
of the airway moving proximally [C0
2
]),
7
asyn-
chronous regional filling and emptying of alveoli
("pendelluft" movement, resulting in smaller gas
volumes effectively reaching more respiratory
units),
8
and gas trapping.
9
With regard to oxygenation, all forms of HFV
should be regarded as a method of achieving a
high positive end-expiratory pressure (PEEP) (or
mean airway pressure) (and therefore minimize
shunt) without the necessity of imposing a large-
volume (and pressure) excursion on top of this to
eliminate C0
2
. It is conceptually similar to the
approach of holding the lung distended with a high
mean airway pressure (and therefore minimizing
shunt) while removing C0
2
by an extracorporeal
membrane.
10
HFV is an umbrella term encompassing differ-
ent delivery systems and respiratory rate ranges.
Current methods used to provide HFV are quite
diverse, but by using the criteria of ventilation rate
and type of gas delivery mechanism, it is possible
to separate HFV into three general categories (Ta-
ble 12-1).
4
"
1. High-Frequency Positive-Pressure
Ventilation
The first HFV category, high-frequency posi-
tive-pressure ventilation (HFPPV), uses a volume-
controlled ventilator (which may simply be a
high-pressure flow generator coupled with a flow
interrupter mechanism)'
2
with a very low internal
compressible volume, which delivers small tidal
volumes at rates of 60 to 100 breaths per minute
(l to 1.7 Hz). The negligible internal compliance
of the ventilator guarantees that the preset tidal
volume (usually approximately equal to the ana-
tomic dead space volume [2 ml/kg]) is the tidal
volume that is actually delivered to the patient.
434 High-Frequency and High-Flow Apneic Ventilation During Thoracic Surgery
Conventional IPPB Can Potentially Interfere with
the Performance of Thoracic Surgery
Table 12-1 CHARACTERISTICS OF THE THREE TYPES OF HFV
Type of
HFV
HFPPV
HFJV
HFOV
Rate/Mi
60 to 100
I 00t o400
400 to 2400
Type of
Ventil ator
Volume
Jet pulsation
Piston pump
Gas
Ent rai nment
No
Yes
Yes
Process
Inspiration
Active*
Active
Active
Exhalation
Passivet
Passive
Active
*Activecaused by the ventilator.
tPassivecaused by elastic recoil of lung.
Abbreviations: HFV = high-frequency ventilation; HFPPV = high-frequency positive-pressure ventilation; HFJV = high-
frequency jet ventilation; HFOV = high-frequency oscillation ventilation.
The low-compression delivery system enhances
gas exchange in the conducting airways by causing
a high instantaneous gas flow.
6
There is no gas
entrainment so that the delivered tidal volume is
all fresh gas and the F,0
2
is the same as the F,0
2
of the compressed gas (Fig. 12-2).'
2
Exhalation is
passive.
13
The inspiratory and expiratory flow pro-
files are the same as with conventional intermittent
positive-pressure breathing (Fig. 12-3).
I4
If a stan-
dard endotracheal tube is used for HFPPV, PEEP
may be added to maximize P
a
0
2
.
8
I 5
The second HFV category, high-frequency jet
ventilation (HFJV), uses pulsation of a small jet of
fresh gas introduced from a high-pressure source
(50 to 60 psi) (a continuous flow is chopped into
square wave pulses [see Fig. 12-3] by a high-
frequency flow interuptor, which is usually a sole-
noid but can be a fluidic or rotating cylinder
valve
12
) into the airway via a small catheter (which
can be passed alone or through an endotracheal
tube) or via an extra lumen in an endotracheal tube
at rates usually between 100 and 400 breaths per
minute (1.7 to 6.7 Hz). Although the tip of the
small catheter may be anywhere from the top to
the bottom of the endotracheal tube, many systems
locate the tip near the carina. (However, HFJV
through catheters located either more distally or
proximally to the carina do not result in as efficient
Figure 12-1 A, Ways in which conventional intermittent positive-pressure breathing (IPPB) can potentially interfere with the
performance of thoracic surgery. The large endotracheal tubes (either single or double lumen) required by conventional IPPB may
interfere with the anastomosis of the airway. The relatively large IPPB tidal movements of the nondependent lung may decrease
surgical exposure (with inspiration) and cause a large amount of operative field movement. The relatively large IPPB tidal
movements of the dependent lung may cause a large amount of mediastinal movement and, therefore, operative field movement. B.
The proposed ways high-frequency ventilation (HFV) and high-flow apneic ventilation can potentially facilitate the performance of
thoracic surgery. The small catheter associated with the use of HFV facilitates the anastomosis of the airway. In addition. HFV of
the nondependent lung increases surgical exposure and minimizes movement of the operative field. HFV of the dependent lung
decreases mediastinal and operative field movement. The low airway pressures associated with HFV may decrease the air leak out
of a bronchopleural fistula.
C0
2
removal.)
13
The jetted fresh gas leaves the
narrow injection cannulas at a very high velocity.
As it exits from the injection cannulas, the high
velocity of the jet flow entrains gas from the en-
dotracheal tube's fresh gas flow or from an injec-
tion cannula side-port gas reservoir (Bernoulli dis-
tribution)
16
(Fig. 12-4).
12
The amount of entrained
gas is uncertain; therefore, quantification of tidal
volume and F,0
2
is difficult. The jetted and en-
trained gases (20 to 40 L/min) impact into the
much larger volume of relatively immobile gas in
the endotracheal tube and conducting airways and
cause the gas to move forward.
11
16
As such, HFJV
retains an element of being a form of pressure-
limited ventilation (e.g., with everything held con-
stant, a decrease in chest wall and/or lung compli-
ance results in a decrease in minute ventilation).'^
Exhalation is passive
13
; consequently, the expira-
tory flow profile is normal (see Fig. 12-3).
14
At
the higher respiratory rate, air trapping may occur,
and it has been shown that, with either the chest
open or closed as respiratory rate increases, airway
pressure (and lung volume) increase progressively
(Fig. 12-5).
18
An example (schematic) of an HFJV system
used clinically for ventilation of the left lung dur-
ing thoracic aorta aneurysm repair is shown in
Figure 12-6.
19
Note that a separate pressure line
(nos. 10 and 11 in Fig. 12-6) measures airway
pressure, which is important in avoiding bara-
trauma in patients with outflow obstruction.
14
As
Figure 12-2 Schematic of high-
frequency positive-pressure ventila-
tion. The gas inlet allows high-fre-
quency rapid velocity, low tidal
volume inspiration toward the pa-
tient. If the respiratory valve were
absent, a pneumatic valve dependent
on the angle at the junction and the
Coanda (wall) effect would provide
inspiratory and expiratory valving.
The addition of the expiratory valve
allows accurate delivery of tidal vol-
umes. (From Wetzel RC, Gioioa FR:
High frequency ventilation. Pediatr
Clin North Am 34:15-38, 1987.
Figure 12- 3 The frequency spec-
trum of high-frequency ventilation
and the flow profiles obtained with
each modality. (IPPV = intermittent
positive-pressure ventilation; HFPPV
= high-frequency positive-pressure
ventilation; HFJV = high-frequency
jet ventilation; HFOV = high-fre-
quency oscillation ventilation.) (From
Smith BE, Hanning CD: Advances in
respiratory support. Br J Anaesth
58:138-150, 1986. Used with per-
mission.)
Figure 12- 4 Schematic of a high-fre-
quency jet ventilator. Note humidification by
saline nebulization and presence of a continu-
ous fresh gas flow, allowing entrainment and
the addition of positive end-expiratory pres-
sure (PEEP). (E.T.T. = endotracheal tube)
(From Wetzel RC, Gioioa FR: High frequency
ventilation. Pediatr Clin North Am 34:15-38,
Figure 12-5 Mean airway pressure observed
at different respiratory rates of high-frequency
jet ventilation of both lungs, with the chest both
open and closed. There was a significant, pro-
gressive increase of P
aw
with higher respiratory
rates. All bars are mean standard deviation.
Data are from 109 patients. (From Howland
WS, Carlon GC, Goldliner PL, et al: High fre-
quency jet ventilation during thoracic surgical
procedures. Anesthesiology 67:1009-1012,
with HFPPV, oxygenation is enhanced with HFJV
when small to moderate amounts of PEEP are
added (via the endotracheal tube).
20
2()a
3. High-Frequency Oscillation
Ventilation
The third HFV category, high-frequency oscil-
lation ventilation (HFOV), uses a piston pump or
loudspeaker that oscillates gas in the airway back
and forth in a sinusoidal fashion at rates of 400 to
2400 breaths per minute (6.7-40 Hz). Characteris-
tically, gas is both actively driven into the lung
and actively withdrawn by the pump stroke,
13
and
the combined inspiratory and expiratory flow pro-
file is sinusoidal (see Fig. 12-3).
14
A fresh gas
flow-by (which is entrained) is located between
the oscillation and the patient
21
(Fig. 12-7).'
2
De-
livered tidal volume is very small (50-80 ml), and
alveolar gas exchange is obtained by enhanced
molecular diffusion and coaxial flow.
4
"
2I
One other type of HFV system, the flow inter-
Figure 12-6 Intraoperative ventilation of the left lung using high-frequency jet ventilation (HFJV) during surgical resection of a
thoracic aneurysm. The patient is in the right lateral position, and a left thoracic incision is performed. HFJV is applied to the left
lung (upper lung) once collapsed and retracted by the surgeon, using a F,0
2
of l, a respiratory frequency of 250 bpm, an
inspiratory/expiratory ratio of 0.43, a driving pressure of 19 3 psi, and a mean airway pressure of 8 mm Hg. Conventional
intermittent positive-pressure ventilation is applied to the right lung using an F,0
2
of 1, tidal volume of 6-10 rnldkg"
1
, and a
frequency of 16 bpm. 1 = Acutronic ventilator VS 600 S; 2 = right lumen of the Carlens tube; 3 = left lumen of the Carlens
tube; 4 = connecting tube; 5 = expiratory line; 6 = additional gases (pure oxygen 10 imin"') 7 = anesthesia bag: 8 =
humidifier; 9 = air oxygen blender; 10 = polyethylene catheter advanced into the left bronchus; 11 = continuous monitoring of
airway pressure; 12 = 0
2
source; 13 = 18-gauge injector cannula. (From Godet et al: High-frequency jet ventilation during one-
lung ventilation in patients undergoing resection of aortic aneurysm by thoracic incision. Anesthesiology 65:A486, 1986, and with
modification by Rouby JJ, Viars P: Clinical use of high frequency ventilation. Acta Anesthesiol Scand (Suppl) 90:134-139. 1989.
Fresh Gas
Source
Patient
Figure 12- 7 Schematic of a high-fre-
quency oscillator. The oscillated gas is in con-
tinuity with the patient and a constant fresh
gas (bias) flow. A low-pass filter directs the
pressure wave toward the endotracheal tube.
(From Wetzel RC, Gioioa FR: High frequency
ventilation. Pediatr Clin North Am 34:15-38,
rupter, has features of the HFPPV, HFJV, and
HFOV categories. Breaths are generated by using
a mechanical valve to interrupt very frequently a
continuous flow of gas into a low-compression
delivery system. Thus, small tidal volumes
(HFPPV) with high flow rates (HFJV) at very
rapid rates (HFOV) are produced. The interrupters
are capable of producing the entire range of HFV
respiratory rates, but rates of 100 to 600 per min-
ute have been found to be most effective.
3
In ad-
dition, although these three broad classes (HFPPV,
HFJV, HFOV) are the most common, many hy-
brids are now appearing, including a large number
of combined modalities. These combined modes
usually superimpose some form of HFV onto
back-up mechanical ventilation, with the "slow"
component operating anywhere from 1 to 60
breaths per minute. The "fast" component may
run at 100 to 3000 cycles/min, and the two-base
frequencies may or may not be synchronized (see,
e.g., Fig. 12-8).
14
Generally, such combined
modes have been used clinically in life-threatening
situations in which conventional mechanical ven-
tilation was believed to be failing. Such a huge
degree of inhomogeneity in system characteristics
and clinical situations make this body of observa-
tions extremely difficult to evaluate.
In the last decade, all three types of HFV have
been shown to be capable of providing adequate
alveolar ventilation and oxygenation in both nor-
mal animals and humans and in those with pul-
monary disease. In anesthetized animals and in
human infants and adults with normal lungs
undergoing surgical procedures, arterial blood
gases demonstrate adequate alveolar ventilation
with HFV.
15
22-26
Studies performed in animals and
humans in acute respiratory failure also showed
that HFV can provide adequate gas exchange.
27-29
In some reports of ventilation during acute respi-
ratory failure, HFV was as good as or even more
effective than IPPV;
29
30
however, these findings
were not uniform with respect to type of HFV,
patient population,
31-34
and incidence of complica-
tions.
35
36
B. Use in Major Conducting Airway
Surgery
The most important thoracic surgery advantage
of HFV is that the small rapid tidal volumes may
be delivered through small-airway tubes; thus, if a
major conducting airway (trachea, carinal area,
main-stem bronchus) has to be divided, the transit
of a small-airway tube through the surgical field
causes much less interference with surgery than
the passage of a large standard or double-lumen
endotracheal tube. The small-airway catheters pre-
sent the surgeon with an unobstructed, accessible
circumference of trachea and bronchus, so that the
ends of a divided airway can be properly aligned
for the construction of an unstressed and airtight
anastomosis. Under these circumstances, and par-
ticularly with HFJV, gases are entrained from the
operating site,
37
which promotes entry of blood
into an open airway.
38
Both HFPPV and HFJV
have been successfully used with small-airway
catheters for three different types of airway sur-
gery (Fig. 12-9).
37
-
46
The first type of airway surgery that has been
Figure 12-8 High-frequency ventilation (HFV) su-
perimposed on intermittent positive-pressure ventila-
tion showing the "pump-up" effect caused by HFV.
Lung volume = ordinate; time = abscissa. (From
Smith BE, Hanning CD: Advances in respiratory sup-
port. Br J Anaesth 58:138-150, 1986. Used with per-
mission.)
High-Frequency and High-Flow Apneic Ventilation During Thoracic Surgery 4 3 9
aided by HFV is carinal resection (in whole or in
part) (Fig. 12-9C and Table 12-2). During sleeve
pneumonectomy (which includes part of carina), a
small-airway catheter passing through an endotra-
cheal tube and then the operative field into the
main-stem bronchus of the dependent lung venti-
lated the dependent lung with either HFJV
3 7 3 8
4 6
or HFPPV.
38
Alternatively, and perhaps more desirable, small
catheters may be passed into both main-stem
bronchi (dependent, nonoperative lung as well as
the nondependent. operative lung; Fig. 12-90.
guided by the surgeons from within the surgical
field and used for HFJV; these authors pointed out
that intravenous anesthesia must be used, the cath-
eters have a tendency to back out into the trachea,
and the surgeons can suction the airway from the
field.
45
In one author's hands,
38
HFV also mini-
mized bronchial and mediastinal movements, and
with HFPPV the continuous outflow of gases
through the open bronchus minimized soiling of
the ventilated lung with blood from the lung under
operation, whereas HFJV created a suction effect
(entrainment) that drew blood and debris into the
open carinal area down into the bronchus of the
lung being ventilated.
38
Additionally, when a left-
sleeve pneumonectomy was being performed, the
small catheter inside the right main-stem bronchus
eliminated the problem of the right upper lobe
collapse associated with the use of right endobron-
chial tubes.
The HFPPV technique described previously for
sleeve pneumonectomy has also been successfully
used for carinal resections with pericardial patch
and very low (distal) tracheal resections.
38
In another case of an almost completely ob-
structing low tracheal tumor in a patient with a
previous left pneumonectomy, HFPPV through
two small bronchoscopically placed catheters (one
for inhalation and one for exhalation) were used to
by-pass the tumor and institute emergency one-
lung HFPPV; the distal trachea containing the tu-
mor was then excised.
44
In a large series of patients
undergoing carinal and tracheal resections, HFJV
was used only when the airway was open, and
IPPV through the more proximal endotracheal
tube was used the rest of the time.
37
In contradis-
tinction to these findings with HFPPV, HFOV de-
livered through a standard single-lumen tube has
been found to cause changes in the diameter of
large airways and to produce a large "mediastinal
bounce'' with each oscillation that "made surgery
on the major airways and around the hilum and
mediastinal structures almost impossible."
47
At present, continued investigation in the use of
HFV to facilitate carinal surgery still seems war-
ranted because older previous techniques for pro-
viding ventilation for major conducting airway
surgery leave a great deal to be desired. Previous
techniques used with sleeve pneumonectomy after
incision of the airway have included use of a con-
ventional double-lumen endotracheal tube, ad-
vancing a single-lumen endotracheal tube through
the end of the opened trachea into the appropriate
main-stem broncus, or using a separate sterile en-
dobronchial tube through the operative field with
a second sterile anesthesia circuit (see chapter 15).
Because repeated endobronchial intubation and ex-
tubation are necessary during construction of the
posterior portion of the anastomosis, the technique
is cumbersome and requires alternating periods of
apnea and ventilation, which may result in poor
gas exchange.
48-51
The second type of airway surgery aided by
HFV is tracheal reconstruction supported by the
tracheal Montgomery T-tube (Fig. \2-9B). The
Montgomery T-tube has two intratracheal limbs
(the proximal limb faces the glottis, and the distal
limb faces the tracheal carina) and one extratra-
cheal limb, which exits from the tracheal and neck
incision to the environment. In one approach, the
HFPPV catheter passes straight down through the
two intratracheal limbs and gas outflow was via
the extratracheal limb.
42
Alternatively, the HFPPV
catheter can be introduced through the extratra-
cheal limb and gently flexed downward to direct
the catheter to lie above the carina.
Previously, it was difficult to establish an ade-
quate airway for the administration of conventional
IPPV with conventional cuffed tracheostomy tubes
using the tracheal Montgomery T-tubes.
52 53
The
use of the extratracheal limb as an airway for the
delivery of large tidal volume IPPV was associated
with a large gas leak through the open proximal
intratracheal limb. To establish adequate ventila-
tion with conventional IPPV, the superior part of
the proximal intratracheal limb had to be occluded
by a Fogarty embolectomy catheter (introduced
through the extratracheal limb) and by a tight pha-
ryngeal pack (introduced through the oral cavity);
this prevented escape of the tidal volume through
the open glottis. IPPV has also been administered
to a pediatric patient through the proximal intratra-
cheal limb of the tracheal T-tube; the proximal
intratracheal limb was intubated with a Cole tube,
and then the extratracheal limb was occluded to
allow for the use of IPPV.
54
These previous con-
ventional IPPV maneuvers and techniques are
cumbersome, are difficult to apply, and can be
dangerous. Blockage of the various T-tube limbs
can, for example, occur suddenly and totally ob-
struct the airway and prevent alveolar ventilation.
In addition, use of these IPPV techniques can im-
pair surgical access and complicate the surgery.
The third type of airway surgery aided by HFV
is resection of a tracheal stenosis (Fig. 12-9).
High-Frequency and High-Flow Apneic Ventilation During Thoracic Surgery
Three Types of Airway Surgery Aided
by Small HFV Catheter
Tracheal
Resections
Small HFV
Catheter
Single Lumen
Endotracheal
Tube
Tracheal
Reconstruction
Supported by
Montgomery T-tube

Sleeve
Pneumonectomy,
Carinal Resections
Small HFV Catheter
Passing From the
Extratracheal Limb
Through the Distal
Intratracheal Limb
or
Small HFV Catheter
Passing Straight Down
the Two Intratracheal
Limbs
One or Two
Small HFV
Catheters
Most of the previous discussion concerning HFV
for carinal resections also applies to tracheal resec-
tions. In HFJV and HFPPV, insufflating catheters
have been inserted through a single-lumen endo-
tracheal tube, which lies above the tracheal steno-
sis, until the distal end of the HFJV and HFPPV
catheters are beyond the tracheal stenosis (usually
in main-stem bronchi irregardless of whether the
patient is an adult, child, or infant).
43
55-57
Thus, in
all cases, tracheal resection and end-to-end anas-
tomosis could be accomplished easily around the
small catheter. This method has the obvious draw-
back that the relation between the tracheal stenosis
and HFJV and HFPPV catheter must permit a suf-
ficiently large passage for exhalation; the passage
must be guaranteed at every moment of the sur-
gical procedure.
43
C. Use in Bronchopleural Fistula and
Tracheobronchial Tree Disruptions
The second proposed advantage of HFV for tho-
racic surgery is that the lower inspiratory pressures
and tidal volumes may result in a smaller gas leak
through pathologic low-resistance pathways such
as bronchopleural fistulas and tracheobronchial
disruptions (Fig. 12-10/4). Consequently, air leaks
(loss of tidal volume) and mediastinal and intersti-
tial emphysema may be minimized with this form
of ventilatory treatment, and HFJV has been suc-
cessfully used in the treatment of major broncho-
pleural fistula
58-63
and tracheobronchial tree disrup-
tions
38
M 65
and in a case of tracheoesophageal
fistula,
66
and HFPPV has been used in a case of a
right main-stem bronchus tear.
67
In most cases of
bronchopleural fistula, P
a
C0
2
was unacceptably
high despite a high minute ventilation on volume-
controlled IPPV, whereas HFJV restored normo-
carbia. However, it appears that the reduction in
air leak flow through the fistula is directly related
to reduced airway pressure with HFV compared
with conventional mechanical ventilation; when
peak and mean tracheal pressures are decreased by
HFV, fistula leak is decreased;
68
when mean air-
way pressure remains constant leak will be
constant;
69
70
and when peak and mean tracheal
pressure are increased by HFV, fistula flow in-
creases.
61
^
63
70
71
Table 12-2 COMBINATIONS OF HIGH-
FREQUENCY VENTILATION USED
FOR THE NONDEPENDENT
(OPERATIVE) AND DEPENDENT
(NONOPERATIVE) LUNG DURING
MAJOR CONDUCTING AIRWAY
SURGERY
Abbreviations: HFJV = high-frequency jet ventilation;
HFPPV = high-frequency positive-pressure ventilation; HFOV
= high-frequency oscillation ventilation.
Thus, in a series of seven consecutive patients
with an average bronchopleural fistula leak greater
than 5 L/min, HFJV (125-150/min) caused only
two patients to have clinically important decreases
in airway pressure and air leak fistula flow, and
none had significant improvement in gas ex-
change.
61
In another series of seven patients, air leak re-
mained constant (at constant mean airway pres-
sure) but oxygenation decreased.
69
Both authors
recommended measurement of tracheal pressure
(and fistula flow) to predict what is happening to
air leak fistula flow.
61
69
In one case of broncho-
pleurocutaneous fistula,
58
intraoperative differen-
tial lung ventilation, consisting of conventional
IPPV for the normal dependent lung and HFJV for
the diseased nondependent lung, was used. HFV
has been used prophylactically to prevent the de-
velopment of a bronchopleural fistula in a patient
who had a very friable bronchial stump due to
mucormycosis (an opportunistic pulmonary paren-
chymal infection that occurs in immunocompro-
mised hosts and diabetics and is associated with a
Figure 12-9 The three types of airway surgery aided by small high-frequency ventilation (HFV) catheters are tracheal resections,
tracheal reconstructions that require support by a Montgomery T-tube, and carinal procedures (sleeve pneumonectomy, carinal
resections). With tracheal resections (A), a simple HFV catheter can be passed beyond the point of airway interruption, but above
the tracheal carina, and used to ventilate both lungs with HFV. With tracheal reconstructions supported by a Montgomery T-tube
(B), the small HFV catheter can be passed from either the extraluminal limb or the proximal intraluminal tracheal limb to the distal
intraluminal tracheal limb and can be used to ventilate both lungs with HFV. With carinal procedures (C), one or two HFV catheters
can be passed into one or both of the main-stem bronchi and can be used to ventilate one or both of the lungs with HFV.
Use of HFV to Treat Bronchopleural Fistula
Low Airway
Pressures
Decrease BPF
Leak
Alternative Methods to Treat Bronchopleural Fistula
Unidirectional
Chest Tube Valve
Which Seats During
Inspiration
Collapse of
Diseased Lung
(One-Lung
Ventilation),
. , * Differential
Application of Positive ,
i r
^
r s n r r n L U n O
Counterpressure = PEEP
w
_ ~
to Pleural Space During
Exhalation
Ventilation
Figure 12-10 The use of high-frequency ventilation (HFV) to treat bronchopleural fistula (A) is based on the supposition thai
HFV results in lower airway pressures, and the lower airway pressures decrease the bronchopleural fistula (BPF) air leak. Th(
alternative methods of treating bronchopleural fistula consist of placing unidirectional valves (which seat during inspiration) in the
chest tube (helps lock in inspiratory tidal volume) (B), applying positive counterpressure to the pleural space during exhalatioi
(helps lock in positive end-expiratory pressure [PEEP]) (C), and collapsing the diseased lung (one-lung ventilation) for a period
time (D). However, chest tube valves often do not work owing to the presence of blood or pus, positive counterpressure to th
pleural space may have major negative hemodynamic implications, and collapse of the diseased lung for an extended period of tim
is technically difficult from the point of view of maintaining proper double-lumen tube position and predisposes the collapsed lun
to infection.
high incidence of postoperative bronchopleural fis-
tula).
72
HFPPV and HFOV have been successfully
used in dogs with bilateral upper lobe broncho-
pleural fistulas,
73
but these two types of HFV have
not been used in humans (except for, perhaps, the
one case of main-stem bronchial tear
67
) with a
bronchopleural fistula. In cases of tracheobronchial
disruptions, mediastinal and interstitial emphy-
sema was believed to be minimized by the use of
HFV
3 8
6 4 , 6 5
6 7 , 7 4
The previous methods of trying to make cor
ventional IPPV more effective in the presence c
pathologic low-resistance airway pathways (in a(
dition to using a very large tidal volume) all ha\
major drawbacks (Fig. 12-1 OB to D). Unidirei
tional valves inserted in chest tubes, which se
during inspiration, frequently malfunction due
interference by blood and pus within the che
tube.
75
Application of positive counterpressure
the pleural space during inspiration causes hem
dynamic interference.
75
Collapse of the diseased
lung with the aid of a double-lumen tube for a
prolonged period of time requires very careful
monitoring, is technically difficult for the intensive
care unit personnel not trained in anesthesia, and
predisposes the collapsed lung to infection.
76
Thus,
it is likely that investigations in the use of HFV to
treat patients with pathologic low-resistance air-
way pathways will continue.
D. Use in Minimizing Movement of the
Operative Field
The third proposed advantage for HFV in tho-
racic surgery is to minimize the tidal movement of
the operative field; theoretically, the lower peak
inspiratory pressures and tidal volumes of HFV
should result in much smaller inflation and defla-
tion movements of the lung. Therefore, HFV of
the nondependent lung might provide a relatively
"quiet" operative lung field, and HFV of the de-
pendent lung could contribute to the "quiet" op-
erative lung field by providing a minimally mov-
ing mediastinum (see Fig. 12-1 and Table 12-3).
In four reports,
77
"
80
HFPPV was used to ventilate
both the nondependent and dependent lungs during
thoracic surgery (HFPPV/HFPPV). With the chest
Table 12-3 COMBINATIONS OF HFV USED
FOR THE NONDEPENDENT
(OPERATIVE) AND DEPENDENT
(NONOPERATIVE) LUNGS TO
MINIMIZE MOVEMENT OF THE
SURGICAL FIELD
Abbreviations: HFV = high-frequency ventilation; HFPPV
= high-frequency positive-pressure ventilation; HFJV = high-
frequency jet ventilation; IPPV = intermittent positive-pres-
sure ventilation.
open, the nondependent lung had minimal ventila-
tion-synchronous movements and limited lung
expansion but still had good aeration and no ate-
lectasis. However, hyperinflation of the lungs
occurred in some patients with chronic obstructive
pulmonary disease, presumably owing to air trap-
ping, and resulted in poor operating conditions.
80
In patients who did not experience hyperinflation,
operating conditions for peripheral lung proce-
dures was adequate. During HFPPV, the airways
could be suctioned without interrupting ventila-
tion. At the conclusion of surgery, the nondepen-
dent lung fully reexpanded as readily as with
conventional techniques. Arterial blood gases, ana-
lyzed frequently in most reports, showed adequate
oxygenation and carbon dioxide removal during
HFPPV despite some brief periods of significant
nondependent-lung compression by the surgeons.
In four reports, HFJV of the nondependent lung
while the dependent lung was ventilated with in-
termittent positive pressure breathing achieved ad-
equate oxygenation ventilation.
I9
81
-
83
The positive
attributes of nondependent-lung HFJV and depen-
dent-lung IPPB are more effective C0
2
removal
(compared with nondependent-lung atelectasis or
continuous positive airway pressure [CPAP]) and.
therefore, a possible decrease in dependent-lung
tidal volume (compared with nondependent-lung
atelectasis or CPAP), adequate oxygenation, good
peripheral lung surgery conditions, and a possible
increase in cardiac output (compared with CPAP).
The negative attributes are that complicated/ ex-
pensive technology is required, and air trapping
and poor central (hilum, mediastinum) operating
conditions may exist (see chapter 11 and Table
11-3).
In one other report, HFJV was successfully used
at 3 Hz to ventilate both lungs, but carbon dioxide
retention became a problem at 6 and 12 Hz.
84
Cu-
riously, in another report, two-lung HFJV at a rate
of 2 Hz, inspiration to expiration ratio of 1:3, and
driving pressure of 1.6 atm resulted in good intra-
operative gas exchange and less postoperative pul-
monary morbidity (decreased hospital stay, de-
creased incidence of chest infection, and increased
postoperative P
a
0
2
) compared with conventional
85
In the two-lung HFJV re-
port referred to, with respect to the direct relation-
ship between respiratory rate and lung volume,
overdetention of the lungs created technical prob-
lems for the surgeons.
18
Finally, in one report, the
dependent lung was selectively ventilated with an
HFV flow interrupter (while the operated lung was
completely collapsed).
86
Compared with selective
dependent-lung IPPV, selective dependent-lung
HFV significantly improved arterial oxygenation,
presumably because of lower dependent-lung pul-
monary vascular resistance (due to either lower
dependent-lung airway pressures
1
or release of de-
pendent-lung vasodilator prostaglandins
2
) and en-
hanced operating conditions (owing to minimal
mediastinal movements). In addition, deflation of
the dependent-lung endobronchial tube cuff further
improved arterial oxygenation owing to overflow of
gases from the dependent lung to the nondependent
lung, which resulted in recruitment of part of the
collapsed lung in gas exchange (Fig. 12-11).
86
-
88
In one report, the use of dependent-lung HFJV
was compared with dependent-lung positive-pres-
sure ventilation during conventional one-lung ven-
tilation (without nondependent-lung CPAP) in
terms of operating conditions, hemodynamics, and
efficiency of gas exchange.
89
Ten patients about to
undergo thoracotomy for carcinoma of the lung
were entered into a crossover trial to compare car-
diovascular and respiratory function during HFJV
and conventional mechanical ventilation for one-
lung anesthesia. All patients were anesthetized
with a standard technique using double-lumen
tubes and placed in the lateral position with the
left chest open. The results showed no significant
differences with regard to ventilation sequence,
but one-lung HFJV gave moderately higher values
than one-lung conventional ventilation for shunt
(32.9 vs. 42.4 per cent, < .01) and positive end-
expiratory pressure (0 vs. 6 mm Hg, < .05) and
lower peak inflation pressure values (p < .01).
There were no significant differences in cardiac
output, pulmonary capillary wedge pressure, arte-
rial carbon dioxide, and available oxygen. Surgical
conditions for mainly peripheral lung surgery were
satisfactory during both methods of ventilation and
satisfactory gas exchange occurred. It was, how-
ever, more difficult to assess adequacy of ventila-
tion during HFJV, and the routine use of this
method of ventilation was not recommended dur-
ing one-lung anesthesia. In addition, the use of
HFV to minimize mediastinal and hilar move-
ments must remain controversial in view of the
report describing a mediastinal "bounce" with
each oscillation, which made surgery around these
structures nearly impossible.
80
In summary, the favorable aspects about HFV
for thoracic surgery are adequate pulmonary gas
exchange (and, perhaps, increased diversion of
blood flow away from the atelectatic lung during
Figure 12-11 Gas kinetics during MOL-HFV are shown by the arrows. The turbulent flow generated at the carina allows for the
participation of the nondependent lung in gas exchange. The P
a
0
2
, P
a
C0
2
, and Q
S
/Q, shown in the box were obtained from 26
patients. The parameters of HFV used are shown in the top box. (F = frequency; DGP = driving gas pressure; IT% = inspiratory
time; PaW = airway pressure.) (From El-Baz N: Application of constant positive airway pressure to the nondependent lung is not
preferable to high-frequency ventilation to optimize oxygenation during pulmonary surgery. J Cardiothorac Anesth 1:589-591,
ventilation of just the dependent lung),
1
good con-
ditions for peripheral lung surgery, possibly good
conditions for major airway conducting surgery,
possibly improved gas exchange in cases of bron-
chopleural fistula, and possible decreased medias-
tinal and pulmonary interstitial emphysema in
cases of tracheobronchial disruption. As previ-
ously mentioned, the first two of these favorable
aspects, in the context of the availability of one-
lung ventilation, seem superfluous and unneces-
sary. The unfavorable aspects of HFV for thoracic
surgery are possibly unsatisfactory surgical condi-
tions for mediastinal and major airway surgery,
hyperinflation in patients with chronic obstructive
pulmonary disease and patients with broncho-
spasm, difficulty in monitoring heart and breath
sounds with an esophageal stethoscope, inability
to judge the adequacy of ventilation from the mo-
tion of the chest wall or lungs, use of high flows
of anesthetic gases, and difficulty in assessing lung
volume. Hence, at present, the exact role of HFV
in thoracic surgery is uncertain. I am certain, how-
ever, that the use of HFV as a routine procedure
in thoracic surgery cannot be recommended at
present.
III. LOW- AND HIGH-FLOW APNEIC
VENTILATION
A. Low-Flow Apneic Ventilation
The need for an absolutely quiet surgical field
for short periods of time often arises during a
thoracotomy in which a standard endotracheal tube
and two-lung ventilation is employed. This can be
accomplished relatively safely using the principle
of apneic mass-movement oxygenation. If ventila-
tion is stopped during the administration of 100
per cent oxygen and the airway is left connected
to a fresh gas supply, oxygen will be drawn into
the lung by mass movement to replace the oxygen
that crossed the alveolar-capillary membrane (the
uptake of oxygen creates a slight subatmospheric
pressure
90
(Fig. 12-124). There is usually no dif-
ficulty in maintaining an adequate arterial P
a
0
2
(especially if 5 to 10 cm H
2
0 CPAP is used) dur-
ing at least 20 min of apneic mass-movement ox-
ygenation. P
a
0
2
will decrease only at approxi-
mately the same rate as the P
a
C0
2
increases (see
later discussion) (C0
2
replaces oxygen in the al-
veolar space in a ratio of one molecule to one
molecule).
90
If the flow of oxygen into the lungs is supracar-
inal and through catheters of internal diameter
greater than 3 mm (i.e., a relatively large internal
diameter catheter results in relatively slow exit
velocity of the gas) and a relatively low flow rate
(less than 0.2 L/kg/min) is used, almost all of the
carbon dioxide produced is retained, and in anes-
thetized subjects the arterial carbon dioxide ten-
sion (P
a
C0
2
) rises at approximately 6 mm Hg in
the first minute owing to the wash in of venous
blood into the arterial compartment (venous blood
has a carbon dioxide tension 6 mm Hg higher than
that of arterial blood) and then 2 to 4 mm Hg each
minute thereafter
91-93
owing to normal C0
2
pro-
duction (see Fig. \2-\2A). In awake apneic human
subjects, the rate of rise of P
a
C0
2
may be much
faster (as much as 10 mm Hg in the first minute
and 6 mm Hg/min thereafter [resulting, in part,
from increased C0
2
production]).
94
On the basis of
these considerations, if a patient had a normal CO
:
production, was hyperventilated to a P
a
C0
2
of 30
mm Hg, and was then made apneic, and oxygen
was insufflated into the lungs with a low flow,
P
a
C0
2
after 10 min of apnea would be 63 to 72
mm Hg. Indeed, one report describes a series of
eight patients in whom apneic mass-movement ox-
ygenation was employed for a period of 18 to 55
min after normal ventilation.
91
Although the lowest
arterial saturation that resulted was 98 per cent, the
arterial P
a
C0
2
in the five patients in whom it was
measured ranged from 103 to 250 mm Hg, and the
pH ranged from 6.72 to 6.97. Although severe
degrees of hypercapnia and respiratory acidosis
may be well tolerated in some healthy patients, it
appears that the safe period of low-flow apneic
oxygenation during thoracotomy would lie well
under 10 min.
If the insufflating flow rate is low (approxi-
mately 0.25 L/kg/min), and the insufflating cathe-
ters are subcarinal (by 3 cm) and have a very small
internal diameter (0.8 mm), then the exit velocity
of the gas is much increased, and it is possible to
maintain P
a
0
2
greater than 45 mm Hg and P
a
CO
:
less than 65 mm Hg for at least 2 hours.
95
B. Interrupted High-Flow Apneic
Ventilation
If the fresh gas flow into the lungs is very high.
it is possible to wash out the carbon dioxide from
the alveolar space (Fig. 12-125). The higher the
flow rate, the greater the effective minute ventila-
tion and carbon dioxide removal. At the same
time, if the fresh gas were oxygen, whatever oxy-
gen was taken up by the pulmonary perfusion
would be replaced, and arterial oxygenation would
not be a problem. These high-flow apneic ventila-
tion concepts have been tested clinically in hu-
mans and experimentally in animals.
1. Low-Frequency Interrupted High
Flow
Interrupted high-flow apneic ventilation is es-
sentially equivalent to very short inspiratory time
Low Flow (<0.2 L/kg/min) Apneic Ventilation
Figure 12-12 Low-flow (< 0.2 L/kg/min) apneic ventilation (A) can provide adequate arterial oxygenation for a period of tim
but does not remove any carbon dioxide so that hypercapnia is a necessary consequence of this method of ventilation. On the othe
hand, high-flow (1 L/kg/min) apneic ventilation (B) maintains arterial oxygenation and, at the same time, is able to remove carbo
dioxide at a rate that is proportional to the oxygen flow rate. High-flow apneic ventilation may be facilitated by use of constar
positive airway pressure (CPAP).
positive-pressure breathing or jet ventilation that
results from short bursts of extremely high flows
of gas through a small ventilating catheter posi-
tioned in a main-stem bronchus (Fig. 12-13).
96
97
The high-pressure (50 psi) gas source is oxygen
(available from portable tanks or central supply
wall outlets) and is capable of delivering a flow of
100 L/min. Since the flows are so great, the short
bursts of oxygen do not cause entrainment of room
air. Depending on airway resistance and lung com-
pliance, the inflation is adjusted (to as low as 30
to 40 L/min) through a reducing valve, so that the
airway pressure is 25 to 40 cm H
2
0. By means of
an on/off release button, inflation rates are main-
tained at 10 to 15 respirations/min in adults and
less than 40 breaths/min in infants,
97
and the ade-
quacy of ventilation is estimated by direct obser-
vation of thoracic cage expansion, mediastinal ex-
cursion, and arterial blood-gas monitoring.
A shortened nasogastric tube, which has been
passed down through the lumen of a single-lumen
tube into a main-stem bronchus (and perhaps
through a divided major conducting airway as
well), has been used as the ventilation catheter in
adults.
96
The appropriate length of the nasogastric
tube is 75 to 40 cm, and the nasogastric tube is
sized according to the patient's weight. Usually, a
steel wire has to be threaded through the lumen of
the nasogastric tube to provide some rigidity.
In infants, the trachea has been intubated orally
with a 14-cm long, 2.5- to 3.5-mm internal diam-
eter, uncuffed polyvinylchloride tube, placing the
tip above the stenotic segment.
97
A 40-cm French
polyethylene tube with a single distal orifice was
Low Frequency Interrupted High Flow Ventilation
50 psi 0
2
Source
(Flow = 100 L/min)
Figure 12-13 Interruption as well as release of a very high fresh gas flow (100 L/min) at a low frequency (10 breaths/min) is
essentially equivalent to intermittent positive-pressure breathing with a very short inspiratory time. A small catheter (nasogastric
tube) can be passed through an indwelling single-lumen tube in the trachea and beyond any airway interruption into one or both of
the main-stem bronchi and used to ventilate one or both of the lungs. The escape of gas on exhalation is around the small catheter
but inside the indwelling single-lumen tube.
used to provide low-frequency interrupted high-
flow ventilation. This was passed through the en-
dotracheal tube and placed just above the stenosis.
The distal airway pressure was monitored via a
130-cm 3 French polyvinylchloride catheter in-
serted with the tip below the stenosis. During jet
ventilation, the proximal end of the endotracheal
tube was left open to the atmosphere to provide an
exit for the continuous outflow of gas. The jet
ventilator used (Mera HFO Jet Ventilator, Senko
Medical Industrial Co., Tokyo, Japan) has a
pneumatic valve system. Oxygen concentration
during jet ventilation was adjusted with an oxy-
gen-nitrous oxide blender. Humidification was
provided by the intermittent infusion of 0.45 per
cent saline directly into the jet stream every 10
to 15 min.
a. USE IN MAJOR CONDUCTING AIRWAY
SURGERY
The value of the interrupted high-flow apneic
ventilation technique was tested in 18 adult pa-
tients
96
and four infants
97
undergoing tracheobron-
chial tree reconstruction. In the adults, the average
duration of catheter ventilation was 35 minutes.
Seven patients had a tracheal resection (six cervi-
cal, one thoracic), nine had a sleeve pneumonec-
tomy, and two additional patients had a carinal
resection. The period of catheter ventilation usu-
ally began after the airways were divided. The
catheter was advanced through either a single-lu-
men or double-lumen endotracheal tube into the
distal airway, and the entire anastomosis was done
around the small nasogastric inflation catheter.
Throughout the period of interrupted high-flow ap-
neic ventilation, the patients were kept fully para-
lyzed and anesthesia maintained by intravenous
narcotics. At completion of the anastomosis, the
catheter was withdrawn, and ventilation was re-
sumed through the existing conventional endotra-
cheal tube.
For patients undergoing carinal resection, the
ventilation was provided by both lungs by using a
separate nasogastric catheter for each bronchus.
Both catheters were introduced through the endo-
tracheal tube, so that the need for intubation across
the operating field was avoided. Pulmonary inter-
stitial and mediastinal emphysema developed dur-
ing the period of high-flow ventilation in one
patient. Although the authors had no clear expla-
nation for this complication, it was possible that
the pressure was too high in the airway or that the
catheter made a tear in the bronchial mucosa.
In the infants, ventilation was initiated manually
via the endotracheal tube, then switched to low-
frequency intermittent high-flow ventilation via
the 5 French catheter using a frequency of less
than 40 breaths/min with an inspiratory to expira-
tory ratio of 1:3. Immediately after resection of the
stenotic area, the jet ventilation catheter and the
airway pressure monitoring catheter were ad-
vanced into the open distal trachea. The catheter
measuring airway pressure was always placed dis-
tal beyond the insufflation catheter. The pattern of
airway pressure provided useful early information
regarding adequacy of ventilation. During resec-
tion and reconstruction, surgical manipulations fre-
quently produced mechanical obstruction by com-
pression of the trachea. This caused gas trapping
as a result of obstruction of the expiratory flow,
which could have caused barotrauma. The contin-
uously monitored airway pressure in the distal tra-
chea detected both elevated pressure caused by gas
trapping and low pressure resulting from inade-
quate ventilation. After anastomosis of the trachea,
the endotracheal tube was advanced into the distal
trachea and manual ventilation was reinstituted.
The inflation nasogastric catheter technique ap-
pears to have specific advantages in regard to sim-
plicity of the equipment and anesthetic technique.
At all times, the anesthesiologist has complete
control of the airway while oxygenation is main-
tained by intermittent inflation of the lungs
through a small, semirigid endotracheal catheter (2
to 6 mm in diameter). The high flow of oxygen
does not entrain air, and the inspired fraction of
oxygen is predictable. Ventilation rates are main-
tained at 10 to 15 respirations/min, and the infla-
tion pressure (15 to 40 cm H
2
0) is regulated by
the direct observation of thoracic cage expansion
and mediastinal excursion and by arterial blood-
gas values. The measurement of high arterial oxy-
gen tension in most patients indicates that a lesser
concentration of inspired oxygen could be used
safely.
Surgical exposure to the trachea is improved by
the interrupted high-flow ventilation because there
is no intrusion of anesthetic apparatus into the
operating field, no endotracheal manipulations
need to be done, and reconstruction can be done
without interruption around the small inflating
nasogastric catheter. This study shows that the
high-flow inflation catheter technique fulfills the
requirements for safe surgical and anesthetic man-
agement during tracheobronchial operations. It ap-
pears to be most useful during operations on the
intrathoracic trachea and should be considered a
valuable alternative to more conventional methods.
2. Continuous High-Flow Apneic
Ventilation
Adequate gas exchange has been accomplished
in dogs without tidal exchange of gases (apnea) if
the airspace is exposed to a very high continuous
flow rate of oxygen; carbon dioxide is washed out
and oxygen is washed in (Fig. 12-12Z?).
98-10
'
Again, as with low-frequency interruption of high
flows, continuous high-flow apneic ventilation re-
quires that the small catheters (2.5-mm internal
diameter) need to be positioned in a main-stem
bronchus (2.5 to 3.0 cm below the carina), and the
flow must approximate 1.0 to 1.2 L/kg/min.
98
"
l 0 1
The catheter position must be precise because su-
pracarinal catheters are much less effective than
subcarinal catheters and the exact depth of the
subcarinal catheter is critical.
102
I03
Insufflation
catheters may be either within or without the en-
dotracheal tube, but in either case the position has
to be stabilized/fixed in some manner. Smaller
bore subcarinal catheters (less than 2.5-mm inter-
nal diameter) at any given insufflation flow appear
to create a higher than optimal exit velocity in
which the exit gas jet may by-pass many of the
branches of the bronchi leading to underventila-
tion. Subcarinal bronchial insufflation tubes of
larger internal diameter (> 2.5 mm) produce too
slow a gas exit velocity, resulting in insufficient
penetration to allow adequate ventilation.
Other observations suggest that there is an opti-
mal gas exit velocity and that it may be possible
to provide this optimal gas exit velocity and ade-
quate ventilation by decreasing both the internal
diameter of the insufflation catheter (e.g., from 2.5
to 1.4 mm, a 44 per cent reduction) and the insuf-
flation flow (e.g., from 1.2 to 0.5 L/kg/min, a 67
per cent reduction) so as to achieve the same gas
exit velocity.
104
Continuous flow rates less than 0.7
L/kg/min have been shown to cause carbon diox-
ide retention
98
mi
as well as fail to decrease P
a
C0
2
as effectively as higher flow rates in a canine
model of ventilatory failure.
105
Continuous high-
flow apneic ventilation is more likely to be suc-
cessful in species with a large collateral ventilation
(dogs) than in species with a low collateral venti-
lation (cats and pigs)
106
(see chapters 2 and 3 and
the following discussion of mechanism of high
flow ventilation). Oxygenation is improved if 4 cm
H
2
0 CPAP is used to the lung that is receiving the
high flow of gases.
99
Without CPAP, the airway
pressure caused by high gas flow does not exceed
2 mm Hg. The fresh gas should be heated and
humidified. Under these conditions, arterial oxy-
genation and carbon dioxide removal are well
maintained. In addition, the method works well
with the chest open as well as closed,
99
and it may
be beneficial in managing a bronchopleural fis-
tula.
104
Gas transport during continuous high-flow ven-
tilation has been explained using a two-zone serial
model of the lung. In the region just distal to the
catheters (zone la), bidirectional flows exist be-
cause, in steady state, all the flow entering the lung
(except for the excess of 0
2
consumption over C0
2
production) must also leave the lungs. Zone lb is
just distal to zone la and is where the turbulence
generated by the flow leaving the catheters (J
ets
)
dominates.
Zone II is the distal lung region in which molec-
ular diffusion and cardiogenic oscillations are the
primary gas transport mechanisms. In addition,
data suggest that another mechanismcollateral
ventilationmight be important during continuous
high-flow ventilation, because recent studies
showed that continuous high-flow ventilation is
completely ineffective in pigs, which are known to
have very high resistances to collateral ventilation.
The relative ineffectiveness of continuous high-
flow ventilation in humans must be explained by
their degree of collateral resistance, which lies
somewhere between that of dogs and pigs.
Continuous high-flow apneic ventilation has
been partially successful in humans.
107
In five fe-
male patients undergoing pelvic procedures,
curved endobronchial catheters (2.5-mm outside
diameter) were placed 2 cm below the carina with
a fiberoptic bronchoscope and secured in place by
inflation of an endotracheal tube cuff. Catheter
position was verified a second time with the fiber-
optic bronchoscope. Continuous high-flow apneic
ventilation was then started with humidified oxy-
gen at total flows between 0.6 and 0.7 L/kg/min
for 30 min. Average control P
a
0
2
was 321 mm Hg
and 30-min P
a
0
2
was 299 mm Hg, whereas control
P
a
C0
2
was 37 mm Hg and 30-min P
a
C0
2
was 55
mm Hg. The rate of rise of P
a
C0
2
was 0.6 mm
Hg/min, which compares favorably with the 3.8
mm Hg/min rise in anesthetized humans exposed
to continuous low-flow apneic ventilation.
91
92
For reasons not well delineated, C0
2
removal
during continuous high-flow apneic ventilation
may be enhanced by using air as the insufflated
gas.
108
In two other human studies, continuous
high-flow ventilation did not achieve normal levels
of P
a
C0
2
but did provide about one third to one
half the normal alveolar ventilation, with P
a
C0
2
rising at a rate roughly half of that observed during
apneic oxygenation.
109
Obviously the technique
will have to be improved (higher flow rates, better
catheter position, more appropriate catheter size,
etc.) before it can find a useful application in tho-
racic surgery.
REFERENCES
1. Hall SM, Strawn WB, Levitsky MG: Effect of high-fre-
quency oscillation on blood flow to an atelectatic lung in
closed-chested dogs. Crit Care Med 12:447-451, 1984.
2. Wetzel RC, Gordon JB, Gregory TJ, et al: High-fre-
quency ventilation attenuation of hypoxic pulmonary
vasoconstriction. Am Rev Respir Dis 132:99-103, 1985.
3. Gillespie DJ: High-frequency ventilation: A new concept
in mechanical ventilation. Mayo Clin Proc 58:187-196,
1983.
CHAPTER
?
13
....... , .
Anesthetic Considerations
(Other Than Management of
Ventilation) During and at the
End of Thoracic Surgery
I. Introduction
II. Management of Bronchospasm
III. Management of Blood Loss
A. Assessment of Blood Loss
1. Direct Observation
2. Physiologic Response to Blood
Loss
B. Minimal Blood Loss (<10 Per Cent
of Blood Volume): Amount of
Crystalloid Fluid Infusion
C. Moderate Blood Loss (10 to 20 Per
Cent of Blood Volume): Should
Blood Be Infused?
D. Severe Blood Loss (>20 Per Cent
of Blood Volume): Diagnostic and
Therapeutic Management Problems
1. Causes of Severe Bleeding
2. Hemostatic Defects Resulting
From Massive Blood Transfusion
3. Management of Massive
Transfusion
IV. Treatment of Nonblood Loss,
Deleterious Hemodynamic Changes
A. Rationale Plan for Managing Low
Cardiac Output State
VI.
B. Pulmonary Artery Hypertension and
Right Ventricular Failure
C. Coronary Artery Disease
1. Monitoring for Ischemia: Further
Considerations
2. Coronary Artery Disease and
Good Ventricular Function
3. Coronary Artery Disease and
Poor Ventricular Function
D. Valvular Heart Disease
E. Arrhythmias
Re-Expansion of Collapsed Lung
During and at the End of Thoracic
Surgery
Transport of Patient
A. Preparation of Patient for Transport
1. Airway
2. Respiration
3. Chest Tube and Drainage
System
4. Circulation
5. Coagulation
6. Miscellaneous
B. Transport
C. Arrival in Intensive Care Unit
454 Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery
I. INTRODUCTION
The management of ventilation is the most com-
mon major anesthesia problem during thoracic sur-
gery. Chapters 11 and 12 were solely concerned
with this problem. One-lung ventilation is abso-
lutely indicated in a number of cases and relatively
indicated in most cases. If hypoxemia should oc-
cur during conventional management of one-lung
ventilation, differential lung management consist-
ing of nonventilated lung continuous positive air-
way pressure (CPAP), with or without ventilated
lung positive end-expiratory pressure (PEEP),
should be instituted. In a few cases involving
opening of a major conducting airway, high-fre-
quency and high-flow apneic ventilation may be
indicated.
There are a number of other major anesthetic
problems that occasionally occur during thoracic
surgery. First, since many of these patients are
long-term smokers, have reactive airways, and will
experience tracheobronchial tree manipulation,
there is an increased incidence of bronchospasm.
Second, blood and fluid infusion may be problem-
atic. Most thoracic surgery patients do not require
replacement of blood loss, and in questionable
cases there are survival data (related to recurrence
of cancer) to suggest that blood transfusion should
be avoided. There are also data on postpneumo-
nectomy pulmonary edema that suggest that crys-
talloid infusion should not be overly aggressive.
However, some thoracic surgery cases, especially
those with extensive local tumor invasion, partic-
ularly to the pleura, may involve a great deal of
blood loss and require massive transfusion; the
blood loss and transfusion requirement may be
compounded by the development of hemostatic
deficits. Third, some thoracic surgery patients have
significant coronary artery disease, and the deter-
minants of myocardial oxygen supply and demand
may need to be controlled within narrow limits.
Finally, the transport of a thoracic surgery patient
from the operating room to a new location (inten-
sive care unit or recovery room) can be a perilous
misadventure. Consequently, the preparation of the
patient for transport at the end of a case and the
transport of the patient to a new location must be
performed with rigorous attention to detail. This
chapter discusses the proper anesthetic manage-
ment of intraoperative bronchospasm, blood loss,
hemodynamic changes, and transportation of criti-
cally ill thoracic surgery patients.
II. MANAGEMENT OF
BRONCHOSPASM
Because 2 to 3 per cent of the population have
asthma, many more have a bronchospastic com-
ponent to their chronic obstructive pulmonary dis-
ease (COPD), and laryngoscopy and endotracheal
intubation are potent stimuli for the development
of bronchospasm, many patients are at risk for the
development of airway narrowing and broncho-
spasm. Airway narrowing and the resultant resis-
tance to airflow have profound effects on pulmo-
nary function. The obstruction is not uniform so
that both air trapping and airway closure occur and
ventilation-perfusion mismatch ensues, with an in-
crease in wasted (dead space) ventilation (caused
by air trapping) and an increase in shunt. Indeed,
air trapping may be so marked in cases of severe
bronchospasm that functional residual capacity ap-
proximates total lung capacity, leaving minimal
inspiratory reserve. Pulmonary hypertension can
also result because the right ventricle is pressure
and volume loaded by the hyperexpanded lung,
whereas the left ventricle is relatively underfilled.
The physical signs normally associated with
acute airflow limitation in an awake, sponta-
neously breathing patient are tachycardia, tachy-
pnea, hyperinflation, wheeze, accessory muscle
use, pulsus paradoxus, and diaphoresis. The use of
the accessory muscles, pulsus paradoxus, and di-
aphoresis are considered the most significant signs
in terms of grading the severity of an attack;
1
ob-
viously some of these signs are relevant to the
anesthetized patient, but auscultation of breath
sounds (bronchospasm), increase in peak and pla-
teau airway pressure, decrease in inspired volume,
and a steep slope to the phase III alveolar plateau
of the capnogram are the usual key intraoperative
diagnostic tests.
The treatment of intraoperative bronchospasm
can be divided into three categories (Fig. 13-1).
The first category contains measures that can be
done immediately, should be done in the majority
of patients, and may correct bronchospasm when
it is mild and not very intense. The second cate-
gory involves administration of major bronchodi-
lating drugs that should be used in patients in
whom the first category of therapeutic measures
does not work and/or in whom the bronchospasm
has a moderate to severe intensity. The third cate-
gory involves administration of drugs that should
be considered as having a minor impact on bron-
chospasm but might make a difference in patients
in whom the first two categories have failed to
resolve the bronchospasm completely.
The first category of therapy involves measures
that should be done in most patients and may
likely reduce the severity of bronchospasm for the
majority of times it occurs during anesthesia (Fig.
13-1, category I). First, the F,0
2
should be in-
creased as a prophylactic measure against the de*
velopment of hypoxemia.
Second, the level of anesthesia should be deep-
Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery 4 5 5
Treatment of Intraoperative Bronchospasm
Figure 13-1 The treatment of intraoperative bronchospasm can be divided into three major categories. The first category consists
of general measures that are used in most patients and may be expected to correct mild to moderate bronchospasm. Many authors
include the use of intravenous lidocaine in this first category. The second category consists of administration of major bronchodilat-
ing drugs, which should be used in patients who have moderate to severe bronchospasm. The third category consists of minor
bronchodilating drugs that may be helpful in a few patients who have specific indications (such as increased vagal tone, release of
histamine), (iv = intravenous; it = intratracheal.)
ened. This may be accomplished in two ways. If
an immediate effect is desired, sodium pentothal,
1 mg/kg, or ketamine, 0.5 mg/kg intravenously,
can be administered. Alternatively, if halothane or
isoflurance are not being used or if they are being
used in low doses, the inspired concentration of
the halogenated drug should be increased. How-
ever, deepening anesthesia may not be effective in
severe bronchospasm.
2
Droperidol (0.22 mg/kg)
may also be a useful way to sedate the broncho-
spastic patient
3
(alpha blockade may result in beta
predominance and bronchodilation).
4
Third, if paralysis is part of or compatible with
the anesthetic plan and the patient is only partially
paralyzed and coughing and straining, then he or
she should be fully paralyzed because exaggerated
expiratory efforts will worsen small-airway ob-
struction.
Fourth, if airway secretions are thought to be
present, they should be suctioned. Finally, because
of carinal irritation by the double-lumen tube, the
cause of the intraoperative forced exhalation/bron-
chospasm, consideration should also be given to
administering intravenous lidocaine, 1 mg/kg, as
well as down the tracheal lumen (which is just
above the carina) to all coughing and/or straining
patients (see the following).
The second category of therapy involves the
administration of major acute bronchodilator drugs
and should be used in any patient in whom serious
bronchospasm develops (see Fig. 13-1, category
II) (see also chapter 6 for preoperative preparation
with bronchodilating drugs and a discussion of the
mechanism of action of bronchodilator drugs and
dosages of the drugs). Because all of the potent
bronchodilator drug can be/should be administered
by a metered-dose inhaler (MDI) (to maximize the
benefit and minimize the risk of the drugs per se),
it is appropriate to first consider how to administer
an MDI to an anesthetized, intubated patient.
The MDI requires some sort of circle system T-
piece with 15-mm ends or fittings that has an ap-
propriately oriented port that will activate the MDI
as well as transmit the aerosol to the airway. Sev-
eral adapters are available to allow MDI use via
the anesthesia circuit or endotracheal tube of a
tracheally intubated patient; for example, an elbow
connector for a mass spectrometer or capnogram
or a universal adapter (Instrumentation Industries
Inc.)
5
are commonly used. However, a significant
problem with the MDIs involves the low percent-
age and variability of actual drug delivery to the
lungs. Much of the drug is deposited in the oro-
pharynx (or on the face) of the nonintubated,
mask-ventilated patient. In an intubated patient,
the majority is deposited in the endotracheal tube
or elbow connector. Estimates of drug delivery to
the lung via MDI circuit adapters have ranged
from 2.5 per cent to 12.3 per cent
6-8
of the dis-
charged dose, with variables including diameter
(direct relationship), length of endotracheal tube
(inverse relationship), and the exact phase of ven-
tilation (prior initiation of inspiration increases the
delivery dose of the drug; see later discussion).
Administration of an MDI via a catheter down
the endotracheal tube ensures more efficient drug
delivery for intubated patients. A simple clinical
system that allows utilization of a catheter within
the endotracheal tube is as follows.
9
Components
of the system include the standard MDI canister, a
35- or 60-ml syringe, a 12-inch, 18-gauge intrave-
nous catheter (different lengths may be required),
and a bronchoscopy elbow port adapter (Fig.
13-2, upper panel). The intravenous catheter is
connected to the syringe. The canister fits easily
into the barrel of the syringe, with the plunger
acting as a trigger. The catheter is passed through
the port of the elbow adapter and down the endo-
tracheal tube (ETT), where the MDI can be dis-
charged by simply pushing on the plunger (see
Fig. 13-2, lower panel). Just as in routine oral use,
bronchodilator discharge is maximally effective
when the canister is in the upright position. Cath-
eter length and
v
placement should allow optimal
drug delivery: outside the distal tip of the endotra-
cheal tube and above the carina. The bronchos-
copy adapter allows repeated administrations of
drug during the inspiratory phase without exces-
sive manipulation of the circuit. The circuit sam-
pling port for the mass spectrometer or end-tidal
C0
2
may allow passage of the catheter to the en-
dotracheal tube tip as well. By minimizing drug
exposure to the circuit and endotracheal tube, drug
delivery should be considerably more efficient. In-
deed, laboratory evaluations using the MDI with
various extensions and catheters have shown im-
proved estimated drug delivery to the lung ranging
from 23.3 per cent
10
to 96.7 per cent.
8
On the basis
of this increased delivery efficiency, some caution
should be exercised in administering a bronchodi-
lator with this system (i.e., titrate to effect).
The manner in which the timed inhalation (to
the activation of the MDI), which carries the bron-
chodilator from the MDI into the respiratory tract,
is delivered is very important. Maximal deposition
of small bronchodilator particles (1-5 ) to the
lower respiratory tract, which is where most of the
therapeutic effect occurs, is obtained by making
the inhalation slow (laminar flow minimizes im-
paction of particles against the walls of the tra-
cheobronchial tree and bifurcations) and deep and
using a two-second inspiratory pause (which al-
lows particles to settle against the walls of the
bronchioles."
Not surprisingly, larger doses of aerosol must
be administered during episodes of severe acute
asthma (usually two- to fourfold increase) to
achieve the maximal effect because high inspira-
tory frequencies and flow rates, low tidal volumes,
and pathologically narrowed airways all conspire
to reduce the delivered dose and the peripheral
distribution of inhaled medication. Given the sim-
plicity and ease of MDI administration and its
accurate dosimetry (as it comes out of the canis-
ter), the use of a nebulizer aerosol generator intra-
operatively does not seem warranted.
The 3
2
-agonist drugs are the most potent and
rapidly acting bronchodilators,
12
and they are ef-
fective during halothane and intravenous anesthe-
sia
13
(i.e., they have additive bronchodilator ef-
fects). Therefore, the
2
^88 should be the
first bronchodilating drug to be administered. The
p
2
-agonists should be delivered by aerosol (MDI
or nebulizer) because of increased bronchial selec-
tivity with decreased side effects (maximal bron-
chodilation can be achieved with barely detectable
blood concentrations), whereas intravenous admin-
istration results in much higher blood levels anc
an increased incidence of side effects. The pressur
ized, metered canister is convenient and provide!
accurate dosimetry (see previous discussion). Thf
various 3
2
-agonists, along with their pertinent clin
ical pharmacology, are listed in Table 13-1.
Responses to -agonists are mediated by ,
receptors, which increase heart rate and cardial
contractility, and ^, which decreasi
Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery 457
Figure 13-2 Upper panel, Individ-
ual components of the metered dose
inhaler (MDI) administration system.
Lower panel, Combined units of the
MDI delivery system. Note the tip of
the catheter extending a short distance
beyond the tip of the endotracheal
tube. (From Dunteman E, Depotis G:
A simple method of MDI administra-
tion in the intubated patient. Anesth
Analg 75:304-305, 1992. Used with
permission.)
TABLE 13-1 BETA-AGONISTS AND TREATMENT OF BRONCHOSPASM
Adrenergic Dura- Inhaled
Beta-Agonist Receptor Bronchodilator Onset Peak tion Via Arrhythmia Side
Trade Name Activity Effectiveness (Min) (Min) (Hours) Cannister Tachycardia Effects Comments
Isoproterenol
Mistometer
Medihaler-Iso
Albuterol
Proventil
Ventolin
Terbutaline
Bre thine
Bricanyl
Metaproterenol
Alupent
Metaprel
Isoetharine
Bronkometer
B, + B,
B, > B,
B, > B,
B, > B,
B, > B,
+ + +
+ + +
2 to 5 5 l to 4 125 ^ + + + + Tachyphylaxis Very potent
ventricular
fibrillation
1 PA
5 to 10 15 4 to 6 90 jig/puff + Muscle tremors Most effective
+ + 5 to 15 25 3 to 4 200 ^ + +
+ 10 to 20 25 3 65 g/puff
+ 10 to 25 25 1 to 2 340 fig/puff + + + Tachyphylaxis Poor choice
Muscle tremors Not to be used
IV or by
inhalation
+ + Muscle tremors Not used
parenterals
Abbreviation: IV = intravenously.
bronchomotor and peripheral vascular tone. Iso-
proterenol is a very potent bronchodilator, but it is
not
2
selective and results in tachycardia and ar-
rhythmias. The propensity toward tachycardia and
arrhythmias is exaggerated by halothane and min-
imized by isoflurane. In addition, isoproterenol is
not as effective by the tracheal route. Conse-
quently, it is fortunate that the following new
drugs, with substitutions in various parts of the
catecholamine structure, have been developed;
they differ from isoproterenol in that they have
higher
2
selectivity, may be administered by the
tracheal route, and have a longer action of dura-
tion.
12
Albuterol (salbutamol) can be administered as a
metered aerosol and, by this route, is as effective
a bronchodilator as isoproterenol is intravenously
but has fewer cardiovascular effects. Nevertheless,
salbutamol may still increase cardiac performance
by decreasing systemic vascular resistance and in-
crease cardiac contractility.
14
Its duration of action
is approximately 6 hours. Terbutaline is now avail-
able as a metered aerosol, and a water-soluble
preparation is available for nebulizer use. Meta-
proterenol and isoetharine are slightly less effec-
tive, have slightly less
2
selectivity, and are
shorter acting than albuterol. They offer no advan-
tage.
Second, although no clinical studies have been
done with anesthetized patients, combining the
anticholinergic inhalant ipratropium (Atrovent) to
P
2
-agonists appears to be a reasonable option to
consider.
15
16
Three studies
17-19
have indicated that
the combination of an anticholinergic agent, ipra-
tropium bromide, to
2
^88 is more effective
in severe acute asthma than either agent alone.
This additive effect is not surprising because the
anticholinergic agents differ in their mode of ac-
tion (blockage of efferent limb of irritant neural
reflex) from 2^88 (blockage/antagonism of
bronchoconstrictor mediator). Figure 13-3 shows
typical results from one study of awake asthmatics
and patients with chronic bronchitis.
17
No increase
in side effects has been observed when ipratro-
pium has been used in combination with other
bronchodilating drugs. Because parasympatholytic
drugs have little or no effect on mediator-induced
airway constriction, they are often more effective
in patients with chronic bronchitis than in those
with asthma
20
perhaps because, generally, they de-
crease the volume of secretions in and, specifi-
cally, in response to irritation/instrumentation of
the airway; therefore, they reduce bronchocon-
stricting reflexes. The
v
bronchodilating effect of
anticholinergic inhalants peak at 1 hour.
Corticosteroids may be extremely useful in
treating the inflammatory basis of bronchospasm,
but it should be realized that the onset of therapeu-
tic action may require 1 to 3 hours,
21
and the
mechanism of action in the modulation of airway
tone is incompletely understood. Proposed mecha-
nisms include stabilization of membranes, de-
creased bronchial mucosal edema and inflamma-
tory cells, and potentiation of catecholamine
action. Despite the 1 to 3 hours of onset of action,
the initiation of intraoperative treatment will ob-
viously have postoperative benefits.
1
22
The advent of an aerosol steroid offers a major
advantage in the treatment of chronic asthma. Side
effects are greatly minimized, but adrenal suppres-
sion may still occur. Beclomethasone dipropionate
(Beclovent) and triamcinolone (Azmacort) appear
to be the best inhaled agents currently available.
Nevertheless, in the acute intraoperative setting,
intravenous steroids are logically more reliable and
quick acting. Hydrocortisone is considered the ste-
roid of choice for parental administration in both
the preoperative preparation of patients with reac-
tive airway disease (1-2 mg/kg) and the treatment
of intraoperative bronchospasm (4 mg/kg). Meth-
ylprednisolone, 2 mg/kg, may also be used.
Finally, theophylline compounds may be bene-
ficial prophylactically in asthmatic patients and
therapeutic in bronchospastic patients. Phosphodi-
esterase inhibition is the major mechanism of ac-
tion of this drug in contrast to direct
2
stimula-
tion; this allows aminophylline to potentially be
useful in the patient who is taking beta blockers or
is already using a
2
^8. Side effects include
hypotension and/or cardiac arrhythmias, especially
during anesthesia with halogenated drugs. The ar-
rhythmias are magnified by halothane and mini-
mized by isoflurane. In addition, aminophylline
may not cause additional bronchodilation in the
presence of 1.5 MAC halothane.
23
Therefore, if
optimal doses of
2
^88 are being given, there
appears to be no advantage and several disadvan-
tages to using aminophylline.
1
Should a patient
require/receive aminophylline during anesthesia, a
loading dose of 5 to 6 mg/kg should be given
intravenously and slowly during 15 to 20 min
(which is a disadvantage compared with the
2
-
agonists) followed by an infusion of 0.5 to 1.0
mg/kg/hour depending on age (inversely propor-
tional). Theophylline blood levels should be ob-
tained on all patients, and the therapeutic range is
10 to 20 ^.
The third category of therapy involves the ad-
ministration of compounds that should be expected
to have a minor impact on bronchospasm but may,
in selected patients, be helpful, especially when
administration of the major bronchodilating drugs
has not completely resolved the problem (see Fig.
13-1, category III). If the increase in airway resis-
tance is thought to be a result of coughing or
straining (reflex induced), intravenous lidocaine, 1
Figure 13-3 FEV, increased in 11 patients with asthma and 10 patients with chronic bronchitis after treatment with ipratropium
(40 g four times a day) alone, salbutamoi (200 g four times a day) alone, and the combination of both drugs. (Reproduced with
permission from Lightbody JM, Ingram CG, Legge JS, Johnston RN: Ipratropium bromide, salbutamoi and prednisolone in bronchial
asthma and chronic bronchitis. Br J Dis Chest 72:181-186, 1978.)
mg/kg, should be administered.
24
Additionally, 4
per cent lidocaine may be administered intra-
cheally (down the tracheal lumen of a double-
lumen tube for anesthesia of the carina and one
main-stem bronchus and down the bronchial lu-
men for anesthesia of the other main-stem bron-
chus). Both forms of lidocaine administration in-
cur little risk and may be of significant benefit
(they decrease coughing reflex and bronchospasm
and protect against or treat ventricular arrhyth-
mias).
Finally, sodium bicarbonate should be adminis-
tered following blood-gas determination if a met-
abolic acidosis is present (acidosis inhibits the
action of catecholamines). Diphenhydramine (a
histamine blocker) may be administered to supple-
ment the level of anesthesia. However, administra-
tion of an antihistamine should not be expected to
have a large effect on the degree of bronchospasm
because histamine is usually not the sole mediator
involved in producing bronchoconstriction, and
the antihistamines do not block reflex airway con-
striction, which is so important during thoracic
surgery. Cromolyn is a mast cell stabilizer and is
thought to inhibit release of mediators from the
mast cells. The action of the drug is entirely pro-
phylactic, and it is not useful in the treatment of
bronchospasm intraoperatively. However, it is a
remarkably benign drug in terms of interaction
with anesthetics, and patients who are receiving
this drug should continue to receive it right up to
the time of surgery. Mucolytics have no proven
value in the treatment of bronchospastic disease.
Indeed, they have been shown to provoke reflex
bronchoconstriction and are best avoided.
III. MANAGEMENT OF BLOOD
LOSS
A. Assessment of Blood Loss
The anesthesiologist should be continuously
aware of the amount of blood loss. The assessment
of blood loss involves both direct observation and
measurement of physiologic responses as indirect
indicators of blood loss (Fig. 13-4).
7. Direct Observation
Surgical swabs and packs should be observed
for the amount of blood staining. Fully stained 4
by 4 gauze sponges and laparotomy pads contain
10 and 50 ml of blood, respectively. If there is a
Assessment of Blood Loss
Figure 134 See legend on opposite page
Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery
question or concern about the exact amount of
blood on the sponges and pads, the sponges and
pads should be weighed; each gram greater than
the known dry weight represents 1 ml of blood
loss. The suction bottles should be frequently ob-
served for blood accumulation. The operative
field, the drapes around the operative field, and the
floor should also be assessed for blood loss. Fur-
thermore, if uncertainty about the amount of blood
loss persists, there should be periodic attempts to
look under the drapes and other concealed places
for blood loss. With all these forms of direct ob-
servation, the anesthesiologist must be aware of
how much saline irrigation was used (to wash out
pleural cavity, test integrity of bronchial stumps)
and how much the irrigation might have contrib-
uted to the observed shed volumes. Similar consid-
erations apply to pleural fluid that was present
preoperatively and then suctioned intraoperatively.
With regard to the aspirated blood in the suction
bottles, the hematocrit of the material in the suc-
tion bottles can always be measured and evaluated
in light of the blood loss, intravenous fluid infu-
sion, and surgical irrigation and suction history.
Usually these data, when processed by the anesthe-
siologist, circulating nurse, and surgeon, result in
a reasonably accurate educated estimate of the
blood loss.
Tying off of the pulmonary vessels for lung
resection results in blood loss due to entrapment
of blood within the specimen. If the pulmonary
veins are tied off first, the specimen may be very
blood-engorged (due to continued inflow) and with
a pneumonectomy the resected lung may contain
500 ml of blood. If the pulmonary arteries are tied
off first, then the entrapped blood loss in the spec-
imen is much less. Postoperative drainage adds to
this blood loss, and after pneumonectomy the
pleural space fills with an unknown but consider-
able volume of blood and plasma.
2. Physiologic Response to Blood Loss
The patient's physiologic response to fluid and
blood loss (Fig. 13-4) is an indirect indicator of
how much has been lost and how adequate re-
placement has been. In the face of continued blood
loss and replacement with just crystalloid infusion,
serial measurement of the hematocrit and hemo-
globin will reveal a progressive decrease. Hemo-
dynamically, blood loss causes a decrease in sys-
temic arterial, central venous, pulmonary arterial
systolic, diastolic, and wedge pressure, cardiac
output, and urine output and an increase in systolic
pressure variation with respiration (i.e., the varia-
tion in the difference between maximal and mini-
mal systolic blood pressure during the respiratory
cycle). The heart rate and systemic vascular resis-
tance usually increase with blood loss if the depth
of anesthesia is not profound. The heart rate may
not increase in patients taking preoperative beta
blockers.
B. Minimal Blood Loss (<10 Per Cent
of Blood Volume): Amount of
Crystalloid Fluid Infusion
The amount of crystalloid infused during pul-
monary resections, particularly during pneumonec-
tomy, requires special consideration. On the one
hand, as for any major operation, insensible fluid,
nasogastric and urine fluid, extracellular (third
space) wound fluid, and blood losses (less than 10
per cent of blood volume) should be replaced by
crystalloid infusion. Failure to do so results in
unacceptable hemodynamic depression due to an-
esthetic drugs and positive-pressure ventilation, in-
ability to tolerate any further sudden loss of blood,
cardiovascular lability with changes in anesthetic
depth and surgical stimulation, and oliguria.
However, there are both experimental
2,5
2h
and
clinical
27-30
data suggesting that overinfusion of
fluids during pneumonectomy may be particularly
hazardous with regard to the development of post-
pneumonectomy pulmonary edema. There are
three interrelated reasons why pneumonectomy
may predispose to the formation of pulmonary
edema (Table 13-2). First, a decreased amount of
pulmonary vascular bed has to accommodate all of
the cardiac output. It has been clearly shown that
if the remaining pulmonary vascular bed is nor-
mal, then pneumonectomy does not cause an in-
crease in pulmonary extravascular fluid, even if
left atrial pressure is deliberately increased to 25
mm Hg.
2V 26
If the remaining pulmonary vascular
bed is nondistensible (diseased), the cardiac output
is greatly increased (as it may be in the early
Figure 13-4 The assessment of blood loss involves both direct observation (top half of figure) and physiologic assessment
(bottom half of figure). Direct observation of blood loss consists of blood on the surgical field and blood that is contained within
suction bottles, on sponges, and within the specimen. Assessment of blood loss through direct observation must be modified by the
amount of irrigation used. The physiologic assessment consists of quantification of decreases in systemic artery (PJ, central venous
(P
cv
), pulmonary artery (P
pa
), and pulmonary artery wedge (P
pa
J pressures and decreases in cardiac output (Q,), hematocrit (HCT),
hemoglobin (Hb), and urine output, and increases in heart rate and systemic vascular resistance (SVR).
TABLE 13-2 THEORETICAL REASONS WHY
PNEUMONECTOMY MAY
PREDISPOSE THE PATIENT
TOWARD THE FORMATION OF
PULMONARY INTERSTITIAL
EDEMA
1. Half the pulmonary vascular bed has to handle all the
pulmonary blood flow and volume.
2. Reduced number of lymphatic channels participate in
pulmonary interstitial fluid clearance.
3. Perioperative positive fluid balance may decrease plasma
oncotic pressure.
stress-ridden postoperative period), and/or exces-
sive fluids were infused intraoperatively, pulmo-
nary hypertension may occur and cause transmem-
brane transudation of fluid.
25
Second, pneumonectomy results in removal of
half of the lymphatics responsible for clearing pul-
monary interstitial fluid. The loss of parenchymal/
hilar/mediastinal lymphatic drainage routes, with
all other considerations equal, but especially with
an increased central venous pressure (CVP) (which
decreases lung lymph flow; see chapter 3), may
cause lymphatic clearance thresholds to be more
easily exceeded, increasing the risk of pulmonary
interstitial fluid accumulation and pulmonary
edema.
Third, although most existing data implicate ex-
cessive fluid administration in the pathogenesis of
this condition, the application of an equivalent he-
modynamic stress by left atrial Foley balloon
inflation
25,26
produces no increased susceptibility
to postpneumonectomy pulmonary edema and
suggests that the decreased oncotic pressures re-
sulting from excessive fluid administration may be
important.
Postpneumonectomy pulmonary edema has
been thought to occur in a number of patients.
27-30
The risk factors for this complication appear to be
high(er) perioperative volume of infused fluid,
right pneumonectomy (leaving a smaller left lung
to cope with the entire cardiac output and, perhaps,
higher pulmonary vascular pressures), and repeat
thoracotomy (increased blood loss).
29
However, as
mentioned previously, simple resection of a single
lung and minimal to moderate increases in hydro-
static pressure are not likely to be the sole cause
of postresection pulmonary edema in patients with
relatively normal hearts and remaining lungs,
26
and
it may be necessary to have also some factor pres-
ent that disposes the patient toward edema (pre-
existing pulmonary vascular disease, decreased on-
cotic pressure, very high cardiac outputs, acute
injury to the perfneability barrier
30
[sepsis, bron-
chopneumonia, or trauma]) in order for postpneu-
monectomy pulmonary edema and the adult respi-
ratory distress syndrome to develop.
Indeed, in a small series of postlung resection
pulmonary edema patients (involving two pneu-
monectomies and three lobectomies), the pulmo-
nary edema fluid protein to serum protein ratio
was 0.6 or greater, which was suggestive of
permeability changes; all five patients had postop-
erative pulmonary infection.
30
The change in
permeability may explain why in some studies the
amount and type of fluid infused seem unrelated
to the development of pulmonary edema.
30
31
The diagnosis of postpneumonectomy pulmo-
nary edema must be made on clinical grounds (se-
vere respiratory distress plus X-ray findings of
pulmonary interstitial edema)
30
because the pul-
monary artery wedge pressure is likely to be mis-
leading. The pulmonary artery occluded pressure
is misleading because balloon inflation in the one
remaining pulmonary artery may cause proximal
pulmonary artery hypertension and distal (pulmo-
nary artery occluded pressure) left-sided hypoten-
sion (caused by the proximal balloon obstruction).
That the distal pulmonary artery occluded pressure
is falsely low is underscored by the fact that pe-
ripheral wedging of a small catheter in a small
pulmonary artery (without balloon inflation) re-
sults in a higher distal occluded pressure than the
usual balloon pulmonary artery occluded pres-
sure.
32,33
A strong clue to the diagnosis of a falsely
low P
pao
is systemic hypotension produced by in-
flation of the pulmonary artery catheter balloon.
In summary, because of the smaller pulmonary
vascular bed remaining after resection and the in-
terruption of lymphatic drainage, it is reasonable
to consider the possibility that patients undergoing
pneumonectomy are at greater risk for the devel-
opment of pulmonary edema from aggressive in-
travenous fluid administration than other types of
surgery patients, especially if the heart and/or the
remaining lung is diseased. A reasonable approach
would be to start out with infusing 10 ml/kg dur-
ing the first hour (mainly to make up insensible
losses and to minimize cardiovascular depression
due to anesthesia) and 5 ml/kg/hour thereafter,
with adjustments made according to the estimated
blood loss and cardiovascular responses. If pul-
monary edema should develop postoperatively,
mechanical ventilation with sufficient PEEP to re-
duce the F,0
2
to nontoxic levels should be insti-
tuted (Table 13-3);
34
however, CPAP with spon-
taneous ventilation has been used successfully to
treat this condition.
35
Next, a thorough search for
a septic focus is warranted; if one is found, it
should be promptly removed or drained after ap-
propriate antibiotic therapy has been instituted (see
Table 13-3).
34
Finally, and most vital, one should
measure pulmonary vascular pressures and exam-
ine both right and left ventricles for evidence of
failure or overload, which, if present, must be
TABLE 13-3 TREATMENT OF
POSTPNEUMONECTOMY
PULMONARY EDEMA
1. Mechanical ventilation with positive end-expiratory
pressure.
2. Search for and eliminate septic focus.
3. Examine both right and left ventricles for failure
(chapter 7).
4. Treat ventricular failure:
a. Inotropic drugs
b. Diuretics
c. Volume restriction
treated with inotropic drugs, diuretics, and volume
restriction (see Table 13-3).
34
In cases of extreme
right ventricular afterloading, it may be appropri-
ate to assess right ventricular performance by
either radioisotopic scanning or thermodilution
ejection fractions.
16
C. Moderate Blood Loss (10 to 20 Per
Cent of Blood Volume): Should Blood
Be Infused?
In the past, the generally accepted safe lower
limit of hemoglobin was 10 g/dl or a hematocrit of
30 per cent. If a patient started out with a normal
hemoglobin level of 14 g/dl, then a level of 10
g/dl represents a loss of about 28 per cent of the
oxygen-carrying capacity of the blood. Certainly,
blood losses less than 20 per cent of the blood
volume (if not instantaneous) can safely be re-
placed with nonsanguineous fluids, even with a
reasonable margin of error in either the estimation
of blood loss or replacement." Furthermore, there
is no evidence in humans that mild to moderate
anemia per se increases wound infection, delays
wound healing, or impairs recovery in any way.
38
In addition, in view of the known potentially lethal
viral diseases that can be transmitted by blood
infusion (hepatitis, autoimmune deficiency syn-
TABLE 13-4 REASONS WHY BLOOD SHOULD
NOT BE TRANSFUSED WHEN
BLOOD LOSS IS LESS THAN 20
PER CENT OF BLOOD VOLUME
1. Risk of transmission of viral diseases (hepatitis, acquired
immunodeficiency syndrome).
2. Reactions (anaphylactic, hemolytic, febrile).
3. Risk of increased cancer recurrence (immunosuppression).
drome) and the possibility of anaphylactic, hemo-
lytic, and febrile reactions (Table 134),
39
it cer-
tainly behooves the anesthesiologist to avoid blood
transfusion to patients who are initially normal
from a cardiopulmonary point of view when losses
are less than 20 per cent of the total blood volume.
Table 13-5 shows the incidences of both immdi-
ate/immunologie and delayed/infectious complica-
tions per unit of blood component infused.
38
Concentrations of hemoglobin below 10 g/dl
may be accompanied by a progressive increase in
the bleeding time,
40
but this is certainly a contro-
versial issue.
38
Decreases below 9 g/dl begin to
require an increased cardiac output to maintain
oxygen transport.
41
However, if the cardiac output
does not increase, as may occur during anesthesia,
and the decrease in hemoglobin concentration is
large, then anaerobic metabolism and lactic aci-
dosis may ensue. In summary, with respect to just
red blood cell mass, when deciding to transfuse a
specific patient, one should consider the duration
of anemia, intravascular volume, extent of the op-
eration, probability for massive blood loss, risk of
immunologic and infectious complications, and
presence of coexisting conditions such as impaired
pulmonary function, inadequate cardiac output,
myocardial ischemia, or cerebrovascular or periph-
eral circulatory disease. These factors are represen-
tative of the considerations that comprise clinical
judgment.
New immunologic and cancer growth data have
emerged that reinforce and further support the no-
TABLE 13-5 MAJOR RISKS OF BLOOD TRANSFUSION*
General Adverse Reaction Specific Adverse Reaction
Estimated Risk Per Unit of Blood
Component Transfused
Immediate/immunologic
Delayed/infectious
Fever or urticaria
Hemolytic, nonfatal
Hemolytic, fatal
Hepatitis
Hepatitis C (formerly non-, non-B)
HIV-l
HTLV-l
CMV
1:100
1:6,000-25,000
< 1:1,000,000
< 1:250,000
1:100-3,000
1:40,000-150,000
< 1:100,000
Varies with recipient immunologic competence
*Data taken in part from Office of Medical Applications of Research, National Institutes of Health.
1S
Abbreviations: HIV = human immunodeficiency syndrome; HTLV = human cell lymphotrophic virus; CMV - cytomega-
lovirus.
tion that blood transfusion should be avoided
whenever blood loss is less than 20 per cent of the
total blood volume and the patient is hemodynam-
ically stable (see Table 13-4).
42-51
These studies
indicate that perioperative blood transfusion is an
important immunosuppressant and that immuno-
suppression may favor growth of cancer.
42-51
The
increased survival of renal allografts after transfu-
sion of whole blood or packed red blood cells
indicates that the blood transfusion before renal
allograft transplant caused immunosuppression.
Furthermore, immunosuppressed patients in
general are at increased risk for the recurrence of
various malignant tumors. For example, it has
been found that perioperative blood transfusion is
associated with decreased survival of patients
undergoing resection for colon and breast cancer
as a result of recurrent or metastatic growths.
43
"*
5,49
In the most recent study in patients with cancer of
the colon, rectum, cervix or prostate, transfusion
of one unit of whole blood or more than three
units of packed red blood cells (within a 1-month
span before or after surgery) was associated with
a markedly increased risk of tumor recurrence.
49
It now appears, on the basis of three new stud-
ies, that intraoperative blood transfusion during
non-oat cell lung cancer resection is also associ-
ated with decreased survival because of cancer
recurrence.
46, 47, 50, 51
In one study,
47
155 patients
undergoing resection for lung carcinoma were ana-
lyzed retrospectively, and it was shown that the
use of blood transfusion was associated with a
significant decrease in survival time in patients
undergoing curative resection of lung carcinoma
despite multivariate adjustments for age, sex, cell
type, right lung versus left lung location, type of
operation, and stage of cancer. In the another
study,
46
surgical techniquerelated variables were
minimized by limiting the study to one surgeon.
Analysis of age, sex, tumor size, histopathology,
admission and discharge hematocrit values, esti-
mated operative blood loss, duration of operation,
extent of resection, anesthetic agents, and blood
transfusion revealed two significant prognostic
factors: extent of resection and use or nonuse of
transfusions. Transfused patients had lower dis-
ease-free rates within 5 years than nontransfused
patients.
In still another study, analysis of long-term sur-
vival of 117 patients surviving curative operation
for stage I lung carcinoma showed that any peri-
operative blood transfusion significantly decreased
prognosis (5-year survival of 53 per cent vs. 81
per cent, = .005).
51
The results of all of these
studies indicate that perioperative blood transfu-
sion of patients with lung cancer undergoing resec-
tion accelerates the appearance of recurrent or
metastatic cancer. The findings are consistent with
the hypothesis that blood transfusion, possibly
through an immunosuppressive mechanism, is re-
sponsible for a poorer prognosis in patients who
undergo resection for carcinoma of the lung. An
editorial,
52
based on one of the latest of the studies
showing a decreased lung cancer survival effect of
perioperative blood transfusion,
50
and a medical
intelligence review article
53
concluded that the
avoidance of perioperative blood transfusion in
cancer patients is a reasonable objective.
A variety of alternatives to homologous trans-
fusions exist. Indeed, Table 13-6 shows every
conceivable procedure that might help to avoid
homologous transfusion in an anemic Jehovah's
Witness.
54
Of the long list in Table 13-6, five
procedures are generally applicable to the usual
thoracic surgery patient.
First, autologous donation/transfusion programs
eliminate the viral transmission and immunologic
risks discussed previously, enhance the erythropoi-
esis by priming the bone marrow, decrease blood
viscosity (increase tissue perfusion), and have psy-
chological benefit. Indeed, decreasing the hemocrit
to approximately 30 volumes per cent may in-
crease oxygen delivery (increased cardiac output
caused by decreased viscosity more than offsets
the decrease in arterial oxygen content [Fig. 13-
5]).
55
Given a 3-week lead time, most patients
given oral iron therapy can predeposit two or three
units of blood.
56
The clinical and laboratory pro-
cedures used in collecting, preserving, and trans-
fusing autologous blood are the same as those used
routinely in providing homologous blood to pa-
tients.
Second, intraoperative blood salvage seems to
be safe in some applications and reduces the re-
quirement for homologous transfusion. Autotrans-
fusion devices fall into three categories.
57
The sim-
plest form is reinfusion of pleural blood, which is
usually defibrinated. Reinfusion of pleural blood ii
probably indicated for major hemothorax. Danger;
of contamination are small. Because the greates
blood loss is usually during initial tube placement
the apparatus needs to be set up in the emergenc;
department. The only real problem is that rela
tively few patients have enough blood loss to jus
tify transfusion; most chest tube placements i)
trauma result in the collection of less than 500 m
of blood. The second category is suction collec
tion. This form of autotransfusion is designed t
be used at operation. Simply, a collection trap i
placed in the suction line, and blood collecte
there is reinfused using a transfusion filter to catc
aggregates, fat particles, and debris. The apparati
is simple, easy to set up, and reasonably priced,
is ideal for bleeding inside the chest. The mo
elaborate category is the blood concentrator. Th
instrument uses a special console that is capable <
TABLE 13-6 MANAGEMENT OF THE ANEMIC JEHOVAH'S WITNESS'
Goal Considered Therapy Effect of Therapy
Minimize blood loss
Maximize oxygen delivery/
consumption ratio
Reduce iatrogenic loss
a. Microchemistry analyzers
b. Pediatric-sized blood samples
c. Analyze necessity of tests
d. Sterile reservoirs
Reduce hemorrhagic loss
a. Hemodilution
b. Red cell scavenging devices
c. Hypotensive anesthesia
d. Desmopressin
e. Progesterone
Maximize blood production
a. Erythropoietin
b. Intravenous iron dextran
c. Nutritional support
Maximize cardiac output
a. Volume expansion
b. Hemodilution
Increase oxygen content
a. Oxygen
b. Fluronated blood substitutest
Decrease metabolic rate
a. Hypothermia
b. Paralysis, sedation
Higher hemoglobin levels and
improved oxygen delivery
Higher hemoglobin levels and
Enhanced cardiac output and
Increased dissolved oxygen content
and improved oxygen delivery
Reduced oxygen consumption
*From Mann MC, Votlo J, Kambe J, McNamee MJ: Management of the severely anemic patient who refuses transfusion:
Lessons learned during the care of a Jehovah's Witness. Ann Intern Med 117:1042-1048, 1992. Used with permission.
tFluronated blood substitutes are of no proven benefit but are often considered as a therapy.
Relative Oxygen
Transport Capacity
Viscosity [cP)
Haematocrit (vol %)
Figure 13-5 The influence of the hematocrit on both viscosity (dashed line) and relative oxygen transport capacity (dashed-
dotted line) as a percentage is displayed. Peak oxygen transport capacity occurs at a hematocrit of 30%. Below a hematocrit of 307r
oxygen transport capacity declines as a result of decreased blood carriage of oxygen. Above this hematocrit, capacity declines as a
result of the altered rhologie characteristics of the blood and the resultant decreased flows in the microcirculation. (From Cosby
ET: Perioperative haemotherapy: I. Indications for blood component transfusion. Can J Anaesth 39:695-707, 1992. Used with
permission.)
washing and concentrating the aspirated blood.
"Concentration" is somewhat relative here; one
does not obtain packed cells from a blood concen-
trator. A hematocrit of 45 per cent is about the
best that can be done. The problem with the con-
centrator is when to use it. There is a delay of 15
to 30 min between collection and reinfusion to run
the washing and concentrating cycles and to col-
lect the blood in a bag for reinfusion.
Third, patient selection for intraoperative isovo-
lemic hemodilution is based on still-emerging cri-
teria. Many of the criteria applicable to the use of
autologous transfusion are comparable to those for
using intraoperative hemodilution. Some patients
can be treated in this way and may avoid homolo-
gous transfusion.
Fourth, maintenance of perfusion is still the pri-
mary problem in dealing with blood loss. There
are a number of cost-effective substitutes for the
treatment of reduced plasma volume and total
blood volume (e.g., crystalloid, dextrans, albumin,
and hetastarch solutions) (see next paragraph).
Fifth, intraoperative, deliberately induced hypoten-
sion may reduce surgical blood loss in some pro-
cedures.
Table 13-7 shows the physiochemical charac-
teristics of the various solutions used for intravas-
cular volume resuscitation.
58
Table 13-8 summa-
rizes the important factors that determine whether
a crystalloid or a colloid solution should be used.
59
In most cases of surgery, the blood loss is easily
witnessed, can be measured, and can/should be
replaced immediately with appropriate volumes of
crystalloid or colloid. However, with this temporal
sequence of events (i.e., immediate observation
and replacement), there is no time for interstitial
fluid shifts and interstitial fluid loss should be min-
imal; because the restoration of blood volume with
crystalloid is transient,
60
both crystalloid and
colloid could/should be used.
59
In patients with
trauma and delayed resuscitation, blood loss is as-
sociated with interstitial fluid loss and crystalloids
are predominantly given to replace this interstitial
loss.
59
Crystalloid administration should be limited if
there is evidence of renal failure or brain or lung
trauma. It is important to emphasize the frequent
assessment of blood loss and hemoglobin to avoid
excessive hemodilution. In addition to some blood
volume and complete interstitial space replace-
ment, crystalloids are given to replace the fasting
deficits, ongoing maintenance requirements, and
interstitial fluid losses resulting from surgical
trauma.
D. Severe Blood Loss ( 2 0 Per Cent of
Blood Volume): Diagnostic and
Therapeutic Management Problems
1. Causes of Severe Bleeding
Hemostasis depends on the interplay of four
components: maintenance of vascular integrity,
normal number and function of platelets, the co-
agulation cascade production of fibrin, and the
eventual removal of fibrin by fibrinolysis (Fig.
13-6). With respect to the first three components
of hemostasis (the vessels, platelets, and the coag-
ulation cascade), damage to the endothelial lining
of a blood vessel exposes subendothelial structures
to blood and results in platelet activation and plate-
let adherence and the formation of a hemostatic
plug. Interactions of the coagulation proteins form
Table 13-7 PHYSIOCHEMICAL CHARACTERISTICS OF THE IMPORTANT CRYSTALLOID AND
COLLOID PREPARATIONS
Variable
Na(mOsmL-' )
CI (mOsm' L-' )
Osmolality (mOsm L~')
Plasma expansion
(per 500 ml) (ml)
Large molecule size
(mean)
Elimination
Duration of expansion
(approximate)
Effect on coagulation
Allergic reactions
Anaphylaxis
Recommended maximum dese
Cost to Emory University
hospital pharmacy ($US)
Normal
Saline
154
154
308
100

transient
(~ 20 min)

0.60 per liter
Lactated
Ringer's
147
109
278
100

transient
( - 20 min)

0.74 per liter
5% Albumin
130-160
300
500
62-69,000
(hepatic)
24 hr

0.011

36.00 per 500 ml
Dextran
(Macrodex)
6% NS
154
154
308
500-700
20-200,000
(70,000)
Plasma hydrolysis, renal
24 hr
Impaired
0.03-4.7
0.08-0.6*
20 ml kg-
1
24 hr
7.22 per 500 ml
Hetastarch
(Hespan)
6% NS
154
154
310
500-700
10-1,000,000
(70,000)
Reticuloendothelial system, renal
36 hr
7
0.85
0.006
20 ml kg"
1
24 hr
38.45 per 500 ml
*From Ramsay JG: Methods of reducing blood loss and non-blood substitutes. Can J Anesth 38:595-612, 1991. Used with permission.
tPreventable by prior administration of dextran 1 (Promit).
Table 13-8 SUMMARY OF FACTORS THAT DETERMINE WHETHER CRYSTALLOID VERSUS
COLLOID IS USED
Factor Crystalloid Colloid
Intravascular volume effect Crystalloid worse Colloid better
Interstitial volume effect Crystalloid better Colloid worse
Pulmonary edema Both have similar potential for pulmonary edema
Peripheral edema Common Less common
Reactions Uncommon More common
Cost Inexpensive Albuminexpensive
Nonalbuminmoderately expensive
*From Davies MJ: Crystalloid vs. colloid: Does it matter? J Clin Anesth 1:464-471. 1989. Used with permission.
fibrin strands that reinforce the platelet plug. For-
mation of a plug stops the bleeding if damage to
the vessel wall is small, vascular constriction has
occurred, and intravascular pressure does not dis-
lodge the friable plug. These three components of
the coagulation processvasoconstriction, platelet
plug, and the formation of coagulation fibrin
strandsoccur simultaneously in response to vas-
cular damage. Derangements in any of these three
components as well as excessive fibrinolysis may
cause excessive bleeding.
In surgery, the cause of severe bleeding is al-
most always loss of vascular integrity, but other
(and multiple) abnormalities may contribute to the
problem. For example, if bleeding resulting from
loss of vascular integrity has been large and sus-
tained, and massive transfusion with old, stored
bank blood has been instituted, the infusion of old,
stored bank blood may, in and of itself, cause a
coagulation defect (see Hemostatic Defects Result-
ing From Massive Blood Transfusion). To institute
curative therapy, it is necessary to identify the
exact cause of the bleeding. Figure 13-6 shows
the systematic workup and logic that should be
used in the diagnosis and treatment of severe
bleeding.
Maintenance of vascular integrity is the primary
responsibility of the surgeon (see Fig. 13-6). Pro-
cedures most likely to incur excessive bleeding
resulting from loss of vascular integrity are resec-
tion of tumors with significant local extension (es-
pecially to chest wall and mediastinum), resection
of tumors that involve bronchial and pulmonary
vessels and the pleura, and procedures that require
surgical dissection of a thickened inflamed vascu-
lar pleura. In order of rapidly decreasing incidence
of cause of intraoperative bleeding are the pulmo-
nary arteries and pulmonary, azygos, and subcla-
vian veins.
61
Of course, failed surgical sutures on
any large blood vessels will result in rapid, mas-
sive blood loss. The loss of vascular integrity is
diagnosed by inspection, and the treatment is sur-
gical correction. If massive blood loss is antici-
pated prior to surgery (on a technical basis), pre-
operative blood donation by the patient (several
weeks before surgery) for personal intraoperative
use,
62
63
transfusion of the patient's own shed
blood aspirated from the surgical field (autotrans-
fusion),
64
and/or preoperative normovolemic he-
modilution (draw off whole blood and simultane-
ously replace with crystalloid) on the day of
surgery
65
may considerably help replacement of
lost blood (Table 13-9).
Hemostasis requires an adequate number of nor-
mally functioning circulating platelets (Fig. 13-6).
Platelets serve as an initial plug for small vascular
defects and also as the substrate on which the
enzymatic reaction of the coagulation cascade oc-
curs. They may be deficient in number (dilution,
destruction, or sequestration) or deficient in func-
tion (drugs, presence of fibrin split products,
uremic plasma, von Willebrand's disease) or both.
The platelet count measures only platelet number.
The bleeding time (Ivy forearm method) evaluates
both platelet quantity and function. The test is
performed by inflating and maintaining an arm
blood pressure cuff at 40 mm Hg and making a
10-mm long, 1-mm deep forearm incision; the
blood is lightly blotted with filter paper at 30-sec
intervals until the clot end point is reached (blood
no longer appears on the filter paper). The normal
forearm bleeding time is 3 to 8 min. Since most
patients undergoing thoracic surgery are in the lat-
eral decubitus position, both forearms are usually
readily available to the anesthesiologist. If the
bleeding time is normal, platelet administration is
not required. If the platelet count is decreased and/
or the bleeding time is prolonged, platelet therapy
will be required, perhaps along with other thera-
pies (see Fig. 13-6). The platelet dose is one plate-
Table 13-9 NONROUTINE MEASURES TO AID
INTRAOPERATIVE BLOOD
TRANSFUSION
1. Preoperative blood donation by the patients for their own
intraoperative use
2. Intraoperative transfusion of the patient's own shed blood
aspirated from surgical field
3. Preoperative normovolemic hemodilution
The Diagnosis of Cause of Bleeding
and the Appropriate Treatment
Figure 13-6 There are four major causes of bleeding, which are indicated in the left-hand column. The diagnosis of the cause
of bleeding and the results of the various diagnostic tests are indicated in the two middle columns. The appropriate treatment for
the cause of bleeding is indicated in the right-hand column. (ACT = activated clotting time; aPTT = activated partial thrombo-
plastin time; PT = prothrombin time; WBCT = whole blood clotting time; FFP = fresh frozen plasma; FSP = fibrin split
products; EACA = epsilon aminocaproic acid.)
let pack unit per 10 kg body weight if the platelet
count is less than 60,000/mm, and one platelet
pack unit can be expected to increase the platelet
count by 10,000/mm.
66
After administration, de-
fects due to storage may prevent the transfused
platelets from functioning normally for several
hours.
66
Defects in the coagulation cascade leading to
fibrin result in excessive bleeding (Fig. 13-7). The
performance of an activated clotting time (ACT)
or activated partial thromboplastin time (aPTT)
screens effectively virtually all of the intrinsic sys-
tem coagulation cascade (factors XII, XI, IX, VIII,
platelet factor 3). The prothrombin time (FT) tests
the extrinsic system (factor XII, platelet tissue fac-
tor, and factor VII, tissue factor) in the same way
that the aPTT tests the intrinsic system.
67
The
ACT, aPTT, and PT all test the final common
pathway sequence of conversion of factor X to
activated factor X, which in turn converts, along
with factor V, prothrombin to activated factor II,
which in turn converts fibrinogen to insoluble fi-
brin. Normal values for the ACT, PTT, and PT are
90 to 120 sec, less than 35 sec, and 12 sec, respec-
tively.
68
Note in Table 13-10 that factor VII is
unique to the extrinsic pathway (tested by the PT),
and factors VIII, IX, and XI are unique to the
intrinsic pathway (tested by the PTT). The PT and
partial thromboplastin time (PTT) can be used,
therefore, to infer the absence of procoagulants.
Unfortunately, loss of the various coagulation fac-
tors must approach 80 per cent before the PT or
PTT become prolonged. The whole blood coagu-
lation time (WBCT), which, with less sensitivity
and more variability, tests the intrinsic system and
final common pathway, can be performed in the
operating room by drawing 1 ml of venous blood
into a test tube maintained at 37C and tilting the
PROLONGED COAGULATION
TESTS
Coagulation Test Factor(s) That May Be
Prolonged Decreased
PT VII
PTT VIII, IX, XI, XII
Both PT and PTT V, X
Abbreviations: PT = prothrombin time; PTT partial
thromboplastin time.
tube 90 back and forth at 30-sec intervals, permit-
ting the liquid blood to run down the side to pro-
duce maximal surface activation. Normal values
for WBCT in the one tube are 2.5 to 4.5 min. A
normal WBCT does not ensure a normal bleeding
time.
Increased fibrinolysis is rarely a cause of exces-
sive bleeding. If the condition is present, the fi-
brinogen level will be markedly decreased, the
production of fibrin split products (FSP) will be
increased, and the FSP test will confirm the diag-
nosis. A negative test for FSP indicates a low level
of FSP because of either little fibrinolysis or de-
pleted fibrin or fibrinogen. Increased FSP is treated
by stopping the cause of FSP production, infusing
clotting factors that have been consumed (platelets
and fresh frozen plasma), and allowing time for
hepatic and reticuloendothelial clearance of FSP.
2. Hemostatic Defects Resulting From
Massive Blood Transfusion
Secondary disorders of hemostasis remain one
of the persistent problems in the area of massive
transfusion (Table 13-11). Factors V and VIII and
platelets may be deficient in stored whole blood.
These factor levels decrease 50 to 80 per cent after
21 days of storage; the remaining levels, 20 to 50
per cent of normal, are above or near the minimal
hemostatic level for factor V (20 per cent) and for
factor VIII (30 per cent). Thus, with transfusion of
whole blood, deficiencies in these two factors are
unlikely to be a primary cause of bleeding. How-
ever, if blood loss greater than 50 per cent of the
blood volume is replaced with nonplasma products
(packed red blood cells, noncoagulation factor
containing protein fluids, crystalloid fluids), then
the level of factors V and VIII may decrease well
below minimum hemostatic levels. An abnormal
aPTT will indicate a deficiency of factors V and
VIII.
69
70
In contrast to the relatively slow decay of coag-
ulation factors, the number of stored platelets rap-
idly decays over 48 to 72 hours. Transfusion of 10
and 20 units of whole blood can be expected to
decrease the platelet count to 100,000 and 50,000/
min,
66
respectively. When dilutional thrombocyto-
penia is the cause of a bleeding disorder (likely to
occur when transfusion exceeds 1.5 times the
blood volume), and there is also a moderate defi-
ciency in either factor V or factor VIII, which is
insufficient to cause bleeding on its own, the mod-
erate deficiency in coagulation factor may possibly
aggravate the thrombocytopenia-induced bleeding.
Thus, when combat casualties who are bleeding
have received more than 20 units of stored blood
and have abnormal aPTT and PT times, infusion
of 500 to 1000 ml of fresh frozen plasma (which
contains factors V and VIII but no platelets) re-
stores the aPTT and PT times to normal but with-
out significant correction of the bleeding disorder.
Subsequent administration of platelets results in
correction of these bleeding disorders.
71
During
acute massive hemorrhage, it is always difficult to
distinguish the amount of continued bleeding that
results from the precipitating loss of vascular in-
tegrity from the subsequent bleeding caused by the
transfusion of massive amounts of stored blood.
Rapid infusion of citrated blood will decrease
ionized calcium. This decrease is transient and rap-
idly reversed as the infused citrate is metabolized.
Ionized calcium is a factor involved in the intrinsic
coagulation cascade and final common pathway.
Since the myocardial depressant effects of a de-
crease in ionized calcium occur long before the
effects on the coagulation mechanism, and the
myocardial depression is treated with calcium
chloride, the anticoagulation effects of hypocal-
cemia are rarely seen.
68
The coagulation system consists of clotting and
fibrinolytic mechanisms. The function of the for-
mer is to prevent excessive blood loss and that of
the latter is to ensure circulation within the vascu-
lature. With disseminated intravascular coagula-
tion (DIC), the clotting system is deranged, and
this leads to disseminated fibrin deposition, which
renders the remaining fluid blood unclottable. The
deposited fibrin may severely alter the microcir-
culation and lead to ischemic necrosis in various
organs, particularly the kidney. The unclottable
blood or circulating serum may induce a severe
hemorrhagic diathesis.
The specific reasons for the development of DIC
Table 13-11 HEMOSTATIC DEFECTS CAUSED
BY MASSIVE TRANSFUSION OF
STORED BANK BLOOD
1. Decreased platelets: definite occurrence
2. Decreased factors V and VIII: possible occurrence
3. Decreased Ca
+ +
: unlikely
4. Disseminated intravascular coagulation: rare
5. Hemolytic transfusion reaction: rare
Table 13-12 COMPLICATIONS OF MASSIVE TRANSFUSION OF STORED BLOOD
Complications
Comment, Clinical Correlates
1. Hemostatic defects (see Table 13-11)
2. Left-shifted Oxy-Hb curve
3. Citrate intoxication
4. Hypothermia if blood not warmed
5. Acidosis
6. Infusion of microaggregates
1. Oozing from all tissue wounds
2. Decreased tissue oxygenation
3. Caused by citrate binding of Ca*~ -+. Decreased Ca~" (decreased
cardiac function). Only occurs with one unit of blood/5 min rate of
infusion. Citrate * HCO," postoperative metabolic alkalosis
4. Increased Vo
2
, postoperative shivering, stress on cardiac function
5. Twenty-one day-old bank blood has pH = 6.9 and Pco
:
= 150-220
mm Hg. No treatment usually necessary because of complication 3
(citrate intoxication).
6. Can contribute to development of adult respiratory distress syndrome.
Use Micropore filters.
are usually not apparent. However, hypoxic aci-
dotic tissues with stagnant blood flow probably
release tissue thromboplastin either directly or
through liberation of some toxin. This triggers the
coagulation process, resulting in consumption of
factors I, II, V, and VIII and platelets. Supposedly,
thrombi and fibrin are deposited in the microcir-
culation of vital organs, interrupting their blood
flow.
Oozing, hypotension, and hemoglobinuria may
indicate a hemolytic transfusion reaction. The sim-
plest screening test is examination of a centrifuged
blood sample for free plasma hemoglobin. If there
is no plasma discoloration, a serious hemolytic
reaction is unlikely. On rare occasions, hemolysis
is the result of mixing of red blood cells with
hypotonic solutions. Therefore, if the plasma is
discolored, all intravenous solution containers
should be examined to ensure that a hypotonic
solution was not used. The many other (nonhe-
mostatic) complications of massive transfusions of
stored blood are shown in Table 13-12.
3. Management of Massive Transfusion
The proper management of massive transfusion
logically follows from the problems that massive
transfusion can cause (Fig. 13-8). First, it is im-
portant to send "stay ahead" orders to the blood
bank for typing and crossing additional units of
blood; the greatest problem in massive transfusion
is the logistical one of always having compatible
blood and blood products to infuse continuously.
For example, if eight units are available and four
are used, and no end to the bleeding is in sight,
four more units should be typed and crossed at the
time the fifth unit is being used. If fully typed and
cross-matched blood is not available, then the pre-
ferred order of using less well-prepared blood is
type specific (ABO-Rh) partially cross-matched
blood (sometimes referred to as an incomplete or
immediate phase cross match, which eliminates
ABO type compatibilities), type specific (ABO-
Rh) un-cross-matched blood, and type Rh-neg-
ative (universal donor) un-cross-matched blood
(which should be used as packed red blood cells
because the plasma of type blood may have anti-
A and anti-B antibodies). All infused blood should
be warmed and passed through a micropore filter.
Each unit of blood (packed red cell units have a
volume of 250-275 ml with a hematocrit of 70-
80 per cent, and whole blood units have a volume
of 500 ml with a hematocrit of 36-44 per cent)
should increase hemoglobin by 1 g/dl and the he-
matocrit by 2 to 3 per cent in the average 70-kg
adult.
38
Many diagnostic tests should be done after
every five units of transfused blood (see Fig.
13-8). Routine hemostatic tests should be platelet
count, ACT or aPTT, and PT. Physiologic studies,
such as arterial blood gases, hematocrit, hemoglo-
bin, electrolytes, calcium, and bicarbonate levels,
should be obtained every five units. Calcium chlo-
ride, in a 500-mg intravenous bolus, should be
administered after every five units, and sodium
bicarbonate should be administered according to
the arterial blood gases.
Several activities should occur after every 10
units of transfused blood (see Fig. 13-8). These
consist of measuring fibrinogen levels and FSPs
and infusing frozen plasma and one platelet pack/
10 kg, according to the National Institutes of
Health Consensus Conference.
38
72
73
In general,
fresh frozen plasma should be administered when
either the PT or PTT is greater than 1.5 normal.
and platelets should be administered when the
platelet count is less than 50,000/1, and in either
circumstance there is clinically significant bleed-
ing. Each unit of fresh frozen plasma will increase
the level of any clotting factor by 2 to 3 per cent
in the average adult.
72
Each unit of platelets should
increase the platelet count in the average adult by
5000/1.
73
Prolonged hypovolemia may lead to re-
fractory oliguria or anuria and should be investi-
gated as to whether it is reversible (prerenal fail-
ure) or irreversible renal failure. Table 13-13
shows the differential diagnosis between prerenal
failure (decrease perfusion) and renal failure (tub-
ular necrosis, which has many causes including
many toxins as well as, and more importantly,
prolonged prerenal failure). Thus, massive trans-
fusion is a busy time and requires staying ahead
with prepared blood, monitoring the hemostatic
and physiologic effects of the massive transfusion,
and treating the hemostatic and physiologic de-
rangements.
IV. TREATMENT OF NONBLOOD
LOSS, DELETERIOUS
HEMODYNAMIC CHANGES
It is possible for any or all of the factors that
determine myocardial oxygen supply and demand
to change a great deal during anesthesia and sur-
gery. Fortunately, the availability of a wide range
of therapeutic maneuvers and options usually al-
lows the anesthesiologist to return any of the de-
terminants of myocardial oxygen supply and de-
mand to normal or to any particular desired value.
Figure 13-9 shows most of the commonly used
options for each hemodynamic aberration. The
precise therapeutic modality chosen depends, of
course, on the abnormality producing the deleteri-
ous hemodynamic change. For example, the heart
rate may be too rapid because anesthesia is too
light (which could be treated with fentanyl, deep-
ening inhalation anesthesia, or infusing a beta
blocker) or the patient is hypovolemic (which
should be treated by increasing vascular volume).
The dosages and infusion information for the he-
modynamically active drugs used in the perioper-
ative period are listed in Table 13-14.
74
It is very unusual for just one of the determi-
nants of myocardial oxygen supply and demand to
change alone without a change in any of the other
determinants of myocardial oxygen supply and de-
mand. The combination of a specific pattern of
change in multiple hemodynamic variables (con-
Figure 13-8 The management of massive transfusion involves several diagnostic and therapeutic activities that must be per-
formed on a regular basis. The diagnostic activities are performed after approximately every five units of transfused blood, and the
therapeutic activities are performed after approximately every 10 units of transfused blood. (ACT = activated clotting time; aPTT
= activated partial thromboplastin time; PT = prothrombin time; ABG = arterial blood gases; Hct = hematocrit; Hb =
hemoglobin.)
Table 13-13 URINARY FINDINGS IN RENAL
HYPOPERFUSION VERSUS
ACUTE TUBULAR NECROSIS
stituting a hemodynamic syndrome or profile)
within a specific clinical context usually makes the
diagnosis. The most important and common he-
modynamic syndromes with which the anesthe-
siologist must be familiar are caused by inadequate
and excessive anesthesia, hyper- and hypovolemia,
sepsis, cardiac failure due to increased afterload,
and cardiac failure due to myocardial ischemia
(Table 13-15).
In order to expand the hemodynamic profile
enough so that a specific pattern becomes recog-
nizable (and distinct from the other hemodynamic
syndromes), it is sometimes necessary to use pul-
monary artery catheter monitoring (which allows
calculation of cardiac output, preload, contractility,
and systemic vascular resistance) (see Table
13-15 and chapter 7). For example, the pulmonary
artery wedge pressure distinguishes hypovolemia
from cardiac failure due to myocardial infarction
(when all other hemodynamic variables are not
dramatically different) (compare row 3 with row
5, Table 13-15), whereas the cardiac output distin-
guishes sepsis from excessive anesthesia (when all
other hemodynamic variables are not dramatically
different) (compare row 6 with row 7, Table
13-15). The abnormal hemodynamic profile in the
context of a specific history allows for diagnosis
of the cause of the abnormal hemodynamic profile
and the right therapeutic modalities to be chosen.
Several abnormal hemodynamic conditions require
special discussion either because they are very
common or they are very difficult to treat.
wmmmmmmmm*
A. Rationale Plan for Managing Low
Cardiac Output State
Several of the conditions listed in Table 13-15
involve a low cardiac output. A rationale sequence
of steps in managing a low cardiac output state
(assuming a normal sinus rhythm) consists of,
in series, preload normalization (fluid challenge),
inotropic support, and afterload normalization
(vasodilator drug) (Fig. 13-10). Of the three, ap-
propriate inotropic support is the least straightfor-
ward and the most complex.
Dopamine is a good initial agent to start on
hypotensive patients who need inotropic support.
At low infusion rates of less than 3 g/kg/min,
mostly dopaminergic receptors are stimulated,
which may not provide much pressure support but
will improve renal perfusion. Infusion rates of 3 to
10 g/kg/min will have both a vasoconstrictive
and a cardiotonic effect. With infusion rates of
greater than 10 to 12 g/kg/min, a,-agonist effects
will predominate. Under these circumstances, sys-
temic blood pressure may be maintained because
of severe peripheral vasoconstriction, but cardiac
output will decrease. In these situations, a vasodi-
lator such as nitroprusside may be used in con-
junction with dopamine to decrease afterload and
increase cardiac output (Fig. 13-10, step 3). Do-
butamine is also a good choice in this situation
because it has a direct inotropic effect and when
used at infusion rates of 10 g/kg/min or greater.
2
stimulation will occur and reduce the vasocon-
stricting effect of dopamine. Dobutamine may be
initially started as the sole inotropic agent on a
patient with low cardiac output and high peripheral
vascular resistance. However, if the patient is also
hypotensive, dobutamine can increase the cardiac
output but may worsen the hypotension because of
its vasodilating effect, especially at higher infusion
rates.
An alternative to using a higher infusion rate of
dopamine coupled with a vasodilator or dobuta-
mine to maintain adequate blood pressure and car-
diac output is to select a sympathomimetic agent
with greater inotropic and vasopressor activity
(epinephrine or norepinephrine) (Table 13-16).
Epinephrine has a much wider range in which
there is a combined alpha- and beta-adrenergic
receptor activity. This can help support the perfu-
sion pressure while maintaining the peripheral cir-
culation, especially to the renal and mesenteric
areas. Epinephrine will also have a bronchodilat-
ing effect, which is not the case with norepineph-
rine. In situations in which cardiac output appears
to be more than adequate (> 10 L/min) and hy-
potension still exists despite high infusion rates of
dopamine or epinephrine, norepinephrine can be
tried. Norepinephrine is the most potent vasocon-
Figure 13- 9 This figure lists the important hemodynamic variables that may change during anesthesia and the surgical procedure
(left-hand column). These hemodynamic variables may change in either direction (middle column), and there are multiple therapeu-
tic options for treating any change in any hemodynamic variable in any direction (right-hand column). See the text for a full
explanation.
Table 13-14 HEMODYNAMICALLY ACTIVE DRUGS COMMONLY ADMINISTERED BY
IV INFUSION IN THE PERIOPERATIVE PERIOD
Abbreviations: VF = ventricular fibrillation; VT
potassium; q = every.
ventricular tachycardia; D
?
W = 5 per cent dextrose in water;
Tabe 13-15 TYPICAL COMMON HEMODYNAMIC SYNDROMES AND THEIR
TREATMENT
strictor of both the venous and arterial vascular
systems. At low infusion rates of less than 2 g/
min (less than 0.03 g/kg/min) there is a mixed
alpha- and beta-adrenergic effect; however, above
this rate, effects from stimulation of the a,-adre-
nergic receptor will predominate. In this situation,
"pure" a,-agonists such as phenylephrine or me-
thoxamine may also be used.
Whenever a sympathomimetic agent other than
dopamine is used to maintain perfusion pressure
and cardiac output, renal blood flow is always
reduced. Maintaining a low infusion rate of dopa-
mine (2-3 g/kg/min) in conjunction with another
vasoactive agent may help maintain urine output.
Sympathomimetic support should always be
weaned as rapidly as possible. Continuous sympa-
thomimetic infusions will decrease receptor sensi-
tivity to catecholamines and potentially produce a
serious cardiomyopathy. There is also the potential
of life-threatening arrhythmias and regional ische-
mia. Finally, sympathomimetic support usually
causes increased heart rate, which is a distinct dis-
advantage in the patient with coronary heart disease.
Amrinone is a noncatecholamine, nonglycoside
alternative in the treatment of perioperative myo-
cardial dysfunction. A type 3 phosphodiesterase
enzyme inhibitor, amrinone combines positive ino-
tropic support with systemic and pulmonary vaso-
dilation. In patients with heart failure, the absence
of drug-induced tachycardia and arrhythmias is of
particular clinical benefit. In these patients, amri-
none's ability to augment cardiac performance
without increasing heart rate or myocardial oxygen
consumption offers significant clinical advantages
during the perioperative period.
75
B. Pulmonary Artery Hypertension and
Right Ventricular Failure
Under normal conditions, right ventricular per-
formance is governed, in order of decreasing im-
portance, by end-diastolic volume (preload),
contractility, pulmonary vascular impedance (af-
terload), and coronary perfusion. However, in
pathologic conditions, such as pulmonary hyper-
tension, the order of right ventricular performance
determinants is reversed, and the most important
factor governing right ventricular function is status
of the pulmonary vascular impedance (i.e., after-
load).
76
Thus, the contractility of the right ventricle
is more dependent on the maintenance of a normal
afterload than is the function of the left ventricle.
The thin-walled right ventricle will not compen-
sate for changes in afterload, particularly if these
changes are acute. In situations in which pulmo-
nary vascular impedance is acutely increased, right
ventricular performance is dramatically affected.
In situations in which this increase in afterload
occurs gradually (i.e., COPD), hypertrophy devel-
ops as a compensatory mechanism, and the right
ventricular output remains normal at rest. How-
ever, in situations in which pulmonary impedance
is increased acutely, these compensatory mecha-
nisms are not operative/adequate and right ventric-
ular failure ensues.
In clinical situations that result in an acutely
increased pulmonary vascular resistance (e.g., sep-
sis, adult respiratory distress syndrome, hypoxia,
and pulmonary emboli), prompt and effective re-
ductions in right ventricular afterload will result in
maintenance of normal right ventricular function.
A variety of intravenous agents are available as
pulmonary vasodilators (Table 13-17), but the re-
sults have been mixed and a decrease in pulmo-
nary and systemic pressure is to be expected (i.e..
Table 13-17 VASODILATOR DRUGS CLINICALLY USED FOR PULMONARY HYPERTENSION*
Drug
Nitroglycerin
Sodium nitroprusside
Hydralazine
Isoproterenol
Nifedipine
Prostaglandin E,
(Alprostadil)
Amrinone
Trade
Name
Nitrostat
Nipride
Apresoline
Isuprel
Procardia
Prosti VR
Dose Range
0.25-1.5 g/kg/min
0.30-5.0 g/kg/min
0.05-0.1 mg/kg bolus
0.02-0.08 g/kg/min
5-10 mg po or si
0.03-0.10 g/kg/mi
5-10 mg/kg/5 min bolus
5-10 g/kg/min
HR
0/+1
0/+1
+ 1
+ 3
0/+1
0/+ 1
0
CO
0
0
+ 1
+ 2/ +3
0/+1
+ 1/ + 2
+ 2
PAP
-2
-1/-2
0/-1
0/-1
-1
-3
-2
Predominant
Vasodilator Effects
Systemic > pulmonary
Systemic > pulmonary
Systemic > > > pulmonary
Systemic = pulmonary
Systemic >> pulmonary
*From Prielipp R: Right-sided heart failure and pulmonary hypertension. In Cheng EY, Kay J (eds): Manual of Anesthesia and
the Medically Compromised Patient. Philadelphia, JB Lippincott Co, 1990, pp 112-124. Used with permission.
Abbreviations: HR = heart rate; CO = cardiac output; PAP = pulmonary artery pressure; po = orally; si = sublingual.
there is no available selective pulmonary vasodi-
lator except for inhaled nitric oxide; at the time of
this writing nitric oxide is an investigational drug).
Occasionally, the decrease in afterload will cause
an increase in cardiac output, which will maintain
the increase in pulmonary artery pressure or even
cause a further increase. However, the use of am-
rinone in this context is most promising.
75
C. Coronary Artery Disease
1. Monitoring for Ischemia (See Chapter
7): Further Considerations
Recently, it has been demonstrated that, in pa-
tients in whom intraoperative ischemia develops,
the risk of postoperative infarction is increased
two- to threefold. These data imply that intraoper-
ative ischemia should be aggressively monitored
and treated. Significant advances in echocardiog-
raphy have enabled this technique to be used in
patients with ischemic heart disease. With ische-
mia, wall-motion abnormalities and systolic thin-
ning develop, followed by decreased left ventricu-
lar compliance (increased P
pao
; for extensive
discussion of this change, see chapter 7) followed
by electrocardiogram changes and, in the awake
patient, angina.
Studies in animals have demonstrated that, very
soon after coronary occlusion (in seconds), the
ischemic myocardium first develops wall-motion
and wall-thickening abnormalities. Both systolic
shortening and thickening of the myocardial fibers
become impaired. These regional effects produce
abnormal contractions of the wall known as dys-
synergia, which can pass through the phases of
hypokinesis, akinesis and, finally, dyskinesis or
aneurysm-type bulging.
The effects of ischemia on wall thickening are
important as well. In fact, wall thickening may be
a more specific index of ischemia than wall mo-
tion. In the area of acute myocardial infarction,
segmental lengths of the involved tissue are in-
creased; paradoxical bulging is noted, and systolic
thinning of the wall occurs.
For those patients at significant risk for intra-
operative ischemia and older patients undergoing
major and prolonged surgery, five electrode elec-
trocardiographic systems are in common use. The
lead systems consist of either a bipolar arrange-
ment, measuring the potential between two elec-
trodes, or a unipolar electrode, measuring the po-
tential between an electrode and a combination of
the remaining electrodes. The standard limb leads
(I, II, IN) are bipolar, whereas the precordial leads
(V,-V
6
) and the augmented limb leads (aVR, aVL,
aVF) are unipolar. For ischemia detection, in an
awake person undergoing exercise stress testing,
lead V
5
is associated with the largest number of
ST-segment abnormalities. Furthermore, the diag-
nostic mode, which filters frequencies below 0.14
Hz, is more useful than the monitor mode (high
pass at 4 Hz) for the detection of these ST-segment
changes. The criteria for ischemic ST changes are
a horizontal or down-sloping depression of 1 mm
or more occurring 80 msec beyond the J point.
This depression can be either new or additive and
is associated with a greater incidence of compli-
cations. The depth of depression of the ST seg-
ment has been related to subsequent coronary
events. Certainly, a 2-mm or more depression car-
ries the most significant prognosis. Recently, the
comparative height of the R wave, used as a nor-
malizing factor, has been shown to give greater
validity to the ST-segment change.
2. Coronary Artery Disease and Good
The first three rows of Table 13-15 represent a
progression of events, beginning with inadequate
anesthesia and ending with cardiac failure secon-
dary to myocardial ischemia, that might occur in a
patient with coronary artery disease and good ven-
tricular function. In this type of patient, angina
pectoris is a primary symptom, hypertension is
frequently present, the cardiac index is normal, the
ejection fraction is greater than 0.5, left ventricular
end-diastolic pressure is less than 12 mm Hg, there
are no ventricular wall-motion abnormalities, and
the cardiovascular system is capable of a hyper-
dynamic response to sympathetic stimulation.
Also, inadequate anesthesia causes all of the deter-
minants of myocardial oxygen consumption to in-
crease; that is, heart rate, preload and afterload,
and contractility (row 1, Table 13-15). These early
changes may be aborted by providing adequate
anesthesia early on, perhaps along with nitroglyc-
erin vasodilatation and beta blockade.
If the left ventricle experiences excessive after-
load, it may become ischemic and begin to fail, as
demonstrated by a decreased cardiac output de-
spite an increased pulmonary artery wedge pres-
sure (row 2 in Table 13-15). These intermediate
changes may be reversed by dilating the arteriolar
side of the circulation with nitroprusside. If the
stress on the left ventricle remains unrelieved, the
left ventricle then may become further ischemic
and fail even more as evidenced by a further de-
crease in cardiac output, despite a further increase
in pulmonary artery wedge pressure, and because
the cardiac output is now so low there is also the
beginning of systemic hypotension (row 3, Table
13-15). Under these late and severe circum-
stances, the heart must be aided with an inotropic
drug. It should be remembered that whenever
vasodilators are used, but particularly with the de-
creased preload effect of nitroglycerin infusion,
concomitant volume infusion may often be re-
quired to maintain systemic and cardiac filling
pressures at normal levels. This scenario empha-
sizes the importance of adequate anesthesia in pa-
tients with coronary artery disease and hyperdy-
namic ventricles.
3. Coronary Artery Disease and Poor
The patient with coronary artery disease and
poor ventricular function may have a history of or
electrocardiographic evidence of previous myocar-
dial infarction and symptoms of congestive heart
failure. There is little or no cardiac reserve, the
cardiac index is less than 2 L/min/m
2
, the ejection
fraction is less than 0.4, left ventricular end-dia-
stolic pressure is often greater than 18 mm Hg,
there may be hypokinetic or dyskinetic segments
of the ventricular wall, and the heart does not
tolerate loss of sympathetic input. In this patient,
deep anesthesia and beta-blocking drugs should be
avoided and cardiac function maintained by using
narcotic anesthesia, nitroglycerin preload and af-
terload reduction, inotropic drugs if cardiac failure
supervenes, and maintenance of preinduction fill-
ing pressures. It should be remembered that, in
order to maintain preinduction filling pressures in
patients with poorly functioning ventricles, nitro-
glycerin increases the oxygen supply/demand ra-
tio, in part, by decreasing preload (diastolic wall
tension). Therefore, nitroglycerin infusion often
needs to be accompanied by volume infusion to
restore simultaneously toward normal both cardiac
output and preload.
D. Valvular Heart Disease
Patients with valvular heart disease are also a
great challenge to successfully and safely anesthe-
tize. For example, aortic stenosis is associated with
a 14-fold increase in perioperative mortality.
77
Ta-
ble 13-18 shows the important hemodynamic
goals to be achieved during anesthesia for patients
with valvular heart disease.
E. Arrhythmias
Thoracic surgery patients are exposed to a great
deal of manipulation of the heart and mediastinum;
these manipulations are associated with a high in-
cidence of supraventricular tachyarrhythmias and
ventricular ectopy.
78
Neck dissections may stimu-
late the carotid sinus reflex, resulting in bradycar-
dia and hypotension. Usually, informing the sur-
geon of the occurrence of these arrhythmias and
terminating the mechanical stimulus causes the
cardiac rhythm to revert back to normal. However,
on occasion the arrhythmias may persist, and it
may be necessary to use one or more of the treat-
ment modalities listed in Table 13-19. Of the hal-
ogenated anesthetics, halothane is most arrhyth-
mogenic and isoflurane the least arrhythmogenic.
V. RE-EXPANSION OF COLLAPSED
LUNG DURING AND AT THE END
OF THORACIC SURGERY (SEE
CHAPTER 11)
Pulmonary edema is well known to occur after
re-expansion of lung following evacuation of a
pneumothorax and a pleural effusion.
79
Less well
known and appreciated is that pulmonary edema
may follow re-expansion of atelectatic lung after a
right main-stem bronchial intubation (with the
chest closed) (left lung becomes edematous) as
well as re-expansion of atelectatic lung during
Table 13-18 HEMODYNAMIC GOALS DURING ANESTHESIA FOR PATIENTS WITH VALVULAR
HEART DISEASE
Variable
Heart rate
Rhythm
Preload
Afterload
Contractility
Global, most
important
AS
Neither increases nor
decreases
Maintain normal
sinus rhythm (do
not lose atrial
kick, cardiovert
atrial fibrillation)
Maintain preload
Maintain blood
pressure
Neither increases nor
decreases
Full heart, maintain
normal sinus
rhythm
IHSS
Avoid increases
Maintain normal
sinus rhythm
Maintain preload
Maintain blood
pressure
Avoid increases
Adequate anesthesia,
avoid
hypertension and
increased
contractility
Valvular Lesion
AR
Avoid decreases
Maintain normal
sinus rhythm
Maintain preload
especially if
vasodilator is used
Systemic
vasodilation may
lead to decreased
regurgitation
Avoid decreases
Avoid decreased
heart rate
MS
Avoid increased heart
rate
Not as important as
in other valvular
lesions
Maintain preload
(however,
increased heart rate
may lead to
increased P
pao
)
May need some
decrease if
pulmonary vascular
resistance requires
a decrease (Table
13-17). If
hypertensive, use
fluid, inotrope.
Amrinone
Avoid increased
pulmonary vascular
resistance and
decreased right
ventricular function
(hypoxemia,
hypercapnia,
acidosis), avoid
increased heart rate
MR
Avoid decreased
heart rate
Maintain normal
sinus rhythm
Maintain preload
Systemic vasodilation
may lead to
decreased
regurgitation
Maintain
Avoid decreased
heart rate, avoid
increased
pulmonary vascular
resistance
(hypoxemia,
hypercapnia,
acidosis)
Abbreviations: AS = aortic stenosis; IHSS = idiopathic hypertrophic subaortic stenosis; AR = aortic regurgitation; MS =
mitral stenosis; MR = mitral regurgitation.
open thoracotomy (see chapter 11). All of these
situations are relevant to the condition of patients
undergoing thoracic surgery and the one-lung ven-
tilation situation. The clinical implications of this
literature on re-expansion pulmonary edema (see
chapter 11) are twofold. First, the collapsed lung
should be expanded slowly and over several non-
staggered but progressively increasing tidal vol-
umes (minimize the development of negative inter-
stitial pressures). The same considerations apply to
demonstrating an airtight bronchial closure just be-
fore closure of the chest after a pneumonectomy
by pressurizing the lung to 35 to 40 cm H
2
0 pres-
sure while the surgeon fills the pleural space
with irrigating fluid. Second, the anesthesiologist
should be aware that the pulmonary edema may
become manifest anywhere between a few minutes
to several hours post re-expansion. The treatment
of re-expansion pufmonary edema is mechanical
ventilation with PEEP, diuretics, and other general
support measures.
VI. TRANSPORT OF PATIENT
80
Some thoracic surgery patients can be extubated
while on the operating room table. These patients
include those who had normal cardiopulmonary
function preoperatively, underwent relatively
short, physiologically nonintrusive surgery, and
have an adequate postoperative vital capacity,
peak inspiratory force, and spontaneous minute
ventilation. These extubated patients should be
sent to the postanesthesia recovery room for close
observation for a period of time. However, in
many thoracic surgery patients, it is obvious, based
on either the preoperative condition of the patient
or the nature of the surgical procedure, that a pe-
riod of postoperative mechanical ventilation and
intensive care will be required; it is most cost
effective to send these intubated patients directly
to the intensive care unit (ICU). In these patients,
preparation for transport to the ICU should be be-
gun toward the end of the procedure.
Anesthetic Considerations (Other Than Management of Ventilation) During and at the End of Thoracic Surgery- 481
Table 13-19 ARRHYTHMIA TREATMENT SUMMARY
Rhythm First Treatment Choice Second Treatment Choice Third Treatment Choice
Ventricular fibrillation
Ventricular tachycardia
Premature ventricular
contractions
Atrial fibrillation
Atrial flutter
Electrical defibrillation: 200
W/sec (3 W/sec/kg) with
increased energy as required
Lidocaine (Xylocaine) l .0-1.5
mg/kg initial bolus followed
by infusion 1^4 mg/min
Verapamil 5.0-10.0 mg IV
Paroxysmal supraventricular Verapamil 5-10 mg IV
(atrial or junctional) Adenosine 0.1-0.2 mg/kg IV
tachycardia
Sinus tachycardia
Sinus bradycardia
Eliminate underlying cause
(e.g.. fever, pain, decreased
blood volume)
Usually no specific treatment
required. If HR^ischemia,
go to second treatment
Eliminate underlying cause
(e.g., hypoxemia, fever, pain)
Usually no specific treatment
required. If hypotension or
ventricular escape beats
present, go to second
treatment
Lidocaine (Xylocaine) 1.0-1.5
mg/kg initial IV bolus,
infusion 1 to 4 mg/min (2
g/500 ml D
5
W)
Propranolol (Inderal) 0.5 mg q 2
min up to 5 mg
Digoxin (Lanoxin) 0.25-0.5 mg
IV
Ouabain 0.1-0.2 mg IV
min up to 5 mg
min up to 5 mg
Digoxin (Lanoxin) 0.25-0.5 mg
IV
Ouabain 0.1-0.2 mg IV
min up to 5 mg
Esmolol 1.0 mg/kg bolus, repeat
q 10 min*
Atropine 0.3-2.0 mg
min up to 5 mg
Fourth choice: Refractory
ventricular fibrillation.
bretylium 5 mg/kg qs 50-100
ml D
S
W up to 30 mg/kg total
dose
Procainamide (Pronestyl) 100
mg q 5 min up to 1000 mg;
may infuse 1-4 mg/min
Cardioversion (first choice if
patient is unstable)
externalsynchronized flutter:
50 W/sec; fibrillation: 200
W/sec; internal: 5-30 W/sec
synchronized
Vagal maneuvers (edrophonium.
carotid sinus massage, Neo-
Synephrine)
Ephedrine 5-10 mg
Fourth choice: isoproterenol 1 4
Fifth choice: atrial pacing if
heart exposed, transvenous
otherwise
*Esmolol is the beta blocker of choice in patients with bronchospasm.
Abbreviations: IV = intravenously; D
5
W = 5 per cent dextrose in water; q every; HR = heart rate.
A. Preparation of Patient for Transport
Once stability of the cardiovascular, respiratory,
and hemostatic systems has been achieved, and the
surgical wounds are being closed, preparations for
transport to the ICU should be under way. The
pretransport preparation check list is shown in Ta-
ble 13-20. Advance notice to the postoperative
ICU should be given 30 to 45 min before leaving
the operating suite. A suggested advance informa-
tion sheet is provided in Table 13-21.
1. Airway
If the patient is going to be intubated and/or
mechanically ventilated postoperatively, the dou-
ble-lumen tube must be changed to a single-lumen
tube (except in rare situations in which differential
lung ventilation is going to be used postoperatively
[see chapter 19]). If the Univent tube is used, the
bronchial blocker should be ^rendered nonfunc-
tional by fully retracting the bronchial blocker,
taping it securely in the fully retracted position.
and cutting the pilot tube to the bronchial blocker
balloon so that the bronchial blocker cuff cannot
be mistaken for the cuff of the main lumen and
inflated. As the end of the operation approaches,
the patient should be given narcotics and left par-
alyzed. After the patient is turned into the supine
position, both lumens of the double-lumen tube.
the oropharynx and nasogastric tube should be
suctioned, and the patient is then ventilated with
100 per cent oxygen. The double-lumen tube
should be visualized entering the larynx by direct
laryngoscopy, and, while continuing to directly
visualize the larynx, the double-lumen tube should
be removed and the single-lumen tube inserted.
Following single-lumen tube insertion, confirma-
tion of proper single-lumen tube placement, and
securement of the single-lumen tube with tape, the
patient should be mechanically ventilated for sev-
eral minutes while other measures are being taken
(see the following). It should be realized that an
esophageal intubation following right pneumonec-
tomy or transposition of the stomach into the chest
may be particularly difficult to diagnose (left-sided
Table 13-20 PRETRANSPORT PREPARATION
CHECK LIST
A. Respiratory system
1. Double-lumen tube changed to single-lumen tube
2. Arterial blood gases on F,0
2
= 1.0
3. Manual transport positive-pressure ventilation system
ready
4. Stethoscope in place
5. Anesthesia mask present
6. Laryngoscope present
B. Chest tube and drainage system
1. Collection compartment below level of patient
2. Blood and air leak not excessive
3. Mediastinum midline
C. Circulatory system
1. Hemodynamically stable
2. EKG on oscilloscope
3. Mean arterial pressure on manometer or phasic
waveform on oscilloscope
4. Vascular catheters untangled, labeled; injection ports
identified; heparin lock and flush on noninfusing lines;
adequate fluid for infusing lines; drug infusions labeled
and in pump meters
5. Coagulation status satisfactory
D. Anesthesia requirements
1. Narcotized
2. Paralyzed
E. Miscellaneous
1. ICU bed present
2. Sufficient personnel available
3. ICU notified
4. Elevator called for
5. Records collected
6. All patient lines that have an operating room
connection (intravascular, Foley, nasogastric, chest
catheters, other monitoring lines, ventilation system)
should be transferred to a transport system connection.
Abbreviations: EKG = electrocardiogram; ICU = intensive
care unit.
or midline breath sounds will be heard, respec-
tively, and left or midline chest movements seen,
respectively, regardless of whether the tube is in
the trachea or esophagus); capnography should be
used to make the differential diagnosis whenever
the single-lumen tube cannot be seen to enter the
trachea directly or there is any question about gas
exchange.
If the laryngeal aperture and entrance of the
double-lumen tube into the larynx cannot be visu-
alized with direct laryngoscopy, then this method
of changing a double-lumen tube to a single-lumen
tube is hazardous and has a high risk of failure.
Inability to visualize the laryngeal aperture is par-
ticularly likely to occur in cases in which massive
amounts of crystalloid fluids and blood have been
infused and the supraglottic area has become
edematous. There are two ways of changing a dou-
ble-lumen tube to a single-lumen tube that utilize
a stylet method and still retain the ability to venti-
late the patient. The first method allows ventilation
during all steps of the tube change, and the second
method allows ventilation most of the time.
The first method involves the use of a jet sty-
let.
81
A jet stylet is a small internal diameter (ID),
hollow, semirigid catheter that is inserted into an
in situ endotracheal tube (in this case, the tracheal
lumen of the double-lumen tube) before extuba-
tion. The proper depth of insertion is to 38 to 39
cm
82
(Fig. 13-11) to ensure that the tip of the jet
stylet remains supracarinal; it is important that the
tip of the jet stylet (tube changer) is supracarinal
before administering any jet ventilation to avoid
barotrauma to the one lung (the one that is venti-
lated if the jet stylet is subcarinal).
When changing a double-lumen tube to a single-
lumen tube, a jet stylet catheter of approximately
100 cm should have an appropriate reserve of
length if the tracheal lumen is uncut
82
and an 80-
cm catheter is long enough if the tracheal lumen is
cut just proximal to the common molding that
binds the two lumens together.
After the double-lumen tube is withdrawn over
the jet stylet, the small-ID hollow catheter may
then be used as a means of ventilation (i.e., the jet
function) and/or as an intratracheal guide for rein-
tubation (i.e., the stylet function) (Fig. 13-12).
81
In some instances, the jet function may safely
allow additional time to assess the need for the
reintubation stylet function. A small Sheridan tube
exchanger (Sheridan Catheter Corp, Argyle, NY)
fits down a 35 French double-lumen tube, and a
medium size tube exchanger will fit down a 41
French double-lumen tube. Both are satisfactory to
be used as a jet stylet in this context (the tracheal
lumen should be cut at the common molding). The
tube exchanger may be connected to the jet injec-
tor by inserting an appropriately sized intravenous
catheter into the proximal end of the stylet (see
Fig. 13-12) or by inserting a female Luer lock
barbed cone adapter into the proximal end of a
Table 13-21 ADVANCE INFORMATION SHEET
FOR ICU
A. Respiratory sytem
1. Endotracheal tube: route, size
2. Ventilator settings: F,0
2
, tidal volume and/or peak
inspiratory pressure, rate, PEEP
3. T-piece: F,0
2
4. Extubated: mask F,0
2
5. Request immediate chest X-ray: yes, no
B. Monitoring system
1. Arterial line: location
2. Central venous pressure: location
3. Pulmonary artery catheter: location
C. Intravenous catheters
1. Site 1: location, fluid, drug
2. Site 2: location, fluid, drug
D. Laboratory studies pending
E. Estimated time of arrival
Abbreviation: ICU = intensive care unit; PEEP = positive
end-expiratory pressure.
Figure 13-11 When a tracheal tube exchanger is inserted to
a depth of 38 to 39 cm down an optimally placed double-lumen
endobronchial tube, its tip will lie proximal to the carina. (From
Hannallah M: Evaluation of tracheal tube exchanges for re-
placement of double-lumen endobronchial tubes. Anesthesiol-
ogy 77:609-610, 1992. Used with permission.)
short length of appropriately sized ETT and the
distal end of the ETT over the tube exchanger.
83
The intratracheal location of the jet stylet may be
preserved while concurrently confirming intratra-
cheal placement of the reintubation tracheal tube
(see Fig. 13-12).
81
83
84
The intratracheal location of the jet stylet may
be preserved while concurrently confirming intra-
tracheal placement of the reintubation tracheal
tube by the following procedure. A standard anes-
thesia circle system is modified by" replacing the
existing elbow connector with a connector incor-
porating a self-sealing diaphragm such as the type
used for bronchoscopy. In addition, the connection
of the proximal end of the lumen of an appropri-
ately sized tracheal tube exchanger to a jet venti-
lator is made immediately available as described
previously. When the new tracheal tube is passed
over the tracheal tube changer, the new tracheal
tube is connected to the anesthesia circuit via the
bronchoscopy elbow by passing the proximal end
of the tracheal tube changer retrograde through the
self-sealing diaphragm of the bronchoscopy elbow
connector (see Fig. 13-12). After conclusively es-
tablishing the intratracheal position and patency of
the new tracheal tube by capnography and auscul-
tation of bilateral breath sounds, the tracheal tube
exchanger is removed by pulling it through the
self-sealing diaphragm of the bronchoscopy elbow
connector.
The second method involves using a fiberoptic
bronchoscope (FOB) as a stylet and cutting away
the double-lumen tube (Fig. 13-13).
85
After ensur-
ing that the patient is completely paralyzed, both
the tracheal and bronchial cuffs of the double-
lumen tube are deflated, and the double-lumen
tube is withdrawn until the bronchial lumen is
above the level of the carina. The tracheal lumen
adapter is cross clamped, and bilateral ventilation
is maintained via the bronchial lumen. With the
tracheal lumen now isolated, a no. 10 scalpel is
used to create a 5 X 5-mm opening in the lateral
wall of the tracheal lumen distal to the reinforced
area connecting the bronchial and tracheal lumina.
A lubricated (inside and outside) 25-cm long. 7-
mm outside diameter single-lumen tube is ad-
vanced over the FOB until the 15-mm adapter is
abutting the handle of the FOB. The distal end of
the single-lumen tube-FOB combination is then
inserted into the tracheal lumen of the double-
lumen tube through the newly created opening and
is advanced until the tracheal mucosa can be visu-
alized.
The new opening in the double-lumen tube is
extended distally with a pair of scissors. The dou-
ble-lumen tube is withdrawn while the tracheal
lumen is cut longitudinally (by a large pair of
sharp scissors or sharp scalpel) until the opening
of the tracheal lumen is out of the mouth. At this
time, ventilation is maintained via the bronchial
lumen of the double-lumen tube, which is resting
in the oropharynx (Fig. 13-13). Once the FOB is
freed from the tracheal lumen of the double-lumen
tube, the single-lumen tube is then passed over the
FOB and positioned in the trachea. Ventilation is
then instituted via the single-lumen tube.
2. Respiration
After 5 to 10 min of ventilation with 100 per
cent oxygen via the single-lumen tube, arterial
Figure 13-12 A method to preserve the intratracheal location of the tracheal tube exchanger during confirmation of intratracheal
placement of a tracheal tube. The tracheal tube exchanger is passed through a self-sealing diaphragm in a fiberoptic elbow adapter.
With this method, positive-pressure ventilation and carbon dioxide sampling may be around the tracheal tube exchanger but within
the tracheal tube. (ETT = endotracheal tube.) (From Benumof JL: Management of the difficult airway: With special emphasis on
awake tracheal intubation. Anesthesiology 75:1087-1110, 1991. Used with permission.)
blood gases should be obtained and the values
known before departure. A simple foolproof man-
ual system for providing controlled positive-pres-
sure ventilation with an inspired oxygen concen-
tration of 100 per cent during transport should be
standing by. The transport ventilating system
should allow PEEP to be provided to the patient.
Manual ventilation during intrahospital transport
of critically ill, mechanically ventilated patients is
safe provided the person performing the manual
ventilation knows the inspired oxygen fraction and
minute ventilation that is required before the trans-
port takes place and is trained to approximate the
minute ventilation during transport.
86
A sufficient supply of oxygen should be avail-
able, remembering that a completely full tank of
oxygen contains enough gas to supply 10 to 12
L/min for approximately 50 min. Monitoring of
ventilation should be provided both by observation
of chest movements and by hearing breath sounds
via an esophageal stethoscope. The anesthesiolo-
gist should take along an appropriately sized an-
esthesia mask (to provide positive-pressure venti-
lation in case of inadvertent extubation), a la-
ryngoscope (to reintubate), and some essential
drugs for cardiovascular resuscitation.
3. Chest Tube and Drainage System
87
The chest tube and drainage system must be
functioning properly before transport. Chest X-ray
confirmation of proper chest tube position is made
by identifying the radiopaque stripe that outlines
the proximal drainage hole (there are multiple
holes) and by noting the absence of a visceral
pleural line.
88
Two tubes are usually required to
drain both air and fluid from the pleural space (an
empty hemithorax after a pneumonectomy requires
only one tube; see later discussion); one tube is
placed at the anterior apex to favor the escape of
air when the subject is sitting upright, and one tube
is placed at the posterior base to drain accumulat-
ing fluid when the subject is supine. They are
usually inserted through separate incisions just be-
fore the chest is closed. As the last stitch closing
the pleural cavity is being tied, the lungs should
be slowly inflated and then held fully inflated to
w
Proximal
clamp
Bronchial lumen
Tracheal lumen
\
- Double-lumen tube
Single-lumen tube over
flexible fiberoptic
bronchoscope
New opening (for bronchoscope)
Longitudinal cut
Figure 13-13 Diagrammatic representation of fiberoptic bronchoscope passed through tracheal lumen of the endobronchial tube
and acting as guide for single lumen tube. The bronchial lumen of the endobronchial tube remains in close proximity to the glottic-
inlet, permitting ventilation of the lungs while the tube's tracheal lumen is freed from the bronchoscope. (From Gatell JA. Barst
SM, Desiderio DP, et al: A new technique for replacing an endobronchial double-lumen tube with an endotracheal single-lumen
tube. Anesthesiology 73:340-341, 1990. Used with permission.)
reverse any remaining atelectasis and to help evac-
uate the pleural space of air and fluid.
Modern chest tube drainage systems have three
compartments (Fig. 13-14). The first compartment
is a simple graduated collection chamber where
the amount of draining blood and fluid can be
measured accurately (see Fig. 13-14, right panel).
The second compartment is an underwater seal
that serves as a simple, but very reliable, one-way
valve, which allows air to escape from the pleural
space to the drainage system but not to enter the
pleural space from the drainage system during the
next inspiration. If the inlet drainage limb to the
underwater seal compartment is put just a short
distance ( 1 to 2 cm) below the level of the water,
there is minimal resistance to the escape of air, but
a huge inspiratory effort would be needed to break
the seal by drawing water from the bottle up the
drain and into the chest (see Fig. 13-14, middle
panel). All drainage systems, whether simple or
complex, require a water seal (i.e., the water seal
compartment is all that is necessary for the treat-
ment of a spontaneous pneumothorax). The third
compartment controls the amount of negative pres-
sure that the suction can generate. The inlet drain-
age limb to the suction control compartment is
exposed to the outlet suction above a level of
water. The level of the water in the outlet limb of
the suction control compartment is vented to at-
mosphere. If the negative pressure above the water
level exceeds the height of the water level, air will
come in from the suction control atmospheric vent,
bubble through the water, and neutralize the exces-
sive negative pressure above the water level (see
Fig. 13-14, left panel).
Two factors determine the level of suction.
First, the central movable vent tube, which can be
raised or lowered to adjust the underwater depth,
determines the negative pressure that the system
will generate. Second, the actual suction pump se-
lected may be a limiting factor. Low-pressure sys-
tems capable of generating between 15 and 20 cm
H
2
0 negative pressure and between 15 and 10
L/min of flow include the Stedman, Gomco, and
Thermovac pumps. Emerson and Sorenson sys-
tems are high-pressure systems capable of pres-
sures of - 60 cm H
2
0 with flows of >20 L/min.
When selecting a suction system, care should be
Figure 13-14 This schematic diagram of a modem
chest tube drainage system shows that the chest tube
drainage system is made up of three compartments. The
first compartment is a collection chamber for fluids
(blood, pus) from within the chest. The second com-
partment contains the origin of the chest tube suction
and a water seal valve that prevents gas or fluid from
being drawn, by a forceful spontaneous inspiration, into
the chest. The third compartment controls, via an at-
mospheric vent, the degree of negative pressure that
can be developed by chest tube drainage system suc-
tion. See the text for a further explanation.
exercised to provide a negative pressure greater
than the possible positive pleural pressure seen on
expiration. If these conditions are not met, a ten-
sion situation could be created despite continuous
suction.
88
In other words, if the flow rate of the
suction system cannot exceed the rate of leakage
of air into the pleural space, a tension pneumo-
thorax will occur.
If functioning lung tissue remains on the oper-
ated side (resection less than a pneumonectomy,
all nonpulmonary surgery), the drains should be
joined to the inlet of an underwater seal as soon as
the pleural cavity has been closed to prevent the
lung from collapsing again (owing to accumula-
tion of air in pleural space) while the remaining
layers of the chest wall are sutured. If separate
bottles are used for the drainage system (see Fig.
13-14, upper panel), the collection compartment
bottle should be below the level of the patient at
all times to prevent the entry of fluid from the
collection compartment and chest tube per se back
into the chest. Modern disposable plastic systems
have hooks on them that can be easily attached to
the lower part of the transport bed. Simple clamp-
ing of the drains is dangerous owing to the possi-
ble development of tension pneumothorax.
Air is expelled through the drain during expira-
tion if the patient is breathing spontaneously, and
the volume escaping increases during coughing.
The drain will bubble continuously if there is a
large leak, particularly if suction is used, and, if
both blood and air are escaping simultaneously, a
prohibitive amount of froth may appear. The froth
can be controlled by adding an antifoaming agent
(such as a few drops of alcohol). If there is no air
leak, the fluid level will move gently in the drain-
age tubing during quiet, spontaneous respiration
but will cease to swing when the lung is fully
expanded and opposed to the chest wall or if the
drain is blocked.
Frequent (every half hour) mechanical stripping
of the chest tubes (a distal movement to create
vacuum within the tubing to suck fluid and/or air
from the chest) or milking of the chest tubes if
blood is being drained (pushing air and fluid and
clots back into the chest) is important because,
once clots have begun to collect within the chest
and chest tubes, chest tube function deteriorates,
and the likelihood of the need for re-exploration
for hemopneumothorax increases. In individuals
who have significant drainage and/or bleeding,
these maneuvers may be required every 5 min.
Nevertheless, the value of these procedures in
postoperative coronary artery by-pass graphing pa-
tients has been questioned.
89
When the functional status of a tube is ques-
tioned, several simple maneuvers can be used to
assess its integrity. Observation of synchronous
water seal and respiratory motion suggest the tube
is still functioning in the pleural space and all
connections are tight. If an air leak is suspected at
a particular point in the system, be it bottle, tube,
or connector, sequential clamping with distal suc-
tion before and after the spot in question should
be performed. Bubbling air through the water seal
when the drainage system is clamped just distal to
the point in question, which disappears when the
clamp is placed just proximal to that point, identi-
fies the site of leakage; the identified component
then can be changed. If the tube is not functioning
and occlusion of the drainage holes is suspected,
the tube can be disconnected and flushed with sa-
line solution in an effort to dislodge obstructing
debris. The evacuation of clots and debris from the
chest tube with the use of a sterile suction catheter
introduced into the chest tube through a sterile cap
has been proposed.
90
After a pneumonectomy but before transport,
the position (laterality) of the mediastinum should
be ascertained.
91
This can be done in two ways:
clinical examination and measurement of pressure
in the hemithorax. Clinical data consist of deter-
mining the position of the trachea and inspecting
the anteroposterior chest roentgenogram. Perhaps
more practical and accurate is the measurement of
pressure in the empty hemithorax by manometer.
If the pressure in the empty hemithorax is signifi-
cantly positive, then it is likely that the medias-
tinum is shifted into the contralateral hemithorax.
If the pressure in the empty hemithorax is signifi-
cantly negative, then it is likely that the medias-
tinum is shifted into the ipsilateral hemithorax. To
get the mediastinum in the midline, air can be
injected into or aspirated from the empty hemitho-
rax, as dictated by the ongoing pressure manome-
ter readings.
Closing the chest without drainage after pneu-
monectomy makes it more difficult to recognize
serious postoperative hemorrhage. Consequently, a
single basal chest drain is usually used and is
joined to an underwater seal bottle without suction
(suction will cause a mediastinal shift to the ipsi-
lateral side and, perhaps, cardiovascular collapse).
It is left undamped until the patient is lying on the
back and breathing spontaneously. Some air usu-
ally escapes through the drain when the patient is
turned from the lateral to the supine position, or if
coughing occurs, when spontaneous ventilation is
being re-established. This air should be allowed to
escape so that pressure does not build up within
the space and cause surgical emphysema around
the drain or the wound site. The mediastinum
should be central at the end of the operation if this
routine is followed, and once this has been con-
firmed, the tube is clamped and remains so while
the patient is returned for postoperative supervi-
sion. Thereafter, it is undamped for 1 to 2 min
every hour for the first 12 to 24 hours so that the
excess fluid may drain, thereby preventing un-
wanted mediastinal shift.
4. Circulation
For the patient to be a candidate for leaving the
operating room, the cardiovascular system must be
stable. Vascular volume status should be optimal
for that particular patient, as judged by the urine
output, systemic pressure, left- or right-sided fill-
ing pressures or both, and cardiac output, if avail-
able. All vascular catheters should be identified by
label and should be free and untangled with at
least one injection port conveniently placed. Ade-
quate fluid for all vascular catheters should be
available. Drugs being administered by continuous
infusion should be reviewed, and the administra-
tion rates and cardiovascular responses confirmed
to be appropriate. The cardiac rate, rhythm, and
mean arterial pressure should be monitored during
transport. Essential cardiovascular resuscitation
drugs should be taken along.
5. Coagulation
If massive transfusion was required intraopera-
tively, the coagulation status should be normal or
as nearly normal as possible at the time of trans-
port. All products to treat coagulation abnormali-
ties should have been administered if possible.
Blood specimens should be sent to the laboratory
from the operating room for the determination of
coagulation status (ACT, aPTT, PT, fibrinogen),
platelet count, hematocrit, serum potassium, and
arterial blood gases. The results of these determi-
nations will therefore be available to the ICU per-
sonnel within moments of the patient's arrival
there.
6. Miscellaneous
A fully functional ICU bed should be available
for direct patient transfer. All records should be
collected in order for them to be transported with
the patient. Sufficient personnel to effect a smooth
transport should be gathered. Prior to moving the
patient, all possible connections of the patient to
the operating room must be disengaged, such as
pressure lines to transducers, electrocardiogram
leads to the room monitor, chest tube and Foley
catheters to the operating room table, and intrave-
nous lines to operating room poles, and must be
transferred to their appropriate places on the trans-
port bed. If an elevator is to be used, it should be
called for so that it is there waiting for the trans-
port team.
In the rare case of transport by air, the pleural
cavity must be decompressed by a chest tube with
a guaranteed one-way flow of gas (from the
pleural cavity to the environment) (by a one-way
valve or underwater seal mechanism) because at
an altitude of 5000 to 6000 feet airspaces will
expand by 20 to 35 per cent (Boyle's Law). Simi-
larly, all other potential airspaces must be decom-
pressed (e.g., stomach, bowel).
92
B. Transport
Preferably, the patient should be moved only
once, and that is directly from the operating room
table to the ICU bed. The move should be per-
formed smoothly and gently to minimize vascular
volume shifts or changes in cardiac function. A
carefully coordinated move by a sufficient number
of personnel should prevent accidents such as dis-
connected intravascular lines and various tubes
being pulled out or malpositioned. The anesthe-
siologist should be responsible only for ventilation
and monitoring, not moving, fetching, or adjusting
parts of the bed.
Monitoring during transport should include at
least an esophageal stethoscope for respiratory and
cardiovascular sounds and an oscillographic dis-
play of the electrocardiogram. In addition, if an
arterial line is in situ, mean arterial pressure should
be monitored, at a minimum, with a sterile anae-
roid gauge directly connected to the arterial line
by a 50-cm segment of sterile intravenous tubing.
Still better, and presently practiced more com-
monly, is an oscillographic display of the electro-
cardiogram and arterial pressure waveform on a
portable monitor for continuous observation dur-
ing transport. A defibrillator should also accom-
pany the very critically ill patient. The longer and
more complex the route to the ICU, the more care
must be exercised in planning and execution of
this transport. If an elevator separates the operating
room from the ICU, it should be equipped with a
separate reserve oxygen system, electrical power
outlet, phone communication system, and if possi-
ble, suction equipment (physicians responsible for
transport should ask themselves what they would
want available in an elevator that is jammed for 2
hours).
C. Arrival In Intensive Care Unit
On arrival in the ICU, the pretransport operating
room take-down should be reversed and a quick
ICU arrival hookup performed. Priorities should
be the same as in a cardiac resuscitation: establish-
ment of the airway, adequate gas exchange, and
circulation. Ventilatory function is assessed as the
patient's airway is connected to a ventilator, set
initially with an F,0
2
of 100 per cent. Other initial
ventilator settings such as tidal volume, respiratory
rate, and PEEP level should duplicate those found
optimal in the operating room (see chapter 20).
The chest must be seen and heard to move bilat-
erally. Vital signs of immediate importance are
cardiac rate and rhythm (electrocardiogram) and
mean arterial and cardiac filling pressures. Infu-
sion rates of potent drugs are verified, and desired
cardiovascular responses are confirmed.
Tasks of lesser importance can be performed as
time permits. Laboratory and other diagnostic pro-
cedures should be performed as soon as practical
(chest film, arterial blood gases, electrolytes, he-
matocrit, 12-lead electrocardiogram, and coagula-
tion screen if indicated). Before the anesthesiolo-
gist leaves the ICU, he or she should ascertain the
present status of the patient, confirm cardiovascu-
lar and respiratory stability, be assured that no
problems of an acute nature exist, and confirm that
the ICU nurses are satisfied with the patient's sta-
tus. A brief early postoperative note as well as a
copy of the anesthetic record should be placed in
the patient's chart.
REFERENCES
1. Fitzgerald JM, Hargreave FE: The assessment and man-
agement of acute life-threatening asthma. Chest 95:888-
894, 1989.
2. Gold MI, Helrich M: Pulmonary mechanics during general
anesthesia: V. Status asthmaticus. Anesthesiology 32:422-
428, 1970.
CHAPTER 14
Anesthesia for Special Elective
Diagnostic Procedures
1. Introduction
II. Bronchoscopy
A. Fiberoptic Bronchoscope
1. Indications
2. Ventilatory Considerations
3. Anesthetic Technique
a. Local Anesthesia
b. General Anesthesia
4. Complications
B. Rigid Ventilating Bronchoscope
1. Indications
a. Sidearm Connection
b. Venturi Connection
a. Local Anesthesia
b. General Anesthesia
4. Complications
III. Mediastinoscopy
A. Indications and Surgical
Considerations
B. Anesthetic Technique
C. Complications
IV. Thoracoscopy
Considerations
C. Complications
492 Anesthesia for Special Elective Diagnostic Procedures
I. INTRODUCTION
There are a number of very common invasive
diagnostic procedures used to evaluate thoracic
surgery patients. These diagnostic procedures all
have special anesthetic implications and problems.
Bronchoscopy requires that the airway be shared
between the anesthesiologist and the endoscopist,
which causes special problems for ventilation and
anesthetic technique. Mediastinoscopy can cause
compression of important vessels and hemorrhage,
and the anesthesiologist's attention must be fo-
cused on recognizing these complications. In ad-
dition, these patients must not be allowed to strain
or cough. Thoracoscopy often involves double-lu-
men tube intubation and one-lung ventilation. This
chapter considers these three common major diag-
nostic techniques in the order just mentioned.
II. BRONCHOSCOPY
Bronchoscopy is the most common invasive
procedure used for the diagnosis and treatment of
chest diseases. There are three general diagnostic
indications for bronchoscopy.
1
First, bronchos-
copy is indicated to investigate the cause of medi-
cally resistant chronic chest symptoms and signs
(cough, pain, wheeze, atelectasis, pneumonia). Se-
cond, bronchoscopy is indicated to determine the
site, extent, and cause of acute changes such as
hemoptysis, acute inhalation injury, and regional
abnormality on the chest X-ray. Third, bronchos-
copy is an essential tool in the evaluation of pa-
tients for thoracic surgery (to evaluate the tracheo-
bronchial tree for tumors, to rule out metastases,
to obtain tissue and specimens [cytology, biopsies,
cultures], to determine presence and exact level of
tracheobronchial esophageal fistula, to palpate the
carinal and subcarinal areas). The therapeutic in-
dications are to aid in the treatment of acute res-
piratory failure (remove secretions, reverse atelec-
tasis, drain abscess, position endotracheal tubes),
to aid in laser resection of bronchogenic carci-
noma, and to allow for removal of foreign bodies.
This chapter discusses anesthesia for diagnostic
bronchoscopy. The anesthetic management of the
various therapeutic bronchoscopies are discussed
elsewhere (laser resection of tumors in chapter 15,
removal of foreign bodies in chapter 17, treatment
of acute respiratory failure and aid in mechanical
ventilation in chapter 20).
There are three types of bronchoscopes in cur-
rent use, and they consist of the flexible fiberoptic,
rigid ventilating, and rigid Venturi (Sanders injec-
tor) types. The manner in which patients are ven-
tilated with these bronchoscopes differs consider-
ably, and the different ventilatory considerations
have important implications for anesthetic tech-
nique (Table 14-1). Flexible fiberoptic bronchos-
copy may be performed with local or general an-
esthesia, whereas rigid ventilating and Venturi
bronchoscopy is best performed under general an-
esthesia. Since the inspired oxygen and inhala-
tional anesthetic concentration is known with a
rigid ventilating bronchoscope but not known with
a rigid Venturi bronchoscope, rigid ventilating
bronchoscopy can be performed with either inha-
lational (including nitrous oxide) or intravenous
general anesthesia, whereas rigid Venturi bron-
choscopy is best performed with intravenous gen-
eral anesthesia and 100 per cent oxygen. Since
minute ventilation is constant with either fiberoptic
or rigid Venturi bronchoscopy (no need to inter-
rupt ventilation) but not with rigid ventilating
bronchoscopy (need to interrupt ventilation when
the eyepiece is removed), the former two can be
comfortably used for long procedures, whereas the
latter should be used for relatively short proce-
dures. In view of these differences in ventilatory
and anesthetic technique, the different broncho-
scopes are discussed separately.
A. Fiberoptic Bronchoscope
1. Indications
Use of the flexible fiberoptic bronchoscope is
especially indicated for bronchoscopy whenever
Table 14-1 VENTILATION CHARACTERISTICS AND ANESTHETIC IMPLICATIONS OF THE
THREE DIFFERENT TYPES OF BRONCHOSCOPES
Type of
Bronchoscope RO,
Concentration
of Inhalation
Anesthesia
Constancy of
Minute Ventilation
Suitable Duration
of Procedure
Preferred Type of
Anesthesia
Flexible fiberoptic
Rigid ventilating
(sidearm
connection)
Rigid Venturi
Known
Knjawn
Unknown
Known
Known

Unknown
Constant
May be inconstant
without packing
Constant
Long
Long (if packing
used)
Long
Local or general
anesthesia
Inhalation or
intravenous general
anesthesia
Intravenous general
anesthesia
Endotracheal Tube Size Without
Fiberoptic Bronchoscope, mm ID
Figure 14-1 This diagram relates the size of an endotracheal tube without an indwelling fiberoptic bronchoscope (x-axis) to the
resultant functional size that the endotracheal tube would have with an indwelling fiberoptic bronchoscope (left-hand y-axis) (three
fiberoptic bronchoscope sizes are shown: 5.0-, 5.7-, and 6.0-mm outside diameter). The cross-sectional area of unoccupied
endotracheal tube in mm
2
that is responsible for the resultant functional size is shown on the right-hand y-axis. (The diagram is a
modification [with permission] of Figure 7 from Lindholm CE, Oilman B, Snyder JV, et al: Cardiorespiratory effect of flexible
fiberoptic bronchoscopy in critically ill patients. Chest 74:362-368, 1978.)
there are mechanical problems with the neck that
would make rigid bronchoscopy especially diffi-
cult; whenever upper lobe and peripheral lung le-
sions need to be visualized and are beyond the
reach and capability of a rigid bronchoscope (the
fiberoptic bronchoscope can examine the third to
possibly the fourth order of bronchi); and when-
ever there is a need to find the site of limited and
perhaps distal hemoptysis, to aid in the treatment
of respiratory failure, to obtain very localized cul-
tures, biopsies, and cytology, and to perform selec-
tive bronchography.
1-5
The many airway manage-
ment and intensive care unit uses of a fiberoptic
bronchoscope are beyond the scope of this chapter
(see chapters 9, 13, and 20).
In a nonintubated patient, flexible fiberoptic
bronchoscopes with outside diameters of 5.0, 5.7,
and 6.0 mm occupy 6, 10, and 11 per cent, respec-
tively, of the cross-sectional area of an average 70-
kg adult trachea (which has a diameter of 18 mm).
Consequently, spontaneous ventilation and removal
of carbon dioxide is not usually significantly im-
paired in a nonintubated, nonbronchospastic, spon-
taneously ventilating patient undergoing fiberoptic
bronchoscopy under local anesthesia. However,
since the suction port is capable of removing large
amounts of air (14.2 L/min at 760 mm Hg through
a 2-mm suction port), which may cause atelectasis
(especially if the bronchoscope becomes wedged
during suctioning), oxygen supplementation is
highly desirable.
6
7
This can be accomplished by
having the patient breathe through nasal cannulas
or, more preferably, through a facial mask (which
provides a higher F,0
2
) that has a hole cut in it to
allow for insertion of the fiberoptic bronchoscope.
In an intubated patient (which would usually be
the case if general anesthesia were used), the fiber-
optic bronchoscope occupies a very significant
amount of the cross-sectional area of the endotra-
cheal tube. Figure 14-1 shows the reduction in
effective size of any sized endotracheal tube with
a 5.0-, 5.7-, and 6.0-mm outside diameter fiberop-
tic bronchoscopy.
8
For example, a 5.7-mm outside
diameter fiberoptic bronchoscope occupies 41 and
52 per cent of the cross-sectional areas of 9- and
8-mm internal diameter endotracheal tubes, re-
spectively, which reduces the functional internal
diameters of these endotracheal tubes to 7.0 and
5.5 mm, respectively.
8
In view of the greatly re-
duced surface area available for ventilation under
these circumstances, it is obvious that ventilation
must either be controlled or vigorously assisted
with positive-pressure ventilation, This may be
easily accomplished by passing the fiberoptic
bronchoscope through a self-sealing rubber dia-
phragm in the elbow connector to the endotracheal
tube; tidal ventilation then occurs around the fiber-
optic bronchoscope but within the endotracheal
tube (Fig. 14-2). However, it must be realized that
if the fiberoptic bronchoscope occupies too much
of the cross-sectional area of the endotracheal
tube, tidal gas may enter the lungs under positive
pressure, but the elastic recoil of the lung may not
force the tidal gas out of the lungs, and distal gas
trapping, high distal positive end-expiratory pres-
sure (PEEP), and barotrauma may result. Conse-
quently, a low inspiratory-to-expiratory time ratio
is desirable. In light of these considerations, it is
apparent that an endotracheal tube with an internal
diameter of 8.0 to 8.5 mm or larger should be used
with any adult-sized fiberoptic bronchoscope.
8
If a
smaller endotracheal tube must be used (e.g., 7.0-
mm internal diameter), the use of helium/oxygen
mixtures should be considered (see chapter 3).
9,10
In addition, periods of hyperventilation with 100
per cent oxygen without the fiberoptic broncho-
scope in place may need to be alternated with short
periods of fiberoptic bronchoscopy. Oxygen can
always be insufflated down the suction port of the
fiberoptic bronchoscope, and it is my practice to
routinely do so."
In anesthetized children and adults and in awake
adults, the fiberoptic bronchoscope may also be
passed through a laryngeal mask airway (Fig.
14-3).
I2
~
14
Table 14-2 shows the relationship be-
tween the outside diameter of the fiberoptic bron-
choscope and internal diameter of the largest en-
dotracheal tube that fits into the shaft of the
various sized laryngeal mask airways.
12
It may be
inferred from Table 14-2 that the internal diameter
of the shaft of the laryngeal mask airway is 3 to 4
Fiberoptic Bronchoscope
Endotracheal Tube Ventilating System
Figure 14-2 This schematic diagram shows that a fiberoptic bronchoscope is ordinarily inserted through a self-sealing diaphragm
in the elbow connector to an endotracheal tube. Inspired gas under positive pressure from the anesthesia circle machine goes around
the outside of the fiberoptic bronchoscope but within the lumen of the endotracheal tube. Exhaled gas is via the same route into the
anesthesia circle system. The seal of the diaphragm in the elbow connector around the fiberoptic bronchoscope ensures the ability
to continue positive-pressure ventilation while the fiberoptic bronchoscope is in use.
Anesthesia for Special Elective Diagnostic Procedures 4 9 5
Figure 14-3 Fiberscope within a 6.0-mm ID endotracheal
tube (ETT) that is within the shaft of a no. 4 laryngeal mask
airway. The fiberscope and ETT pass through the central com-
partment of the grille at the end shaft of the laryngeal mask
airway, and the cuff of the endotracheal tube is inflated to
increase the distinctiveness of the ETT. (From Benumof JL:
Use of the laryngeal mask airway to facilitate fiberoptic-endos-
copy intubation. Anesth Analg 74:313-314, 1992. Used with
permission.)
mm greater than the internal diameter of the cor-
responding endotracheal tube (e.g., the internal di-
ameter of the no. 4 laryngeal mask airway is ap-
proximately 9-10 mm). The difference between
the internal diameters of the laryngeal mask air-
way and fiberoptic bronchoscope determines the
resistance to airflow, and the clinical implications
of an increase of airflow resistance are the same as
passing a relatively large fiberoptic bronchoscope
through a relatively small endotracheal tube.
It is apparent from the previous discussion that
passage of a fiberoptic bronchoscope through an
endotracheal tube in generally anesthetized pediatric
patients is problematic with respect to ventilation.
Fortunately, in generally anesthetized children (as
well as adults), the fiberoptic bronchoscope may
be passed into the trachea without going through
an endotracheal tube in the following way.
15
Gen-
eral anesthesia is induced, and the patient is venti-
lated (either spontaneously or controlled with pos-
itive pressure) via a mask. The fiberoptic
bronchoscope can be introduced into the airway
down the central slit in the diaphragm of a bron-
choscopy elbow adaptor (Fig. 144)
15
or through
an anesthesia mask with diaphragm (Fig. 14-5)."
In either case, the airway can thus be examined
with little interference with ventilation. These
methods avoid creating a small breathing space
between the fiberoptic bronchoscope and the con-
duit for the fiberoptic bronchoscope (either an en-
dotracheal tube or laryngeal mask airway), avoid
interfering with the view of the vocal cords, and
perhaps avoid further edema in an already nar-
rowed airway.
In the rare and unusual situation of high tracheal
and laryngeal obstructions in infants, the institu-
tion of transtracheal jet ventilation may be life-
saving
16
(irrespective of whether a fiberoptic or
rigid instrument is used for the definitive proce-
dure).
a. LOCAL ANESTHESIA
Patients with reactive airways may need preop-
erative bronchodilation (,-agonists, aminophyl-
line, steroids; see chapter 13 and Fig. 13-1). In
addition, atropine premedication
17
IH
is useful to
decrease the volume of secretions and to cause
some bronchodilation. Diazepam and barbiturates
are useful to decrease anxiety and systemic toxic-
ity of local anesthetics.
The fiberoptic bronchoscope is relatively easily
Table 14-2 LARGEST SIZE ENDOTRACHEAL
TUBE AND FIBERSCOPE THAT
CAN FIT CONCENTRICALLY INTO
THE LARYNGEAL MASK AIRWAY*
Largest
Fiberscope That
LMA Largest ETT That Fits Into ETT
Size Fits Into LMA (Middle Column)
Number (ID in mm) (OD in mm)
4 6.0. cuffed 5.0+
3 6.0, cuffed 5.0+
2 4.5, without cuff 3.5+
l 3.5, without cuff 2.7
*From Benumof JL: Use of the laryngeal mask airway to
facilitate hberoptic-endoscopy intubation. Anesth Analg
+01ympus BF-P20D.
t-Pentax FB-10Hor F1-10P.
01ympus PF-27M.
Abbreviations: LMA = laryngeal mask airway: ETT =
endotracheal tube; ID = inside diameter; OD = outside di-
ameter.
Figure 144 Fiberoptic bronchos-
copy in an anesthetized child using
the Olympus BF 3C10 scope (exter-
nal diameter 3.6 mm). (From Khoo
ST: Anaesthesia for fiberoptic bron-
choscopy in children. Anaesthesia
42:248-249, 1990. Used with per-
mission.)
Figure 14- 5 Use of the anesthesia
mask with diaphragm and oral air-
way intubator as aids to fiberoptic
tracheal intubation in an anesthetized
(and paralyzed) patient. (Reproduced
with permission from Benumof JL:
Management of the difficult airway:
with special emphasis on the awake
intubation. Anesthesiology 75:1087-
1110, 1991.)
Table 14-3 LOCAL ANESTHETIC
TECHNIQUES FOR
NASOTRACHEAL FIBEROPTIC
BRONCHOSCOPY
1. Cocaine (10 per cent) to nasal mucosa
2. Soft nasopharyngeal airways liberally coated with lidocaine
ointment
3. Local anesthetic spray to nasopharyngeal, oropharyngeal,
laryngeal, and tracheal mucosal surfaces
a. Tetracaine 0.5 per cent with epinephrine
b. Lidocaine 4 per cent
4. Block of the lingual branch of IX
5. Superior laryngeal nerve block
a. Externally by needle
b. Internally by swab soaked in local anesthetic
6. Transtracheal block
7. Local anesthetic spray down suction channel of fiberoptic
bronchoscope
and atraumatically passed through the nose into
the trachea when the patient has had adequate top-
ical anesthesia and sedation. The naris selected is
the one that the patient can breathe through most
easily (presumably, this is the naris that is the
larger or less obstructed one). There are several
local anesthetic techniques that clearly facilitate
nasotracheal fiberoptic bronchoscopy; all tech-
niques or combination of techniques must anesthe-
tize the nose, pharynx, larynx, and trachea (Table
14-3).
The nose may be anesthetized by spraying either
0.5 per cent tetracaine with epinephrine or 4 per
cent lidocaine with epinephrine, by topically ap-
plying cocaine by cotton-tipped wooden pledgets
(total dose over 15 min in a 70-kg patient not to
exceed 300 to 350 mg [3 to 4 ml of 10 per cent
solution or 3.3 mg/kg]), or by passing progres-
sively larger soft nasopharyngeal airways that are
liberally coated with lidocaine ointment. The first
two nasal techniques both anesthetize the nose and
shrink the nasal mucosa by active vasoconstric-
tion, whereas the last technique anesthetizes the
nose and dilates the nasal cavity by mechanical
means.
The pharynx may be anesthetized by spraying
local anesthetic through the oral cavity over the
tongue and pharynx, by gargling viscous lidocaine,
or by blocking the lingual branch of IX bilaterally.
This easily performed block is effective in elimi-
nating the gag reflex and hemodynamic response
to a laryngoscopy."
I9
The patient's tongue is
gently retracted laterally (by pulling the tip of the
tongue with gauze and by pushing it with a tongue
blade), exposing the palatoglossal arch (also called
the anterior tonsillar pillar (Fig. 14-6). The base
of the palatoglossal arch forms a U- or J-shaped
band of tissue or bridge starting from the soft
palate, running along the lateral pharyngeal wall
to the lateral margin of the base of the tongue. The
palatoglossal arch is pierced approximately 0.5 cm
from the lateral margin of the root of the tongue at
the point where it joins the floor of the mouth (at
the trough of the U- or J-shaped band of tissue)
using a 25-gauge spinal needle. The length of the
spinal needle allows the local anesthetic syringe to
be outside of the mouth and therefore not in the
line of vision. The needle is inserted 0.5 cm and
an aspiration test is performed. Air will enter the
syringe if needle placement is too deep, causing
the tip of the needle to exit from the posterior
aspect of the palatoglossal arch and enter the oro-
pharynx.
An aspiration test is also helpful in reducing the
possibility of an intravascular injection. Two milli-
liters of 2 per cent lidocaine are slowly injected.
The procedure is then repeated on the opposite
side. Because the injection is made into loose sub-
lingual tissue, there should be minimal resistance
to injection. Within a few minutes, the posterior
third of the tongue, the pharynx, and the pharyn-
geal side of the epiglottis (vallecula) should be
adequately anesthetized to allow direct laryngos-
copy with a Macintosh blade with minimal dis-
comfort or gagging.
The larynx may be additionally anesthetized by
Figure 14-6 The lingual branch of the glossopharyngeal
nerve (IX) can be blocked by instilling l to 2 ml of a local
anesthetic at the trough of the glossopalatal arch (this band of
tissue sweeps upward from the lateral base of the tongue to the
palate; it is commonly called the anterior pillar). The trough of
the glossopalatal arch may be best visualized by retracting the
tongue with a tongue blade (in this example a gloved finger is
being used). (From Benumof JL: Management of the difficult
airway: With special emphasis on the awake intubation. Anes-
thesiology 75:1087-1110, 1991. Used with permission.)
continuing the local anesthetic spray via the mouth
and nose, by superior laryngeal nerve blocks with
2 ml of 2 per cent lidocaine, or by transtracheal
block with 4 ml of 4 per cent lidocaine during
exhalation. The superior laryngeal nerve block
technique .consists of needle application of local
anesthetic to the thyrohyoid membrane between
the superior lateral cornu of the thyroid cartilage
and the inferior lateral margin of the cornu of the
hyoid bone (Fig. 147).
20
*
21
An internal superior
laryngeal nerve block technique consists of paint-
ing the pyriform fossae with sponges that are
soaked with local anesthetic. Superior laryngeal
nerve block anesthetizes the lower pharynx, laryn-
geal epiglottis, vallecula, vestibule, aryepiglottic
fold, and posterior rima glottis.
The trachea may be anesthetized by the local
anesthetic spray as the patient breathes the nebu-
lized material, by the transtracheal block, or by
local anesthetic sprayed down the suction channel
of the fiberoptic bronchoscope (spray as you go).
In one study that compared 100 mg of transcricoid
lidocaine with 240 mg of spray-as-you-go lido-
caine, the transcricoid method was significantly
more effective in decreasing the number of coughs
per minute.
22
Once the fiberoptic bronchoscope is introduced
into the trachea, topicalization of the distal tra-
cheobronchial tree can be accomplished by contin-
uing to spray local anesthetic down the suction
channel of the fiberoptic bronchoscope. Transtra-
cheal block is the least performed maneuver be-
Figure 14-7 This schematic diagram shows the point of
emergence of the superior laryngeal nerve through the thyro-
hyoid membrane between the superior lateral cornu of the
thyroid cartilage and the inferior aspect of the hyoid bone.
Local anesthetic injected at this point will cause a superior
laryngeal nerve block.
cause of the impressive anterior and posterior mu-
cosal bruise that can be caused by the block and
regularly observed with the fiberoptic broncho-
scope after the block (personal observation).
As can be seen from the preceding list of tech-
niques, the one that can anesthetize everything
(naris, pharynx, larynx, trachea) is simply spraying
local anesthetic through the nose and the oral cav-
ity. In my experience, 0.5 per cent tetracaine with
epinephrine provides a more complete block than
4 per cent lidocaine. A very simple but effective
system that produces a very dense cloud or mist of
local anesthetic is shown in Figure 14-8. In my
experience, this system has made nebulization of
local anesthetic to all the mucosal surfaces the
single most effective local anesthetic maneuver
and can provide adequate anesthesia by itself
alone. It is extremely important for the anesthetist
to be unhurried and complete (10-sec spraying pe-
riods should be alternated with 10- to 20-sec rest
periods); this approach usually requires at least 15
min to result in adequate anesthesia. The total dose
of tetracaine over 15 min in a 70-kg patient should
not exceed 100 mg (20 ml of 0.5 per cent solu-
tion), and the total dose of lidocaine over 15 min
in a 70-kg patient should not exceed 400 mg (10
ml of 4 per cent solution) or 6 mg/kg. Nebulized
tetracaine appears to be more potent and to have a
significantly longer duration of action than lido-
caine. Table 14-4 shows the most important prop-
erties of the major agents used for topical anesthe-
sia of the respiratory tract.
It cannot be stressed enough that tracheal intu-
bation with a fiberoptic bronchoscope, is a difficult
procedure if anesthesia is inadequate (the operator
will have a rapidly and widely moving field of
vision), whereas the procedure is relatively easy if
the patient is adequately anesthetized (the operator
has a quiet field). Consequently, the author uses,
in addition to the local anesthetic spray of all the
mucosal surfaces (Fig. 14-8), cocaine to the nose
or soft nasopharyngeal airways coated with lido-
caine ointment (if the nose is going to be used),
pharyngeal gargle of viscous lidocaine, and, per-
haps, bilateral superior laryngeal nerve blocks.
The topical agents can cause local tissue irrita-
tion, and in susceptible patients (asthmatics) at-
tempted anesthesia of the respiratory tract may
cause bronchospasm; the bronchospasm is unre-
lated to airway histamine responsiveness.
23
Bron-
chospasm in these patients can be eliminated by
prophylactically administering an inhaled p
2
-ago-
nist.
24
If too large a dose of a topical anesthetic is
given too rapidly, absorption into the systemic cir-
culation may occur to a degree sufficient to pro-
duce toxic reactions. The initial effect is usually
one of central excitation, which may be manifested
by agitation, yawning, laughing, looking around
Figure 14-8
connected to an
the green tubing
of spread of the
This schematic diagram shows the system for creating a fine mist of local anesthetic. Oxygen green tubing is
oxygen tank. A hole is cut in the oxygen green tubing near the nebuhzation chamber. When oxygen is flowing in
and a finger is placed over the hole, a fine dense mist from the nebuhzation chamber results. The size and velocity
mist are proportional to the oxygen flow rate.
the room, nausea, vomiting, and twitching; higher
serum levels cause seizures. The risk of these toxic
central nervous system effects may be minimized
by premedicating the patient with small doses of a
benzodiazepine and/or a barbiturate. Other possi-
ble adverse effects include cardiovascular depres-
sion, hypotension, syncope, and arrhythmias.
Rarely, allergic or idiosyncratic reactions occur,
including rashes, eczema, edema, bronchospasm,
methemoglobinemia, and agranulocytosis. Hyper-
sensitivity reactions seem to be more common
with the ester type drugs (see Table 14-4).
It is apparent from Table 14-3 that there are
several other permutations of local anesthetic tech-
niques that may be used to anesthetize all of the
upper airway mucosal surfaces. However, no mat-
ter which technique or combination of techniques
is used, unless the anesthetist is thorough and pa-
tient and allows for a sufficient amount of time
(approximately 20 min) to achieve adequate anes-
thesia, any approach may result in spotty anesthe-
sia. With all techniques, intravenous lidocaine is
useful in depressing the cough reflex as well as
decreasing the incidence of arrhythmias.
25-27
b. GENERAL ANESTHESIA
It is important to realize at the outset that the
amount of general anesthesia required for bron-
choscopy can be greatly diminished if some topi-
cal local anesthesia is used. Again, the single most
effective maneuver that one can do to anesthetize
the entire upper respiratory passages is to nebulize
local anesthetic to all the upper airway mucosal
surfaces. Two types of general anesthesia, or a
combination of the two, may be used with a fiber-
optic bronchoscope. Either an oxygen-nitrous ox-
ide, intravenous barbiturate/narcotic, short-acting
muscle relaxant (succinylcholine drip, atracurium
or vecuronium), and/or a halogenated drug anes-
thesia technique may be used. The factors that
determine the choice between the two techniques
are extensively discussed in chapter 8. It is impor-
tant to emphasize that with either type of general
anesthetic adequate anesthesia to prevent laryngo-
spasm and bronchospasm is of paramount impor-
tance. Administration of intravenous lidocaine will
significantly depress the cough reflex and decrease
the incidence of premature ventricular contractions
during either type of general anesthesia.
25-27
Post-
operatively, the patients should breathe an elevated
F,0
2
for several hours (see complications).
17
28
4. Complications
Intraoperative laryngospasm and bronchospasm
due to inadequate anesthesia are the most common
intraoperative emergencies during fiberoptic bron-
choscopy.
29
30
Intraoperative hypoxemia is also
common
6
-
7
29
30
and can be a result of either poor
ventilation (due to bronchospasm or using a fiber-
optic bronchoscope that is too large) or atelectasis
(due to suctioning from a wedged position). Major
cardiac arrhythmias may develop in as many as 11
per cent of patients during fiberoptic bronchos-
copy, and the occurrence of the arrhythmias is
often associated with the presence of a P
a
0
2
< 60
mm Hg.
31
If hypoxemia is corrected, the arrhyth-
mias are usually self-limiting and do not need to
be treated. Deleterious hemodynamic effects of
PEEP and parenchymal barotrauma may result in-
traoperatively if the cross-sectional area of the en-
dotracheal tube is reduced greatly enough by the
fiberoptic bronchoscope, thereby preventing the
escape of expired gases. A chest X-ray should be
obtained postoperatively to rule out mediastinal
emphysema and pneumothorax if high airway pres-
sures were noted during the procedure.
3
8
A chest
X-ray does not appear indicated if the periopera-
tive period was completely uncomplicated.
32
33
Bleeding is an infrequent complication of fiberop-
tic bronchoscopy (may follow transbronchial bi-
opsy).
31
Hypoxemia develops during fiberoptic bron-
choscopy in both healthy volunteers and sick pa-
tients, with an average decline of the P
a
0
2
by 20
mm Hg, and lasts for 1 to 4 hours following the
procedure.
28
34-36
Hypoxemia occurring after fiber-
optic bronchoscopy is mainly due to atelectasis
caused by suctioning while the bronchoscope was
in a wedged position. Consequently, in patients
with an endotracheal tube, the anesthesiologist
should sigh the patient with positive pressure at
the end of the procedure. In addition, patients may
develop increased airway obstruction after fiber-
optic bronchoscopy,
28
probably secondary to direct
mechanical activation of cough and irritative re-
flexes in the airway and possibly by direct trauma-
induced mucosal edema.
17
Use of helium oxygen
mixtures can greatly decrease the resistance to
airflow.
9
I0
B. Rigid Ventilating Bronchoscope
1. Indications
There are several relative indications for the use
of a rigid ventilating bronchoscope. First, a venti-
lating bronchoscope has a hole near the end of the
bronchoscope that allows ventilation of the contra-
lateral lung when the tip of the bronchoscope is ir
the other lung; therefore, the ventilating broncho
scope permits bilateral lung ventilation when th<
tip of the bronchoscope is inside one of the main
stem bronchi. The ventilating bronchoscope ma;
be attached to the anesthesia circle system vi;
a sidearm (Fig. 14-9, bottom panel,
37
and Fig.
14-10) or to a jet injector for Venturi ventilation
(see Fig. 14-9, top panel
37
and Fig. 14-11). Se-
cond, it is the instrument of choice for foreign
body removal (even in broncholithiasis).
38
Third,
massive hemoptysis (500 ml per 24 hours) must
be assessed with an open-tube bronchoscope (see
chapter 17). Adequate suctioning, removal of
blood clots with a large forceps, and packing of a
major bleeding site can be much more readily ac-
complished with a rigid rather than a fiberoptic
bronchoscope. In addition, control of the airway
can be maintained during all these therapeutic ma-
neuvers. On rare occasion, it may be necessary to
position the bronchoscope in one main-stem bron-
chus to provide ventilation to one lung (and seal
off the bleeding lung) while the patient is taken to
the operating room. Fourth, constricting and/or
bleeding lesions may be by-passed by an open-
tube bronchoscope if the airway becomes seriously
Figure 14-9 The Storz adult bronchoscope. Top panel, With Sanders jet injector attachment. The anesthesia circle system
attachment is plugged. Bottom panel. With connection to anesthesia circle system. The jet injector attachment is plugged. (From
Vaughan RS: Endobronchial intubation. In Latto IP, Rosen M (eds): Difficulties in Tracheal Intubation. London. Bailliere Tindall.
Rigid Ventilating Bronchoscope
Figure 14-10 This schematic diagram shows a rigid ventilating bronchoscope system, which consists of the anesthesia circle
system attached to a flexible connector that is attached to the sidearm of the bronchoscope. With the proximal eyepiece in place,
most of the inspired gas goes into the patient. However, since the bronchoscope cannot fully fill the area of the trachea, there is a
variable leak around the distal end of the bronchoscope. Exhaled gases are through the anesthesia circle system. When the eyepiece
is removed, there is a very large leak out the proximal end of the bronchoscope.
compromised. Fifth, the open-tube bronchoscope
allows a large biopsy of a main bronchial neo-
plasm to be taken. Indeed, one report has de-
scribed using the rigid bronchoscope to core out
(by biopsy and by screwing the bronchoscope)
obstructing lesions in patients with terminal malig-
nancy.
39
However, the use of the rigid broncho-
scope in this manner has been condemned.
40
Sixth,
the rigid bronchoscope is necessary for the place-
ment of tracheobronchial tree stents.
41
Seventh, the
rigid bronchoscope allows the surgeon to palpate
the carina preoperatively to assess operability and
to identify extension of subcarinal disease. Finally,
the rigid bronchoscope is necessary for small chil-
dren (see chapter 18).
a. SIDEARM CONNECTION
The rigid ventilating bronchoscope (see Fig.
14-9, bottom panel, the Storz or Negus rigid bron-
choscope) may be attached directly to the anesthe-
sia machine circuit via a sidearm adapter (see Fig.
14-9, bottom panel, and Fig. 14-10). Conse-
quently, inspired oxygen and anesthetic gas con-
centrations are known and can be administered
with conventional or high-frequency positive-pres-
sure ventilation. Although spontaneous ventilation
may be allowed, the dangers of inadequate venti-
lation due to bronchospasm and a tight chest wall
(too light anesthesia), or hypoventilation (too deep
anesthesia) and tracheobronchial tree damage due
to unexpected coughing usually cause the risk/ben-
efit ratio of spontaneous ventilation to be high.
Consequently, these patients should usually be
paralyzed, and ventilation should be controlled
with positive pressure.
Because the usual rigid bronchoscope has an
external diameter of 11 mm, there is normally a
variable leak around the distal end of the broncho-
scope (with the proximal eyepiece in place), but
this may be compensated for by using a high gas
flow rate (greater than 10 L/min) (first choice) or
by packing the pharynx with saline-soaked gauze
Rigid Venturi Bronchoscope
Figure 14-11 This schematic diagram of a rigid Venturi bronchoscope shows that the jet of gas exiting from a Venturi needle
placed within the lumen and parallel to the long axis of the bronchoscope entrains gas from the environment. The jetted gas comes
from a high-pressure source and an intermittent (I2 L/min) injector. The flow of gas from the tube into the patient is equal to the
volume of gas through the jet plus the air entrained.
(second choice). It is possible to provide effective
ventilation in the vast majority of patients with
this system. However, the proximal eyepiece must
be removed during suctioning, foreign body ma-
nipulations, or the taking of biopsy specimens.
Since ventilation must be interrupted when the sur-
geon removes the occluding eyepiece (all the in-
spired gas including the inhalational anesthetics,
which can sedate the endoscopist, go into the room
out the open proximal end), a high F,0
2
should be
used if the surgeon requires removal of the eye-
piece for a considerable period of time. One to 2
min maximum of apnea may be allowed at any
one time before ventilation must be resumed, with
a shorter time allowed for obese patients and pa-
tients with lung disease.
42
b. VENTURI CONNECTION
The rigid Venturi-effect bronchoscope relies on
an intermittent (10-20 L/min) high-pressure oxy-
gen jet to entrain air and ventilate the lungs with
an air-oxygen mixture (see Fig. 14-9, top panel,
and Fig. 14-11). The Venturi jet is delivered via a
reducing valve (Sanders injector) to a 1.0- to 1.5-
inch, 18- or 16-gauge needle that is inside and
parallel to the lumen of the bronchoscope. The
high velocity of the Venturi jet exiting from the
end of the needle creates a negative pressure just
outside the open end of the needle that draws en-
vironmental air into the jet stream.
Oxygen jets (from a 50-psi source) usually en-
train two to three times their own volume of am-
bient air.
43
The Venturi jet plus the entrained air
creates a positive intraluminal tracheal pressure
and a tidal volume. At any given reducing valve
pressure, the exact amount of tracheal pressure and
tidal volume depends on the driving pressure from
the reducing valve, the size of the needle, and the
diameter, length, and type of bronchoscope. With
the usual Venturi bronchoscope and an 18-gauge
needle orifice, a 50-psi source causes a flow of 160
L/min and a peak airway pressure of 27 cm H-,0.
44
The jet reducing-valve pressure can be adjusted
according to the tracheal pressure and observed
chest movements; usually a tracheal pressure of 30
cm H
2
0 results in normocarbia. In comparison
with the intermittent gas exchange often required
by the rigid ventilating bronchoscope, the rigid
Venturi bronchoscope provides more constant and
adequate ventilation.
45
Use of high-frequency jet
ventilation at rates of 150 to 300 breaths/min
(using a commercial high-frequency ventilator)
through the rigid Venturi bronchoscope has re-
sulted in adequate gas exchange (but still less ef-
fective than jet ventilation).
46
a. LOCAL ANESTHESIA
Passing a rigid bronchoscope into the trachea
requires a considerable amount of pressure, mu-
cosal stimulation, and significant neck extension
that is not required for fiberoptic bronchoscopy. In
addition, there is no guarantee in an awake patient
that the patient will not suddenly move at a critical
point as a result of some noxious or unpleasant
stimulus. Consequently, rigid bronchoscopy is
usually performed under general anesthesia. If lo-
cal anesthesia must be used, the premedication and
various local blocks, as previously described for
fiberoptic bronchoscopy and used collectively, will
usually render rigid bronchoscopy tolerable.
b. GENERAL ANESTHESIA
Bronchospastic and asthmatic patients must
have good pharmacologic control of bronchomotor
tone before rigid bronchoscopy. General anesthe-
sia for rigid bronchoscopy should also be preceded
by some topical local anesthesia of the larynx.
Thus, less general anesthesia will be required, per-
mitting easier, more rapid awakening and return of
laryngeal reflexes. Both nitrous oxide/intravenous
anesthesia/short-acting muscle relaxant and halo-
genated drug/short-acting muscle relaxant anesthe-
sia may be used with ventilating bronchoscopes.
Intravenous and intratracheal lidocaine with either
type of general anesthetic minimizes straining,
coughing, and arrhythmias.
25-27
Of course, the choice of anesthesia will also be
based on the overall medical status of the patient
and the skill and speed of the surgeon. Because
the Venturi principle renders the inspired oxygen
concentration and inhalational anesthetic concen-
tration uncertain, an oxygen-narcotic-thiopental-
muscle relaxant general anesthetic is most often
used for bronchoscopy with a Venturi broncho-
scope. Paralysis is a helpful part of the Venturi
technique because a compliant thorax with mini-
mal resistance is sometimes necessary to ensure
adequate ventilation.
4. Complications
Experiencing the passing of a rigid broncho-
scope can be very unpleasant for an awake patient.
The rigid bronchoscope may fracture teeth, and it
is occasionally necessary to use extreme extension
of the neck to insert the bronchoscope, which may
cause vasovagal reactions. Massive hemorrhage
from directly traumatized lesions may occur.
30
33
The tip of the rigid bronchoscope can perforate the
mucosa, causing pneumomediastinum and subcu-
taneous emphysema. Because a large leak may
occur around the distal end of the bronchoscope,
which may prevent adequate ventilation, and peri-
ods of apnea may be required, there is an increased
risk of hypoxemia and hypercarbia.
30
33
In view of
the inherent intense stimulation associated with the
procedure, arrhythmias (especially if either hypox-
emia, hypercarbia, inadequate anesthesia, or halo-
thane anesthesia is present) may occur.
30
33
In addition, there are several potential special
disadvantages in using the rigid Venturi broncho-
scope. First, there is no continuous documentation
of delivered oxygen and inhaled anesthetic drug
concentration. Second, the high gas flow rates re-
sulting from the jet ventilation may cause blood or
tumor particles to be accidentally blown down into
more peripheral bronchi. Third, care must be taken
not to generate excessive carinal airway pressures
in children owing to a tight glottic fit, which may
prevent escape of gas (exhalation) around the
bronchoscope.
III. MEDIASTINOSCOPY
Considerations
Mediastinoscopy is commonly performed be
fore thoracotomy to establish a diagnosis and/or t<
determine resectability of a lung carcinoma
47
am
to establish a correct diagnosis in patients sus
pected of having a lymphoma and in those with
mediastinal mass.
48
The importance of determinin
cell type is most simply and dramatically undei
scored by pointing out the example that lymphe
mas require radiation, whereas thymomas requir
resection.
After a suprasternal notch incision (cervical mt
diastinoscopy), a tunnel is created (through th
pretracheal fascia) by blunt dissection along th
anterior and lateral walls of the trachea into th
mediastinum, behind (posterior to) the aortic arcl
and down to the subcarinal area (Fig. 14-12). Th
procedure allows for direct inspection and biops
of the superior mediastinal lymph nodes, which 1:
posterior to the aortic arch (i.e., the anterior ar
lateral para-main-stem bronchial, anterior subcai
nal, anterior, and lateral paratracheal lymph node
(see Fig. 14-12). Left-sided central and apical 1<
sions are much more easily observed via a le
Anesthesia for Special Elective Diagnostic Procedures
505
Mediastinoscopy
Mediastinoscope
Right Common
Carotid
Right Subclavian
Artery
Innominate
Artery (pinched)
Anterior and Lateral
Para-Tracheal
and Para-Mainstem
Bronchial Nodes
Left Recurrent
Laryngeal
Nerve
Esophagus
Trachea
Left
Subclavian
Artery
Left Common
Carotid Artery
Anterior Subcarinal
Lymph Nodes
Superior
Vena Cava
Figure 14-12 This schematic diagram shows the placement of a mediastinoscope into the superior mediastinum. The mediasti-
noscope passes in front of the trachea but behind the thoracic aorta. This location of the mediastinoscope allows for sampling of
anterior and lateral para-mainstem bronchial lymph nodes, anterior subcarinal lymph nodes, and anterior and lateral para-tracheal
lymph nodes. Anatomic structures that can be compressed by the mediastinoscope (see areas marked by an *) and that can cause
major complications are the thoracic aorta (rupture, reflex bradycardia), innominate artery (decreased right carotid blood flow can
cause cerebral vascular symptoms, and decreased right subclavian flow can cause loss of right radial pulse), trachea (inability to
ventilate), and vena cava (risk of hemorrhage with superior vena cava syndrome).
anterior approach (this anterior mediastinotomy
procedure uses a second rib interspace incision and
is ordinarily a much less complicated procedure
than mediastinoscopy but not always).
49
Tumors of the thymus and anterior mediastinum
are also not examined by the usual diagnostic cer-
vical mediastinoscopic approach because they are
anterior to the great vessels; thus, an anterior me-
diastinotomy is also required to examine this area.
Table 14-5 summarizes one author's approach to
mediastinoscopy for lung regions.
47
A parasternal
mediastinoscopy can also be used for the evalua-
tion of left upper lobe lesions.
50
Contralateral positive nodes (to a lung carci-
noma) are considered an absolute contraindication
to thoracotomy, whereas if only ipsilateral nodes
are involved, thoracotomy may be performed de-
pending on the expected resectability of the tumor.
Previous mediastinoscopy is a strong relative con-
traindication to a repeat procedure because scar-
ring eliminates the plane of dissection. "Strong
relative contraindication" is used rather than the
classical "absolute" term because two studies/re-
views of 101
51
and 140
52
repeat mediastinoscopies
found a 0 per cent mortality rate and 23 per cent
and 7 per cent complication rate, respectively, all
of which were successfully treated. Relative con-
traindications to mediastinoscopy include superior
vena cava syndrome, severe tracheal deviation,
cerebrovascular disease, and thoracic aortic aneu-
rysm.
53
With the advent of computed tomography,
and magnetic resonance imaging, the role of me-
diastinoscopy may diminish considerably in the
future in the sense that negative scans may obviate
the need for invasive staging.
In addition to the usual preanesthetic evaluation
of patients, one should specifically look for the
Table 14-5 ACTUAL GUIDELINES FOR THE
USE OF MEDIASTINOSCOPY FOR
DIAGNOSIS OF LUNG LESIONS*
No mediastinoscopy
Peripheral squamous cell carcinoma
Undiagnosed peripheral nodule <3 cm
Cervical mediastinoscopy
Right-sided centrally located squamous cell carcinoma
Every undifferentiated carcinoma and adenocarcinoma
(except for left upper lobe and left central tumors)
Anterior mediastinoscopy
Left upper lobe tumors
Left-sided centrally located tumors (when anterior
mediastinoscopy is negative, a cervical approach should
be performed)
*From Van Schil PEY, Van Hee RHGG, Schoofs ELG: The
value of mediastinoscopy in preoperative staging of broncho-
genic carcinoma. J Thorac Cardiovasc Surg 97:240-244, 1989.
Used with permission.
signs and symptoms of the relative contraindica-
tions to mediastinoscopy, such as obstruction or
distortion of the upper airway and superior vena
caval obstruction, signs and symptoms of impaired
cerebral circulation (which may be compounded
during mediastinoscopy by compression of carotid
vessels), and evidence of the myasthenic syndrome
due to lung carcinoma. Since there is the potential
risk of hemorrhage, a large-bore intravenous cath-
eter should be inserted and blood should be im-
mediately available during mediastinoscopy. If the
superior vena cava is obstructed, a lower extremity
intravenous catheter is highly desirable.
Although mediastinoscopy can be performed
under local anesthesia,
54
55
general anesthesia with
controlled positive-pressure ventilation is preferred
because it allows the surgeon more flexibility in
his or her dissection, minimizes the potential for
air embolus (see the following), and facilitates
management of major complications such as mas-
sive hemorrhage.
56
However, local anesthesia may
be considered for patients with active cerebrovas-
cular disease to monitor cerebral function contin-
uously in the awake state.
Following the intravenous induction of general
anesthesia and paralysis with either succinylcho-
line, atracurium, or vecuronium, lidocaine is
sprayed directly on the trachea and given intrave-
nously in order to minimize coughing during oro-
tracheal intubation and during the procedure itself.
A nonkinking tube should be considered in cases
in which tracheomalacia is a possibility. A short-
acting muscle relaxant (succinylcholine drip, atra-
curium, vecuronium) is also used during the pro-
cedure to prevent coughing, venous engorgemem
from straining, and movement during the proce-
dure.
51
The head-up position minimizes the venous
engorgement but maximizes venous air embolism
The choice of anesthetic agent used is dtermine
by the patient's overall medical condition (set
chapter 8). During mediastinoscopy the anesthe
siologist's attention is primarily focused on detect
ing the occurrence of the complications of mdias
tinoscopy. This is performed by palpating the righ
radial pulse (vessels most commonly compresse
are the innominate and right subclavian and ca
rotid arteries) and by observing for reflex brady
cardia (compression of the aorta), arrhythmia
(mechanical stimulation of the aorta, ventricles)
hypovolemia, tension pneumothorax, and com
pression of trachea (see the following). Blooi
pressure and oxygen saturation should be meas
ured using the left arm; if an arterial line is placed
the right wrist may be preferable because it ca
immediately signal when compression of the in
nominate artery has occurred. Immediately follow
ing the procedure patients can usually be extu
bated. The patient should be nursed in a head-u
position to minimize venous engorgement.
The anesthetic considerations for anterior medi-
astinotomy are very similar to those for mediasti-
noscopy except that the incision is larger, the
incidence of complications is lower because struc-
tures can be visualized and controlled more read-
ily, the position of the head is unimportant, and
the blood pressure can be measured on either arm
because there is little chance of compression of the
innominate artery.
C. Compl i cati ons
Although overall mortality from the procedure
is low (0.1 per cent),
57
and even one institution has
undertaken ambulatory (outpatient) mediastinos-
copies in a hospital-based surgical suite,
58
serious
complications can occur that the anesthesiologist
must be prepared to diagnose and to treat (Table
14-6). In a series of 6490 patients undergoing
mediastinoscopy, the major complications reported
(numbers of patients in parentheses) consisted of
hemorrhage (48), pneumothorax (43), recurrent la-
ryngeal nerve injury (22), infection (22), tumor
implantation in the wound (8), phrenic nerve in-
jury (3), esophageal injury (1), chylothorax (1), air
embolism (1), and transient hemiparesis (l).
57
The
overall complication rate in various studies has
been 1.5 to 3.0 per cent.
57
59
~
67
At the time of
occurrence, most of these complications required a
specific anesthetic management response.
Significant (occasionally massive) hemorrhage
has been the most frequent major problem encoun-
tered during mediastinoscopy. If it occurs, thora-
cotomy must be performed immediately; while
these rapid preparations are in progress, the sur-
geon should attempt to control hemorrhage by
compressing the bleeding site with a sponge for-
ceps or a small pack, although the relative inacces-
sibility of the operative field may make this ma-
neuver difficult or ineffective. The anesthesiologist
should (1) rapidly begin volume replacement
through one (or more) large-bore intravenous can-
nulas that have been placed prior to the induction
Table 14-6 MAJOR COMPLICATIONS OF
MEDIASTINOSCOPY
1. Hemorrhage
2. Pneumothorax
3. Recurrent laryngeal nerve injury
4. Air embolism
5. Compression of vessels
a. Aorta reflex bradycardia
b. Innominate artery
Right carotid hemiparesis
Right subclavian loss of right radial pulse
6. Compression of trachea
7. Infection, tumor spread
of anesthesia; (2) send for blood that was reserved
for the patient preoperatively; (3) support the cir-
culation pharmacologically until volume replace-
ment is achieved; (4) ensure adequate oxygenation
and ventilation; (5) administer atropine for reflex
bradycardia from aortic compression (if it occurs);
and (6) discontinue or reduce the dose of all anes-
thetic drugs until normovolemia is reestablished.
Rarely, it may be necessary to induce deliberate
hypotension to control bleeding in this setting.
60
Should hemorrhage originate from a superior vena
cava tear, volume replacement and drug treatment
may be lost into the surgical field unless they are
administered via a peripheral intravenous line
placed rapidly in the lower extremity.
60
Pneumothorax is another relatively frequently
encountered complication of mediastinoscopy. It is
usually not apparent until the postoperative period,
and the majority of patients do not require chest
tube decompression.
57
All patients should be mon-
itored for signs of pneumothorax in the postoper-
ative period and a chest roentgenogram obtained
when doubt exists. Pneumothorax that occurs in-
traoperatively, as evidenced by increased peak in-
spiratory pressure, tracheal shift, distant breath
sounds, hypotension, and cyanosis, requires im-
mediate treatment by chest tube decompression.
65
When mediastinoscopy causes a recurrent laryn-
geal nerve injury (see Fig. 14-12) it is permanent
in approximately 50 per cent of patients.
57
If injury
to the recurrent laryngeal nerve is suspected, the
vocal cords should be visualized while the patient
is spontaneously breathing (usually at the time of
extubation). If the vocal cords are nonmoving and/
or are in a midline position, the problem of post-
operative laryngeal obstruction must be consid-
ered.
During mediastinoscopy the mediastinoscope
tip is located intrathoracically and therefore di-
rectly exposed to pleural pressure. Venous air em-
bolism (when venous bleeding is present) can oc-
cur much more easily if patients are breathing
spontaneously because of the development of neg-
ative intrathoracic pressure during inspiration;
therefore, controlled positive-pressure ventilation
during this procedure minimizes the risk of air
embolism.
The mediastinoscope can exert pressure against
the innominate artery and result in diminished
blood flow to the right carotid and right subclavian
arteries (see Fig. 14-12). This phenomenon may
be of special significance in patients with pre-ex-
isting compromised cerebral circulation. Compres-
sion of the right carotid artery has been proposed
as the cause of a left hemiparesis that occurred in
one patient and that subsequently cleared 48 hours
after the procedure.
57
In another patient compres-
sion of the right subclavian artery caused the loss
of the pulse and blood pressure in the right arm
and was misdiagnosed as an intraoperative cardiac
arrest.
66
In another study, blood pressure in the
right arm was significantly decreased for periods
of 15 to 350 sec in four of seven patients who
underwent mediastinoscopy.
67
This last report,
therefore, recommended that blood pressure be
measured in the left arm and that the right radial
artery be continuously monitored by palpation or
finger plethysmography during mediastinoscopy.
A right radial arterial line, of course, would very
sensitively and continuously monitor the occur-
rence of innominate or right subclavian artery
compression. An oxygen saturation monitor would
do this task less sensitively. Any decrease in right
radial artery pressure requires repositioning of the
mediastinoscope, especially in patients with cere-
bral vascular insufficiency. Avoiding excessive ex-
tension of the neck vessels, which might contrib-
ute to pinching of neck vessels, is also important
in this group of patients.
Autonomic reflexes may occur as a result of
compression or stretching of the trachea, vagus
nerve, or great vessels. Sudden changes in pulse
and/or blood pressure during mediastinoscopy may
initially be empirically treated by repositioning of
the mediastinoscope. Atropine is given for persist-
ent bradycardia.
IV. THORACOSCOPY
Considerations
Diagnostic thoracoscopy (pleuroscopy) permits
a valuable examination of the thorax and is most
commonly performed to aid in the diagnosis of
pleural and parenchymal disease, to help establish
the staging of suspected neoplasms, and, most im-
portantly and frequently, to determine the cause of
recurrent pleural effusions.
68-74
Approximately 5 to
10 per cent of patients with lung cancer present
with an ipsilateral pleural effusion; pleuroscopy
should always precede thoracotomy in otherwise
operable patients with bronchogenic carcinoma
and ipsilateral cytologically negative pleural effu-
sion.
74
Thoracoscopy will obtain a definite diag-
nosis in 90 per cent of these patients.
74-76
Diagnos-
tic thoracoscopy is most often done after
thoracocentesis or closed-chest pleural or lung bi-
opsy has been performed and a diagnosis has still
not been established.
The development of endoscopic video systems
and instrumentation (stapling devices, dissectors,
coagulators, autotyping sutures, and the neodym-
ium:yttrium aluminum garnet laser has permitted
therapeutic thoracoscopy for a wide variety of ma-
jor thoracic procedures. These procedures include
biopsy of many intrathoracic structures, peripheral
wedge and sublobar resection, lobectomy, removal
of mediastinal cysts, closure of persistent/recurrent
pneumothorax and leaking blebs, pleurodesis, dorsal
thoracic sympathectomy, drainage of spinal abscess,
resection of posterior mediastinal neurogenic tu-
mor, retrieval of intrathoracic (pleural) foreign
bodies, definitive treatment of postpneumonec-
tomy chylothorax, and pericardectomy.
73,77-82
Diagnostic thoracoscopy is done by making a
small incision (2-3 cm) in the lateral thoracic wall
(usually at the level of the sixth intercostal space
and usually with the patient in the lateral decubitus
position), digitally examining the pleural space,
and then introducing a thoracoscope, laparoscope,
or mediastinoscope into the pleural cavity. The
procedure allows for complete inspection of the
pleural space of a hemithorax and diaphragm (es-
pecially if a double-lumen tube and one-lung ven-
tilation are used). Fluid and biopsies can usually
be easily obtained through the incision, although
many surgeons prefer to use a separate incision for
additional instruments such as biopsy forceps (in-
cluding cryobiopsy), staplers, and coagulators (see
discussion of therapeutic thoracoscopy immedi-
ately following).
Figure 14-13 shows the surgical approach to
therapeutic thoracoscopy. Depending on the type
of operation, usually three incisions ranging from
15 mm to 35 mm are necessary. One incision is
made in the sixth or seventh intercostal space in
the midaxillary line for insertion of the scope, and
the second and third incisions are made in the third
to the sixth intercostal spaces along the anterior
and posterior axillary lines. The latter two inci-
sions are used for insertion and manipulation of
instruments. The instruments consist of lasers, cau-
tery, cryoprobes, staplers, retractors, and graspers.
The hemostatic properties and the quality of the
pleural and lung biopsies obtained using the cry-
oprobe are found to be useful by some.
83
At the conclusion of the procedure, the chest
may be irrigated with saline, the lung expanded,
and a chest tube placed through the pleuroscope
introduction site (and perhaps through an instru-
ment site as well). If diagnostic thoracentesis is
positive for malignant cells, palliative treatment
(chemotherapy, radiation therapy, or both) only is
indicated.
74
Thoracoscopy can be done with either local (es-
pecially if only pleural exploration is to be per-
formed),
76
8 3
8 4
regional, or general anesthesia with
one-lung ventilation (especially if therapeutic thor-
Figure 14-13 Incisions for thora-
coscope and instruments. The instru-
ments may consist of staplers, lasers,
cautery, cryoprobe, retractors, and
graspers. (From Lewis RJ, Caccavale
RJ, Sisler GE, Mackensie JW: One
hundred consecutive patients un-
dergoing video-assisted thoracic op-
erations. Ann Thorac Surg 54:421-
acoscopy is to be performed).
73
74 7 7
~
8 2
8 5
86
The
choice of anesthesia is determined by considering
(1) the expected extent and length of the proce-
dure; (2) the gentleness and technical abilities of
the surgeon; (3) the expertise of the anesthesiolo-
gists; and (4) the physical condition and psycho-
logical state of the patient.
Local anesthetic infiltration of the lateral tho-
racic wall and parietal pleura is the simplest way
to provide anesthesia for the patient,
87
although
some patients may experience considerable dis-
comfort when this method is used. Partial collapse
of the lung on the operated side occurs when air
enters the pleural cavity. This allows for good
visualization of the pleural space by the surgeon.
It is both hazardous and unnecessary to insufflate
gases under pressure into the hemithorax in ques-
tion in order to increase visualization of the pleural
space. Surprisingly, even though many of these
patients suffer from advanced pulmonary disease,
changes in P
a
0
2
, P
a
C0
2
, and cardiac rhythm are
usually minimal during the procedure when it is
performed under local anesthesia and the patient is
breathing spontaneously.
88
89
However, it is pru-
dent to use a high F,0
2
to overcome the loss in
lung volume caused by the unavoidable pneumo-
thorax.
Intercostal nerve blocks performed at the level
of the incision and at two interspaces above and
below may provide more complete analgesia for
thoracoscopy, especially if they are placed far
enough posteriorly to anesthetize the parietal
pleura. The addition of an ipsilateral stellate gan-
glion block helps to prevent the cough reflex that
is sometimes elicited during visualization and ma-
nipulation of the hilum.
When general anesthesia is performed for diag-
nostic and therapeutic thoracoscopy, a double-lumen
endotracheal tube should be used. Positive-pres-
sure ventilation seriously interferes with visualiza-
tion of thoracic contents (although, on occasion,
positive-pressure ventilation affords the possibility
of freeing trapped lung and to determine whether
apposition of the pleural surfaces will be possi-
ble,
85
and diagnostic and therapeutic thoracoscopy
is therefore a relative strong indication for one-
lung ventilation.
73
74
77
~
82
85
86
If the procedure is
short and the ipsilateral lung needs to be deflated
for only a brief period, blood gases are not rou-
tinely required during the procedure. However, for
patients with marginal pulmonary status in whom
the period of one-lung anesthesia lasts for more
than a few minutes, the usual monitoring precau-
tions should be taken, as discussed in chapter 7.
C. Complications
Complications are very rare during this simple
procedure, although it is possible that any structure
that the surgeon has to manipulate may be dam-
aged. A few critically ill, spontaneously breathing
patients may experience impaired gas exchange
during the procedure. With the availability of ade-
quate coagulation techniques, the most common
complication is tumor seeding at the thoracostomy
site. Tumor seeding is treated effectively with su-
perficial beam radiation.
REFERENCES
1. Landa JF: Indications for bronchoscopy. Chest 73:686-
690, 1978.
2. Barrett CR: Flexible fiberoptic bronchoscopy in the criti-
cally ill patient. Chest 73:746-749, 1978.
3. Raj PP, Forestner J, Watson RD, et al: Techniques for
fiberoptic laryngoscopy in anesthesia. Anesth Analg
53:708-714, 1974.
4. Snider GL: When not to use the bronchoscope for hemop-
tysis. Chest 76:1-2, 1979.
5. Simelaro JP, Marks B, Meals R, et al: Selective bronchog-
raphy following fiberoptic bronchoscopy. Chest 70:240-
241, 1976.
6. Lampton LM: Bronchoscopy: Caution! JAMA 73:138,
1978.
7. Miller EJ: Hypoxemia during fiberoptic bronchoscopy.
Chest 75:103, 1979.
8. Lindholm CE, Oilman B, Snyder JV, et al: Cardiorespira-
tory effect of flexible fiberoptic bronchoscopy in critically
ill patients. Chest 74:362-368, 1978.
9. Pingleton SK, Bone CR, Ruth WC: Helium-oxygen mix-
tures during bronchoscopy. Crit Care Med 8:50-53, 1980.
10. Fleming MD, Weigelt JA, Brewer V, Mclntire D: Effect
of helium and oxygen on airflow in a narrowed airway.
Arch Surg 127:956-960, 1992.
11. Benumof JL: Management of the difficult airway: With
special emphasis on the awake intubation. Anesthesiology
75:1087-1110,1991.
12. Benumof JL: Use of the laryngeal mask airway to facilitate
fiberoptic-endoscopy intubation. Anesth Analg 74:313-
314, 1992.
13. Benumof JL: The laryngeal mask airway: Indication and
contraindications. Anesthesiology 77:843-846, 1992.
14. Brimacombe J, Swainston R, Thompson J: A potential new
technique for awake fiberoptic bronchoscopyUse of the
laryngeal mask airway. Med J Aust 156:876-877, 1992.
15. Khoo ST: Anaesthesia for fiberoptic bronchoscopy in chil-
dren. Anaesthesia 45:248-249, 1990.
16. Ravussin P, Bayer-Berger M, Monnier , Savary M, Free-
man J: Percutaneous transtracheal ventilation for laser en-
doscopic procedures in infants and small children with
laryngeal obstruction: Report of two cases. Can J Anaesth
34:83-86, 1987.
17. Neuhaus A, Markowitz D, Rotman HH, et al: The effects
of fiberoptic bronchoscopy with and without atropine pre-
medication on pulmonary function in humans. Ann Thorac
Surg 25:393-398, 1978.
18. Belen J, Neuhaus A, Markowitz D, et al: Modification of
the effect of fiberoptic bronchoscopy on pulmonary me-
chanics. Chest 79:516-519, 1981.
19. Woods AM, Lander CJ: Abolition of gagging and hemo-
dynamic response to awake laryngoscopy. Anesthesiology
67:A220, 1987.
20. Cooper M, Watson RL: An improved regional anesthetic
technique for peroral endoscopy. Anesthesiology 43:273-
374, 1975.
21. Gotta AW, Sullivan CA: Anaesthesia of the upper airway
using topical anaesthetic and superior laryngeal nerve
block. Br J Anaesth 53:1055-1057, 1981.
22. Webb AR, Fernando SSD, Dalton HR, Arrowsmith JE,
Woodhead MA, Cummin ARC: Local anaesthesia for fi-
breoptic bronchoscopy: Transcricoid injection or the
"spray as you go" technique? Thorax 45:474^177, 1990.
23. Kirkpatrick MB: Lidocaine topical anesthesia for flexible
bronchoscopy. Chest 96:965-967, 1989.
24. McAlpine LG, Thomson NC: Lidocaine-induced broncho-
constriction in asthmatic patients: Relation to histamine
airway responsiveness and effect of preservative. Chest
96:1012-1015,1989.
25. Christensen V, Ladegaard-Pedersen HJ, Skovsted P: Intra-
venous lidocaine as a suppressant of persistent cough
caused by bronchoscopy. Acta Anaesth Scand 67(Suppl):
84-86, 1978.
26. Elguindi AS, Harrison GN, Abdulla AM, et al: Cardiac
rhythm disturbances during fiberoptic bronchoscopy: A
prospective study. J Thorac Cardiovasc Surg 77:557-561,
1979.
27. Luck JC, Messeder OH, Rubenstein MJ: Arrhythmias from
fiberoptic bronchoscopy. Chest 74:139-143, 1978.
28. Salisbury BG, Metzger CF, Altose MD, et al: Effect of
fiberoptic bronchoscopy on respiratory performance in pa-
tients with chronic airway obstruction. Thorax 30:441-
446, 1975.
29. Suratt PM, Smiddy JF, Gruber B: Deaths and complica-
tions associated with fiberoptic bronchoscopy. Chest
69:747-751, 1976.
30. Lukomsky GI, Ovchinnikov AA, Bilal A: Complications
of bronchoscopy: Comparison of rigid bronchoscopy under
general anesthesia and flexible fiberoptic bronchoscopy un-
der topical anesthesia. Chest 79:316-321, 1981.
31. Shrader DL, Lakshminararyan S: The effect of fiberoptic
bronchoscopy on cardiac rhythm. Chest 73:821-824, 1978.
32. Milam MG, Evins AE, Sahn SA: Immediate chest roent-
genography following fiberoptic bronchoscopy. Chest
96:477^179, 1989.
33. Zavala DC: Flexible Fiberoptic Bronchoscopy: A Training
Handbook. Iowa City, University of Iowa Publications De-
partment, 1978.
34. Zavala DC: Complications following fiberoptic bronchos-
copy: The "good news" and the "bad news." Chest
73:783-785, 1978.
35. Albertini RE, Harrell JH, Kurihara N, et al: Arterial hypox-
emia induced by fiberoptic bronchoscopy. JAMA
230:1666-1667, 1974.
36. Dubrawsky C, Awe RJ, Jenkins DE: The effect of bron-
chofiberoptic examination on oxygen status. Chest 67:137-
140, 1975.
37. Vaughan RS: Endobronchial intubation. In Latto IP, Rosen
CHAPTER 15
Anesthesia for Special Elective
Therapeutic Procedures
I. Introduction a. Central Congenital
II. Laser Resection of Major Airway Intrapulmonary
Obstructing Tumors Bronchogenic Cyst
A. General Considerations b. Peripheral Congenital Cysts
B. Various Types of Laser and of the Lungs
Adjunctive Therapies c. Congenital Cystic
1. Nd-YAG Laser Resection Adenomatoid Malformation
a. General Considerations d. Bronchopulmonary
b. Anesthetic Considerations Sequestration
c. Ventilation/Bronchoscopy 2. Traumatic Lung Cyst
Systems 3. Hydatid Cyst of the Lung
d. Suggested Anesthesia and 4. Pneumatocele
Ventilation Technique for 5. Bullous Emphysema
Laser Resection of Airway B. Anesthetic Considerations
Tumors VI. Pulmonary Resection in Patients
2. Hematoporphyrin Derivative/ After Pneumonectomy
Laser-Beam Photodynamic A. General Considerations
Therapy of Bronchogenic B. Anesthetic Considerations
Carcinoma VII. Unilateral Bronchopulmonary
3. Carbon Dioxide Laser Lavage
a. General Considerations A. General Considerations
b. Anesthetic Considerations B. Anesthetic Considerations
4. Adjunctive Endobronchial VIII. Pulmonary Arteriovenous
Radiotherapy Malformations
a. General Considerations A. General and Surgical
b. Anesthetic Considerations Considerations
5. Other Adjunctive B. Anesthetic Considerations
Bronchoscopic/Laser Therapies IX. Lung Transplantation
of Airway Strictures A. General and Surgical
a. Cryotherapy of Strictures Considerations
b. Balloon Catheter Dilation of 1. Types of Lung Transplantation
Strictures and Their Indications
III. Tracheal Resection 2. Patient Evaluation and
A. General Considerations Selection Criteria
B. Anesthetic Considerations 3. Donor Organ
1. Orotracheal Intubation Considerations
2. Insertion of Tube Into Opened 4. Surgical Operation
Trachea Distal to Area of B. Anesthetic Considerations
Resection 1. Anesthetic Drugs
3. High-Frequency Jet Ventilation 2. Lung Separation
4. High-Frequency Positive- 3. Monitoring
Pressure Ventilation 4. Major Intraoperative Anesthetic
5. Cardiopulmonary By-Pass Problems
IV. Broncholithiasis a. Initiation of One-Lung
A. General and Surgical Ventilation
Considerations b. Clamping of the Pulmonary
B. Anesthetic Considerations Artery
V. Giant Bullous Emphysema and Air c. Perfusion of the
Cysts Transplanted Lung Without
A. General Considerations Ventilation
1. Developmental Lung Cysts C. Postoperative Considerations
514 Anesthesia for Special Elective Therapeutic Procedures
X. Tumors at the Confluence of the
Superior, Anterior, and Middle
Mediastina
B. Compression of the
Tracheobronchial Tree
1. Anesthetic Management
C. Compression of the Pulmonary
Artery and Heart
D. Superior Vena Cava Syndrome
XI. Repair of Thoracic Aortic
Aneurysms
XII. Thymectomy for Myasthenia Gravis
I. INTRODUCTION
This chapter describes the anesthetic manage-
ment of special elective thoracic surgery proce-
dures. Each of these special thoracic surgery pro-
cedures involves anesthetic considerations that are
distinct from those in all other types of thoracic
surgery. The presentation of the special cases is
organized according to anatomic location; the
chapter considers diseases and problems of the
proximal airway to the alveolar space (laser resec-
tion of airway lesions, tracheal resection, treatment
of broncholithiasis, giant bullous emphysema and
air cysts, progressive serial resections, lung lavage,
resection of pulmonary arteriovenous malforma-
tion, and lung transplantation), resection of me-
diastinal masses, repair of thoracic aortic aneu-
rysms, and, finally, thymectomy for myasthenia
gravis, one-lung ventilation in morbidly obese pa-
tients, and resection of superior sulcus tumors. An-
esthesia for esophageal surgery is presented in
chapter 16. Thoracic surgery procedures that do
not have special anesthetic considerations (beyond
such usual considerations as one-lung ventilation
and massive blood loss, which have already been
presented [see section III]) are not discussed here.
II. LASER RESECTION OF MAJOR
AIRWAY OBSTRUCTING TUMORS
Malignant tumors (bronchogenic and metastatic)
of the tracheobronchial tree progressively obstruct
major conducting airways, resulting in slow as-
phyxiation of the patient. Treatment with external
radiation therapy and chemotherapy requires sev-
eral weeks to result in temporary resolution of the
obstructive lesions, and sometimes the progressive
obstruction cannot be reversed by this means. For-
tunately, in recent years, the technology to permit
B. Anesthetic Considerations
1. Preoperative Considerations
2. Intraoperative and
Postoperative Considerations
XIII. One-Lung Anesthesia in Morbidly
Obese Patients
XIV. Thoracic Outlet Syndromes
A. General and Surgical
Considerations
1. Benign Causes
2. Superior Sulcus (Pancoast)
Tumor
passage of laser ("laser" is an acronym for light
amplification by simulated emission of radiation)
energy to ablate and cause necrosis of these pre-
viously inoperable airway tumor obstructions has
been developed. Creating a channel through a
completely obstructing lesion or widening an ex-
isting channel that is partially obstructed by a le-
sion has resulted in a great improvement in symp-
toms, an increase in exercise tolerance, improved
pulmonary function test results (e.g., Fig. 15-1),
an increase in P
a
0
2
, a sense of well-being, and an
improvement in the appearance of ventilation-per-
fusion scans.
1-7
The mean increase in ventilation
to the treated lung was 50 per cent and was accom-
panied by an increase in perfusion of 24 per cent
(Fig. 15-2) and suggests release of hypoxic pul-
monary vasoconstriction.''
Partially obstructing lesions with a visible free
bronchial wall and visible lumen are technically
much easier to approach. This is because a laser
beam can hit a free margin tangentially, whereas a
laser beam must hit a fully obstructing lesion per-
pendicularly (blind penetration of the mass),
which, therefore, involves a much greater risk of
hemorrhage. Thus, the best results have been ob-
tained with partially obstructing lesions, and in a
high percentage of cases (greater than 85 per cent)
the relief is immediate and dramatic and is associ-
ated with a low incidence of bleeding.
2 3
When the
obstructing lesion is complete, the improvement is
more limited, occurs in a lesser percentage of pa-
tients (30 to 50 per cent), and is associated with a
higher incidence of bleeding.
2,3
8
The most dramatic improvement is seen in those
patients with intraluminal obstructions of the tra-
chea and main-stem bronchi (tumor and fibrous
stenoses) because most of the lung benefits from
improvement in ventilation, whereas ablation oi
creation of a channel through a peripheral lesior
can be expected to cause only a limited increase in
ventilation-perfusion ratio of the small amount oj
lung distal to the lesion.
3
5
Serial flow-volume
Anesthesia for Special Elective Therapeutic Procedures 5 1 5
Figure 15-1 Maximal expiratory and inspiratory flow-volume loops before (solid line) and after (dashed line) laser treatment or
surgery (dotted line) in 11 patients: three with (patients 7, 8, and 9) and eight (patients 1-6, 10, II) without underlying emphysema.
Patients 3, 5, 6, 8, and 10 had concomitant complete right upper lobe (RUL) bronchial obstruction, and patient 4 had a 6-mm right
main-stem bronchial obstruction in addition to complete obstruction of the left main-stem bronchus. The diameter of the orifice at
the site of maximal obstruction of the main-stem bronchus both before and after treatment is shown in relation to the corresponding
flow-volume loop. Note the relatively parallel rightward shift of the descending limb of the maximal expiratory flow-volume curve
after treatment. (COPD = chronic obstructive pulmonary disease.) (From Gelb AF, Tashkin DP, Epstein JD, et al: Physiologic
characteristics of malignant unilateral main-stem bronchial obstruction. Am Rev Respir Dis 38:1382-1385, 1988. Used with
permission.)
loops are a simple but effective way of following
changes in the caliber of the airway.
6
7
Use of the laser in cases of extrinsic compres-
sion with internal narrowing is an absolute con-
traindication to laser treatment because the laser
would result in destruction of the bronchial wall,
leading to subsequent cicatrization and causing
further obstruction.
9
Two other relative medical
Figure 152 Radionuclide ventilation and perfusion scans immediately before and 4 days after laser treatment in a patient with
a tumor (squamous cell carcinoma) obstructing the right intermediate bronchus. Regional changes in the treated lung have been
identified by computer subtraction, in which the pretreatment image has been subtracted from the posttreatment image. A marked
improvement in ventilation and perfusion is seen at the site that was treated. (From George PJM, Clarke G, Tolfree S, et al: Changes
in regional ventilation and perfusion of the lung after endoscopic laser treatment. Thorax 45:248, 1990. Used with permission.)
contraindications include total obstruction over 4
to 6 weeks with no possible/probable distal airway
opening being present and a bleeding diathesis.
10
Use of the laser beam itself (as opposed to the
ventilating system; see following sections) in-
volves several risks. Airway and endotracheal
tubes fires are the most feared hazard during laser
surgery of the airway."
12
The intense heat field
surrounding the laser beam can dry out and ignite
(as well as completely penetrate) most materials.
The risk of fire depends on the nature of the ma-
terial, the gaseous milieu, the beam wattage, and
the mode of operation.
All endotracheal tubes (rubber or plastic) can be
ignited by a laser beam in 50 to 100 per cent
oxygen; therefore, an F,0
2
of less than 0.5 in nitro-
gen or helium (nitrous oxide supports combustion)
should be used during the firing of the laser
beam.
8, 13
~
16
Helium is an especially attractive
choice for decreasing the F,0
2
when it is also de-
sired to decrease airway resistance.
17-19
In addition to decreasing the F,0
2
to less than
0.5 in nitrogen or helium, there are several meth-
ods for decreasing the risk of laser-induced endo-
tracheal tube fires. First, one can avoid using an
endotracheal tube altogether by using jet ventila-
tion through a metal laryngoscope. However, the
jetted gas, although it may entrain room air (Ven-
turi effect), is still 100 per cent oxygen; in all
cases, the amount of mixing of entrained gas with
100 per cent oxygen is not known. Consequently,
jet ventilation should occur between, not during,
periods of laser vaporization.
Second, metal tubes (Figs. 15-3 to 15-5)
20
"
2
-
1
may be used. Table 15-1 lists the many advan-
tages and disadvantages of these tubes. Of note
most of the metal tubes cannot offer protectior
against the neodymium:yttrium aluminum game
(Nd:YAG) laser (see Table 15-1).
Third, and because of the deficiencies in th(
metal tubes (see Table 15-1), the use of metal foi
has been advocated as a better way to protect com
bustible endotracheal tubes (red rubber and poly
vinylchloride) from laser impaction.
20-22
Self-ad
hesive aluminum tape can be obtained from th
3M Company (St. Paul, MN). Copper foil tape ca
be purchased from the Venture Tape Compan
(Rockland, MA). An endotracheal tube should b
wrapped using a single length of 0.25-inch-widt
tape in a spiral fashion, being careful not to crai
Figure 15-3 From left to right are the
Xomed. Bivona, and Mallinckrodt laser endotra-
cheal tubes. (From Sosis M: Anesthesia for laser
surgery. J Voice 3:163-174, 1989. Used with
permission.)
rough or sharp edges or points that could damage
the upper airway or tracheal mucosa (Fig. 15-6).
24
The tape should be wrapped in an overlapping
fashion so that bending of the tube will not expose
uncovered areas.
A new product, the Metrocel Laser-Guard, is an
easily placed external covering for sizes 5- to 8-
mm internal diameter (ID) endotracheal tubes that
contains a pure silver foil and is said (according to
the manufacturer, Americal Corporation, Mystic,
Figure 15-4 Two Norton endotracheal tubes are shown. (From Sosis M: Anesthesia for laser surgery. J Voice 3:163-174. 1989.
m
9
Figure 155 The Xomed Laser-Shield II endotracheal tube. A, The laser-resistant wrapping. B, The unprotected silicone proximal
to the cuff. C, The cuff inflated with saline and methylene blue. D, The unprotected silicone distal to the cuff. (From Green JM,
Gonzalez RM, Sonbolian N, Rehkopf P: The resistance to carbon dioxide laser ignition of a new endotracheal tube: Xomed Laser-
Shield II. J Clin Anesth 4:89-92, 1992. Used with permission.)
CT) to be resistant to C0
2
, YAG, KTP, and argon
lasers; no formal studies of this type of endotra-
cheal tube covering have been done to date.
Once combustible tape is applied to an endotra-
cheal tube, the wrapped portion of the tube is
considered to be protected from a direct hit by a
C0
2
laser. However, the endotracheal tube is still
subject to damage by the laser in two ways. First,
the inside of the endotracheal tube is still subject
to indirect combustion resulting from high temper-
ature in proximity to the tube or hot sparks, which
can cause the combustion of the tube even though
the tube is foil wrapped on the outside.
20
22
Second, the wrapping of endotracheal tubes
with foil still leaves the cuff unprotected. The
cuffs on the polyvinylchloride and red rubber en-
dotracheal tubes can be punctured when struck by
a C0
2
laser with a power of 10 W in only 0.1
sec.
20
22
It has been suggested that normal saline should
be used to fill the cuff; a C0
2
laser impact on the
cuff will produce a fine spray that will serve as a
"built-in automatic sprinkler system."
25
The sa-
line will serve as a heat sink that will absorb the
laser's energy, thus preventing combustion of the
cuff. A small quantity of methylene blue or indigo
carmine should be added to the saline in the cuff
to make perforation more obvious. In addition,
saline-soaked pads (pledgets, cottonoids) should
be passed between the vocal cords and the cuff of
the endotracheal tube.
The possibility of a fire in the airway during
laser surgery makes it necessary that a plan of
action be developed before this catastrophe occurs
(Table 15-2). At the first sign of airway fire, the
anesthesiologist should stop ventilation and termi-
nate the flow of all anesthetic gases, including 0
2
,
because they promote combustion and the endotra-
cheal tube should be quickly removed. The patient
should be ventilated via a mask. The patient's air-
way should be thoroughly examined. If the fire
was quickly extinguished, little or no damage may
have occurred. In such a case, the decision can be
made to proceed with surgery. Extensive airway
burns may necessitate mechanical ventilation, an-
tibiotics, and steroids. In the absence of an airway
fire, use of a laser beam in the tracheobronchial
tree does not increase carboxyhemoglobin levels,
and S
p
0
2
is very nearly equal to S
a
0
2
.
26
The surgeon also has a responsibility for pre-
venting airway fires. The laser should not be used
as a cautery. The surgeon should use the lasei
intermittently, at moderate wattage (45 W), and ir
a noncontinuous mode (durations less than 0.5-1
sec).
14
This will help prevent excessive heat fiek
build-up and tissue desiccation and will decreasi
Table 15-1 ADVANTAGES AND DISADVANTAGES OF VARIOUS METAL TUBES USED FOR LASER (PRIMARILY C0
2
) THERAPY
Metal Tube Physical Characteristics
Cost
($)
Reusable
Sizes
mm ID Ventilation Advantages Ventilation Disadvantages Other Clinical/Laser Concerns
Bivona
Fome-Cuff
(Fig. 15-3)
Norton
(Fig. 15-4)
Xomed
Laser Shield I
(Fig. 15-3)
Laser Shield II
(Fig. 15-5)
Mallinckrodt
Laser-Flex
(Fig. 15-3)
Aluminum core covered with
silicone, a polyurethane foam cuff
(self-expanding when exposed*to
atmosphere pressure) covered
with a silicone envelope; external
pilot tube
Flexible spiral-wound, all stainless
steel, no cuff, somewhat rough
sandblasted or matte finished
Manufactured from a nonreflective
silicone elastomer containing
metallic material
Same design but shaft is wrapped
with aluminum foil which is
overwrapped with Teflon
Flexible stainless-steel shaft, 2 PVC
cuffs with 2 separate pilot tubes
that run inside the steel shaft
80 No
199 Yes
75
49
No
No
4, 5, 6, 7
4.0, 4.8,
6.4
4, 5, 6, 7
4.5, 6.0
Cuff can be inflated with
saline, pilot tube can be
positioned away from the
area of laser impact;
functions as a standard
endotracheal tube
Functions as a standard
endotracheal tube
Functions as a standard
endotracheal tube
If cuff or pilot tube damaged, tube
has to be withdrawn with the cuff
inflated; not recommended for
postoperative ventilation
Leak with positive-pressure
ventilation, thick walls (large OD
for a given ID); the 4.8-mm ID
size has 1.3-mm wall thickness
Only small sizes available
Cannot be used with Nd:YAG (only
C02); cuff is made of combustible
material; fill cuff with saline
Completely nonflammable
Offers protection only against the
CO, laser, cuff is thin and easily
perforated (need wet cottonoids
around cuff); silicone
(unwrapped) area is combustible
Proximal cuff is supposed to protect
distal cuff; plastic portions of tube
are combustible; cannot be used
with Nd:YAG
Abbreviations: ID = internal diameter; Nd:YAG = neodymium:yttrium aluminum garnet; CO, - carbon dioxide; OD = outside diameter; PVC = polyvinylchloride.
Figure 15-6 Cuff wrapping technique. If
a wrapped endotracheal tube is the chosen
method for laser protection, the technique
for wrapping is critical in ensuring protec-
tion from both ignition and foil-induced mu-
cosal abrasions. It is often helpful to first
sparingly paint the tube with a medical ad-
hesive like benzoin or Mastisol. The end of
the tape should be cut with a scalpel to ap-
proximately 60 degrees. Begin wrapping by
aligning the cut end of the tape with the
junction of the tube and the proximal end of
the cuff. Wrap in a spiral with a 30 to 50 per
cent overlap between layers. Wrap, includ-
ing the inflation tube for the cuff, and con-
tinue until just short of the pilot balloon.
Take care not to wrinkle the tape at any
point. (From Rampil IJ: Anesthetic consid-
eration for laser surgery. Anesth Analg
tissue penetration, which decreases the risk of he-
morrhage and damage to underlying or adjacent
normal tissue.
The hazards of laser surgery also include inad-
vertent exposure of operating personnel to laser-
beam energy. The eye is most susceptible to in-
jury. All operating room personnel must wear ap-
propriate safety glasses during laser use. The pa-
tient's eyes must be taped closed and be well
covered with aluminum foil during the procedure.
Dermacare Corporation has designed for both C0
2
and Nd:YAG lasers aluminized, nonflammable
protective shielding for the patient's eye, the an-
esthesia circuit, and a utility barrier drape.
27
Expected pathophysiologic complications are
hypoxemia (obstruction of functioning airways
with blood and debris), postoperative respiratory
failure, and myocardial infarct in poorly anesthe-
tized/sedated patients with coronary artery disease.
In addition, puncture of the tracheobronchial tree
(mediastinal emphysema and pneumothorax) and
hemorrhage are always high risks if the target site
is not still.
28
Occasionally, it may be expected that
a distal abscess may be encountered after relief of
Table 15-2 AIRWAY FIRE PROTOCOL
Cease ventilation and turn off all anesthetic gases including
oxygen*
Extinguish flames with saline*
Remove endotrachealjube*
Ventilate by mask
Evaluate airway for burns (laryngoscopy, bronchoscopy)
*These steps should be taken simultaneously by the anesthe-
siologist and surgeon.
a total obstruction; this complication requires ag-
gressive evacuation of pus and support of ventila-
tion. The overall perioperative mortality for pallia-
tive procedures is presently about 0.4 to 2.0 per
cent,
29
but some earlier series have reported mor-
tality as high as 10 per cent.
2
30-32
B. Various Types of Laser and
Adjunctive Therapies
1. Nd-YAG Laser Resection
a. GENERAL CONSIDERATIONS
For several reasons, the Nd:YAG (neodymium
is the actual active laser material and the yttrium-
aluminum-garnet is the passive host matrix)
24
is by
far the most commonly used laser to resect ob-
structing tracheobronchial tree tumors.
1
"* First, be-
cause tissue is destroyed by a coagulation-vapori-
zation sequence, bleeding is minimal and
remaining tissue margins do not become edema-
tous or scar. Second, the Nd-YAG laser is capable
of destroying a large amount of tissue because it
has good tissue penetration (is poorly absorbed by
hemoglobin). Third, the transmission characteris-
tics of the Nd-YAG laser system are suitable for
tracheobronchial tree tumors. The Nd-YAG laser
has a wave length (1064 nm) that is readily con-
ducted through a flexible quartz monofilament and
that can be readily passed down either a fiberoptic
or a rigid bronchoscope. Since the laser beam is
invisible, a companion light beam that projects a
visible spot is simultaneously transmitted down the
fiber to allow for accurate aiming. The aiming spot
is in the same position and is the same size as the
invisible laser beam. This pilot light permits accu-
rate aiming with a laser fiber-to-target beam dis-
tance of only 5 to 10 mm. At this distance, the
laser beam divergence is less than 10, with an
area of photovaporization 1 to 2 mm in diameter
and up to 4 mm in depth. A continuous 3 L/min
flow of air is passed simultaneously through a
coaxial Teflon sheath to keep the fiber tip cool and
free of debris. Fourth, since the Nd-YAG laser can
be used with a fiberoptic bronchoscope, it is suita-
ble for resection of airway lesions beyond the di-
rect range of a rigid bronchoscope, such as upper
lobe and peripheral lesions; nevertheless, lesions
in such remote locations should be approached
with great caution.
b. ANESTHETIC CONSIDERATIONS
The anesthetic technique used depends greatly
on the ventilation/bronchoscopy technique (see
later discussion). However, in all cases, complete
stillness is mandatory to provide accurate aiming
of the laser beam toward the lesion. Consequently,
if local anesthesia is used, it must be complete and
solid and usually requires that a fiberoptic bron-
choscope (as opposed to a rigid bronchoscope) and
supplemental intravenous sedation be used.
1-3
Because the nature of this procedure requires
instrumentation of the airway, general anesthesia
is most commonly used to provide a still target
area. Short-acting intravenous drugs such as pro-
pofol and alfentanil work well. Certainly, neuro-
muscular blockade will guarantee stillness but it
also requires intermittent positive-pressure breath-
ing (IPPB).
33
'
34
Occasionally, IPPB may result in
bits of tissue and blood being blown around and
down the airways; if this complication occurs,
spontaneous ventilation is obviously preferable.
However, if spontaneous ventilation is allowed, it
must be realized that the window of anesthetic
depth between a nonmoving patient and an ade-
quately spontaneously ventilating patient may be
narrow and difficult to find.
If spontaneous ventilation is allowed, it is im-
portant to have syringes of an ultra-short-acting
barbiturate and/or propofol and lidocaine in the
intravenous catheter so that anesthetic response to
any sudden patient reaction to the surgical stimu-
lation may be immediate. Usually a reaction to the
stimulation is first signaled by an invagination of
the posterior membranous part of the trachea and
a simultaneous feeling of outward tension on the
bronchoscopy/ventilation system as the patient be-
gins to bronchoconstrict.
C. VENTILATION/BRONCHOSCOPY SYSTEMS
There are several ventilatory procedures that
must be followed with use of any ventilation/bron-
choscopy system. First, arterial oxygenation
should be continuously monitored using a pulse
oximeter. Second, if the pulse oximeter shows de-
creasing values (less than 90 per cent in previously
fully saturated patients) the laser resection should
be interrupted, and ventilation with high concen-
trations of oxygen should be resumed until an ad-
equate oxygen saturation is obtained. During 100
per cent oxygen breathing, obstructed airways
should be lavaged and aspirated via a rigid (see
Figs. 15-9 to 15-12) and perhaps a fiberoptic
bronchoscope.
If periods of desaturation are persistent, the sur-
geon must be asked to resect smaller amounts of
tissue, resect for shorter periods of time, control
bleeding better, remove more necrotic tissue with
forceps, and minimize suction. During actual laser
resections, the inspired oxygen concentration
should always be less than 50 per cent, with the
balance being nitrogen (nitrous oxide supports
combustion of oxygen). Third, a fiberoptic bron-
choscope must be used to reach upper lobe and
peripheral lesions. A fiberoptic bronchoscope can
be passed down a rigid bronchoscope; thus, the
two bronchoscopes do not have mutually exclusive
uses.
The simplest of all ventilation/bronchoscopy
systems is to pass a fiberoptic bronchoscope alone
under local anesthesia into the tracheobronchial
tree.
1, 2
The patient then spontaneously breathes
supplemental oxygen around the fiberoptic bron-
choscope. The operator must always be cognizant
that the patient may move even though liberal
amounts of local anesthesia and sedation were
used. A flexible fiberoptic bronchoscope is a rea-
sonable approach to laser treatment when the le-
sion is small and peripheral.
The fiberoptic bronchoscope can be passed
through an indwelling endotracheal tube
8
l3
(see
Fig. 14-2). Use of an endotracheal tube to intro-
duce the fiberoptic bronchoscope almost always
requires general anesthesia. As described in chap-
ter 14, the endotracheal tube must be larger than 8
mm in internal diameter, since positive-pressure
ventilation must occur around the fiberoptic bron-
choscope but within the endotracheal tube lumen
(see Figs. 14-1 and 14-2); adequate spontaneous
ventilation through this restricted surface area is
not possible in most patients. The fiberoptic bron-
choscope must be passed through a self-sealing
diaphragm to allow administration of positive-
pressure ventilation. However, even with positive-
pressure ventilation, hypercarbia may be a problem,
and in one series hypoventilation (P
a
C0
2
45-60
mm Hg) occurred in 30 of 32 procedures done in
this fashion.
8
Furthermore, use of a flexible fiberoptic bron-
choscope for major lesions results in the inability
to control hemorrhage, keep the lens clear, and
establish an airway. In general, and because of
these problems, it seems that a flexible fiberoptic
bronchoscope (whether used by itself or in com-
bination with an endotracheal tube) is now used
only for small peripheral lesions. If a fiberoptic
bronchoscope is used for a major lesion, it is prob-
ably used by an endoscopist who is inexperienced
with a rigid bronchoscope.
29,35
'
36
Indeed, endosco-
pists who once used just flexible fiberoptic bron-
choscopy now use a rigid bronchoscope most of
the time, and the flexible fiberoptic bronchoscope
(e.g., when used for upper lobe and peripheral le-
sions) is used through the rigid bronchoscope.
28,33,37
Most endoscopists now perform laser resections
using a rigid open-tube bronchoscope
3,4
l 6,28,29 35,
36 38
(see Figs. 14-10, 14-11, 15-9 to 15-12).
There are two reasons for this preference. First,
from the standpoint of effectively dealing with he-
morrhage and hypoxemia as well as the operator's
comfort and facility of manipulation of the laser
beam, there is no doubt that the rigid scope is the
first choice compared with the fiberoptic broncho-
scope. The laser beam can be manipulated better
with the rigid bronchoscope because of better field
of vision, better suction, easy cleaning of the tele-
scope, and easy removal of tumor fragments by
forceps. Hemorrhage can be better controlled be-
cause the laser can be better manipulated for co-
agulation, relatively large suction catheters can
clear the airway, and there is the possibility of
application of adrenaline cottonoids for hemosta-
sis. In addition, the rigid bronchoscope can be
used to establish an airway. This is not the case
when a fiberoptic bronchoscope is used, and the
procedure is much more tedious.
A flexible fiberoptic bronchoscope can be used
in conjunction with a rigid bronchoscope, can be
passed through a self-sealing diaphragm down the
lumen of the rigid bronchoscope, and can be used
intermittently (interchanged with large suction
catheters or forceps). The flexible fiberoptic bron-
choscope may facilitate aiming the laser beam by
allowing the operator to visualize both the tip of
the laser and the mass simultaneously (Figs. 15-7
and 15-8).
39, 40
Second, the open-tube technique
minimizes the possibility of endobronchial and en-
dotracheal tube fires. In distinction to the flamma-
ble coating of the fiberoptic bronchoscope, "steel
does not burn."
3
Thus, the rigid/flexible broncho-
scope approach combines the safety of the rigid
bronchoscope with the maneuverability of the fi-
berscope.
It must be realized that the rigid bronchoscope
cannot visualize upper lobe lesions and, if the
bronchoscope is used below the tracheal carina, it
must have side holes to ventilate the other lung. If
the bronchoscope does not have side holes, there
must be some additional method of ventilating the
other lung, such as jet ventilation, high-frequency
ventilation, or high-flow apneic ventilation. How-
ever, and fortunately, a fiberoptic bronchoscope
can be easily introduced through a tracheally po-
sitioned (above the carina) rigid bronchoscope to
reach airways below the tracheal carina; in this
instance, ventilation is around the fiberoptic bron-
choscope but within the walls of the rigid broncho-
scope. Passage of the fiberoptic bronchoscope
through the rigid bronchoscope is easily accom-
plished, since the internal diameter of adult rigid
bronchoscopes in general use today is 10.5 mm.
As described in chapter 14, ventilation through a
rigid bronchoscope can be with IPPB, spontaneous
ventilation, jet ventilation,
35,38 40
or high-frequency
ventilation.
d. SUGGESTED ANESTHESIA AND
VENTILATION TECHNIQUE FOR LASER
RESECTION OF AIRWAY TUMORS
The following is my anesthetic management of
laser resection cases involving the use of a rigid
bronchoscope (Table 15-3). Anesthesia is induced
with a short-acting barbiturate or propofol (2-3
mg/kg) and/or alfentanil (approximately 20 /
kg). Controlled positive-pressure breathing by
mask is instituted, and either propofol is adminis-
tered as a constant infusion (100-200 g/kg/min)
or isoflurane is administered until surgical levels
of anesthesia are achieved (5-10 min). Isoflurane
is chosen because these patients may become hy-
percarbic and acidotic; compared with halothane,
isoflurane minimizes the risk of arrhythmias. Lar-
yngoscopy is then performed (sometimes without
paralysis), and the tracheobronchial tree is sprayed
with local anesthetic (laryngotracheal kit). Con-
trolled ventilation via a mask with 100 per cent
oxygen and isoflurane is resumed for a short pe-
riod of time; then the surgeon and assistants se-
quentially insert a mouth guard, the rigid broncho-
scope, petroleum jelly-gauze packing into the
nose, saline-wetted gauze into the mouth, and a
transparent dressing over the mouth and around
half the bronchoscope.
41
The gauze and transparent
dressing greatly decrease the leak around the bron-
choscope (see Fig. 14-10).
Even though the mouth and nasal packing and
the transparent dressing over the mouth greatly
decrease the leak, high flows of oxygen (10 L/
min) may still be necessary to deliver effective
positive-pressure ventilation. If it is expected that
the rigid bronchoscope will not have to be moved
once it is placed in the trachea, and the airway
lesion is not high in the trachea, then the rigid
bronchoscope may be fitted with a reusable endo-
tracheal tube cuff.
42
After passage of the broncho-
scope, the anesthesia/ventilation system is then
Figure 15-7 Laser resection by combined use of rigid and
fiberoptic bronchoscopes. The laser is aimed by passing it
through the biopsy channel of the fiberoptic bronchoscope,
which has been fitted with a laser safety filter in the eyepiece.
The large suction hood removes smoke exhausted from the
rigid bronchoscope. (From Hetzel MR, Smith SGT: Endoscopic
palliation of tracheobronchial malignancies. Thorax 46:325-
333, I99l. Used with permission.)
connected to the side arm of the bronchoscope.
Connection to a coaxial Maplison D circuit (Bain
circuit) is desirable because the circuit is light and
will not place a strong pull on the rigid broncho-
scope. A large syringe of both an ultra-short-acting
barbiturate or propofol and/or alfentanil and lido-
caine is placed in the intravenous catheter. If any
tensing or outward pushing on the bronchoscope
is felt or movement of the tracheobronchial tree
(particularly invagination of the posterior membra-
nous part) is observed, small doses of the short-
acting barbiturate (50-100 mg) and/or propofol
(2-3 ml) and lidocaine ( 1 mg/kg) are administered.
If the endoscopist thinks that spontaneous ven-
tilation is indicated or is essential for the broncho-
scopic procedure, spontaneous ventilation is al-
lowed to return at this point. If spontaneous
ventilation is not an issue, ventilation is usually
controlled with IPPB, and relaxants (mivacurium,
succinyldicholine) may be administered (by bolus
or infusion) to appropriately facilitate this mode of
ventilation. A pulse oximeter is always used, 100
per cent oxygen is used during nonresection peri-
ods, and less than 50 per cent oxygen in nitrogen
is used during resection periods (e.g., 8 L/min of
nitrogen plus 3 L/min of oxygen equals F,0
2
of
0.45 and 10 L/min of nitrogen plus 2 L/min of
oxygen equals F,0
2
of 0.33).
Whenever moderate hypoxemia occurs (relative
to the initial S
p
0
2
), it is extremely important to
stop the procedure, ventilate the patient with F, 0
;
of 1.0, and clear up the airway (i.e., control the
bleeding, suction the airway clear, remove necrotic
tissue by forceps) (see Table 15-3, step 12). If
severe hypoxemia resulting from tracheal obstruc-
tion (without hemorrhage) occurs, the laser resec-
tion must be immediately suspended and all efforts
concentrated on re-establishing tracheal patency.
An immediate improvement in ventilation can
be achieved by passing the bronchoscope tip
through the obstruction (Fig. 15-9). In this posi-
tion, it must also be used to perform a tracheo-
bronchial toilet and remove any debris. If severe
hypoxemia occurs during treatment of a unilateral
main-stem obstructing lesion (without hemor-
rhage) (Fig. 15-10, left panel), the bronchoscope
must be withdrawn above the carina level and the
healthy side cleaned up (Fig. 15-10, right panel).
If tracheal hemorrhage occurs, then, as with tra-
cheal obstruction, the foremost priority is to re-
establish tracheobronchial patency by passing the
bronchoscope through the hemorrhaging area (Fig.
15-11). In doing so, the hemorrhage is controlled
by tamponade with the bronchoscope, and it is
possible to remove distal blood accumulation and
ventilate the patient.
As soon as the patient's ventilation is secured.
attention can then be directed to the tumor site.
Treatment and/or coagulation should be carried
out from distal to proximal, using the suction tube
to maintain a dry field and a clear telescopic view.
With massive bronchial hemorrhage, the patient's
healthy lung can be secured by putting the healthy
lung in a nondependent position. In this safety
position, blood and secretions can be evacuated
via the bronchoscope to achieve patency (Fig. 15-
12).
Because the laser resection of the tumor usually
makes ventilation and gas exchange better, every
effort is made to extubate the patient after resec-
tion. If the patient can be extubated, there is a
strong tendency to treat laser resection bronchos-
copy as an outpatient procedure. Approaching the
procedure on an outpatient basis minimizes the
financial impact and maximizes the amount of
good-quality postresection lifetime.
Jet Ventilator
LASER Fiber
Emerging from
Suction Channel
of Flexible
Bronchoscope ^LASER Fiber
Entering
Suction Channel
of Flexible
Bronchoscope
Figure 15-8 The instrumentation set-up. The flexible bronchoscope with the laser fiber in the suction channel is passed down
through the rigid bronchoscope. The jet ventilator is connected to the proximal end of the rigid scope. (From Bloomquist S,
Algotsson L, Karlsson SE: Anaesthesia for resection of tumours in the trachea and central bronchii using the Nd-YAG laser
technique. Acta Anaesthesiol Scand 34:506-510, 1990. 1990. Munksgaard Intl Publishers Ltd., Copenhagen, Denmark.)
Table 15-3 SUGGESTED ANESTHESIA AND
VENTILATION TECHNIQUE FOR
LASER RESECTION OF AIRWAY
TUMORS
1. Preoxygenation, pulse oximetry
2. Short-acting barbiturates and/or propofol and/or alfentanil
3. Isoflurane via mask IPPB surgical anesthesia
4. Laryngoscopy, spray tracheobronchial tree with local
anesthetic
5. Isoflurane via mask IPPB or propofol 100-200 g/kg/min
6. Insert rigid bronchoscope
7. Pack nose and mouth
8. Short-acting barbiturate or propofol and/or alfentanil and
lidocaine in-line to control any small sudden reaction to
stimulation
9. Spontaneous versus controlled ventilation (see text)
10. No paralysis versus paralysis (see text)
11. F,0
2
<0.5 in N
2
during laser firing (no N
2
0)
12. If % saturation decreases excessively
Ventilate with 100% 0
2
Control bleeding
Suction blood
Remove necrotic* tissue
13. Extubate if possible
Abbreviation: IPPB intermittent positive-pressure breath-
ing.
2. Hematoporphyrin Derivative/Laser-
Beam Photodynamic Therapy of
Bronchogenic Carcinoma
During the last several years, it has become ap-
parent that complete remission and, in some cases,
apparent cure of bronchogenic carcinomas (29 per
cent in the latest series of patients)
43
can be ob-
tained by use of hematoporphyrin derivative/laser-
beam photodynamic therapy of bronchogenic car-
cinoma.
43
"*
9
In the noncured patients, clinical
improvement almost always occurs. Phototherapy
(laser) in the presence of a sensitizer, hematopor-
phyrin derivative, produces cytotoxicity through
two distinct mechanisms: One is energy depen-
dent, and the other is power dependent. The en-
ergy-dependent mechanism is the result of selec-
tive photon absorption by the tumor and selective
energy transfer to the tumor by hematoporphyrin
derivative. When tumor cells have been sensitized
in this manner, a laser beam, which can have a
low energy level, then produces activated oxygen
in the tumor cells; this results in photodynamic
chemical reactions that impair tumor cell mem-
J
Tamponade
Bleeding Site
With Rigid
Bronchoscope
Rigid
Bronchoscope
(with ventilation
slots)
Figure 159 To treat hypoxemia resulting from major tracheal obstruction (left panel), the operator should pass the rigid bronchoscope through the
tracheal obstruction (middle and right panels) and perform bronchial toilet. (Modified with permission from Dumas JF, Shapshay S, Bourcereau J, et al:
Principles for safety in application of neodymium-YAG laser in bronchology. Chest 86:163-168, 1984.)
Figure 1510 To treat hypoxemia resulting from a unilateral main-stem bronchus obstruction (left panel), the operator shoul
withdraw the bronchoscope above the carina and perform tracheobronchial toilet of the healthy side (right panel). (Modified wit
permission from Dumas JF, Shapshay S, Bourcereau J, et al: Principles for safety in application of neodymium-YAG laser i
bronchology. Chest 86:163-168, 1984.)
brane function and cause tumor cell death within
24 to 48 hours. In contrast, the power-dependent
mechanism has the direct thermal laser-beam ef-
fect of causing coagulation, vaporization, and cut-
ting and excision. The cellular death in this in-
stance is related to the power of the light and is
independent of photochemical reactions mediated
by the hematoporphyrin derivative. Normal tissue
is not harmed by the photodynamic therapy.
The patients receive hematoporphyrin derivative
intravenously 3 to 5 days before light irradiation.
The drug circulates into all cells but clears faster
from normal cells than from cancer cells. If a blue-
violet light is shown through a fiberoptic broncho-
scope after 3 to 4 hours, the tumor may be diag-
nostically located because the target neoplastic
tissue will be fluorescent.
50
During this period of time (up to 2 weeks post-
injection), sunburn will occur if the patient is ex-
posed to sunlight.
43
The argon laser is carried by a
medical-grade quartz optical fiber, which can be
passed through either a flexible fiberoptic broncho-
scope or a rigid, open-tube bronchoscope. The
same anesthesia technique and ventilation consid-
erations that apply to Nd:YAG laser therapy apply
here. In the reported studies to date, the optical
fiber has been'introduced through a flexible fiber-
optic bronchoscope, and both local and general
anesthesia techniques have been used.
In contradistinction to the palliative nature of
Nd:YAG laser therapy, photodynamic therapy ca
be curative of small, young, radiologically occu
(<3 cm
2
) lesions. Patients best suited for this at
proach are those with small carcinomas in situ c
early invasive carcinoma lesions and for whom a
operation is not feasible either physiologically (
technically. In addition, it may be used for patien
who have failed to respond to standard therap
and have residual local disease. Photodynam
therapy may be also used in patients in whom it
desirable to reduce the extent of surgical resectio
with preoperative photodynamic therapy, select*
patients are able to undergo sleeve lobecton
rather than pneumonectomy. Finally, the techniqi
can be used in a palliative way to open obstruct)
bronchi.
51
An early danger of this technique
bleeding within a large mass, and a late danger
obstruction of the tracheobronchial tree with
erotic tissue (which may require mechanical ve
tilation).
43
3. Carbon Dioxide Laser
The carbon dioxide laser (C0
2
is the actual ;
tive laser material) is absorbed by tissues to
much greater extent than the Nd:YAG laser. Cc
sequently, if cutting and excision or vaporizati
with a shallow depth of penetration with minir
Tamponade
Bleeding Site
With Rigid
Bronchoscope
Rigid
\ Bronchoscope
(with ventilation
slots)
Fiberoptic
Bronchoscope
Suction
Catheter
Figure 15-11 To treat tracheal hemorrhage (left panel), the operator must pass the bronchoscope distal to the bleeding (middle and right panels) and
thereby tamponade the bleeding with the bronchoscope, and evacuate blood and other secretions. With ventilation secure, one may proceed to eliminating the
bleeding by coagulation. (Modified with permission from Dumas JF, Shapshay S, Bourcereau J, et al: Principles for safety in application of neodymium-YAG
laser in bronchology. Chest 86:163-168, 1984.)
Fiberoptic
Bronchoscope.
Laser Fiber
Rigid
Bronchoscope
(with ventilation
slots)
Suction Catheter
Figure 15-12 To treat bronchial hemorrhage (left panel), the operator should use suction to keep the field clear (right panel) to
identify and coagulate the bleeding site. (Modified from Dumas JF, Shapshay S, Bourcereau J, et al: Principles for safety in
application of neodymium-YAG laser in bronchology. Chest 86:163-168, 1984.)
scattering (which unfortunately precludes good co-
agulation) are required, the C0
2
laser is the tool of
choice. Unfortunately, the C0
2
laser requires
straight-line rigid optics. Therefore, the C0
2
laser
is the most commonly used for removal of lesions
that can be directly visualized in the laryngeal and
supralaryngeal areas.
The removal of laryngeal lesions by a C0
2
laser
requires rigid instrumentation.
47
If the lesion is
subglottic, a bronchoscope must be used. The an-
esthetic considerations for these kinds of cases are
similar to those for procedures utilizing the Nd-
YAG laser. If the lesion is supraglottic, an endotra-
cheal tube can be used to ventilate the patient.
However, the chance that a laser beam could strike
and ignite the oxygen-containing endotracheal
tube and result in a catastrophic intraluminal blow-
torch-type airway fire is much greater. Conse-
quently, the endotracheal tube needs to be meticu-
lously protected from the laser (see prior
discussion of metal tubes; Figs. 15-3 to 156 and
Table 15-1). In brief, this may be done by wrap-
ping the endotracheal tube circumferentially with
metallic tape above and below the endotracheal
tube cuff. Protection of the cuff of the endotra-
cheal tube is provided by placing saline-saturated
sponges above the endotracheal tube in the sub-
glottic larynx and filling the cuff with saline. Fi-
nally, an operating platform (a flat, circular metal
instrument) can be inserted into the subglottic lar-
ynx above the level of the packed sponges to act
as a "catcher's mitt" to further protect the
sponges, endotracheal tube and cuff, and tissues of
the subglottic larynx from laser-beam irradiation.
47
Finally, and alternatively, flexible metal tubes
have been developed (see Table 15-1). Depending
on the site of surgery, a seal may be obtained with
a cuffless metal tube by a saline-soaked pharyn-
geal pack (surgery on the nose, tongue, palate, oral
cavity) or subglottic saline-soaked swabs attached
to wires (surgery on the larynx, trachea). Cuffed
(to be filled with saline), metal-impregnated, laser-
beam-resistant tubes are available commercially
(see Table 15-1), but they are for single use only,
are expensive, and retain some vulnerability to the
carbon dioxide laser beam (the tube can be dam-
aged after approximately 10 hits). The metal-im-
pregnated tube offers no resistance at all to the
Nd:YAG laser beam. A metal tube for use with
the C0
2
laser in infants has been developed.
52,
"
Alternatively, but similarly, a metal suction tube
(1.2 mm ID, 2.0 mm outside diameter [OD] may
be passed subglottically through the suspension
laryngoscope) and used for jet ventilation.
54
4. Adjunctive Endobronchial
Radiotherapy
The advent of the laser has made it possible to
reopen airways that had been completely or par-
1
tially obstructed by recurrent carcinoma. However,
the limitations of the laser are now being appreci-
ated. Since only the intraluminal extent of the tu-
mor can be treated, recurrences of tumor are rather
frequent, even expected, and they necessitate re-
peat laser treatments at a potentially higher risk of
hemorrhage each time.
4
16
55
~
57
Thus, laser therapy
has a relatively short duration of therapeutic effect.
In addition, laser therapy is limited to treatment of
endobronchial tumors and is not suitable for treat-
ment of bronchial obstruction due to extrinsic
(peribronchial) malignant disease. To circumvent
some of these problems and attempt to offer pallia-
tion to patients not suitable for laser therapy, inser-
tion of temporary endobronchial catheters for en-
dobronchial radiotherapy has been tried and
appears promising. Endobronchial radiotherapy
places a radioactive source near the tumor and
delivers a relatively high dose (compared with ex-
ternal radiation) without damage to normal tis-
sue.
58
"*
0
When this approach was first tried, and only a
relatively low-intensity radioactive source was
available, a flexible fiberoptic bronchoscope was
first inserted through the self-sealing diaphragm of
the elbow connector to an endotracheal tube, usu-
ally under general anesthesia.
58
59
If complete
bronchial obstruction was encountered, the Nd-
YAG laser fiber was passed through the biopsy
channel of the fiberoptic bronchoscope and used
to reopen the airway. Immediately after comple-
tion of the Nd-YAG laser therapy, the endobron-
chial radiotherapy catheters were inserted percuta-
neously into the trachea via cricothyroid
membrane puncture (see later discussion) and po-
sitioned under the direct vision of the fiberoptic
bronchoscope. If only partial obstruction of the
bronchial airway was observed, the endobronchial
radiotherapy catheter was inserted into the trachea
via cricothyroid membrane puncture (see later, dis-
cussion) right away for endobronchial irradiation.
Similarly, distal segmental obstructions were also
treated with a distally positioned endobronchial
catheter instead of the laser. All patients were ex-
tubated at the completion of the catheter insertion
procedure; this is not surprising in view of pre-
vious Nd-YAG laser experience providing symp-
tomatic respiratory relief that is sometimes imme-
diate and dramatic. The symptomatic relief due to
the reopening of a large segment of the lung was
accompanied by improvements in pulmonary func-
tion test results, arterial blood-gas values, and ven-
tilation-perfusion lung scans.
The endobronchial radiotherapy catheter had a
percutaneous subglottic placement because the
catheter had to be left in situ for 1 to 4 days, the
subglottic placement diminished irritative cough
reflexes and reduced the risk of catheter dislodg-
ment, and it allowed the patient to eat normally
while being treated with radiation for several
days.
58
The subglottic placement was effected
either by a small incision over the cricothyroid
membrane followed by direct placement of the
catheter into the trachea or by percutaneously
passing a 12 French introducer through the crico-
thyroid membrane and using the introducer as a
sheath through which the endobronchial catheter
can be passed.
58
The endobronchial radiation catheter was placed
in the correct position within the tracheobronchial
tree under direct vision with the fiberoptic bron-
choscope. When the endobronchial radiotherapy
catheter is in the correct position, the catheter can
be loaded with the radiation (iridium-192 seeds)
so that the length of the radiation source is 2 cm
more than the length of the obstruction. The radius
of the cylindric volume treated with radiation is
5.0 mm for endobronchial lesions and 7.5 to 15.0
mm for peribronchial disease. Once treatment is
complete, the endobronchial catheter is removed.
The tiny cricothyroid incision heals readily, and
persistent air leaks have not been a problem. Alter-
natively,
l98
Au seeds could be implanted in the
tracheobronchial tree via a 2.6-mm ID ejection
cable passed through a fiberoptic bronchoscope.
61
Recently, the technology has improved so that
the radioactivity can be delivered to the tumor in
a matter of minutes by use of a high-intensity
radioactivity remote afterloader.
59
60
62
~
64
This de-
vice consists of a lead safe containing a high-
intensity radioactive source. The source is attached
to the end of a long radioactivity-transferring cable
(4 mm in diameter) (Gamma Med
59
or Selectron
25
catheters), which can be advanced into the appro-
priate treatment position through or alongside a
fiberoptic bronchoscope.
62
The various studies have used two different
ways to deliver a short course of high-intensity
radiation. In the first method,
59
the position of the
treatment cable was determined as follows. An
endotracheal and/or endobronchial tube was ad-
vanced to a position close to the tumor. A fiber-
optic bronchoscope was passed through the endo-
tracheal/endobronchial tube and measured the
proximal and distal limits of the tumor. The radio-
activity-transferring catheter was inserted into the
endotracheal/endobronchial tube and positioned
against the distal end of the tumor. After roentgen-
ographic verification of the proper position of the
endotracheal/endobronchial tube and the radioac-
tivity-transferring cable, the length of the tumor
and dosage of radioactivity to be used were en-
tered into the computer and the treatment was be-
gun. The cable remained in each position for a few
seconds, as determined by the computer, and then
was retracted proximally in 0.5- to 1.0-cm steps to
the next position until the treatment was com-
pleted. The total treatment time is usually approx-
imately 3 to 5 min.
In the second method,
60,62
a transnasal fiberoptic
bronchoscope, passed under local anesthesia, di-
rected the radioactivity-transferring catheter into a
position near the tumor. With the cable in this
position, high-intensity radiation was administered
in various advancement/retraction steps for various
dwell times, with total treatment times of several
minutes.
62
Because the procedure is so short and
physiologically unintrusive, it can be done on an
outpatient basis and multiple times.
62
In three
studies of patients with stage III lung carcinoma
disease, 80 to 90 per cent have had markedly im-
proved symptoms and endoscopic and tomo-
graphic findings and improved pulmonary function
tests, P
a
0
2
, and V/Q scans.
62-64
Very late hemop-
tysis (days to months after treatment) may be a
complication related to this therapy.
62
""
64
The major anesthetic consideration for these
cases is the same as that for laser resection with a
fiberoptic bronchoscope through an endotracheal
tube, namely, a narrowed airway (because of both
the fiberoptic bronchoscope and the radiotherapy
cable) with which to ventilate the patient. The en-
dotracheal tube is inserted using topical anesthesia
and intravenous sedation. The techniques dis-
cussed previously for intubation with the patient
awake are applicable for this procedure. Use of a
spiral-wire endotracheal tube is advisable since
it visualizes well on the roentgenographic films
needed to confirm proper placement. The endotra-
cheal tube cuff is not inflated, and the patient
breathes around and through the tube. For bron-
chial lesions a smaller tube is required and, again,
the cuff is not inflated, and the "intubated" lung
breathes through and around the endobronchial
tube while the "unintubated" lung breathes inde-
pendently around the tube. In one advanced case,
58
a lung abscess was encountered distal to a total
obstruction after using the laser to vaporize the
total occlusion. Aspiration of purulent material
into the contralateral lung followed, necessitating
evacuation of pus with a rigid bronchoscope, im-
mediate reintubation with a double-lumen tube,
and differential lung ventilation for 1 week while
the abscess and aspiration pneumonia were treated
aggressively. While the patient was intubated, an
endobronchial catheter was inserted through the
double-lumen tube into the right main-stem bron-
chus for radiotherapy.
5. Other Adjunctive Bronchoscope/
Laser Therapies of Airway Strictures
a. CRYOTHERAPY OF STRICTURES
The palliation of refractory airway strictures and
malignant tumors of the tracheobronchial tree by
cryotherapy is new treatment when radiation and
surgical therapy are no longer possible.
65
The prin-
cipal feature of the 55 cm long and 3-mm OD
cryoprobe, which uses nitrous oxide as the coolant
source, is its nonrigidity, thus enabling one to
reach tumors obstructing the main bronchus or the
upper lobe. Indeed, most tumors that are visible
through a fiberoptic bronchoscope are now
reached with a flexible cryoprobe inserted into a
flexible bronchoscope.
The temperature obtained on the tip of the
cryoprobe is - 80C, but the tissues are frozen at
30C or -40C; therefore, it is a problem to
control the cryodestruction. An impedance metric
method, in which one electrode is represented by
the tip of the cryoprobe and the other is inserted
into the skin of the patient, has been used to con-
trol the amount of freezing.
65
Cryotherapy in the bronchial tree is capable of
inducing coagulation necrosis of tissue and de-
struction of tumors. Indications for cryotherapy
appear sometimes to be the same as laser therapy
for tracheobronchial lesions (when surgical or ra-
diation therapy is no longer feasible), but the cost
of a cryoprobe is much less. Nevertheless, the re-
sult is delayed, and the technique is not useful for
acute respiratory distress.
b. BALLOON CATHETER DILATION OF
STRICTURES
Present therapeutic modalities for the relief of
an acquired bronchial stenosis include surgical re-
section, cryotherapy, and laser photoresection. One
report describes the dilation of such stenosis using
angioplasty or valvuloplasty balloon catheters.
66
In
this report, after induction of general anesthesia
and passage of a rigid bronchoscope, and in one
case the application of an Nd:YAG laser to the
stenotic area, a 4-mm balloon angioplasty or val-
vuloplasty catheter was introduced through the
rigid bronchoscope and positioned in the stenotic
areas. The balloon was inflated with normal saline
four to five times its maximum dimensions, as
visualized through the bronchoscope and held in-
flated for 30 sec. The treatment greatly alleviated
the patient's symptoms. Clearly, this method of
dilation can delay, or even prevent, surgery and
can be performed when surgery is not feasible.
III. TRACHEAL RESECTION
Tracheal resection is indicated, if technically
feasible, in patients who have tracheal obstruction
caused by prior tracheal trauma (e.g., stenosis from
prolonged intubation, blunt or penetrating trauma)
(most common indication), primary tracheal tu-
mors, the majority of which are carcinomas (se-
cond most common indication), congenital anom-
alies, and vascular lesions. In a series of acute
tracheal injuries, 50 per cent of the patients had
blunt trauma to the trachea and tracheolaryngeal
area (mainly motor vehicle accidents), 40 per cent
had penetrating injuries such as gunshot or knife
wounds, and 10 per cent had iatrogenic injuries
from intubation attempts.
67
Intubation injuries can occur directly from the
intubation itself or secondarily from the endotra-
cheal tube cuff.
68
The injuries can be located in the
intra- or extrathoracic trachea. In one series of 365
patients, 279 had postintubation injuries: 263 were
stenotic lesions, 6 had malacia alone, and 10 had
combinations of stenosis and malacia. Figure 15-
13 shows the most frequent types and locations of
tracheal tube-induced injury in the adult popula-
tion.
69
The most frequent cause of tracheal stenosis
in children is also endotracheal intubation.
Tumors are the second most frequent cause of
tracheal stenosis. In the series reported by Grillo
et al.,
70
56 of 365 patients had primary tumors of
the trachea: 19 had squamous cell carcinomas, 18
had cystic adenomas (cylindromas), and 19 had
other tumor types including carcinoids or chondro-
sarcomas. Metastatic cancer and invasion of the
trachea by other nontracheal mediastinal tumors
contributed another 30 cases in this series. The
most frequent tumor in this group was thyroid
cancer that either caused symptoms from extrinsic
compression of the trachea or actual tracheal dis-
ease. Unfortunately, primary tracheal tumors are
frequently diagnosed late because they are not
readily apparent on the plain chest film, and in
many cases the slowly progressive symptoms of
upper airway obstruction are misdiagnosed as
asthma or chronic bronchitis for long periods of
time.
Generally, the airway must be narrowed to 5 to
6 mm in cross-sectional diameter before signs and
symptoms become clinically evident. The general-
ized and nonspecific symptoms of tracheal disease
(stridor, hoarseness, wheezing, progressive dysp-
nea, and decreased exercise tolerance) are similar
for benign and malignant lesions.
ENDOTRACHEAL TUBES

CUFFED
TRACHEOSTOMY TUBES
VOCAL CORDS, CRICOID:
gronuloma
stenosis
CUFF SI TE:
stenosis,t.e. f i st ul a
TUBE TIP SITE
gronuloma
esophageal or,, fistula
art eri al /

STOMAL SITE:
anterior stenosis
granuloma
malacia
CUFF SITE:
- stenosis, t.e. fistula
TUBE TIP SI TE:
gronuloma
fistula /esophageal on
varterial t
*r*s
Figure 15-13 The most frequent types and locations of injuries in the adult trachea after intubation and tracheostomy, (t.e. =
tracheoesophageal.) (Reproduced with permission from Grillo HC: Surgery of the trachea. Curr Prob Surg 3:59, I970.)
Congenital abnormalities of the trachea requir-
ing tracheal reconstruction are rare. Abnormalities
such as tracheal agenesis and atresia are incompat-
ible with life. Congenital abnormalities caused by
vascular rings and slings from aberrant segments
of the embryonic aortic arch are, however, fre-
quently correctable. The most common such lesion
is an anomalous innominate artery. This vascular
anomaly usually results in only mild tracheal com-
pression. A double aortic arch is the next most
frequent vascular anomaly that results in tracheal
compression. Another abnormality involves an
aberrant origin for the left pulmonary artery from
the right pulmonary artery. This lesion is associ-
ated with complete tracheal rings. Therefore, both
the vascular and tracheal abnormality require re-
pair to obtain a competent airway.
Of patients who have operable disease, approx-
imately 80 per cent have a segmental resection
(which may include carina or larynx) with primary
anastomosis; 10 per cent have segmental resection
with prosthetic reconstruction; and 10 per cent
have insertion of a T-tube stent. Adjuncts to sur-
gical extirpation include pre- and postoperative ex-
ternal radiation, internal radioactive seed radiation
(transferred by endobronchial catheter as described
previously or directly placed by thoracotomy), and
preoperative laser debulking therapy.
The four possible incisions for approaches to
tracheal lesions are shown in Figure 15-14.
71
Up-
per and midtracheal lesions are approached
through a cervical collar incision with an upper
sternal split if necessary (Fig. 15-14 and 15-
\4B). This incision requires that the patient be in
a modified reverse-Trendelenburg position with
the neck hyperextended. Lesions of the lower tra-
chea and the carina are approached through a right
posterior thoracotomy (Fig. 15-14D). With this
approach, the patient is placed in a standard left
lateral decubitus position. Figure 15-14C shows
an incision that has rarely been used recently
secondary to improved surgical techniques with
the other approaches.
Median sternotomy with or without a cervical
collar incision
72
or a right posterior lateral thora-
cotomy
73
can provide adequate exposure for most
cases and was extensively used in the past. How-
ever, exposure of the carina and main bronchi
through a median sternotomy requires a transperi-
cardial approach. The anterior pericardium is di-
vided vertically to permit circumferential mobili-
zation of the ascending aortic arch, which is then
retracted laterally to the left. The superior vena
cava is displaced laterally to the right, and the right
main pulmonary *artery is exposed and displaced
inferiorly (see Figs. 2-30 and 14-12). The poste-
rior pericardium, which is displayed by these latter
maneuvers, is then divided vertically, and the en-
^
tire mediastinum, trachea, and carina are clearly
exposed and accessible. Occasionally, additional
posterolateral thoracotomy may be necessary to
gain additional exposure.
Whenever a primary anastomosis is performed,
the anesthesiologist may be asked to flex the pa-
tient's head to reduce the tension on the anasto-
mosis; if undesirable tension is still present, it may
be expected that further cervical incisions and
proximal laryngeal release or thoracotomy and dis-
tal main-stem bronchi release, or both, will have
to be done.
Many previous technical limitations to the per-
formance of tracheal surgery can now be over-
come by careful preoperative delineation of the
site and degree of obstruction, close intraoperative
communication between the surgeon and anesthe-
siologist, improved anesthetic management tech-
niques, and meticulous postoperative care. All of
these components contribute to the ability to pro-
vide adequate ventilation throughout the perioper-
ative period. Although the results of this compli-
cated surgery on primary tracheal tumors depend
on tumor cell type, location, and method of resec-
tion, it is generally accepted that, in the few insti-
tutions with a reasonable degree of surgical expe-
rience, worthwhile survival can be obtained in the
majority of patients. The next section discusses the
care of patients undergoing tracheal resection and
describes several different methods of airway man-
agement (Table 15-4).
Unless airway obstruction is imminent, pulmo-
nary function should routinely be studied preoper-
atively. The presence of preoperative lung disease
that is severe enough to indicate postoperative
ventilatory support is a relative contraindication to
tracheal resection since the trauma of positive air-
way pressure and an endotracheal tube cuff at the
tracheal suture line may cause wound dehis-
cence.
77
Obtaining a history of position-dependent
airway obstruction is important because the induc-
tion of anesthesia should be accomplished with
such patients in a position that does not cause
airway obstruction.
Preoperative evaluation should also include tra-
cheal tomograms, computed tomography (to define
the exact position of the lesion), bronchoscopy
(usually deferred until the time of operation so as
not to precipitate airway obstruction from edema
or hemorrhage), arterial blood gases and flow-vol-
ume loops (upper airway obstructions have char-
acteristic shapes to the loop: Variable lesions [such
as a pedunculated tumor or tracheomalacia] may
be extra- or intrathoracic; extrathoracic variable
Figure 15-14 Incisions used for tracheal resection and reconstruction. A, Cervical collar incision for upper tracheal lesions. H.
Cervical collar incision with a sternal split for complicated upper and benign lower tracheal lesions. The dashed line is a flap that
can be raised at the fourth interspace to expose the entire trachea. C, Incisions to be made if a mediastinal tracheostomy is
anticipated and total exposure of the trachea is required. D, Malignant lower tracheal and carinal resections require a posterior right
thoracotomy that is usually made along the fourth intercostal space. (From Grillo HC: Surgical approaches to the trachea. Surg
Gynecol Obstet 129:347-352, 1969. Used with permission.)
obstructions cause an inspiratory limb plateau, and
intrathoracic variable obstructions cause an expi-
ratory limb plateau; fixed lesions (e.g., circumfer-
ential scar) cause both an inspiratory and expira-
tory plateau (Fig. 15-15). Steroids may be given
when preoperative tracheal edema is a contributing
factor to the decreased size of the airway lumen.
During surgery, all patients should have an ar-
terial catheter placed to facilitate frequent* deter-
mination of arterial blood gases. It should be
placed in the left radial artery because the innomi-
nate artery (which supplies the right radial artery)
crosses the trachea and may be compressed during
surgery (see Figs. 2-30 and
1
412). Other appro-
priate monitoring, as described in chapter 7,
should be placed before the induction of anesthe-
sia. A variety of methods for providing adequate
oxygenation and carbon dioxide elimination have
been utilized during tracheal resection and are de-
scribed in Table 15-4. These can be divided into
five approaches: (1) standard orotracheal intuba-
tion, (2) insertion of a tube into the opened trachea
distal to the area of resection, (3) high-frequency
jet ventilation through the stenotic area, (4) high-
frequency positive-pressure ventilation (HFPPV),
and (5) cardiopulmonary by-pass.
Special equipment that should be available be-
fore the induction of the patient for tracheal recon-
struction and that will enable the anesthesiologist
to provide most of the ventilatory options noted
above are listed in Table 15-5. First, an anesthesia
machine that is able to deliver high flows of oxy-
gen (up to 20 L/min) is needed. High flows may
be needed during rigid bronchoscopy without
packing and when the tracheobronchial tree is
open (incised).
A wide assortment of endotracheal tubes must
be available. This includes polyvinylchloride tubes
ranging from size 4- to size 8-mm ID; cuffed.
unsterile flexible armored tubes from size 20
French to 34 French; and extra long endotracheal
tubes (these are constructed by fitting an endotra-
cheal tube one size larger than the primary endo-
tracheal tube over the primary endotracheal tube
or by using the base of the endotracheal tube
adapter (the male end) as a bridge between two
equal-sized (with respect to internal diameter/out-
side diameter) endotracheal tubes (Fig. 15-16)
94
or
by using the Wilson tube (see chapter 9).
95
In ad-
dition to the unsterile tubes, a complete set of
sterile armored tubes for endotracheal and endo-
bronchial intubation from the field should be avail-
able. A set of sterile corrugated tubing and a sterile
Y-piece that can be passed to the anesthesiologist
Table 15-4 APPROACHES TO AIRWAY MANAGEMENT DURING TRACHEAL RESECTION
General Approach Year Surgical Approach Technique Remarks
/. Orotracheal Intubation
A. Belsey
79
B. Kamvyssi-Dea et
al.
80
C. Beyer & Wilson
95
//. Insertion of Tube Into
Opened Tracheal
Distal to Area of
Resection
A. Geffin et al.
77
1950 Right thoracotomy
1975 Right thoracotomy
1988
Orotracheal tube
Orotracheal tube
Orotracheal tube
1969
Various combinations of
right thoracotomy,
cervical collar, and
median sternotomy
Cervical incision for high Endotracheal tube
lesion; right Endobronchial tube
thoracotomy for low
lesion
B. Debrand et al.
81
C. Boyan & Privitera
8
1979
1976
Cervical incision
Median sternotomy
D. Abou-Madi et al.
83
1979 Median sternotomy
E. Lippmann & Mok
84
1977
F. Akdikmen& 1965
Landmesser
85
G. Dodge et al.
86
1977
H. Theman et al.
78
1976
I. Beyer & Wilson
95
1988
J. Neville et al.
97
1990
Right thoracotomy
Right thoracotomy
Right thoracotomy
Right thoracotomy
right thoracotomy,
median sternotomy
right thoracotomy,
median sternotomy
4-mm endotracheal tube
into distal trachea
Cuffed endotracheal tube
into distal tracheal
stump
28 Foley catheter into
distal tracheal stump
Left endobronchial tube
Right endobronchial tube
Right and left
endobronchial tubes
Right and left
endobronchial tubes
Right and left
endobronchial tubes
Right and left
endobronchial tubes
First description of this type of
managementtube advanced
distally as resection carried
out
Tube advanced distally as
resection carried out
Summarizes the literature
(particularly more recent)
using this approach
Excellent review article
described experience with 31
patients; stressed avoiding
tracheostomy and
cardiopulmonary by-pass for
this procedure
Patient 4 months old
Stressed preoperative forced
vital capacity, loop of airflow
versus volume to confirm
diagnosis of airway
obstruction
End cut off just distal to balloon
allowed bilateral lung
ventilation through short
tracheal stump
Tube quickly inserted following
intraoperative tracheal
resection
Patient died postoperatively;
authors suggested use of
cardiopulmonary by-pass for
this procedure
Two anesthesia circuits used
Two ventilation systems used;
selective occlusion of
contralateral pulmonary artery
Summarizes the literature
(particularly more recent)
using this approach
Surgery involved tracheal
reconstruction with a tracheal
prosthesis
to connect to the anesthesia machine for ventila-
tion during the reconstruction should also be avail-
able.
Availability of a high-frequency positive-pres-
sure ventilator is desirable. It should be capable of
rates of 60 to 120 bpm with variable inspira-
tory: expiratory (I:E) ratios and variable pressure
settings. A second anesthesia machine should be
available for the resection of lesions involving the
distal trachea and carina. This machine can be
used instead of the high-frequency positive-pres-
sure ventilator.
1. Orotracheal Intubation
The first ventilating technique uses a standard
but uncut long orotracheal tube; this is placed
4-
Table 15-4 APPROACHES TO AIRWAY MANAGEMENT DURING TRACHEAL RESECTION Continued
General Approach
III.
IV.
V.
High-Frequency Jet
Ventilation
A. Lee & English
87
B. McNaughton
C. Baraka
89
D. Ellis et a l
w
E. Dartevelle et al."
h
F. Baraka et al.*
HFPPV
A. Eriksson et al.
76
. El-Bazet al."
C. El-Bazet al.
74
Cardiopulmonary
By-pass
A. Woods et al.
92
B. Coles et al
93
C. Bencaetal.
9
*
Year
1974
1975
1977
1976
1988
1986
1975
1982
1982
I96l
1976
1988
Surgical Approach
Median sternotomy
Not stated
Right thoracotomy
Not stated
Ipsilateral thoracotomy
Right thoracotomy
Cervical incision
Cervical incision
Right thoracotomy
Right thoracotomy
Right thoracotomy
Median sternotomy
Technique
No. 8 French suction
catheter passed
through stenotic area;
bronchoscope injector
No. 12 catheter; jet
ventilation
5-mm cuffed
endotracheal tube; jet
ventilation
1.52-mm diameter
manometer line;
bronchoscope injector
Small tube
6.0-mm ID cuffed tube
No. 14 catheter; rate
60/min
2-mm catheter; rate
150/min
2-mm catheter; rate
150/min
Cardiopulmonary by-pass
Remarks
Patient 8 years old
No significant effects on the
circulation noted
Tube pushed (from above) past
lesion during anastomoses
Patient 13 years old
Surgeries were sleeve
pneumonectomies
Technique includes F,0
:
= 1.0
(no air entrainment);
summarizes more recent
literature
First report of HFPPV for
tracheal resection
Used with uncuffed tracheal
T-tube stent
Two patients for carinal
resection and one for tracheal
resection; only left lung
ventilated
First reported case; instituted
after chest was open
By-pass instituted under local
anesthesia before operation
Summarizes the airway
management approaches used
for tracheal reconstruction in
neonates (especially
cardiopulmonary by-pass)
Abbreviations: ID = internal diameter; HFPPV = high-frequency positive-pressure ventilation.
above the tracheal lesion after the induction of
general anesthesia and is merely insinuated, by the
surgeon, past (distal to) the area of stenosis or
mass.
79
80
As the trachea and carina are being dis-
sected, gentle ventilation by hand facilitates expo-
sure. Although this method is relatively easy to
employ, a large tube may traumatize the lesion and
cause bleeding or dislodgment of tissue, resulting
in further airway obstruction. Furthermore, the
technique is limited to cases with a relatively mild
stenosis, and the presence of an endotracheal tube
in the surgical field makes the tracheal anastomo-
sis more difficult to complete.
2. Insertion of Tube Into Opened
Trachea Distal to Area of Resection
To overcome these problems, endotracheal or
endobronchial tubes have been inserted into the
opened trachea distal to the site of resection
(second approach).
77
8I
~
87
With this second ap-
proach, initially either a small endotracheal tube is
passed distal to the obstruction or a standard en-
dotracheal tube is placed proximal to it (Figs. 15-
17, 15-18A and 15-19). All further tracheal
and endobronchial intubations are performed with
armored tubes passed into the airway, which is
surgically opened distal to the lesion. The surgeon
must have a complete set of sizes of endotracheal
tubes to choose from since either main-stem bron-
chus or any lobar bronchus may need to be intu-
bated.
With a high tracheal lesion, a cervical incision,
possibly combined with a median sternotomy, pro-
vides adequate surgical exposure. An opening is
made in the trachea distal to the area to be re-
sected, and a sterile endotracheal tube is inserted
Figure 15-15 Maximal inspiratory and expiratory flow-volume curves in (A) fixed obstruction (intrathoracic or extrathoracic), in
which airway diameter does not change with either inspiration or expiration, (B) extrathoracic variable obstruction, and (O
intrathoracic variable obstruction. The dotted line indicates 50 per cent of the vital capacity (VC); the ratio of expired to inspired
flow at this point is the mid-VC ratio and is normally 0.9 to l .0. A, With a fixed obstruction, both expiratory (exp) and inspiratory
(insp) flows are equally altered and the mid-VC ratio remains normal. B, With a variable extrathoracic obstruction, forced expiration
results in a slightly positive ( + ) intratracheal pressure that is greater than the pressure around the airway (atmospheric or 0),
resulting in a decrease of the obstruction (airway dilates). During forced inspiration when pressure around the airway (0) exceeds
the intratracheal pressure ( ), the obstruction is increased (airway narrows). Since the expiratory curve is normal and the inspiratory
curve is altered, the mid-VC ratio is much greater than normal. C, With a variable intrathoracic obstruction, forced expiration
results in a very positive ( + + ) pleural pressure that is greater than the slightly positive ( + ) intratracheal pressure, resulting in an
increase of the obstruction (airway narrows). During forced inspiration, the intratracheal pressure ( - ) is greater than the pleural
pressure ( ), thus decreasing the obstruction (airway dilates). Since the inspiratory curve is normal and the expiratory curve is
attenuated, the mid-VC ratio is much less than normal. A normal flow-volume curve is a composite of the expiratory curve in
and the inspiratory curve in C. (TLC = total lung capacity; RV = residual volume.)
by the surgeon into the distal trachea
77 8 l
8 2
(Fig.
15-17#). This second endotracheal tube is con-
nected to a Y-piece and a second set of anesthetic
hoses and is handed off to the anesthesiologist in
order to continue ventilation. After excision of the
tracheal lesion and placement of the posterior tra-
cheal sutures, the second (distal) endotracheal tube
is removed from the trachea, the original (first)
endotracheal tube is advanced past the anastomosis
line and reconnected to the anesthetic circuit, and
the anastomosis is completed (Fig. 15-17C and
D). The anesthetic approach is similar in infants
with bronchopleural dysplasia and strictures
secondary to mechanical ventilation with an endo-
tracheal tube; however, a tracheostomy tube is
Table 15-5 EQUIPMENT REQUIRED FOR
TRACHEAL RECONSTRUCTION
1. Anesthesia machine capable of delivering up to 20 L/min
of 0
2
2. Assortment of endotracheal tubes
a. Polyvinylchloride tubes, size 4 mm uncuffed to size 8
mm cuffed
b. Unsterile flexible armored tubes, 20-34 F
c. Extra-long endotracheal tubes, 20-30 F
d. Sterile armored tubes, 20-34 F
3. Sterile, corrugated tubing attached to a sterile Y-piece
4. High-frequency positive-pressure ventilator (optional)
5. A second anesthesia machine
usually present before the induction of anesthe-
sia.
100
101
With a low tracheal lesion, a right thoracotomy
provides the necessary surgical exposure. If there
is sufficient trachea distal to the area of resection,
a Foley catheter with the tip cut off just distal to
the balloon may be used as a single-lumen endo-
tracheal tube; it is inserted by the surgeon and
secured just above the carina, avoiding endobron-
chial intubation and the need for one-lung anesthe-
sia.
83
Otherwise, if there is not enough distance
between the tracheal lesion and the carina to pro-
vide placement of even this homemade endotra-
cheal tube, endobronchial intubation and one-lung
ventilation are necessary
84-86
(Fig. 15-185). If
oxygenation or ventilation is inadequate, it may be
possible to decrease blood flow to the atelectatic
lung by tightening reversible snares around the
pulmonary artery of the nonventilated lung.
77
78
However, this maneuver may be technically diffi-
cult, and an alternative technique is to pass a sec-
ond endobronchial tube into the other bronchus to
provide ventilation to both lungs
86
(see later dis
cussion). As with the technique for the high tra
cheal lesion, after the posterior anastomosis is
completed, the endobronchial tube(s) is removed
and the original endotracheal tube is pushed past
the site of resection; in this situation, however, it
is likely that an endobronchial intubation may
Figure 15-16 Extended tracheal tube illustrating site of incision at the 15-mm connector and location of the cut segment within
the modified tube. This part is used as a bridge from the proximal original tracheal tube to the extension segment of a second
tracheal tube. (From Holzman RS: A tracheal tube extension for emergency tracheal reanastomosis. Anesthesiology 70:170 171.
again be required to complete the anastomosis
(Fig. l 5-18CandD).
Several methods have been described for man-
aging the airway during carinal resection.
77
78
97
'
l02
While the affected segment is being resected, left
lung ventilation may be carried out via an endo-
bronchial intubation of the left main-stem bron-
chus below this lesion (Fig. 15-195).
77 I02
After
the right main-stem bronchus and trachea have
been reattached, the left endobronchial tube is re-
Figure 15-17 Airway management and surgical procedure for resection of a high tracheal lesion. A, Initial intubation above the
lesion. B, Second endotracheal intubation distal to the lesion after the trachea has been opened. C, Placement of sutures for the
posterior anastomosis. D, The second endotracheal tube has been removed, and the original endotracheal tube has been advanced
distal to the anterior anastomosis. (Redrawn with permission from Geffin B, Bland J, Grillo HC: Anesthetic management of tracheal
resection and reconstruction. Anesth Analg 48:884. 1969.)
Figure 15-18 Airway management and surgical procedure for resection of a low tracheal lesion. A, Initial intubation above the
lesion. B, Left endobronchial intubation distal to the lesion after the trachea has been opened. C, Placement of sutures for the
posterior anastomosis. D, The endobronchial tube has been removed, and the original endotracheal tube has been advanced distal
to the anterior anastomosis into an endobronchial position. (Redrawn with permission from Geffin B, Bland J, Grillo HC: Anesthetic
management of tracheal resection and reconstruction. Anesth Analg 48:884, 1969.)
moved, and the original endotracheal tube is ad-
vanced distal to the suture line.
77, l02
The right lung
is then ventilated via this tube, while the left main
bronchus is reanastomosed to the trachea at a dif-
ferent point of origin (Fig. 15-19C and D). Again,
blood flow to the nonventilated lung can be re-
duced by tightening ties placed around the appro-
priate pulmonary artery. An alternate method of
ventilation during carinal resection is to perform
endobronchial intubation of both severed bronchi,
a technique that enables two-lung ventilation for a
much longer period during the procedure.
78
This
latter technique requires the use of two ventilating
systems. As the posterior anastomosis is being
made, ventilation is achieved through both of the
distal main-stem bronchi. During repair of the an-
terior wall, ventilation is done via the original en-
dotracheal tube (above the anastomosis site). The
air leak that initially occurs through the anasto-
mosis site diminishes progressively as the anterior
sutures are placed. If the anastomosis site leak is
excessive, the endotracheal tube can be pushed
into an endobronchial position until the placement
of additional sutures reduces the leak.
Figure 15-19 Airway management and surgical procedure for resection of a carinal lesion. A, Initial intubation above the lesion.
B, Left endobronchial" intubation distal to the lesion after the left main-stem bronchus has been severed. C, The trachea is
anastomosed to the right main-stem bronchus. D, The left endobronchial tube has been removed to allow for anastomosis between
the trachea and the left main-stem bronchus. Ventilation during (D) is accomplished via the original endotracheal tube. (Redrawn
with permission from Geffin B, Bland J, Grillo HC: Anesthetic management of tracheal resection and reconstruction. Anesth Analg
48:884, 1969.)
It is apparent that these conventional techniques
of airway management for tracheal resection are
fraught with hazard. During operation a slight
head-down tilt helps to minimize aspiration of
blood and secretions. Intermittent sighs help pre-
vent bronchiolar obstruction and atelectasis. A
high inspired oxygen concentration is employed,
since an oxygen-filled functional residual capacity
(FRC) permits a few extra minutes to correct rela-
tively common episodes of airway obstruction
and/or tube displacement. Ventilation is continu-
ously monitored by auscultation and observation
of the chest, measurement of compliance (peak
inspiratory pressure), and arterial blood-gas deter-
minations. Several different sizes of armored en-
dotracheal tubes must be available for use through-
out the procedure. Finally, close communication
must exist between the surgery and anesthesia
teams. Disadvantages of these complex airway
techniques include soilage of the lung with blood
and debris, the presence of tubes in the surgical
field, and the occasional necessity of using one-
lung (or less) ventilation.
In an effort to overcome these problems, a third
approach to airway management during tracheal
resection consists of high-flow jet ventilation
through small-bore endotracheal tubes or catheters
(see Fig. 12-9).
87
"
91
96
With this technique, a small-
bore, uncuffed catheter is placed through the ste-
notic area, and ventilation is accomplished by ex-
posing the lung to rapid, intermittent, high-flow,
fresh gas from the catheter. Oxygen jets entrain
ambient air, which, in turn, provides the volume
necessary for adequate ventilation. Using this tech-
nique, acceptable blood-gas values have been
maintained, and there have been no deleterious
effects on the circulation. With only a small cath-
eter in the field, the surgeon can more easily per-
form the tracheal resection and anastomosis.
The disadvantages of this technique, however,
include the following: entrainment of blood and
secretions into the open distal airway by the Ven-
turi effect; possible inadequate escape or air
around the jet catheter during exhalation when the
catheter is passed through a tight stenotic area,
leading to air trapping; plugging of the catheter
with blood; displacement of the catheter; aspira-
tion of blood; and technical difficulties with high-
pressure injectors.
This technique of jet ventilation for tracheal ste-
nosis has been modified with good success in two
ways.
91
" First, a small, cuffed endotracheal tube
has been used for the entire procedure (Fig. 15-
20).
99
A small endotracheal tube (6-mm internal
diameter) is placed above the stenosis, and jet ven-
tilation around the mass is begun. When the lesion
has been excised, the small jet-ventilation endotra-
cheal tube is passed into a main-stem bronchus,
and jet ventilation of just the one lung is contin-
ued. When the anastomosis has been completed
around the small endotracheal tube, the endotra-
cheal tube is brought back above the anastomotic
line.
Second, the jet ventilation through the small,
cuffed endotracheal tube may be done with an
F,0
2
of 1.0 (no entrainment of room air, blood, or
secretions).
99
The trachea is intubated with a long.
cuffed endotracheal tube (6.0-mm ID), the distal
end of which is kept above the level of the tracheal
tumor (Figs. 15-20 and 15-21).
99
A Venturi injec-
tor is inserted into the proximal end of the tracheal
tube through a self-sealing diaphragm. The orifice
of the injector is in continuity with the anesthesia
circuit, which delivers 0
2
-halothane (Fig. 15-21).
The pop-off valve of the anesthesia circuit is kept
wide open. A flow of oxygen at a pressure of 50
psi is delivered via the injector. The oxygen flow
is automatically interrupted by a solenoid valve
controlled by an electronic timer. Intermittent ox-
ygen jets are delivered at a rate of 20 per minute
with an I:E ratio of 1:2. The oxygen jets entrain
0
2
-halothane from the anesthesia circuit to main-
tain a high level of inspired oxygen concentration
and to administer the inhalation anesthetic. The
concentration of halothane delivered by the vapor-
izer varies between 2 and 3 per cent.
4. High-Frequency Positive-Pressure
Ventilation
The fourth approach to airway management dur-
ing tracheal resection utilizes HFPPV
76
(see Fig.
12-9). This technique utilizes small tidal volumes
(50 to 250 ml) delivered via a small catheter at
relatively rapid rates (50 to 150/min). Advantages
of using HFPPV for tracheal resection include a
relatively unobstructed surgical field, no interrup-
tion of ventilation during operation, minimized
contamination of the lungs from blood and debris
by a continuous outflow of gas, minimized lung
and mediastinal movement, and production of con-
tinuous positive airway pressure (CPAP), which
lessens the risk of alveolar collapse. This tech-
nique has been used by others successfully for
both tracheal
74 75
and carinal resections.
74
During
carinal operations, HFPPV to the left lung alone
generally provided adequate oxygenation and ven-
tilation, although a system of two catheters and
bilateral HFPPV could be used if necessary.
76
5. Cardiopulmonary By-Pass
The fifth approach to airway management dur-
ing tracheal resection, especially in cases of carinal
Figure 15-20 Diagram depicting
the position of the tracheal tube be-
fore resection of the tracheal tumor
(A), during tracheal reconstruction
(), and after completion of the anas-
tomosis (Q. (From Baraka A, Man-
sour R, Jaoude CA, et al: Entrap-
ment of oxygen and halothane during
jet ventilation in patients undergoing
excision of tracheal and bronchial tu-
mors. Anesth Analg 65:191-194,
TIMER
Figure 15-21 Diagram of system
used for jet ventilation. The injector
is connected to the cuffed tracheal
tube on one side and to the anesthe-
sia circuit on the other side. The in-
jector is driven by intermittent jets of
oxygen (50 psi) that are automati-
cally interrupted by an electronic
timer controlling a solenoid valve.
(From Baraka A, Mansour R, Jaoude
CA, et al: Entrainment of oxygen and
halothane during jet ventilation in
patients undergoing excision of tra-
cheal and bronchial tumors. Anesth
Analg 65:191-194, 1986. Used with
permission.)
SOLENOI D
VALVE
resection, has been the institution of cardiopulmo-
nary by-pass either at the time of resection
92
or
prior to the start of surgery.
93
Following resection,
anesthesia can be continued by conventional tech-
niques using a standard endotracheal tube. Al-
though many surgical teams perform difficult tra-
cheal resections with the cardiopulmonary by-pass
team standing by, the risk of intrapulmonary he-
morrhage due to heparinization precludes its use
in most cases.
77
Helium-oxygen breathing mixtures can signifi-
cantly decrease the resistance to gas flow through
stenotic areas and may be considered preopera-
t i ve^ for some of these patients.
93
However, the
use of helium-oxygen mixtures prevents the use of
high inspired oxygen concentration during anes-
thesia, and it is therefore not often recommended.
77
Postoperatively, most patients are kept in a po-
sition of head flexion in order to reduce tension on
the suture line. If ventilatory support is necessary
postoperatively, the endotracheal tube must be po-
sitioned so that the cuff does not rest on any suture
line. Early extubation is highly desirable in order
to minimize the compromise of blood flow to the
trachea, which might be caused by an inflated tra-
cheal cuff. Chest physiotherapy to remove secre-
tions should not be too vigorous and may need
augmentation by fiberoptic bronchoscopy. Sys-
temic antibiotics and steroids are not routinely
given unless infection or excessive edema, respec-
tively, is strongly anticipated. If massive bleeding
into the airway or chest occurs postoperatively, it
is likely due to erosion into the pulmonary artery
or aorta (high incidence following insertion of a
tracheal prosthesis) and is usually not treatable.
IV. BRONCHOLITHIASIS
A. General and Surgical Considerations
Although the term broncholithiasis implies the
presence of calcified material within the tracheo-
bronchial tree, the definition includes external
compression of a bronchus by calcified lymph
glands.
I03
-
106
Broncholiths almost invariably origi-
nate as peribronchial lymph nodes, which calcify
after an inflammatory process. Subsequent erosion
into the bronchial lumen (and potentially the pul-
monary artery and esophagus) by the contant mo-
j tion of respiration and the beating of the heart
produces the characteristic symptoms of hemopty-
sis and lithoptysis. Bronchial obstruction may also
be complicated by recurring pneumonia and ab-
scess formation. Lithoptysis, while pathognomonic
* of the condition, is seen only rarely, having been
reported in only two of 43 patients
104
and six of 41
patients.
107
The most common presenting symptom
is cough, which was reported in 67 to 100 per cent
of patients by various authors.
104,105
l07
Hemoptysis of varying degrees has been re-
ported in 37 to 85 per cent of cases.
104
I05
l07
Ate-
lectasis and/or features of obstructive pneumonia
have been described in approximately 25 per cent
of cases in two reports.
104
I05
Carcinoma of the
lungs has also been reported in association with
broncholithiasis.
107 I(,s
Most commonly, the diag-
nosis of broncholithiasis is based on symptoms
(cough, hemoptysis, and expectoration of stones)
and roentgenographic and bronchoscopic findings.
Hemoptysis in this condition may be coming
from granulation tissue, which occurs in associa-
tion with bronchial erosion, but may also be due
to a blood leak from erosion into a branch of a
pulmonary artery. The latter is particularly likely
in a patient who presents with major hemoptysis,
and in this situation endoscopic removal of the stone
may be complicated by major hemorrhage.
103 I09
Consequently, proximal control of the pulmonary
artery has been recommended as a first maneuver
to prevent fatal hemorrhage.
104
"
Conservative resection should be the intraoper-
ative goal. This can occasionally be accomplished
by a local resection with or without a bronchoplas-
tic procedure.
104
"" Adhesions generally make a
lobectomy the most popular procedure. Rarely, in-
jury to the pulmonary artery necessitates a pneu-
monectomy.
104
" Rigid bronchoscopy is a reason-
able option only when the bleeding is slight and
provided that, in case the bleeding becomes major/
uncontrollable, surgical intervention is immedi-
ately available.
1031
"
4
The two dominant anesthetic concerns are he-
morrhage and distal infection; both problems re-
quire separation of the lungs (preferably with a
double-lumen tube, which maximizes suctioning
capability). Hemorrhage and infection require lung
separation to prevent further contamination of the
unaffected side as well as to clean up the unaf-
fected side (especially before turning the unaf-
fected side into the dependent position).
Hemorrhage also requires preparation for large-
volume infusion. In addition, lymph node calcifi-
cation in this disease is frequently associated with
an intense inflammatory response, which makes
surgical dissection difficult and increases signifi-
cantly the likelihood of intraoperative complica-
tions, specifically, hemorrhage and esophageal
damage. Consequently, preoperative preparation
by having large intravenous catheters in place and
extra blood available is necessary.
V. GIANT BULLOUS EMPHYSEMA
AND AIR CYSTS
Cystic disease of the lung is a relatively com-
mon disease. The majority of the lung cysts are
acquired, and a minority result from developmen-
tal defects in the tracheobronchial tree. Although a
satisfactory classification of cystic disease of the
lungs is difficult, a practical one has been proposed
in Table 15-6 based on the pathogenesis of the
cyst.
1
" Developmental lung cysts usually become
symptomatic in infancy and childhood; these are
discussed further in chapter 18. Those that do not
become symptomatic until adulthood do so be-
cause they become infected/abscessed; therefore,
the main anesthetic consideration is lung separa-
tion/one-lung ventilation.
Similar concepts apply to traumatic and hydatid
lung cysts. Otherwise, from the point of view of
anesthesia for thoracic surgery for adults, the im-
portant lung cysts are degenerative bullae, and the
anesthetic management of bullectomy is discussed
in detail. For the sake for completeness, the term
bleb usually connotes a subpleural collection of air
within the layers of visceral pleura caused by a
ruptured alveolus; the ruptured alveolus could re-
sult from any of the entities listed in Table 15-5,
but bullae-induced blebs are by far the most com-
mon. The air dissects through the interstitial tissue
into the thin, fibrous layer of visceral pleura, where
it accumulates to form a bleb. Rupture of a bleb is
often associated with the development of a spon-
taneous pneumothorax."
2
1. Developmental Lung Cysts
a. CENTRAL CONGENITAL INTRAPULMONARY
BRONCHOGENIC CYST
This is a rare, relatively simple anomaly result-
ing from abnormal budding and branching of the
Table 15-6 CLASSIFICATION OF CYSTIC
DISEASE OF THE LUNGS*
1. Developmental
Central congenital cyst-intrapulmonary bronchogenic
cyst
Peripheral congenital cyst of lung
Congenital cystic adenomatoid malformation
Pulmonary sequestration
2. Traumatic
Traumatic lung cyst
3. Parasitic
Pulmonary hydatid cyst
4. Inflammatory
Pneumatocele ..
5. Degenerative
Bullous emphysema
Honeycomb lung
*Based on data from Shamji et al.
1
"
tracheal diverticulum between the third and sixth
weeks of gestation. Small groups of cells that ulti-
mately develop as a cyst separate from either the
primitive main lung buds or their lobar and seg-
mental bud branches. Bronchogenic cysts may be
found in the lung or in the mediastinum. The ma-
jority of uninfected bronchogenic cysts cause no
symptoms and are discovered incidentally on a
routine chest radiograph.
When symptoms do occur, which are usually
hemoptysis, fever, cough, and expectoration of pu-
rulent secretions, they indicate the presence of
bronchial communication and infection. In neo-
nates, communication between the cyst and the
tracheobronchial tree may result in a rapid expan-
sion of the cystso-called tension cystproduc-
ing acute cardiorespiratory embarrassment and
death. Surgical excision is the treatment of choice
for both the infected and the uninfected broncho-
genic cyst. Excision is justified in an asymptomatic
patient to establish diagnosis and to prevent infec-
tion.
b. PERIPHERAL CONGENITAL CYSTS OF THE
LUNGS
These represent a disorder of bronchial growth,
occurring during the sixth to 16th weeks of gesta-
tion, in which the distal ramifications of the bron-
chial tree become separated from the main bron-
chial branchings. The resulting cysts may or may
not communicate with the parent bronchus and
tend to be multiple, either limited to a lobe or
generalized, involving one or both lungs. Whereas
limited cystic disease does not interfere with nor-
mal growth, the generalized variety is responsible
for the honeycomb lung seen in infants dying soon
after birth. The absence of bronchial communica-
tion produces asymptomatic lung cysts that are
discovered on routine chest X-ray. However, the
bronchial communication is often present, setting
the stage for two complications: infection and ten-
sion cysts. If the cyst becomes infected, it may
resemble a lung abscess, with cough, fever, chills,
mucopurulent sputum, hemoptysis, chest pain, and
an air-fluid level on chest X-ray.
A tension cyst causing acute cardiorespiratory
embarrassment (displaced mediastinum and de-
pressed ipsilateral diaphragm) may require urgent
decompression by needle aspiration or chest tube
insertion. A quiescent cyst subsequently should be
removed by the most conservative resection pos-
sible: cystectomy, segmentectomy, or lobectomy.
C. CONGENITAL CYSTIC ADENOMATOID
MALFORMATION
This is a less common form of cystic lung dis-
ease, characterized by a large, rubbery multicystic
mass of pulmonary tissue produced by abnormal
proliferation of the terminal bronchioles at the ex-
pense of alveoli. This malformation is usually lo-
calized to one lobe. A diagnosis of congenital cys-
tic adenomatoid malformation requires surgical
excision by lobectomy. The lesion is discussed in
chapter 18.
d. BRONCHOPULMONARY SEQUESTRATION
This is an uncommon congenital malformation
characterized by a mass, usually cystic, of non-
functioning lung tissue that is detached from the
normal communication with the tracheobronchial
tree and receives most or all of its arterial blood
supply from an anomalous systemic artery. Bron-
chopulmonary sequestration probably occurs as a
result of an accessory lung bud arising from the
foregut caudad to the normal lung bud around the
fifth to sixth week of gestation. If this accessory
lung bud arises relatively early, the abnormal non-
functioning lung tissue will be sequestered within
a normal pulmonary lobe inside its visceral pleura
and is called an intralobar sequestration.
Later development of the accessory lung bud
results in complete structural separation of the ab-
normal, nonfunctioning lung tissue (accessory lobe
or segment) from the rest of the normal function-
ing lung but is within the pleural cavity and is
invested with its own visceral pleura. This is called
an extralobar sequestration. Both forms of seques-
tration, extralobar and intralobar, have an anoma-
lous systemic arterial blood supply, which usually
arises directly from the aorta either above or below
the diaphragm, occasionally from an intercostal
artery, and rarely from the brachiocephalic artery.
Differences between intralobar and extralobar se-
questrations are outlined in Table 15-7. Note that
intralobar sequestrations are much more likely to
present in adulthood (with infection).
The treatment of symptomatic intralobar seques-
tration is usually lobectomy, because recurrent in-
fection obliterates the normal segmental planes
within the lobe, making safe segmental resection
only occasionally possible. The treatment for ex-
tralobar sequestration is sequestrectomy (resection
of just the sequestration and no normal tissue). The
only serious technical hazard of the operation for
sequestration is accidental transection of the anom-
alous systemic artery, which may then retract to
an inaccessible point within the mediastinum or
below the diaphragm, resulting in uncontrollable
bleeding.
2. Traumatic Lung Cyst
The development of a traumatic lung cyst is
rare. Recognition of a traumatic lung cyst, how-
ever, is important because it resolves sponta-
neously and requires no special treatment. The
proposed mechanism of traumatic lung cysts is
sudden compression of an area of the lung against
a closed airway (either bronchus or glottis), caus-
ing increased pulmonary pressure in the com-
pressed area of the lung and rupturing the lung
tissue. Consequently, the cyst may contain only air
or may be more or less filled with blood derived
from torn alveolar capillaries. The symptoms are
often not severe, and spontaneous resolution usu-
ally occurs between 2 to 16 weeks.
3. Hydatid Cyst of the Lung
Hydatid (watery) cysts result from the develop-
ment of the intermediate larval stage of the dog
tapeworm, Echinococcus granulosus, in humans or
other intermediate hosts (sheep, cow, or pig). Hy-
datid disease is frequently encountered in the
Table 15-7 CHARACTERISTICS OF PULMONARY SEQUESTRATION
Characteristic Intralobar Extralobar
Incidence
Sex
Laterality
Location
Pleural investment
Age at presentation and symptoms
Arterial supply
Venous drainage
Associated anomalies
Foregut connection
Bronchial communication
Uncommon
Equal
60%, left
Posterior basal segment
Not separate; sequestration part of
normal lobe
50% > 20 years, adolescent to
young adult; recurrent pulmonary
infection
Systemicfrom aorta; often a
single, large vessel
Pulmonary-inferior pulmonary
vein
Uncommon
Very rare
Present, small
Rare
Male (80%)
90%. left
Above or below diaphragm
Has its own separate pleural
investment-visceral pleura
60% < l year, neonate; respiratory
distress
Systemicfrom pulmonary artery or
aorta; usually small vessels
Systemicazygos or hemiazygos
vein; less often portal vein
Common (>50%), e.g., congenital
diaphragmatic hernia (30%)
More common
None
sheep- and cattle-raising regions of the world (i.e.,
Australia, New Zealand, South Africa, South
America, and the Mediterranean countries of Eu-
rope, Asia, and Africa). Indeed, in one hospital (in
Ankara, Turkey) over a 16-year period, 1115 pa-
tients with pulmonary hydatidosis were treated
surgically.
113
A human may become infected, usually in
childhood, when his or her hands or food becomes
contaminated with infected canine (primary host)
fecal material containing eggs, which are ingested.
The eggs hatch in the stomach and proximal small
intestine, releasing the embryos (larval stage),
which penetrate the proximal intestinal mucosa to
enter the portal venous circulation. Most of the
embryos are filtered out in the liver, but some
escape the liver to be disseminated, the common
sites being the lungs, brain, and bones. The hyda-
tid fluid resembles crystal-clear water: It is odor-
less and colorless. It suspends the daughter cysts
in solution.
A number of significant complications are found
with hydatid cyst of the lung. The cyst progres-
sively increases in size at a rate varying from a
few millimeters to 5 cm a year, growing more
rapidly in children. Pressure effects occur owing
to compression by the hydatid cyst of adjacent
structures (i.e., bronchus, superior vena cava,
esophagus, or neurovascular structures of the tho-
racic inlet). Spontaneous rupture of the hydatid
cyst into a bronchus may cause any of the follow-
ing: (1) Dramatic expectoration of cyst fluid and
fragments of the laminated membrane (a sponta-
neous cure may follow the evacuation of cyst con-
tents and collapse of the cyst wall); (2) sudden
death from asphyxiation caused by flooding of the
bronchial tree with hydatid fluid; (3) sudden death
from anaphylactic reaction to the cyst contents; (4)
secondary infection forming a chronic lung ab-
scess or a localized area of bronchiectasis mani-
fested by chronic cough, mucopurulent or dark
bloody sputum, and repeated febrile episodes; (5)
no symptoms.
Diagnosis of pulmonary hydatid cyst should be
suspected in any patient who lives in an endemic
area, who is older than 3 years, and who presents
with single or multiple opacities on the chest
radiograph. The indirect hemagglutination test is
the clinical test of choice, with a false-positive rate
of only 1 to 2 per cent. The diagnosis of a pulmo-
nary hydatid cyst is an indication for surgery.
Complete excision of the disease process with
maximum conservation of lung tissue is desirable.
The principles of surgical excision are as fol-
lows. The intact cyst may be removed by simple
enucleation without needle aspiration. During de-
livery of the cyst, it should not be touched or
grasped to avoid rupture. The surgical field must
always be protected from the possibility of spillage
of cyst fluid. The intact cyst may be rendered
sterile by first instilling hypertonic saline solution
(10 ml of 10 per cent NaCl solution) into the cyst.
The intact cyst may also be aspirated before re-
moval.
After the endocyst and its laminated membrane
have been removed, one should locate the patent
bronchial openings, which are invariably present
in the fibrous wall of the ectocyst or pericyst.
These openings must be carefully closed with fine
suture. At times, segmental resection or lobectomy
may be necessary for removal of pulmonary hy-
datid cyst if the lung tissue has been destroyed by
prolonged compression or infection.
4. Pneumatocele
Pneumatoceles are inflammatory in origin, char-
acteristically appearing during the healing phase of
pneumonia. They are radiolucent zones in the lung
fields that become more clearly visible as the
pneumonia resolves and then disappear sponta-
neously.
5. Bullous Emphysema
The same mechanisms of dilation and destruc-
tion that are responsible for the development of
emphysema are responsible for the formation of
the bullae. In fact, just as emphysema may occur
on a familial basis, so may bullae (and repeated
episodes of spontaneous pneumothorax)."
4
De-
struction of lung parenchyma from inflammation
results in loss of its elasticity and formation of an
air-filled thin-walled space within the lung or
cyst."
5
"
6
The walls of the bullae may be formed
by connective tissue septa and/or compressed lung
parenchyma and/or pleura.
Morphologically, bullae may be classified into
three types: type I or narrow-necked bullae; type
II or superficial (a pleural surface) broad-based
bullae; and type III or deep (surrounded by em-
physematous tissue) broad-based bullae.
There are several complications of bullae. Ex-
piratory obstruction, with progressive air trapping,
causes the space-occupying bulla to enlarge
slowly. Occasionally a "tension bulla" may pro-
duce acute cardiorespiratory embarrassment. The
development of a spontaneous pneumothorax
caused by rupture of a bulla may constitute an
emergency indication for surgical treatment if the
pneumothorax is recurrent, persistent, or life-en-
dangering. Bullae sometimes develop an air-fluid
level and appear to have a clinical picture of lung
infection and abscess; however, surgery is not al-
ways necessary"
2
because complete resolution
may be obtained medically."
7
Alternatively, in-
fected bullae may be drained percutaneously."
8
Bullectomy is the surgical resection of one or
more bullae and is performed only in selected pa-
tients. The most important physiologic effect of
bullae is to compress the surrounding lung, and
the release of the lung compression after bullec-
tomy is considered to be the most important factor
in symptom relief rather than eliminating dead
space ventilation."
1
Indications for bullectomy in chronic obstruc-
tive pulmonary disease patients include intolerable
breathlessness even after full medical therapy, rap-
idly enlarging bullae, or the repeated occurrence
of pneumothorax (a history of which most patients
have).
119
The goal of surgery is to remove the
bullae while preserving as much functioning lung
as possible. Therefore, lobectomy is not only un-
desirable but often unnecessary. Patients who are
considered too compromised to tolerate an open
approach may tolerate a closed thoracoscopic ap-
proach.
120
The compression of a large area of lung can be
visualized on the chest roentgenogram (as well as
on an angiogram) (e.g., Fig. 15-22)"
2
as the
crowding together of pulmonary vessels. However,
computed tomography is a more accurate way to
localize the bulla as well as to determine its func-
tional characteristics and the nature of the sur-
rounding lung architecture (Fig. 15-23).
121-123
In
addition, an equally strong case for functional im-
pairment can be made if radioisotope studies show
that the compressed area has good perfusion and
some, but reduced, ventilation
124
125
(see chapter 5
and Fig. 5-22). Whole-lung pulmonary function
tests are not useful for determining candidates for
bullectomy.
121
The long-term results of bullectomy for giant
bullae in emphysema are encouraging.
126
In 27 em-
physematous patients who had either unilateral (10
patients) or bilateral (17 patients) bullae that oc-
cupied more than 50 per cent of the hemithorax,
bullectomy significantly decreased dyspnea, they
were no longer functionally disabled, and mean
survival time was greater than 7 years. In fact, the
larger the bullae (and presumably the amount of
compressed normal lung), the better are the func-
tional results.
127
The postoperative spirographic improvement
depends on the type of bullae (e.g., resection of
bullae with open communication with the bron-
chial tree results predominantly in improvement of
forced expiratory flow in 1 sec [FEV,] as a per-
centage of vital capacity). In addition, the more
normal the compressed lung, the more dramatic,
positive, and enduring are the results.
128
Thoracoscopic C0
2
laser treatment of bullous
emphysema in 20 of 22 patients has been success-
Figure 15-22 A, Chest roentgenogram demonstrates large
^biapical bullae. B, Pulmonary angiogram demonstrates intact
but crowded vasculature. (From Klingman RR. Angelillo VA.
DeMeester TR: Cystic and bullous lung disease. Ann Thorac
Surg 52:576-580, I99l. Used with permission.)
ful in moderately increasing forced vital capacity.
FEV,, and maximal exercise treadmill times. Thor-
acoscopic carbon dioxide laser oblation of bullae
has been so successful because it maximizes pres-
ervation of normal lung better than any open tho-
racotomy approach. It is likely that this approach
to bullectomy will be used with increasing fre-
quency in the next few years.
129
13()
Intracavitary drainage by tube thoracostomy
(which requires rib resection, a purse-string pleu-
rotomy, and a large balloon catheter) is another
method of preserving normal lung tissue.
131
"
133
The
long-term prognosis after any surgical treatment of
bullous emphysema is, of course, diminished in
any patient who has chronic purulent bronchitis
preoperatively and/or who does not abandon
smoking.
INSP EXP
Figure 15-23 Three-dimensional
reconstruction (posterior oblique
view) of the computed tomography
scan of the lungs of a patient with a
large bulla in the left upper lobe. The
measured volume of the bulla does
not change from inspiration (INSP)
to expiration (EXP), but there is a
change in shape as a result of distor-
tion. This would invalidate the meas-
urement of ventilation from the
change in area of the bulla in a single
slice. (From Morgan MDL, Denison
DM, Strickland B: Value of com-
puted tomography for selecting pa-
tients with bullous lung disease for
surgery. Thorax 41:855-862, 1986.
Anesthesia for the removal of bullae involves
several specific ventilation hazards (Table 15-8).
First, most of these patients have severe gen-
eralized chronic lung disease with little or no ven-
tilatory reserve. Thus, ventilation (which needs to
be controlled once the chest is opened) of one
severely diseased lung (the one without the giant
bulla) may be hazardous and runs the risks of
hypoxemia, hypercarbia, and pneumothorax on the
ventilated side. If the ventilated lung also contains
a bulla, the risks are obviously even greater. In
addition, since general anesthesia is necessary for
the procedure, it is likely that many patients with
severe lung disease will be committed to at least a
short period of postoperative mechanical ventila-
tory support.
Second, when a bulla or air cyst is in commu-
nication with a bronchus, positive-pressure venti-
lation may cause it to increase in size.
134
If a sig-
nificant portion of the tidal volume enters the
bullous cavity, alveolar dead space ventilation will
be greatly increased, and unless there is an equiv-
alent increase in minute ventilation, the rest of the
lung may be inadequately ventilated. This compli-
cation is most likely to occur when the chest is
opened because the chest wall no longer limits the
expansion of the bulla.
Third, because of the rapidity with which closed
airspaces take up nitrous oxide and expand in
size,
135
this agent is best avoided (especially in Table 15-8
patients whose bullae are thought to have poor
communication with the bronchial system).
136
Fourth, if a check valve is present in the airway
that communicates with the cavity, overinflation
and air trapping may occur within the cavity.
Fifth, and most important, positive pressure
within the bulla might cause it to rupture, creating
a pneumothorax, which would likely be under ten-
sion if the chest were closed (especially in patients
whose bullae are thought to have good communi-
cation with the bronchial system).
134
Tension pneu-
mothorax in these patients is usually a catastrophic
event owing to the impairment of venous return
and cardiac output as well as further compromise
of ventilation. Insertion of a chest tube at this point
would create, in effect, a large bronchopleural cu-
taneous fistula that could divert much of the ven-
tilation out through the chest tube. Recently, high-
frequency ventilation with low tidal volumes and
airway pressure has been used successfully to
avoid positive-pressure rupture of a bulla.
137
The cornerstone of the anesthetic management
of patients with giant bullae or cysts, whether they
are undergoing thoracoscopic or open procedures,
is the insertion of a double-lumen tube to allow
differential treatment of the two lungs. Thus, in
patients with unilateral disease, the double-lumen
tube can allow adequate ventilation of the nondis-
eased side while preventing rupture of the diseased
side.
In patients with bilateral disease, the double-
lumen tube still allows differential lung treatment
to maximize gas exchange as well as provides an
increased capability to deal with the complications
of a ruptured bulla. For example, a double-lumen
tube allows all possible permutations of high-fre-
VENTILATION HAZARDS DURING
BULLECTOMY
I. Rest of lung is diseased
II. Bullae may increase in size owing to:
A. Intermittent positive-pressure breathing
B. N
2
0
C. One-way check-valve
III. Bulla may rupture pneumothorax
quency ventilation, CPAP, positive end-expiratory
pressure (PEEP), and zero end-expiratory pressure
to the two lungs, depending on the pathology in
each lung (see Differential Lung Ventilation, chap-
ter 20). In addition, as each bulla is resected, the
double-lumen tube allows ventilation to the oper-
ated lung to be re-established for short periods,
enabling the surgeon to identify and suture any air
leaks that may be present (identification may in-
volve filling the hemithorax with saline and
searching for bubbles).
129
A ruptured bleb is usu-
ally small and cannot be seen as a hole and is
always covered with fibrin materials. A ruptured
emphysematous bulla is usually large enough to
be identified as a hole.
129
The versatility of the double-lumen tube is par-
ticularly important in thoracoscopic procedures,
which may average 3 hours in duration.
120
A double-lumen endotracheal tube can be in-
serted with the patient either awake and the airway
topically anesthetized (for those with histories of
repeated pneumothorax and/or bilateral bullae) or
under general anesthesia (for the majority of pa-
tients) in order to isolate the affected lung and
provide positive-pressure ventilation to the contra-
lateral lung.
138
139
While the depth of general an-
esthesia is being increased, spontaneous ventila-
tion may be maintained (most indicated in patients
with a history of repeated pneumothorax or bilat-
eral bullae), but it should be realized that sponta-
neously breathing patients with significant pulmo-
nary disease under general anesthesia probably
will not be able to ventilate themselves adequately.
Alternatively, and preferably in the majority of
patients, the patient can be anesthetized and both
lungs ventilated using a limited amount of positive
airway pressure; gentle ventilation by hand is the
best way to ensure low airway pressures. If a ma-
jor air leak develops intraoperatively while posi-
tive-pressure ventilation is being used with a vol-
ume-cycled ventilator, inadequate ventilation may
occur (the machine delivers the preset volume but
it does not go into the lungs). Under these circum-
stances, a pressure-cycled high inspiratory flow
ventilator must be used (e.g., Siemens Servo 900C,
Solna, Sweden).
120
If positive-pressure ventilation is used while the
chest is closed (always the case for the nonopera-
tive thorax), it is important that the anesthesiolo-
gist be able to diagnose and treat a pneumothorax
rapidly. External stethoscopes should be attached
over each hemithorax at the points where breath
sounds are maximal in order to monitor for pneu-
mothorax on each side.
140
However, it should be
realized that advanced bullous disease may com-
pletely prevent breath sounds from being heard
externally at all. In addition to a decrease in breath
sounds, a pneumothorax or check-valve mecha-
nism in a cyst may be signaled by an increase in
airway pressure, tracheal shift to the opposite side,
or hypotension out of proportion to the depth of
anesthesia.
Equipment for chest tube placement must be
immediately available for these patients. However,
chest tube placement for a ruptured bulla poses
two problems. First, rupture of only one of several
bullae present in a lung may result in a somewhat
localized pneumothorax that is decompressed only
by precise localization of the chest tube. Second,
as discussed previously, a large bronchopleural cu-
taneous fistula may be created when the chest tube
is placed, making ventilation difficult (see prior
discussion and Bronchopleural Fistula, chapter
17).
Theoretically, after bullectomy, the patient's
pulmonary status should be improved because the
"healthy
1
' lung tissue that was previously com-
pressed by the bullae should now be able to ex-
pand. However, in my and others' experience,
120
the weaning and extubation process, especially the
spontaneous subsidence of air leaks, may take up
to several days in some patients with advanced
disease. When mechanical ventilation is required
postoperatively, positive airway pressure should be
minimized, if possible, to decrease the chance of
producing a pneumothorax from rupture suture
lines and/or residual bullae.
141
When large air leaks do occur, a pressure-cycled
ventilator may be required. If the leak is greater
than 50 per cent the tidal volume, then the
patient may require re-exploration.
120
When the air
leak is large, the chest tube should be put to water
seal only (active suction greatly increases the
leak).
120
When the air leak is large and refractory
to all manipulations, including surgical explora-
tion, percutaneous partial extracorporeal carbon
dioxide removal and return to spontaneous venti-
lation may be tried (and has been used success-
fully).
142
*
When bilateral bullectomy is done because of
extensive disease in both lungs, a sternal splitting
incision with the patient supine is usually used.
143
Indeed, some surgeons routinely use a mediasti-
notomy and bilateral approach/exploration because
of the high frequency of bilateral problems (e.g..
33-60 per cent spontaneous occurrence of unilateral
pneumothorax after unilateral operations).
144 I4S
However, sequential posterolateral thoracotomies
may be planned if it is desired to see how the
patient responds to the first bullectomy. With bi-
lateral bullectomy, the same anesthetic principles
apply as with unilateral bullectomy. A double-lu-
men endotracheal tube is again the cornerstone of
management, allowing differential lung treatment
throughout the operation. If sequential posterolat-
eral thoracotomies are planned, provision for gain-
ing access to the closed chest should be made in
the way the skin is prepared and the patient is
draped in case suspicion of a pneumothorax in the
closed chest arises.
VI. PULMONARY RESECTION IN
PATIENTS AFTER
PNEUMONECTOMY
Patients who have had a pneumonectomy may
have a new lesion appear in the remaining lung;
the density, if it persists and increases in size, is
almost always malignant. Pneumonectomy has
been generally considered in the past to be a con-
traindication to further pulmonary resection. How-
ever, six case reports involving eight patients
146-151
and one recent series of 15 patients
152
have pro-
vided evidence that limited pulmonary resection
after pneumonectomy is feasible with a low oper-
ative mortality. The long-term results show that
resection of these "secondary" tumors can result
in a reasonably prolonged and worthwhile survival
in some patients.
These cases must be done with a single-lumen
endotracheal tube. Fluid balance must be carefully
regulated because the size of the pulmonary vas-
cular bed will be decreased further from an already
small size and the risk of pulmonary hypertension
will be increased by the resection. Thus, all aspects
of postoperative care, such as maintenance of fluid
balance (risk of pulmonary edema and pulmonary
hypertension), positive-pressure ventilation (risk
of right-sided heart failure), and adequate evacua-
tion of the pleural space (risk of lung compres-
sion), must be undertaken with increased care be-
cause of the patient's increased sensitivity to
derangements in any of these factors.
VII. UNILATERAL
BRONCHOPULMONARY LAVAGE
153
Unilateral bronchopulmonary lavage or massive
irrigation of the tracheobronchial tree of one lung
has been employed with good success in patients
with pulmonary alveolar proteinosis as a means of
removing the enormous accumulations of alveolar
lipoproteinaceous material that these patients char-
acteristically have.
154-161
The lipoproteinaceous
material is thought to be surfactant,
162
and the ab-
normal accumulation is due to failure of clearance
mechanisms rather than enhanced formation.
163
The abnormal accumulation of alveolar lipopro-
teinaceous material is bilateral and symmetric (but
can be unilateral)
164
and causes the classic chest
roentgenographic picture of airspace consolidation
with patchy, poorly defined shadows throughout
the lung.
165
Very commonly, the distribution of X-
ray opacities is "butterfly" or "bat wing"; the
roentgenographic picture parallels the course of
the disease. Computed tomographic scan of the
chest may show the extent of the disease more
clearly than plain radiography.
166
The diagnosis is made in most patients between
the ages of 20 and 50 years (although the disease
has been described in patients from 0.5 to 72
years).
164
Most patients have had symptoms for 2
to 3 years before the definitive diagnosis is made.
The male to female ratio is 4 to 1, and many
patients (approximately 50 per cent) have a history
of exposure to dusts and chemicals.
164
The occur-
rence of the disease in several siblings suggests a
genetic predisposition. The differential diagnosis
consists of sarcoidosis, extrinsic allergic alveolitis,
tuberculosis, and pulmonary fibrosis.
The airspace consolidation causes progressive
hypoxemia and shortness of breath (first on exer-
tion and then finally at rest),
165
167
and the lungs
have a low compliance. Cough, fever, and chest
pains are other common symptoms. Pulmonary
function tests reveal a restrictive pattern
164
and are
depressed the most in patients with severe X-ray
changes. The diagnosis of the alveolar proteinosis
is made by correlating these clinical, roentgeno-
logic, and laboratory data, but the definitive diag-
nosis rests with the results of a lung biopsy (usu-
ally transbronchial at the time of fiberoptic
bronchoscopy),
168
169
but an open thoracotomy
may be required to obtain an adequate tissue spec-
imen.
164
The transbronchial specimens should be
multiple and from the area of greatest change on
X-ray.
169
The indications for lavage consist of P
a
0
2
less
than 60 mm Hg at rest or hypoxemic limitation of
normal activity.
170
I71
Infrequently, lung lavage may
be performed in patients with asthma, cystic fibro-
sis, and radioactive dust inhalation.
154,159
l72
m
Unilateral lung lavage is performed under gen-
eral anesthesia with a double-lumen endotracheal
tube, allowing lavage of one lung while the other
lung is ventilated (Fig. 15-24). In patients with
alveolar proteinosis, lavage is performed on one
lung and then after a few days rest on the other
lung. After lung lavage, these patients usually have
marked subjective improvement, which correlates
with increases in P
a
0
2
during rest and exercise,
vital capacity and diffusion capacity, and clearing
Anesthesia for Special Elect'v ' erapeutic Procedures 549
Normal saline
30 cm above lung
Clamp
Left side conn ted
to anesthesia circuit
Left side double-lumen tube
PA Catheter
in Right
Lung
-Drainage by gravity
Figure 15-24 Technique for providing unilateral pulmonary lavage. A left clear plastic double-lumen endotracheal tube allows
ventilation to the left lung during lavage of the right lung (and vice versa). Normal saline is infused and drained by gravity; clamps
on the connection tubes determine direction of fluid flow. (PA = pulmonary artery.)
of the chest roentgenogram.
164 1 7 4
l 7 5
Some patients
require lavage every few months, whereas others
remain in remission for several years, and the dis-
ease may even eventually completely remit.
164 I68
174. 175 Prolonged spontaneous remission is also
possible.
164176
153
This section discusses the technique for manag-
ing bronchopulmonary lavage in patients with pul-
monary alveolar proteinosis. When the patient is
admitted to the hospital, ventilation-perfusion
scans of the lung are obtained. Ventilation can be
maximized during lung lavage by performing the
first lavage on the most severely affected lung,
allowing the "better" lung to provide gas ex-
change. If the scan indicates relatively equal in-
volvement (as is usually the case), the left lung is
lavaged first, leaving the larger right lung to sup-
port gas exchange.
Unilateral lung lavage is performed in the oper-
ating room, where the appropriate amount and
type of equipment and ancillary personnel are
present to enhance safety. Patient safety is also
enhanced if a relatively constant team, composed
of members of the departments of anesthesia and
pulmonary medicine, becomes familiar with the
nuances and technique of unilateral lung lavage.
Patients are usually cooperative and require only
light premedication. Since many of these patients
are hypoxemic at rest, they are given oxygen by
face mask following premedication and during
transport to the warmed operating room.
The procedure takes several hours to complete,
and lavage fluid temperature cannot always be pre-
cisely controlled. Thus, some patients require ex-
ternal warming (e.g., with heating blanket) to
maintain normal body temperature. The monitor-
ing system consists of a blood pressure cuff, elec-
trocardiogram, precordial stethoscope, temperature
probe, pulse oximeter, and peripheral arterial and
central venous pressure catheters. Patients with
compromised cardiovascular function are moni-
tored with a pulmonary artery catheter in place of
the central venous catheter.
After several minutes of preoxygenation (see
later discussion), general anesthesia is induced
with 3 to 4 mg/kg of thiopental in divided doses
and inhalation of either isoflurane or halothane in
100 per cent oxygen. Isoflurane is relatively indi-
cated in patients in whom therapeutic levels of
theophylline (and the risk of arrhythmias) are
present. Neuromuscular blockade is monitored
with a peripheral nerve stimulator and is induced
with a nondepolarizing muscle relaxant. When a
suitable level of anesthesia has been reached, the
trachea is topically anesthetized with lidocaine and
intubated with the largest size left-sided double-
lumen endotracheal tube that can be passed atrau-
matically through the glottis (see chapter 9). A
clear, plastic, disposable left-sided double-lumen
tube is used because of the ease and certainty with
which it is correctly positioned, the reliable left-
cuff seal obtained (the right endobronchial cuff is
small and inflates asymmetrically), and the ability
to continuously observe the tidal movements of
respiratory gas moisture (ventilated lung) and the
lavage drainage fluid for leaking air bubbles (la-
vaged lung). The largest size tube is used because
the left endobronchial cuff will make contact over
a greater bronchial mucosal area with less air in
the left cuff (compared with a small double-lumen
tube). In addition, a large tube facilitates suction-
ing, which is an important consideration at the end
of the case when the lungs need to be made as
clear as possible.
Precise placement of the tube and detection of
leaks are essential because of the serious hazard of
spillage during the lavage procedure. The position
of the double-lumen tube must be confirmed with
a fiberoptic bronchoscope as described in chapter
9, and cuff seal must be demonstrated to hold
against 50 cm H
2
0 pressure using the catheter-
under-water technique described in chapter 9. The
eyes should be protected with a lubricant and eye
pads.
The question of patient position during unilat-
eral lung lavage is important for there are major
advantages and disadvantages related to each po-
sition (Table 15-9). The lateral decubitus position
with the lavaged lung dependent minimizes the
possibility of accidental spillage of lavage fluid
from the dependent, lavaged lung to the nondepen-
dent, ventilated lung. However, during periods of
lavage fluid drainage, pulmonary blood flow,
which is gravity dependent, would preferentially
perfuse the nonventilated, dependent lung, and the
right-to-left transpulmonary shunt would be maxi-
mal. The lateral decubitus position with the la-
vaged lung nondependent minimizes blood flow to
the nonventilated lung but, on the other hand, in-
Table 15-9 UNILATERAL LUNG LAVAGE:
PATIENT POSITION
I. Lateral decubitus with lavaged lung nondependent
Advantage: minimizes blood flow to the nonventilated
lung
Disadvantage: maximizes the possibility of spillage
II. Lateral decubitus with lavaged lung dependent
Advantage: minimizes the possibility of spillage
Disadvantage: maximizes blood flow to the
nonventilated lung
III. Supine position
Balances spillage and blood flow distribution problems
551
creases the possibility of accidental spillage of la-
vage fluid from the lavaged lung to the dependent,
ventilated lung. As a compromise, the supine po-
sition is used in order to balance the risk of aspi-
ration against the risk of hypoxemia.
Following insertion and checking of the double-
lumen endotracheal tube and positioning of the
patient, baseline total and individual lung compli-
ance should be measured. Airway pressure can be
electronically transduced and continuously re-
corded on a paper write-out, and a Wright spirom-
eter should be placed in the expiratory limb of the
anesthesia circle system in order to accurately
measure tidal volume. A volume ventilator that
can deliver relatively high inflation pressures is
required, since these patients have diseased and
noncompliant lungs. Prlavage total dynamic com-
pliance (chest wall and lung) of both lungs to-
gether (using 15 ml/kg of breath) and then of each
lung separately (using 10 ml/kg of breath) is meas-
ured. Following measurement of total and individ-
ual lung compliance, and with the patient breath-
ing 100 per cent oxygen, baseline arterial blood
gases are measured.
The patients are completely preoxygenated prior
to the induction of anesthesia and lavage for two
reasons. First, as with the induction of general
anesthesia in any patient, an oxygen-filled FRC
greatly minimizes the risk of hypoxemia during
the apneic period required for laryngoscopy and
endotracheal intubation. This consideration has in-
creased importance for patients with alveolar pro-
teinosis since they are already severely hypox-
emic. Second, preoxygenation eliminates nitrogen
from the lung that is to be lavaged. Alveolar gas
is then composed only of oxygen and carbon diox-
ide. During fluid filling, these gases will be ab-
sorbed, which allows the lavage fluid maximal ac-
cess to the alveolar space. Failure to remove
nitrogen from the lung prior to filling with lavage
fluid may leave peripheral nitrogen bubbles in the
alveoli and thus limit the effectiveness of the la-
vage.
Warmed isotonic saline is used as the lavage
fluid and is infused by gravity from a height of 30
cm above the midaxillary line. After the lavage
fluid ceases to flow (i.e., lung filling is complete),
drainage is accomplished by clamping the inflow
line and unclamping the drainage line, which runs
to a collection bottle placed 20 cm below the mid-
axillary line (Fig. 15-24). The inflow and outflow
fluid lines are connected to the appropriate endo-
tracheal tube lumen by a Y-adapter. Each tidal
lavage filling is accompanied by mechanical chest
percussion and vibration to the lavaged hemitho-
rax prior ^to drainage. The lavage fluid that is
drained is typically light brown, and the sediment
layers out at the bottom of the collection bottle
Figure 15-26 Changes in shunt (Q
s
/Q,[%]) and cardiac output (CO) associated with each of the seven bronchopulmonary
lavages. (From Cohen E, Eisenkraft JB: Bronchopulmonary lavage: Effects on oxygenation and hemodynamics. J Car-
diothorac Anesth 4:609-615, 1990. Used with permission.)
tension and cardiac output are in opposite direc-
tions and have different phase lags, the trend in
mixed venous oxygen saturation remains constant
(see Fig. 15-25).
177
With respect to measuring pulmonary vascular
pressures and central venous pressures, the situa-
tion is analogous to measuring these intrathoracic
vascular pressures during PEEP (absolute intralu-
minal vascular pressure relative to atmospheric
pressure increases, whereas transmural pressure
[relative to pleural pressure] remains constant or
decreases [see chapters 7 and 20]). An adequate
degree of neuromuscular blockade must be main-
tained because unexpected vigorous coughing dur-
ing the procedure could alter double-lumen endo-
tracheal tube position.
If a small leak should occur during lavage, the
following may be observed sequentially: (1) the
appearance of bubbles in the lavage fluid draining
from the lavaged lung, (2) rales and rhonchi in the
ventilated lung, (3) a difference between lavage
volumes administered and those drained from the
lavaged lung (the former exceeds the latter), and
(4) a fall in arterial oxygen saturation. If a small
leak is suspected or detected by any of these signs
and the lavaged lung has been only minimally
treated, the lavaged lung should be drained of all
fluid, and the position of the double-lumen endo-
tracheal tube, the adequacy of cuff seal, and sepa-
ration of the lungs should be rechecked with a
fiberoptic bronchoscope. Before beginning the la-
vage procedure again, and no matter what the dou-
ble-lumen tube malposition was, the functional
separation of the two lungs and adequacy of cuff
seal should be tested and found adequate by using
the previously described air bubble leak-detection
method.
Massive spillage of fluid from the lavaged lung
to the ventilated lung is not a subtle event and
results in a dramatic decrease in ventilated lung
compliance and a rapid and large decrease in arte-
rial oxygen saturation. Under these circumstances,
the lavage procedure must be terminated no matter
how much treatment has been accomplished. The
patient should be moved quickly to the lateral de-
cubitus position with the lavaged side dependent,
and the operating room table should be placed in
a head-down position in order to facilitate removal
of lavage fluid. Vigorous suctioning and inflation
of both lungs should be carried out. The double-
lumen tube should be changed to a standard sin-
gle-lumen tube, and the patient should be addition-
ally treated with a period of mechanical ventilatory
support with PEEP. Timing of further unilateral
lung lavage attempts will be dictated by the
patient's subsequent clinical course and gas-
exchange status.
After the effluent lavage fluid becomes clear,
the procedure is terminated. The lavaged lung is
thoroughly suctioned, and ventilation is begun.
Since the compliance of the lavaged side will be
much less than that of the ventilated side at this
time, large tidal ventilations (sighs) (15 to 20 ml/
kg) to that side alone (with the nonlavaged side
temporarily nonventilated) are necessary to re-ex-
pand alveoli. Arterial blood oxygenation may de-
crease precipitously during this time, but this can
be minimized by clamping the nonlavaged side
after a large inspiration of 100 per cent oxygen.
After lavage, the recovery procedure consists of
repetitive periods of large tidal ventilations, suc-
tioning and chest wall percussion to the previously
lavaged side, conventional two-lung ventilation
with PEEP, and bilateral suctioning and postural
drainage while intermittently measuring combined
(total) and individual lung dynamic compliance.
As the compliance of the lavaged lung returns
toward prlavage values, ventilation with an air-
oxygen mixture may help lavaged lung alveoli
with low ventilation-perfusion ratios to remain
open. When the compliance of the hemithorax of
the lavaged side returns to its prlavage value, the
neuromuscular blockade is reversed. Mechanical
ventilation and extubation guidelines are the same
as for any patient with pulmonary disease (see
chapter 20); most patients are able to be extubated
while still in the operating room. If the patient is
not considered a candidate for extubation, the dou-
ble-lumen tube is changed to a single-lumen tube,
and the patient is mechanically ventilated with
PEEP in a conventional manner.
In the immediate postlavage period, deep
breathing (incentive spirometry), coughing exer-
cises, chest percussion, and postural drainage are
used to remove remaining fluid and secretions and
to re-expand the lavaged lung. After 3 to 5 days of
recovery, the patient is returned to the operating
room to have the opposite side lavaged. The anes-
thetic considerations for the second lavage are the
same as those for the first lavage, although oxy-
genation is usually not nearly as severe a problem
as during the first lavage because the treated and
now near-normal lung will be used to support gas
exchange.
There are two special problems associated with
pulmonary lavage that may be encountered. First,
a few critically ill patients may be unable to toler-
ate the conventional procedure. Second, unilateral
lavage through a double-lumen tube is not possible
in children and small adults.
There are three alternative (and more compli-
cated) ways of accomplishing lung lavage in
patients who simply cannot tolerate one-lung ven-
tilation under any circumstance. First, extracorpo-
real membrane oxygenation (ECMO) has been
utilized to provide support of gas exchange during
standard unilateral lung lavage. Partial venoarterial
cardiopulmonary by-pass has been used for a few
hours during unilateral or bilateral lung lavage,
179-181
but the distribution of oxygenated blood from the
venoarterial by-pass can be markedly nonho-
mogeneous and dependent on the site of blood
return
182
l83
and requires major arterial cannulation.
Venovenous by-pass allows uniform arterial dis-
tribution and the safety of not requiring major ar-
terial cannulation.
184
I8S
In one of these latter
cases,
185
partial venovenous (femoral veininfe-
rior vena cava to internal jugular veinsuperior
vena cava connection) cardiopulmonary by-pass
with a membrane oxygenator was especially suc-
cessfully used to exchange respiratory gases dur-
ing and after bilateral lung lavage in a patient with
severe hypoxemia due to alveolar proteinosis (the
patient had previously been unable to tolerate oth-
erwise conventional unilateral lung lavage). Dur-
ing by-pass, the lungs were mechanically venti-
as a second try in patients who had unacceptable
oxygenation during lung drainage on a first try.
Lavage in the conventional manner is not pos-
sible in children (or small adults) in whom double-
lumen endotracheal tubes are too large to be in-
serted. This problem routinely occurs in persons
weighing less than 25 to 30 kg, since the small-
est double-lumen endotracheal tube made is 28
French, with each lumen being slightly less than
4.5 mm. In this situation, partial cardiopulmonary
by-pass has been successfully used to provide ox-
ygenation during unilateral or bilateral lavage.
179

189. 190 j
n o n e 0
f
m e s e
reports, the technique was
used in two brothers aged 4 and 2.5 years.
189
Both
these patients underwent whole-lung lavage, dur-
ing which time blood was removed from both fem-
oral veins, oxygenated, and then returned to the
left femoral artery. Both patients were eventually
discharged from the hospital, although they contin-
ued to require supplemental oxygen by face mask.
In another report, the technique was used to sup-
port gas exchange during whole-lung lavage in a
ventilator-dependent, 3.7-kg, 8-month-old child.
190
The extracorporeal oxygenation system was again
venoarterial (right internal jugularright atrium
catheter to right axillary artery catheter). Marked
improvement in pulmonary function was noted
after lavage (total 420 ml/kg), by-pass was able to
be discontinued 3 hours after lavage, and the pa-
tient was extubated 48 hours following lavage. In
the last report,
189
partial venoarterial by-pass with
a bubble oxygenator permitted bilateral simultane-
ous lung lavage in two siblings aged 3 and 4 years.
By-pass in these patients was carried out with fem-
oral vein and femoral artery cannulation. During
by-pass, radial artery P
a
0
2
ranged between 25 and
30 mm Hg, which may, in part, have been related
to continued cardiac output of desaturated blood
into the proximal aorta. No neurologic sequelae
were noted. Oxygenation and functional levels
were improved following lavage. These various
efforts to treat pulmonary alveolar proteinosis in
children with bilateral lung lavage support by ex-
tracorporeal oxygenation must be regarded as suc-
cessful, considering that in children the average
survival from the time of severe symptoms without
this kind of treatment is less than 1 year.
Finally, one report describes a 7-year-old patient
who was lavaged via a single-lumen tube placed
endobronchially with a fiberoptic bronchoscope,
while the nonlavaged lung was ventilated from
above via an anesthesia mask or nasopharyngeal
airway (the mouth was sealed with a transparent
dressing [Opsite]).
191
In this case, a 3.5-mm flexi-
ble bronchoscope was passed through a 4.5-mm
ID, cuffed endobronchial tube 50 cm long. The
bronchoscope and tube were passed through a
bronchoscope adapter attached to a tight-fitting an-
esthesia mask. Using the bronchoscope, the endo-
bronchial tube was guided into the right main-stem
bronchus with its tip just above the right middle-
lobe bronchus. The bronchoscope was then with-
drawn.
In theory, placement of the endobronchial cath-
eter in the right main-stem bronchus with its bal-
loon cuff below the right upper lobe bronchus al-
lowed ventilation of the right upper lobe as well
as the entire left lung. While the patient was ven-
tilated through the mask, the right middle and
lower lobes were lavaged with 250- to 400-ml
aliquots of saline to a total of 8 L, using gravity to
fill and drain the lung. During this ventilation pro-
cedure/setup, the stomach may gradually become
distended, and an orogastic tube should be passed
for gastric decompression.
The success of this technique depends on sev-
eral factors. First, the endobronchial catheter must
be positioned accurately; the flexible broncho-
scope greatly facilitates this. The catheter must
remain fixed in the proper position; this is facili-
tated by neuromuscular blockade. Vigilance in
monitoring the breath sounds and the position of
the tube is necessary because if the tube were to
slip proximally, or the balloon cuff failed, fluid
could flood both lungs. A second requirement is
that the endobronchial catheter must be large
enough to allow effective lavage yet small enough
to allow effective ventilation around it. The ge-
ometry of a cylinder within a cylinder is such that
there is much more cross-sectional area available
for ventilation around a single tube than would be
available through two parallel tubes (Fig. 15-
27).
I91
Figure 15-27 shows an example of the con-
siderations involved in the selection of the size of
the endobronchial lavage catheter.
VIII. PULMONARY
ARTERIOVENOUS
MALFORMATIONS
192
Although most cases of pulmonary arteriove-
nous malformations (including aneurysms) are
congenital, the condition is often not recognized
until the second decade of life.
192
It has been sug-
gested that the aneurysm progressively enlarges
with age in response to increasing flow, with even-
tual necrosis of the vessel wall. This may suddenly
increase the magnitude of the right-to-left shunt or
cause hemorrhage, which then leads to clinical de-
tection. There are four presentations that appear to
be sufficient to raise the suspicion of the diagnosis
even when they are found in isolation: hemoptysis,
abnormalities on chest roentgenogram that are
Figure 15-27 Geometry of a tube within the tra-
chea. The tracheal diameter is 8 mm, and that of the
4.5-mm internal diameter endobronchial tube is 5.5
mm. The cross-sectional area of the endobronchial
tube lumen (A) is 15.9 mm
2
, whereas the cross-
sectional area of the trachea remaining for ventila-
tion around the tube (B) is 26.3 mm
2
. If the patient
were to be ventilated through the 4.5-mm tube and
a second tube (for lavage) were passed alongside,
th largest internal cross-sectional area (Q the tra-
chea could accommodate would be 1.76 mm
2
, thus
wasting 21.4 mm
2
of area (#') potentially available
for ventilation. (From MacKenzie B, Wood RE,
Bailey A: Airway management for unilateral lung
lavage in children. Anesthesiology 70:550-553,
|tKTr.itvv
consistent with pulmonary arteriovenous malfor-
mation, evidence of a right-to-left shunt, and cen-
tral nervous system phenomena attributed to in-
fected or noninfected emboli without an obvious
source.
193
The association between pulmonary
arteriovenous malformations and Osler-Weber-
Rendu disease is well known and pulmonary arte-
riovenous formation occurs most commonly in pa-
tients with the Osier-Weber-Rendu disease. The
presence of symptoms and abnormal physical
aigris is correlated with the size of the pulmonary
arteriovenous malformation and not with the num-
ber of lesions.
Because the essential defect is right-to-left
shunting from the pulmonary artery to pulmonary
vein, it is not surprising that arterial oxygen ten-
sion is reduced in more than 80 per cent of cases,
and shunt magnitudes of as much as 79 per cent
of the cardiac output have been recorded.
194
In the
presence of significant shunting and arterial desat-
uration, a secondary polycythemia results. In gen-
eral, hemodynamic variables including intracar-
diac, pulmonary arterial, and pulmonary artery
occluded pressures and cardiac output are normal.
! Although the chest radiograph is abnormal in
approximately 98 per cent of patients with pulmo-
nary arteriovenous malformations (a single periph-
eral, circumscribed, noncalcified lesion that is con-
nected by blood vessels to the hilus of the lung is
the most common finding), pulmonary angiogra-
phy remains the diagnostic standard for these le-
sions. In surgical candidates, angiography is man-
datory to establish the number, extent, and location
of the lesions and to delineate the arterial supply
and venous drainage.
I Therapeutic options include surgery, emboliza-
tion, or regular follow-up without invasive treat-
ment in those patients with small lesions that are
clinically silent. However, and most important, the
natural history of pulmonary arteriovenous malfor-
mations is far from benign, and treatment by
embolization or surgical incision should be consid-
ered in all cases. In other words, the complications
from the disease are much greater than the compli-
cations from the treatment.
When an aneurysm is distal to the main pulmo-
nary trunk, resection becomes a therapeutic option.
In the past, resection was often the only therapeu-
tic option for these lesions.
193
Some surgeons have
opted for pneumonectomy as the intervention for
aneurysms of a main pulmonary artery or multiple
more peripheral aneurysms.
193
Peripheral aneu-
rysms have also been treated with lobectomy or
with resections of lesser amounts of pulmonary
parenchyma.
193
Although such resections can be
performed with low morbidity/mortality, surgery
has limitations as a therapeutic approach in pa-
tients with limited pulmonary function or in ten-
uous medical condition.
Over the last few years, embolotherapy (percu-
taneous localization and occlusion of vascular le-
sions) under local anesthesia has become the treat-
ment of choice for pulmonary arteriovenous
malformation not involving the pulmonary trunk
or main pulmonary arteries. The dual circulation
of the lungs allows interruption of pulmonary ar-
teries with minimal necrosis. This fact, coupled
with advances in interventional radiology, has dra-
matically altered the approach to pulmonary arte-
riovenous malformation located distal to the main
pulmonary arteries over the past few years. Em-
bolectomy can now be performed with minimal
loss of lung tissue and minimal morbidity, and no
mortality has been reported to date.
193
There are three main anesthetic considerations.
First, lung separation is necessary for two reasons.
As with all major resections, lung exposure is fa-
cilitated by collapse of the operative lung. In ad-
dition, in these cases, the surgeon will need to gain
control of the vasculature by gaining control of the
hilar vessels. Thus, whether a pneumonectomy is
being performed or not, collapse of the operative
lung will facilitate exposure of the hilar structures.
Second, multiple large-bore intravenous routes
of access are necessary because of the potential for
hemorrhage with these cases. In addition, an index
of left ventricle preload is desirable (i.e., central
venous pressure and/or pulmonary artery occluded
pressure).
Third, hypoxemia may be profound in those pa-
tients with a high degree of right-to-left shunting.
Hypoxemia may be minimized by using an in-
spired oxygen concentration of 100 per cent and
operative lung CPAP. However, early control of
hilar vessels is the most definitive treatment of
shunting through the operative lung in these cases.
Blood gases may need to be measured from pe-
ripheral arteries as well as from various pulmonary
veins to prove that the resection of the pulmonary
arteriovenous malformation has been complete.
195
IX. LUNG TRANSPLANTATION
1. Types of Lung Transplantation and
Their Indications
Four types of lung transplantation are currently
being performed. Table 15-10 lists the four types,
the indications for each type (by category of disease
as well as by specific examples/problems),
196-207
whether cardiopulmonary by-pass is required, and
the major limitations to each type of transplanta-
tion or technique. The first operation, single-lung
transplantation with bronchial anastomosis to the
transplanted lung, is widely used when the remain-
ing native lung is not infected and cardiac function
is not a problem. Chronic idiopathic pulmonary
interstitial fibrosis, emphysema (especially a,-anti-
trypsin deficiency), and pulmonary hypertension
with satisfactory right ventricular function are
three specific examples of this general category of
patient.
200
Idiopathic pulmonary fibrosis is associ-
ated with a 40 to 80 per cent mortality within 5
years of diagnosis.
208
Severe life-threatening em-
physema may occur in the fourth and fifth decades
of life in individuals with a,-antitrypsin deficiency
of the homozygous type variant. The prevalence
of homozygous type is estimated to range from
1 in 1600 to 1 in 5000, but only a minority, almost
always smokers, present with end-stage obstruc-
tive respiratory disease.
209
Deficiency of a,-pro-
tease inhibitors is associated with a lack of protec-
tion against neutrophil elastase in the lower
respiratory tract. Transplantation of a normal lung
in juxtaposition to a lung with either pulmonary
fibrosis
206
or obstructive airways disease
207
results
in the majority of ventilation and perfusion divert-
ing to the new lung, resulting in good V/Q match
and excellent oxygenation.
206,207
The natural history of pulmonary hypertension
is quite variable, but the typical mean survival is 2
to 3 years from the time of diagnosis.
210
The
course of this disease is one of progressive right-
sided heart failure that ultimately results in the
patient's death. The major issue in assessing pa-
tients with pulmonary hypertension for transplan-
tation has been the variability in the natural history
of the disorder. In some or many patients with
pulmonary hypertension and reasonable right ven-
tricular function, right ventricular and pulmonary
vascular function has significantly improved after
the insertion of a new low-load pulmonary circuit
in the form of a new single-lung transplant.
198,204-206
For example, in one report, mean pulmonary artery
pressure decreased from 71 to 15 mm Hg and
mean right ventricular ejection fraction increased
from 26 to 41 per cent.
205
However, the single-
lung transplant for pulmonary hypertension does
not completely relieve hypoxemia because vir-
tually all blood flow is diverted to the new trans-
planted lung, whereas ventilation is equally split
between the native and transplanted lung, both of
which have a normal compliance.
206
Patients with these three disorders are otherwise
excellent lung-transplant candidates because of
their relatively young age and lack of other organ
system involvement.
The second type of lung transplantation, bilat-
eral, sequential single-lung transplantation with bi-
lateral bronchial anastomosis to each transplanted
lung, is essentially equivalent to two sequential
single-lung transplantations.
193,199,211,212
This oper-
ation is indicated when both lungs are infected and
therefore must be removed (if one of the two in-
fected lungs remained, it would infect the newly
transplanted lung; indeed, foci of infection in a
remaining lung will likely be exacerbated by
necessary immunotherapy). Examples of patients
whose main disability is secondary to conditions
that result in chronic sepsis are cystic fibrosis and
end-stage emphysema/bronchitis. Indeed, more
than 98 per cent of patients with cystic fibrosis die
from pulmonary-related complications by the third
to fourth decade of life.
213
The third type of lung transplantation, double
lung with tracheal anastomosis, is most often per-
formed on full cardiopulmonary by-pass. The in-
dications for this operation are rapidly changing at
the time of this writing, but the operation has been
done for virtually every disease that might benefit
from lung transplantation; however, because of the
disappointment with the healing of the tracheal
anastomosis and the satisfaction with the healing
of the bronchial anastomosis of the single-lung
Table 15-10 TYPES OF LUNG TRANSPLANTATION AND THEIR INDICATIONS, WHETHER CPB IS
UTILIZED, AND THE MAJOR LIMITATIONS TO THE TRANSPLANTATION TECHNIQUE*
Type of
Transplantation Category of Problem Specific Examples CPB
Major Limitations to
Transplantation
Technique
Single lung, bronchial
anastomosis
Bilateral sequential
single lung, bronchial
anastomosis
Double lung, tracheal
anastomosis
Double-lung plus heart
Terminal lung function,
no infection,
satisfactory cardiac
function
Terminal lung function,
bilateral infection,
good cardiac function
All of the above
Terminal lung and
cardiac function
Chronic interstitial
fibrosis, emphysema
(all types but
especially a,-
antitrypsin
deficiency),
pulmonary
hypertension with
satisfactory RV
function (new
application)
Cystic fibrosis, end-
stage infection, and
emphysema
All of the above
No, partial femoral
arteryfemoral vein
standby
No, partial femoral
arteryfemoral vein
standby
Yes
Severe pulmonary
hypertension with
right ventricular
failure, Eisenmenger's
syndrome
Yes
1. Requires patient to
tolerate 1LV
2. Requires patient to
tolerate unilateral
pulmonary artery
clamping
Same as above (single
lung)
1. High incidence
(approximately 50%)
of poor tracheal
healing
2. Hemorrhage caused
by heparinization for
CPB
1. Hemorrhage resulting
from heparinization
for CPB
2. Separate independent
risk of heart rejection
3. High incidence of
advanced coronary
artery sclerosis
4. High incidence of
obliterative
bronchiolitis (chronic
rejection)
*Based on data from Raju et al.,
196
Bisson & Bennette,
197
Kaiser & Cooper,
198
Kaiser et al.,
199,20
Calhoun et al.,
2
Levine et al.,
204
Starnes et al.,
205
Kramer et al.,
206
and Yacoub et al.
207
Abbreviations: CPB = cardiopulmonary by-pass; RV = right ventricular; 1LV = one-lung ventilation.
Cooper,
transplant, the double-lung procedure is falling out
of favor.
199
200
202
The tracheal anastomosis does
poorly because it does not have the mediastinal
collaterals (particularly from the coronary arte-
ries)
214
that the double-lung plus heart transplant
does.
The fourth type, combined double-lung and
heart transplantation, is used for patients with both
very severe/terminal cardiac and lung disease. The
lung disease may be primary (e.g., primary pul-
monary hypertension) and cardiac disease secon-
dary (e.g., right ventricular failure), or the cardiac
disease may be primary (Eisenmenger's syndrome
with septal defect) and the lung disease secondary
(e.g., marked increases in pulmonary artery pres-
sure). Unfortunately, this operation has the intra-
operative complications related to cardiopulmo-
nary by-pass as well as the postoperative
complications of rejection episodes of both the
heart and the lung (which may occur completely
independently of one another),
215
sclerosis of the
transplanted coronary arteries,
203
and late oblitera-
tive bronchiolitis (resulting from a chronic state of
rejection).
203
The third and fourth lung transplan-
tation techniques (double lung and double lung
plus heart) are not discussed here because they are
outside the scope of this textbook (i.e., they re-
quire cardiopulmonary by-pass).
The overall results of lung transplantation are
very good. Aside from intraoperative problems
(see major limitation to transplantation techniques
in Table 15-10 and anesthetic considerations
later), the major immediate/subacute concern is the
healing of the airway anastomosis. Lung allografts
are unique among solid organ transplants in that
their systemic bronchial arterial supply is not re-
constructed at the time of transplantation, and
blood supply to the anastomosis depends on the
development of retrograde collateral flow from the
transplanted pulmonary artery.
216
217
Although
omentopexy (omental wrap) provides immediate
support/augmentation of collateral blood flow to
the bronchial anastomosis (and therefore good
healing), omentopexy has not had a great impact
on the healing of tracheal anastomosis.
217
Never-
theless, the overall 1-year survival worldwide for
all types of lung transplantations is 60 per cent,
although individual centers report survival of 75
per cent.
218
Although the lungs remain deinnervated after
transplantation, there are no long-term differences
in patterns of breathing in recipients either when
awake or asleep compared with that of normal
individuals. The absence of a peripheral lung
cough reflex has not proven to be problematic (the
patient can still consciously cough). Rehabilitation
of patients after successful surgery is excellent
with restoration of a normal lifestyle with little or
no functional restriction. Although reoccurrence of
disease in transplanted lungs remains a worry,
there is to date no published evidence to suggest
that it has caused clinical problems for all the
diseases listed in Table 15-10.
2. Patient Evaluation and Selection
Criteria
It is obvious that candidates for lung transplan-
tation should have end-stage disease with signifi-
cant functional impairment that interferes with ac-
tivities of daily living. Most of these patients, with
the exception of the pulmonary hypertensive, re-
quire oxygen 24 hours a day. To support the need
for transplantation further, there should be docu-
mented evidence of disease progression with an
anticipated life expectancy of 12 to 18 months or
less without transplantation intervention. Potential
candidates need to be extremely well motivated to
cope with the stresses associated with the preop-
erative and postoperative periods and with the
lifelong care required after transplantation. The
complete selection criteria of one institution are
summarized in Table 15-11.
198
One of the critical physiologic insults of single-
lung and bilateral sequential single-lung trans-
plants into the recipient is the clamping of the
pulmonary artery before pneumonectomy. This
can lead to marked increases in pulmonary arterial
pressure and may signal or indicate the onset of
pulmonary edema and right-sided heart failure.
The factor on which the transplant operation
hinges, therefore, is the ability of the recipient's
right ventricle to maintain sufficient cardiac output
in spite of the acute increase in pulmonary vascu-
lar resistance. Thus\ it is necessary to obtain some
idea of right ventricular function before operation
(see chapter 5). Several groups have used right
ventricular ejection fraction, obtained from a mul-
Table 15-11 SELECTION CRITERIA FOR
LUNG-TRANSPLANT
RECIPIENTS*
I. End-stage lung disease with evidence of progression of
disease and life expectancy <12-18 months
II. No other systemic disease
III. No significant coronary artery disease
IV. Demonstrated compliance with medical regimens
V. No contraindication to immunosuppression
VI. Psychologically stable; no history of alcohol or drug
abuse
VII. Must be ambulatory with 0
2
as required
VIII. Must not be on systemic steroids
IX. Single-lung transplant
A. Age < 60 years
B. No chronic infectious lung disease such as chronic
bronchitis, bronchiectasis, or cystic fibrosis
X. Bilateral lung transplant
A. Age < 50 years
*From Kaiser LR, Cooper JD: The current status of lung
transplantation. Adv Surg 25:259-307, 1992. Used with per-
mission.
tiple-gated technetium scan of the right ventricle,
as their index.
198, 2I9
Arbitrarily, until more is
known, their cutoff point has been a value of 20
per cent; those with a right ventricle incapable of
ejecting this proportion of the diastolic volume are
considered unfit for single-lung transplantation.
Systemic steroids must be discontinued before
consideration for lung transplantation because of
the adverse affects of these agents on bronchial
healing.
220
It is often a lengthy and difficult process
to wean these patients off steroids, and in some
cases it may not be possible.
Patients who are accepted for transplantation
move to the area and enroll in an outpatient pul-
monary rehabilitation program. While awaiting
transplantation, patients often have doubled their
distance on the 6-min walk test. It has been the
experience of many that the conditioning provided
by a pulmonary rehabilitation program allows for
earlier weaning from the ventilator and facilitates
ambulation after transplantation. Every effort is
made to transplant only those patients who are
well enough to function in an outpatient setting.
Contraindications to lung transplantation include
advanced age, the presence of active systemic or
significant coronary artery disease, renal insuffi-
ciency, psychological instability, or perceived in-
ability to comply with the complex postoperative
medication regimen.
3. Donor Organ Considerations
One of the limiting factors to the widespread
application of lung transplantation (particularly for
the millions of patients with emphysema who are
in the fifth and sixth decades of life) remains the
Table 15-12 LUNG DONOR CRITERIA*
Age < 55 years
Normal chest X-ray, no infiltrates or consolidation
P
a
0
2
> 250 mm Hg on F,0
2
l.O, positive end-expiratory
pressure 5 cm H-,0 for 5 minutes
No evidence of purulent secretions or aspirated
gastrointestinal contents seen at bronchoscopy
No significant chest trauma or pulmonary contusion
No previous major thoracic surgery
Smoking history?
mission.
scarcity of suitable lungs for implantation. Proba-
bly only 5 to 10 per cent of available donors have
lungs that currently are considered acceptable for
transplantation, even when the heart, liver, kidney,
and pancreas are suitable. Of these, it is estimated
that organs are retrieved from only 20 to 30 per
cent of these potential organ donors because of
religious, social, or other reasons.
I98
Suitable donors are young patients (younger
than 40 years) diagnosed as brain dead and desig-
nated as potential organ donors, who have had a
short period of ventilatory support, without post-
catastrophe pulmonary injury or barotrauma, and
who do not have a history of significant pre-exist-
ing lung disease. The criteria for the acceptance of
donor lungs are, by design, somewhat rigid,
221
and
are summarized in Table 15-12.
198
The issue of
the smoking history of a potential donor is a diffi-
cult one. Lungs from nonsmokers are preferred,
but the reality of donor supply does not often al-
low this luxury. In general, a donor lung may be
accepted if the smoking history is less than 10
pack-years.
198
In general, an attempt is made to provide a
prospective recipient with donor lungs that would
be an appropriate size for that individual in the
absence of their lung disease. For patients with
pulmonary fibrosis, a larger lung than their native
lung is chosen because considerable mediastinal
shift occurs, the diaphragm lowers, and the thorax
expands to accommodate the new normal lung. In
contrast, patients with chronic obstructive pulmo-
nary disease have an overexpanded chest cavity
that will decrease in size when air trapping is no
longer a clinical problem; hence, lungs that are
smaller, by as much as 5 to 10 cm in vertical
height and 5 cm in horizontal chest diameter, have
been used.
198
As with other organs transplants,
lung transplantation requires ABO compatibility.
Guidelines for the management of the multiple-
organ donor are summarized in Table 15-13.
198
The technical aspects of donor pneumonectomy
are mainly surgical considerations. The lung is
removed intact with as long a piece of main bron-
chus and of main pulmonary artery as possible.
The pulmonary veins are removed from the heart
with a cuff of left atrium (Fig. 15-28).
201
Techni-
cally, the left lung is easier to extract and reim-
plant (see Fig. 15-28). For double-lung trans-
plants, the donor operation is similar, except the
pulmonary veins from both lungs share a common
atrial cuff and the trachea is removed (Fig. 15-
29) 196 Yh
e
double-lung block is easily split for the
bilateral sequential single-lung transplant proce-
dure.'
99
With current preservation techniques, total do-
nor ischemic time should be kept under 6 hours.
This constraint currently limits flying time be-
tween donor and recipient hospitals to 2Vi hours.
The lung may be preserved by flushing cold elec-
trolyte solution through the pulmonary artery to
affect rapid cooling and gain a degree of cytopro-
tection. To allow more uniform distribution of the
perfusate, the lungs are ventilated during the pul-
monary artery flush. The specimen is then im-
mersed in cold solution and packed for transport.
Table 15-14 provides a complete listing of lung-
preservation methods used to date.
198
It is currently
believed that injury to donor lung results not only
from ischemic insult but also from injury occur-
ring at the time of reperfusion of blood through
the preserved organ. Several experimental models
of acute lung injury implicate oxygen free radicals
in the genesis of reperfusion injury.
222
223
A variety
of mediators, including prostaglandin metabolites.
Table 15-13 GUIDELINES FOR MANAGEMENT
OF MULITPLE-ORGAN DONORS*
Maintain mean arterial blood pressure > 70 mm Hg
Central venous pressure not to exceed 10 cm H
:
0 and/or
pulmonary capillary wedge pressure not to exceed 12 mm
Hg
Vasopressor support to maintain blood pressure with either
dopamine 2.5-10 g/kg/min or phenylephrine 0.06-0.18
mg/min
Treat diabetes inspidus with vasopression (5-10 units every 8
hours intravenously) or DDAVP (0.3 g/kg intravenously
over 30 minutes)
Fluid replacementprevious hour's urine output minus 100
ml
Maintain urine output at 1-2 ml/kg/hr
Replace electrolyte losses
Maintain normothermia (35 C-37 C)
P
a
G\ > 100 on lowest F,0
:
possible
Maintain 5 cm H.,0 positive end-expiratory pressure
Frequent tracheobronchial suctioning; specimen for Gram's
stain
Arterial blood gases every 2 hours
Current chest radiograph
transplantation. Adv Surg 25:259-307. 1992. Used with per-
mission.
Abbreviation: DDAVP = l-deamino-8-D-arginine vaso-
pressin.
LMB
Figure 15-28 Donor lung with the pulmonary veins (PV),
pulmonary artery (PA), and the stapled left main bronchus
(LMB). The lung is triple-bagged in a sterile plastic container
with iced saline slush for transport. (From Calhoun JH,
Groover FL, Gibbons WJ, et al: Single lung transplantation:
Alternative indications and technique. J Thorac Cardiovasc
Surg 101:816-824, 1991. Used with permission.)
Figure 15-29 Donor double lung
with retained pleuropericardial flaps and
atrial cuff around the^ pulmonary veins.
(From Raju S, Heath BJ, Warren ET,
Hardy JD: Single and double-lung trans-
plantation. Ann Surg 211:681-692,
Table 15-14 METHODS OF LUNG
PRESERVATION*
I. Hypothermic immersion
A. With atelectasis
B. With inflation
C. With ventilation
D. With hyperbaric oxygen
II. Normothermic perfusion
A. Autoperfused heart-lung preparation
III. Hypothermic perfusion
A. Core cooling, with or without perfusion
IV. Flushing
A. Extracellular fluid solutions
1. Low potassium dextran
2. Fujimura
3. Ringer's, saline
4. Plasma, blood
5. Rheomacrodex
B. Intracellular fluid solutions
1. Collins
2. Sacks
V. Pharmacologic additives
A. Prostaglandins
1. PGI
2
B. Calcium channel blockers
C. Free radical scavengers
1. Superoxide dismutase
2. Catalase
3. Glutathione
4. Dimethyl sulfoxide
*From Kaiser LR. Cooper JD: The current status of lung
mission.
play a role in the vascular and bronchial smooth-
muscle response to injury in the lung. It is likely
that some of these mediators are also important in
the response of the lung to ischemia.
4. Surgical Operation
The surgical technique for single-lung trans-
plantation, although requiring great attention to
numerous details, is still relatively straightforward.
With all considerations being equal, most surgeons
chose the technically easier left side to transplant.
However, the side to transplant may depend on
several other factors. If the recipient has had a
major thoracotomy or previous pleurodesis, it is
far easier to transplant the opposite side. If a sig-
nificant proportion of arterial perfusion, as meas-
ured by a radionuclide lung perfusion scan, goes
to one side, the other side should be transplanted,
which allows some additional protection during
the early postoperative period when the function
of the transplanted lung may be somewhat com-
promised. Occasionally, the side for transplanta-
tion is chosen based on donor considerations such
as a unilateral infiltrate seen on the donor chest
radiograph or the availability of only one lung
when the other was being used by another center.
The recipient is subjected to three surgical pro-
cedures in sequence, one of which (the second
described later) is now only occasionally per-
formed (optional) and is done only for safety rea-
sons. The first procedure, with the patient supine,
is a small midline laparotomy through which the
omentum is mobilized. With its vascular pedicle,
part of the omentum is pushed up through the
diaphragm into the thorax behind the xiphisterum
from where it can be retrieved and wrapped
around the donor-recipient bronchial or tracheal
anastomosis (Fig. 15-30).'
96
The omental wrap
provides early vascularization to the airway suture
line. However, it should be noted that the problem
of healing of the bronchial anastomosis has been
solved in one institution by use of nonabsorbable
prolene sutures and a telescoping anastomosis (in-
stead of using an omental wrap and thereby avoid-
ing an abdominal incision).
201
Second, in some high-risk patients, after closure
of the abdomen, a femoral vein and artery can be
exposed and the skin closed to facilitate the speedy
implementation of femoral artery to femoral vein
by-pass should an untoward event occur. More
commonly, the groin area is just simply prepared
and left exposed in case rapid access is needed for
the institution of partial cardiopulmonary by-pass.
If the by-pass is established, it may be used for
rewarming of the patient at the end of the proce-
dure.
2
"
Third, in the case of a single-lung transplant, the
patient is then repositioned for a posterolateral tho-
racotomy and a pneumonectomy is carried out. In
the case of bilateral sequential single-lung trans-
plantation, the patient is left supine and a bilateral
fourth interspace anterior thoracotomy and sternot-
omy are performed. The pulmonary artery and
veins are divided as far distally from the hilum as
feasible, and for single-lung transplants the bron-
chus is divided near the origin of the upper lobe.
196
Connection of donor lung to recipient is in se-
quence: pulmonary veins, on their atrial pedicle, to
a zone of recipient left atrium isolated in a vascu-
lar clamp; pulmonary artery anastomosis; and fi-
nally, the bronchial anastomosis (Figs. 15-31 and
15-32) and omental wrap (see Fig. 15-30).
224
The
bilateral sequential single-lung transplantation op-
eration is essentially the same as two single-lung
transplants in series.
197
Double-lung and combined
heart and lung transplantation with cardiopulmo-
nary by-pass are beyond the scope of this textbook
and are not discussed here.
1. Anesthetic Drugs
In practice, single-lung transplantation is a
pneumonectomy in a patient who, under normal
"r }
Omental Patch
Figure 15-30 Omental wrap around tracheal su-
ture line. (From Raju S, Heath BJ, Warren ET, Hardy
JD: Single- and double-lung transplantation. Ann
Surg 211:681-692, 1990. Used with permission.)
Figure 15-31 Single left-lung transplant with anastomosis of the donor pulmonary veins (PV) to the recipient left atrium (LA).
(PA = pulmonary artery; LMB = left main bronchus.) (From Calhoun JH, Groover FL, Gibbons WJ, et al: Single lung
transplantation: Alternative indications and technique. J Thorac Cardiovasc Surg 101:816-824, 1991. Used with permission.)
Figure 15-32 Completion of a single left-lung transplant with a running 4-0 Prolene suture of the pulmonary artery (PA). (PV
= pulmonary vein; LMB = left main bronchus.) From Calhoun JH, Groover FL, Gibbons WJ, et al: Single lung transplantation:
Alternative indications and technique. J Thorac Cardiovasc Surg 101:816-824, 1991. Used with permission.)
circumstances, would be judged unfit for such an
operation. Scrupulous attention must be paid to
asepsis, and premedication should be restricted to
those drugs that decrease airway secretions rather
than those with analgesic or sedative properties (to
which recipients are likely to be very sensitive and
thereby hypoventilate). Drugs selected for induc-
tion, and for the provision of analgesia and neuro-
muscular blockade, should be neutral in action on
the cardiac and respiratory systems. This means
that the usual anesthetics used for patients
undergoing cardiac surgery such as high-dose syn-
thetic narcotics, perhaps combined with low doses
of halogenated drugs as well as nitrous oxide, and
vecuronium for neuromuscular blockade are per-
fectly acceptable.
2. Lung Separation
As with any pulmonary operation, the procedure
is made easier if the airway is divided and a clear
surgical area for operation is produced. There have
been many reports of anesthesia for lung trans-
plantation; some institutions use double-lumen
tubes, whereas others use bronchial blockers. Both
lung-separation techniques have been used with
great success, but, in terms of independent lung
suctioning and rapid independent lung inflation
and deflation, use of a double-lumen tube is more
versatile. If a left-sided double-lumen tube is used
for a left lung transplant, then it is important to be
sure, with fiberoptic bronchoscopy, that the left
endobronchial cuff is as proximal as possible to
allow mobilization and anastomosis of the bron-
chus to the transplanted lung.
225
226
3. Monitoring
The monitoring must be comprehensive as it
must be in any major operation, but use of capnog-
raphy and pulmonary artery catheter monitoring
deserves special comment. Capnography is impor-
tant in two respects. First, as noted in chapter 7,
the PETC0
2
tracing may demonstrate a considera-
ble slope to the phase III plateau in patients with
airway obstruction and emphysema. If the expira-
tory time is relatively short (i.e., less than the time
needed to reach a true PETC0
2
in emphysematous
lungs), then a large and inconsistent P
a
C0
2
-
PETC0
2
gradient may be present throughout the
anesthetic. Therefore, during the anesthetic, inter-
mittent arterial blood-gas analysis may be required
to optimize ventilation.
211
Second, but related to
the out-of-phase ventilation of the native lung
(high airway resistance, low compliance) and the
donor lung (low airway resistance, normal compli-
ance), is the occurrence of a biphasic capno-
gram.
227
Biphasic capnogram waveform secondary
to severe kyphosis, scoliosis,
228
and main-stem
bronchial intubation
229
has been observed previ-
ously in patients in whom the pattern of expiration,
as determined by alveolar time constants, was dis-
tinctly different in each lung.
Similarily, single-lung transplantations may also
demonstrate a biphasic capnograph waveform that
is produced by two different populations of alveoli
(one in the remaining emphysematous lung and
one in the newly transplanted lung) (Fig. 15-
33).
227
The first peak represents expired C0
2
from
the allograft, which has normal compliance, good
perfusion, and normal ventilation-perfusion ratios
(V/Q). The second peak most likely reflects ex-
pired C0
2
from the native lung, because the
slanted upstroke is characteristic of the mis-
matched V/Q ratios and differing alveolar time
constants in emphysema. Independent of the per-
fusion characteristics of the two lungs, the differ-
ences in their resistances and compliance alone
could account for the observed biphasic capno-
gram, with the low-resistance transplanted lung
emptying more rapidly and the more compliant
native lung emptying last. After an increase in
respiratory rate, the capnogram will show only a
single curve (Fig. 15-34).
227
This probably repre-
sents preferential exhalation of C0
2
from the trans-
planted allograft, with minimal contribution of ex-
pired C0
2
from the native emphysematous lung.
227
A pulmonary artery catheter, situated in the
main pulmonary artery, is probably mandatory.
During perfusion of just one lung, the wedge pres-
sure should be interpreted with caution because
there is some evidence that these results may be
spurious when measured in patients who have had
a pneumonectomy (the inflated balloon can block
blood flow to the one remaining lung, resulting in
an erroneously low value).
230
The index of partic-
ular acute interest is actually the pulmonary artery
pressure because it is a measure of the load on the
right ventricle. The use of intraoperative trans-
esophageal echocardiography also permits contin-
uous assessment of right ventricular function.
199
The wedge pressure is of less relevance initially,
provided that left ventricular function is known to
be adequate. Later, during recovery, the pulmonary
artery occluded pressure may be more important,
as may the ability to measure cardiac output by
thermodilution or other indicator technique.
4. Major Intraoperative Anesthetic
Problems
Potentially, and predictably, major problems
may occur when elements of the physiology of
lung ventilation and pulmonary blood flow are
radically altered by operative maneuvers. In prac-
tice, and in the usual temporal sequence, the sig-
nificant problem periods are the institution of one-
lung ventilation, the clamping of the pulmonary
artery before pneumonectomy, and the time pe-
riod, if allowed, during which the donor lung
is perfused but not ventilated. The detectable
changes for these three events in terms of moni-
tored data are shown in Table 15-15, in addition
to suggested directions for corrective therapy.
219
a. INITIATION OF ONE-LUNG VENTILATION
With respect to the initiation of one-lung venti-
lation, severe hypoxemia and hypercapnia may oc-
cur not only because of the large shunt created
through the nonventilated lung, but also because
the ventilated lung may not support ventilation as
well as a normal lung might. Obvious and simple
therapeutic maneuvers consist of increasing the in-
spirated oxygen concentration to 100. per cent (if
Figure 15-33 C0
2
tracings at a respiratory rate of eight breaths per min and a volume of 700 ml. Left, The trend of ETco
2
ove
the previous 20 min. Right, Biphasic capnogram at an ETco
2
of 31 mm Hg and a graph speed of 0.5 cm/s. (From Williams L
Jellish WS, Modica PA, et al: Capnography in a patient after single lung transplantation. Anesthesiology 74:621-622, 1991. Use
with permission.)
Figure 15-34 CO, tracings at a respiratory rate of 13 breaths per min and a tidal volume of 700 ml. Left, The 20-min trend of
ETco
:
, which has been decreasing for approximately 10 min. Right, A normal-appearing capnogram at an ETco
:
of 25 mm Hg and
a graph speed of 0.5 cm/s. (From Williams L, Jellish WS, Modica PA, et al: Capnography in a patient after single lung
transplantation. Anesthesiology 74:621-622, 1991. Used with permission.)
F,0
2
= 1.0 was not in use before the initiation of
one-lung ventilation) and increasing/changing the
minute volume of ventilation (change tidal volume
and/or respiratory rate). In general, these conven-
tional maneuvers are adequate to obtain acceptable
gas exchange even during ventilation of one se-
verely emphysematous lung.
231
If these simple
conventional maneuvers fail, a search for the ap-
propriate nonventilated lung CPAP and ventilated
lung PEEP levels should be quickly instituted
(chapter 11). In the event of supervening circula-
tory deterioration (severe pulmonary hypertension
and/or systemic hypotension) or failure of ventila-
tion (hypoxemia and/or hypercapnia) despite these
corrective maneuvers, partial femoral artery to
femoral vein cardiopulmonary by-pass may be
needed. The problem of hypercapnia and hypox-
emia during one-lung ventilation should be de-
creased when the ipsilateral pulmonary artery is
ligated.
b. CLAMPING OF THE PULMONARY ARTERY
With respect to pulmonary artery clamping, the
crux of the operation is the ability of the right
ventricle to withstand the acute increase in pul-
monary artery pressure induced by clamping the
pulmonary artery on the side of the pneumonec-
tomy. Hence, the preoperative assessment of right
ventricle function is emphasized. The increase is
sudden, but not usually sustained, and the pulmo-
nary vascular resistance in the perfused lung usu-
ally decreases after several minutes so that some
of the strain is taken off the right ventricle. In the
event of right ventricular failure, inotropic support
may be required. For excessive increases in pul-
monary artery pressure, drugs with pulmonary
vasodilatory activity are indicated: Nitroglycerin,
nitroprusside, and hydralazine are drugs that have
been successfully used in this context.
2
"
219
225
In
addition, inotropic support may be necessary (e.g.,
dopamine, 5-10 g/kg/min, dobutamine, 5-15 g/
Table 15-
Event
-15 OPERATIVE EVENTS AND THEIR MANAGEMENT*
Parameter Therapy
Onset of one-lung anesthesia P
a
0
2
decreases
P,CO^ increases
Pulmonary artery clamped
Donor lung perfused but not
ventilated
Pulmonary artery pressure
P.O, decreases
Increase inspired 0
2
(if not using F,0
2
= l.O)
Increase minute volume
Recruit nonventilated lung
CPB
Inotropes
Vasodilators
Diuretics
CPB
Pulmonary artery clamped:
Partial
Total
Low-disturbance lung
ventilation
Problem eliminated when
> ipsilateral pulmonary artery is
i gated
*From Conacher ID: Isolated lung transplantation: A review of problems and guide to anaesthesia. Br J Anaesth 61:468-474,
1988. Used with permission.
Abbreviations: CPB = partial (femorofemoral) cardiopulmonary by-pass.
kg/min).
219
225
232
Experience is not sufficient to
support a definitive opinion on this aspect of ther-
apeutic invention, and the use of prostacyclins is
also conjectural. Once again, if these therapeutic
steps fail to improve the recipient's condition, sup-
port with partial femoral artery to femoral vein
cardiopulmonary by-pass may be required.
196
C. PERFUSION OF THE TRANSPLANTED LUNG
WITHOUT VENTILATION
Finally, if vascular integrity of the transplant is
achieved before the completion of the bronchial
anastomosis, and blood flow is allowed to occur, a
shunt may be created. During this period of perfu-
sion without ventilation, ST-segment changes and
tachycardia/other arrhythmias have been noted in
a few patients; it is possible that the institution of
transplanted lung blood perfusion causes the wash-
out of cold transplant perfusate and/or air and/or a
bolus of potassium from the transplant (Eurocol-
lins has a potassium concentration of 107 mmol/L,
and high serum potassium has been documented
in some of these patients).
233
This unstable phase
is usually short lived, lasting the time it takes to
complete the bronchial anastomosis and reinflate
the transplanted lung. During this unstable phase,
it may be necessary to reduce/eliminate the perfus-
ate injection/shunt by clamping of the pulmonary
artery to the transplant or by separate ventilation
of the transplanted lung.
C. Postoperative Considerations
Patients undergoing lung transplantation face a
multitude of hemodynamic, respiratory, metabolic,
infectious, immunologic, and other complications.
A high level of intensive care is necessary for the
first few weeks, and sometimes much longer, if
success is to be achieved. Protocols for surveil-
lance and monitoring of patients undergoing a
complex cardiopulmonary procedure and trans-
plantation should be in place. After the critical
postoperative period, these patients do surprisingly
well in the long term, with levels of energy and
activity very similar to those of the general popu-
lation.
Many patients emerge from the operation mod-
erately severely hypothermic (32C)
233
in spite of
the use of warmed gases and intravenous fluids
and warming blankets. Lung-transplant recipients
are then subjected to donor-transmitted infections
in the early postoperative period and later from
nondonor sources. Donor-transmitted infections
are associated wifh a high mortality rate, whereas
later infections of nondonor source appear to be
tolerated better.
196
The recipient may be treated
expectantly with appropriate antibotics initially
based on donor Gram's stain findings and later by
actual recipient culture.
196
Intermittent fiberoptic
bronchoscopy may be used to evacuate the airway,
obtain culture specimens, as well as monitor/doc-
ument a satisfactory anastomotic lumen.
The transplanted lung is readily susceptible to
fluid overload, as well as large fluid loss from
exudation into the plerual cavity, mainly because
of the total disruption of pulmonary lymphatics
combined with the increase in extravascular lung
water that accompanies preservation, reimplanta-
tion, and reperfusion, as well as perhaps loss of a
peripheral lung cough reflex. Low-dose vasocon-
strictor blood pressure support rather than volume
infusion, once adequate filling pressures have been
established, should be used in an effort to mini-
mize lung water. Diuretics are frequently used.
199
Despite these measures, the transplanted lung,
especially when only single-lung transplantation is
carried out, frequently demonstrates a diffuse in-
terstitial infiltrate on chest radiograph. Patients are
weaned from ventilatory support as early as possi-
ble, and this may be facilitated by using a rigorous
preoperative rehabilitation program and an epi-
dural narcotic analgesia.
199
225
Clearly, the longer
that endotracheal intubation is required, the higher
is the likelihood of pneumonia developing and the
greater is the likelihood of problems with airway
healing. This is a particular concern in patients
with infective end-stage lung disease whose upper
and lower airways are colonized at the time of
transplantation. From the point of view of treating
both lung infection and lung water accumulation
with intermittent fiberoptic bronchoscopy, ventila-
tion with a large-diameter single-lumen tube is
desirable.
In patients with emphysema, the remaining na-
tive lung may show evidence of severe air trapping
in the emphysematous lung and may need to be
treated with deliberate hypoventilation (in an at-
tempt to prevent the air trapping and preferential
ventilation of the highly compliant native lung). It
may be necessary to use differential lung ventila-
tion postoperatively for the highly compliant na-
tive lung and the relatively normal single trans-
planted lung (see chapter 20).
226
234
~
236
The air
trapping, if severe, may cause hypotension, a phe-
nomenon that one group has come to call "pul-
monary tamponade."
226,235
Occasionally, a prolonged ileus may occur that
mandates the use of postoperative intravenous total
parenteral nutrition. Patients should walk as early
as possible, using supplemental oxygen as neces-
sary and monitoring transcutaneous oxygen satu-
ration. Frequent chest physical therapy is essential
for optimal mobilization of secretions and is begun
while the patient is still intubated.
The usual immunosuppression protocol consists
of using cyclosporine (4 mg/hr), azathioprine (2
mg/kg/day), and antilymphocyte globulin (2 mg/
kg/day) intravenously. The intravenous cyclospor-
ine is started at the conclusion of the transplanta-
tion operation. The antilymphocyte globulin, cy-
closporine, and azathioprine are continued for the
first 10 to 14 days after the transplantation; pred-
nisolone is then added (second to third postopera-
tive week; steroids before this time interfere with
the healing of the airway anastomosis) and contin-
ued with azathioprine and cyclosporine for the
next 6 months. Subsequently, patients are weaned
from prednisolone.
All patients continue on azathioprine and cyclo-
sporine for life (no human leukocyte antigen
matching of donor and recipient is possible; there-
fore, all patients require immunosuppression for
life). Suspected acute episodes of rejection may be
controlled by pulsed methylprednisolone, 10 mg/
kg intravenously for 3 days, followed by aug-
mented oral corticosteroids for 1 month. Patients
who do not respond to methylprednisolone may be
given a monoclonal antibody directed against
cells intravenously over 7 to 10 days. However,
the diagnosis of rejection is a difficult one and
presumptive and consists of observing developing
pulmonary infiltrates on radiography, fever, and
leukocytosis and a decrease in arterial oxygen ten-
sion (unfortunately, these changes are also sugges-
tive of infection). Early acute rejection has been
almost a constant feature of isolated lung trans-
plantation, with two or three rejection episodes
usually occurring within the first month. Bron-
choscopy is usually performed to obtain material
for culture to rule out infection, and transbronchial
lung biopsies may be performed during suspected
rejection episodes in an attempt to obtain tissue
confirmation of rejection. The episodes of rejec-
tion are usually readily managed with pulsed ste-
roid therapy, as described previously.
With the replacement of the diseased lung by
the transplanted lung, resetting of the receptors
that control oxygenation, pH, and carbon dioxide
tension appears to take several weeks.
196
Most pa-
tients have experienced mild to moderate hyper-
carbia for the first several weeks after transplanta-
tion that gradually normalizes later. Either because
of inadequate feedback signals from respiratory
muscles as a result of the reduced respiratory effort
required after transplantation or because there is
delay in resetting the central chemoreceptors, pa-
tients frequently experience a sense of hypoxia in
the postoperative period and will often respond by
hyperventilation. Repeated blood-gas determina-
tions during these episodes are frequently in the
normal range and have not shown hypoxemia.
These distortions in respiratory mechanisms sub-
side gradually with time.
X. TUMORS AT THE CONFLUENCE
OF THE SUPERIOR, ANTERIOR,
AND MIDDLE MEDIASTINA
The mediastinum is divided arbitrarily, for the
purposes of description, into upper and lower parts
at the upper level of the pericardium by a plane
that extends from the sternal angle to the lower
border of the fourth thoracic vertebra. The upper
part is named the superior mediastinum, and the
lower part is again subdivided into three parts: the
anterior mediastinum, in front of the pericardium:
the middle mediastinum, containing the heart and
pericardium; and the posterior mediastinum, be-
hind the pericardium. At the confluence of the
superior, anterior, and middle mediastina are the
middle portion of the superior vena cava, the tra-
cheal bifurcation, the main pulmonary artery, the
aortic arch, and parts of the cephalad surface of
the heart (see Figs. 2-23 to 2-25, 14-7, 15-38,
and 15-39). In adults, the majority of tumors in
this region originate from involvement of the hilar
lymph nodes with bronchial carcinoma or lym-
phoma, whereas in infants the masses are most
often benign bronchial cysts, esophageal duplica-
tion, or teratoma. Tumors of this region can cause
compression and obstruction of three of the vital
mediastinal structures: the tracheobronchial tree in
the region of the tracheal carina, the main pulmo-
nary artery and atria, and the superior vena cava.
The spectrum of clinical syndromes that can be
caused by sclerosing mediastinitis (an encasing fi-
brotic reaction to chronic inflammation) well illus-
trates this point.
237
Computed tomographic scan of
the chest is probably the single most important
diagnostic procedure because it defines the size
and degree of compression of vital structures.
The most common complication to occur during
anesthesia for masses involving the three vital me-
diastinal structures (tracheobronchial tree, pulmo-
nary artery and vein, and superior vena cava) is
airway obstruction; this was a feature in 20 of 22
patients traced (1969-1983) and succinctly sum-
marized.
238
Although airway obstruction has been
predominant in terms of symptomatology, it is not
uncommon for compression of two or three of
these three major organs to be present in varying
degrees in the same patient.
238
239
Each of these
complications is life-threatening and can cause
acute deterioration and death during anesthesia if
not handled with the most extreme caution and
expertise. Each of these three major complications
and anesthetic management problems is discussed
separately. Figure 15-35 shows the overall strat-
egy for managing these three problems.
B. Compression of the
Tracheobronchial Tree
Most anterior mediastinal masses that cause air-
way obstruction are lymphomatous in origin.
However, a number of benign conditions such as
cystic hygroma, teratoma, and thymoma and thy-
roid tumors can occur in a similar fashion. A tissue
diagnosis, therefore, usually (see later discussion)
must be made before radiation or chemotherapy
can be undertaken. Thus, most patients with a me-
diastinal mass causing airway obstruction will first
require anesthesia for a diagnostic procedure (cer-
vical or scalene node biopsy, staging laparotomy
for Hodgkin's disease).
However, it should be noted that general anes-
thesia before radiation treatment is associated with
a 16 per cent incidence of life-threatening respira-
tory complications; indeed, 70 per cent of those
who experienced life-threatening respiratory com-
plications were symptomatic preoperatively.
240
241
Not all patients in whom severe intraoperative
problems developed had respiratory systems and
signs preoperatively (significant tracheal obstruc-
tion may not cause symptoms if the onset has been
insidious and chronic).
The strong concern about the wisdom of using
general anesthesia before radiation therapy, espe-
cially when the radiation therapy field can create a
window for later accurate tissue diagnosis,
242,243
has been voiced.
241
244
Nevertheless, some pathol-
ogists insist that any prebiopsy irradiation of tumor
area distorts cellular morphology to the extent that
it prevents recognition of cell type and use of
specific cell-type chemotherapeutic agents.
240
In
fact, in one institution, in patients with mediastinal
lymphomas, 2.2 procedures per patient were re-
quired to obtain enough tissue for immunopheno-
typing.
245
The anesthetic management of these patients is
based on two overriding considerations. First, the
obstruction of major airways by a tumor is usually
life-threatening because the obstruction usually oc-
curs around the bifurcation of the tracheobronchial
tree and is, therefore, distal to the endotracheal
Figure 15-36 In noncritical ste-
nosis (a), the normal laminar flow
profiles (as are produced by sponta-
neous ventilation) are compressed
but resume their usual pattern distal
to the stenosis. In critical stenosis
(b), laminar flow profiles are se-
verely compressed. If the gas flow
reaches critical velocity (as produced
by vigorous positive-pressure venti-
lation), flow will be disrupted into
turbulent eddies, and laminar flow
cannot be restored. With critical ste-
nosis, if either the density or the ve-
locity of the gas is reduced (c), gas
flow may be maintained with much
less disruption of the laminar flow
profiles. (From Sibert K. Biondi JW,
Hirsch NP: Spontaneous respiration
during thoracotomy in a patient with
a mediastinal mass. Anesth Analg
66:904-907, 1987. Used with per-
mission.)
tube. Inhalation induction precipitated obstruction
in three cases, and in none of these did intubation
completely relieve it. In one case, intubation made
the obstruction complete until a long, thin tube
was passed, producing partial improvement.
238
In
another case,
246
ventilation was difficult until spon-
taneous ventilation returned with the tube still in
place. In another case,
246
only a 4.5-mm rigid
bronchoscope eventually secured a patent airway
in an 11-year-old boy. In eight cases, intubation
itself precipitated or exacerbated a partial main-
stem bronchial obstruction.
246
"
249
The obstruction
was thought to be relieved by the return of spon-
taneous respiration in some of the patients.
246-248
It
may be that loss of chest wall tone and the distend-
ing forces of active inspiration (loss of negative
intrathoracic pressure gradient, which increases
airway diameter and holds the airway open) after
administration of muscle relaxants release extrinsic
support of a critically narrowed airway.
Alternatively, intubation in the presence of dis-
tortion or compression of the trachea may cause
complete obstruction if the orifice of the tube im-
pinges on the tracheal wall, or if the lumen of the
tube is occluded where it passes a narrowed sec-
tion or turns a sharp angle. Such obstruction has
developed in many patients and was relieved only
by a long, thin tube or bronchoscope passed be-
yond the stenosis.
246
249
"
253
Other reasons for tracheobronchial tree com-
pression after the induction of general anesthesia
include loss of lung volume (the tumor remains
constant in size), the presence of tracheomalacia,
increase in central blood volume in the supine
position (the tumor remains constant in size),
edema, bleeding and hematoma formation in the
tumor after it has been manipulated, injury to the
recurrent laryngeal nerve (vocal cord paralysis),
and finally the creation of very high flow rates and
turbulent flow profiles with vigorous positive-pres-
sure ventilation (Fig. 15-36 and Table 15-16).
In view of this reported seriousness of airway ob-
struction during general anesthesia, all possible at-
Table 15-16 POSSIBLE CONTRIBUTORY
CAUSES OF
TRACHEOBRONCHIAL TREE
OBSTRUCTION FOLLOWING THE
INDUCTION OF GENERAL
ANESTHESIA AND TRACHEAL
INTUBATION IN PATIENTS WITH
MEDIASTINAL MASSES
1. Loss of lung volume, mass remains constant in size,
pressure on vital structures increases
2. Increase in central blood volume, mass remains constant in
size, pressure on vital structures increases
Loss of negative pleural pressure and distending transmural
pressure gradient
4. Obstruction of tip of endotracheal tube by wall of
tracheobronchial tree resulting from distortion (bending,
unusual curves, etc.)
5. Presence of tracheobronchial tree malacia
6. Increase in size of tumor as a result of surgical
manipulation (edema, hematoma)
7. Creation of turbulent flow caused by vigorous positive-
pressure ventilation
8. Injury to the recurrent laryngeal nerve
3.
Table 15-17 IMPORTANT MANAGEMENT
PRINCIPLES FOR TUMORS AT
THE CONFLUENCE OF THE
SUPERIOR, ANTERIOR, AND
MIDDLE MEDIASTINA
1. Perform all procedures under local anesthesia if at all
possible
2. Use radiation and/or chemotherapy if at all possible prior
to general anesthesia
3. If general anesthesia, then consider inspection of
tracheobronchial tree with fiberoptic bronchoscope and
intubate awake
4. If general anesthesia, then maintain spontaneous ventilation
tempts to perform the procedure under local anes-
thesia should be undertaken.
Second, the response of lymphomatous tumors
to radiation or chemotherapy is normally dramatic.
Chest roentgenograms reveal a marked decrease in
tumor size, and the symptoms are usually im-
proved. Consequently, it behooves the treating
physicians to use radio- or chemotherapy if at all
possible (sometimes the cell type can be known
with a reasonable degree of certainty without a
biopsy) before general anesthesia. It should be
noted that, during radiation, a small window can
be created to spare some tissue for adequate histo-
logic diagnosis.
242,243
The following management plan is based on the
two previous principles
189
(Table 15-17 and Fig.
15-37). If a patient with a mediastinal mass near
the confluence of the superior, anterior, and middle
mediastina exhibits dyspnea and/or intolerance of
the supine position and a biopsy is scheduled, it
should be done under local anesthesia, if at all
possible. If the cell type is thought to be radiosen-
sitive or chemosensitive, then appropriate types of
therapy should be undertaken before any further
surgery is performed.
As noted previously, an anatomic radiation
treatment window can be created to spare tissue
for accurate histologic diagnosis.
242,243
Following
these types of therapy, the radiologic and com-
puted tomogram appearance of the tumor must be
reviewed along with a dynamic evaluation of pul-
monary function (see later discussion).
If the patient does not have dyspnea or intoler-
ance of the supine position (i.e., is asymptomatic),
a series of noninvasive tests may be performed to
evaluate the anatomic and functional position of
the tumor. First, a flow-volume loop should be
performed in the upright and supine positions. The
flow-volume loop is an extremely sensitive tool
for evaluating obstructive lesions of the major air-
ways
255
and can differentiate between extrathoracic
and intrathoracic airway obstructions. With extra-
thoracic obstruction, the inspiratory limb of the
flow-volume loop will show a plateau, and with
intrathoracic airway obstruction, the expiratory
limb of the flow-volume loop will show a plateau
(see Fig. 15-15). A disproportionate reduction in
maximal expiratory flows should also alert the
physician to the presence of tracheomalacia and its
inherent risk of precipitating dynamic airway col-
lapse after tracheal extubation.
Second, the chest computed tomographic scan
will best reveal the anatomic location of the tumor
and, perhaps, show a static picture of airway ob-
struction. Third, echocardiography may be per-
formed in both the upright and supine positions to
determine the impact of the tumor on the geogra-
phy of the heart.
If any of these three tests have positive results,
then local anesthesia should be used for biopsy
even if the patient is asymptomatic. If all three of
these major noninvasive diagnostic tests have neg-
ative results, the patient may be anesthetized with
general anesthesia if necessary, but local anesthe-
sia is still preferable. Again, once the biopsy has
been taken and the tissue shown to be radio- or
chemosensitive, appropriate therapies should be
instituted and the patient re-evaluated radiographi-
cally and functionally before further surgery is at-
tempted.
If general anesthesia is to be used in adults, the!
airway should be evaluated by fiberoptic bron-
choscopy with topical anesthesia before the induc-
tion of general anesthesia, if possible.
25
^
258
The
fiberoptic bronchoscope should be jacketed with
an endotracheal tube, and after the fiberoptic bron-
choscopy examination has been completed, the pa-
tient may be intubated.
It is not advisable to pass a flexible broncho-
scope through an extremely narrowed airway seg-
ment because this may precipitate total obstruction
(the patient must exhale around the bronchoscope,
even if oxygen is delivered through the small suc-
tion channel). Also, even brief introduction and
prompt withdrawal of a fiberscope from a narrow
passage may produce enough local edema to con-
vert a partial obstruction into total occlusion. In
this situation, it is best to intubate, keeping the
fiberoptic bronchoscope proximal to the obstruc-
tion. The endotracheal tube (or double-lumen tube
with main-stem bronchial obstructions or rigic
bronchoscope), however, should be placed distai
to any tracheobronchial tree obstruction whenevei
it is inserted.
253
Indeed, in one report, the only wa>
both lungs could be ventilated at the same time
was to cannulate both main-stem bronchi indepen-
dently.
259
In some extremely compromised pa-
tients, and depending on the exact location of the
mediastinal mass, a surgical airway with the pa-
tient awake may be the best first choice of airway
General anesthesia should be induced with th(
571
Conservative approach
to Extubation
Figure 15-37 Algorithm for anesthetic management of a patient with a mediastinal mass and suspected tracheobronchial tree
obstruction. (CPB = cardiopulmonary bypass.) (Adapted with permission from Neuman GG, Weingarten AE, Abramowitz RM, et
al: The anesthetic management of the patient with an anterior mediastinal mass. Anesthesiology 60:144-147, 1984. Used with
permission.)
patient in the semi-Fowler's position, the patient
should be allowed to breathe spontaneously
throughout the procedure, and muscle relaxants
should be avoided
256
(even with a thoracotomy, if
at all possible).
254
In children, slow intravenous
administration of ketamine and administration of
nitrous oxide with spontaneous ventilation fulfills
these goals
260
(this is certainly possible if the
thorax does not have to be entered; if the thorax
does have to be entered, positive-pressure ventila-
tion may be required).
Spontaneous ventilation, however, does not
guarantee that airway obstruction will not occur
(half the reported cases of perioperative obstruction
have been during spontaneous ventilation).
242
261
Large swings in intrathoracic pressure, which may
promote collapse of a weakened tracheobronchial
tree, must be avoided. The operating room team
should retain the capability of changing the pa-
tient's position rapidly to the sitting or lateral or
prone position (e.g., Fig. 15-38). In one case, the
fortuitous application of a laryngoscope blade re-
lieved a near-complete obstruction of the tracheo-
bronchial tree perhaps by applying tension to one
end of the trachea, which straightened out the tra-
chea.
262
A rigid ventilating bronchoscope should
be on hand, and the appropriate personnel and
equipment for cardiopulmonary by-pass also
should be available.
These patients must be closely monitored in the
first few hours. Airway obstruction requiring rein-
tubation and mechanical ventilation has occurred
(40 per cent of all perioperative obstruction
cases)
242
261
possibly secondary to an increase in
Figure 15-38 Fiberoptic bronchoscopic appearance of lower trachea with anesthetized patient with large anterior mediastinal
mass in supine position (A), exhibiting almost total obstruction of the trachea in anteroposterior plane. With patient in sitting
position (B), lumen appears normal. (From Prakash UBS, Abel MD, Hubmayr RD: Mediastinal mass and tracheal obstruction
during general anesthesia. Mayo Clin Proc 63:1004-1011, 1988. Used with permission.)
tumor size resulting from tumor edema after in-
strumentation. In some patients, especially those
requiring reintubation, it may be best to irradiate
the central core of the tumor and put the patient
on steroids before extubation is attempted again.
256
C. Compression of the Pulmonary
Artery and Heart
Compression of the pulmonary artery and heart
is rare because the pulmonary trunk is more or less
protected by the aortic arch and tracheobronchial
tree; there are only four definite case reports in the
literature.
263-266
However, in view of the lethal na-
ture of this complication, it warrants discussion.
Although experience with this problem is ex-
tremely limited, the cell type has been lymphoma-
tous in three reports and a cyst in the fourth pa-
tient. In the patient with a large mediastinal cyst,
it was thought that a double-lumen tube caused the
left main-stem bronchus to put pressure on the
cyst, which, in turn, compressed the pulmonary
artery.
266
In an animal model of an expansile ante-
rior mediastinal mass (800-ml intravenous infusion
bag), the most significant cardiopulmonary effect
of the mass, irrespective of the type of ventilation
used (positive- or negative-pressure breathing) was
obstruction of the pulmonary artery, causing right
ventricular enlargement, decrease in left ventricu-
lar volume through ventricular interdependence
(septal shifting), and decrease in cardiac index and
arterial oxygenation.
267
Large mediastinal lympho-
mas have also been associated with arrhythmias
under anesthesia owing to pericardial or myocar-i
dial involvement.
Similar principles for compression of the tra-
cheobronchial tree apply to compression of the
pulmonary artery. Most patients have their firsi
anesthetic experience because they require a diag-
nostic procedure (e.g., a biopsy). All diagnostic
procedures should be performed under local anes
thesia if at all possible. Since the symptoms usu
ally worsen when the patient assumes the supin<
position, an unusual position may need to be uti
lized (see later discussion). Lymphomatous tumor;
are usually radiosensitive.
These patients should be evaluated preopera
tively in a manner similar to that for those witl
compression of the tracheobronchial tree. If then
is absolutely no indication of the cell type, loca
anesthesia should be tried, if at all possible, ti
perform the biopsy. If the cell type is known or i
highly suspected, preoperative irradiation shouli
be seriously considered. Some tissue should b
excluded from the radiation field for accurate his
tologic diagnosis.
242
256
If general anesthesia is required, and the tumc
is anterior to the heart, the sitting, leaning forwarc
or even face-down position is advised (because th
tumor will not press on the pulmonary artery c
heart [as it would in the supine position]), and
spontaneous ventilation should be maintained
throughout the procedure (as with the intrathoracic
airways, negative pleural pressure will distend the
vessels and keep them open).
Measures to maintain venous return, pulmonary
artery pressure, and cardiac output, such as volume
loading and use of ketamine, should be considered.
Arrangements for extracorporeal oxygenation
should be completed preoperatively. The anesthe-
tist has to be aware of the danger of air embolism
in a patient in the sitting position having a verti-
cally nondependent surgical procedure.
D. Superior Vena Cava Syndrome
The superior vena cava syndrome is caused by
mechanical obstruction (compression, invasion,
thrombosis) of the superior vena cava. The causes
of superior vena caval obstruction, in order of rap-
idly decreasing incidence, are bronchial carcinoma
(40-90 per cent) (Fig. 15-39),
268
malignant lym-
phomas (5-15 per cent), and benign causes (1-10
per cent), such as pulmonary artery, central ve-
nous, hyperalimentation, and pacemaker-catheter-
induced thrombosis of the superior vena cava,
269
idiopathic mediastinal fibrosis, mediastinal granu-
loma, and multinodular goiter.
270-273
The classic features of superior vena cava syn-
drome include dilated distended veins in the upper
half of the body due to increased peripheral ve-
nous pressure (which can be as high as 40 mm
Hg); edema of the head, neck, and upper extremi-
ties; dilated venous collateral channels in the chest
wall; and cyanosis. Venous distension is most
prominent in the recumbent position, but in most
instances the veins do not collapse in the normal
manner with the patient upright. The majority of
patients have respiratory symptoms (shortness of
breath, cough, orthopnea, nasal stuffiness), which
are due to obstruction of the airways by engorged
veins and mucosal edema and are ominous signs.
Similarly, a change in mentation and the pres-
ence of headaches, caused by cerebral venous hy-
pertension and edema, are also ominous signs. In
some cases, the superior vena cava becomes oc-
cluded quite slowly, collateral vessels have a
chance to develop, and the signs and symptoms
may be insidious in onset.
272
Collaterals drain to
the azygos vein in cases of high obstruction or to
the hemiazygos or through chest wall veins to the
inferior vena cava in cases of obstruction below
the azygos confluence point.
272
When the occlu-
sion occurs relatively rapidly, all clinical manifes-
tations are more prominent; in this setting, facial
edema may be so severe that it prevents patients
from opening their eyes. Moreover, rapidly in-
creasing venous pressure in the cerebral circulation
may lead to neurologic impairment as cerebral per-
fusion pressure is decreased.
The most common radiologic sign is widening
of the superior mediastinum.
272
273
Venography
confirms the diagnosis (but not the cause).
274
De-
termination of cause may require thoracotomy,
sternotomy, bronchoscopy, lymph node biopsy,
and so on.
272-274
Transesophageal echocardiogra-
phy has been used to delineate the mechanism of
superior vena cava obstruction.
275
Most patients with superior vena cava syndrome
(i.e., those with incomplete obstruction) caused by
a malignant process are treated with irradiation
and chemotherapy.
274
However, in patients with
near-complete to complete obstruction (who usu-
ally have signs of cerebral venous hypertension
and/or airway obstruction), or when irradiation or
chemotherapy proves ineffective, surgical by-pass
or resection of the lesion via median sternotomy is
indicated.
274
These operations are usually techni-
cally quite difficult because tissue planes are
poorly delineated, anatomy is grossly distorted,
and varying degrees of fibrosis are present. Good
temporary results have been obtained with balloon
angioplasty.
276
-
The preoperative anesthetic evaluation of a pa-
tient for superior vena caval decompression should
include careful assessment of the airway. The
same degree of edema that is present externally in
the face and neck can be expected to be present in
the mouth, oropharynx, and hypopharynx. In ad-
dition, the airway may be compromised by exter-
nal compression, fibrosis limiting normal move-
ment, or recurrent laryngeal nerve involvement. If
tracheal compression is suspected, it should be
evaluated by computed tomography.
Premedication for these patients is light or de-
leted when there is concern about the integrity of
the airway. A drying agent is helpful if a difficult
intubation is anticipated, and it should be admin-
istered to all patients having difficulty swallowing.
The patient is transported to the operating room in
the head-up position in order to minimize airway
edema. A radial artery catheter is inserted in all
patients and, depending on the medical condition
of the patient, a central venous or pulmonary ar-
tery catheter is inserted via the femoral vein prior
to the induction of anesthesia. On one occasion
atrial pacing via a pulmonary artery catheter was
utilized in a patient with suspected cardiac in-
volvement from a mediastinal tumor who was ex-
periencing significant bradycardia preoperatively.
At least one large-bore intravenous cannula should
be inserted in the leg or femoral vein prior to
Figure 15-39 A common cause of superior vena cava (SVC) obstruction is thoracic malignancy. In the patient described in thi
report, an adenocarcinoma arose from the left main bronchus, involving the tracheal cartilage, hilar lymph nodes, and aorta,
central venous catheter may have also contributed to the thrombosis in conjunction with the tumor. (From Smith S, Vacek JL, Dun
MI: A sudden change in central venous pressure. Hosp Physician 41-45, 1987. Used with permission.)
operation; the upper extremities are not used for
intravenous infusions because of the long and un-
predictable circulation time that results from the
superior vena caval obstruction. Cross-matched
blood should be available in the operating room at
the time of sternotomy.
The method chosen for induction of anesthesia
and intubation depends on the preoperative airway
evaluation. If it is necessary for the patient to
maintain the sitting position in order to achieve
adequate ventilation prior to induction, intubation
with the patient awake may be facilitated by using
a fiberoptic laryngoscope
277
or bronchoscope.
278
The most significant intraoperative problem en-
countered is bleeding. Substantial venous blood
loss results from the abnormally high central ve-
nous pressure. Further, unexpected arterial bleed-
ing may occur because of the difficulty of dissect-
ing in a distorted surgical field. Cross-matche
blood should therefore be available in the opera
ing room at the time of sternotomy.
Postoperatively, especially after diagnostic pre
cedures such as mediastinoscopy and broncho!
copy, wherein the superior vena caval obstrucie
has not been relieved, acute severe respiratory fai
ure requiring intubation and mechanical ventili
tion may occur.
271
279
~
283
The mechanisms of tlj
acute respiratory failure are obscure, but the mq
likely ones that are unique to the superior vei
cava syndrome are acute laryngospasm and/)
acute bronchospasm (both due to continued an
perhaps, increased obstruction of the superior vei
cava), impaired respiratory muscle function (p
tients with malignant disease may have an abnc
mal response to muscle relaxants),
283
and increas
airway obstruction by the tumor (due to turn
swelling). Consequently, these patients must be
closely monitored in the first few postoperative
hours.
XI. REPAIR OF THORACIC AORTIC
ANEURYSMS
Patients with thoracic aortic aneurysms may
come to the operating room on either an elective
or an emergency basis. Thoracic aortic aneurysms
in patients who are hemodynamically stable and
free of significant symptoms are repaired on an
elective basis. Patients with thoracic aortic aneu-
rysms that were caused by trauma, or those who
are hemodynamically unstable with acute symp-
toms (e.g., chest pain) due to an acute dissection,
come to the operating room on an emergency
basis. Consequently, the anesthetic management of
patients with thoracic aortic aneurysms could be
logically discussed as either an elective therapeutic
procedure or an emergency procedure.
The anesthetic management of patients with
thoracic aortic aneurysms is discussed with special
emergency procedures (see Thoracic Aortic Aneu-
rysms and Dissections/Disruptions, chapter 17).
The reason for this is twofold. First, patients re-
quiring emergency repair of thoracic aortic aneu-
rysms outnumber patients requiring elective re-
pair.
284
Second, the emergency cases are more
challenging, involve higher risks, and entail a
greater number of anesthetic considerations. In-
deed, there is a 16 per cent surgical mortality for
emergency resection compared with a 5 per cent
surgical mortality for elective resection. The dis-
cussion of the anesthetic management of patients
with thoracic aortic aneurysms in chapter 17 in-
cludes all the anesthetic considerations that would
apply to patients undergoing elective resection of
thoracic aortic aneurysms.
XII. THYMECTOMY FOR
MYASTHENIA GRAVIS
Myasthenia gravis is a disease of neuromuscular
transmission characterized by weakness and easy
fatigability of the voluntary muscles. The disease
has a prevalence of about 1 in 30,000. If untreated,
myasthenia gravis has a 40 per cent mortality rate
in 10 years, but with modern treatment death is
rare.
285
Evidence indicates that the disease is caused by
an autoimmune attack by the immunoglobulin G
autoantibody on the postsynaptic acetylcholine re-
ceptors in skeletal muscle. There appears to be a
causal connection between the binding of antibody
to receptors, a decrease in number of the receptors
in motor end-plates, and interference with neuro-
muscular transmission
286
(Fig. 15-40). The aver-
age life of an acetylcholine receptor in a normal
individual is 7 days, whereas in a myasthenic in-
dividual it is 1 day. The place of thymectomy in
the treatment of myasthenia gravis is well estab-
lished. It has been reported to result in remission
or clinical improvement in at least 80 per cent of
patients.
287
The diagnosis can usually be established by the
typical historic and neuromuscular examination
findings.
288
The most frequent complaint of myas-
thenic patients is diplopia; ptosis, the second most
common sign, may be missed if it is mild. Ptosis
may be unilateral and characteristically may shift
from side to side. Dysarthria is an early symptom
of bulbar involvement, followed by difficulties in
chewing and swallowing, leading to weight loss.
In fact, almost all myasthenia gravis patients have
abnormal esophageal manometric studies.
288
In ad-
vanced myasthenia gravis, the most frequent bul-
bar sign is facial weakness. In 15 to 20 per cent of
myasthenic patients, the chief complaint is extrem-
ity weakness and easy fatigability; the arms are
affected more frequently. Respiratory muscle
weakness may prompt the patient to visit a physi-
cian, but this is rarely the first symptom.
Various tests can be used to confirm the diag-
nosis. Anti-acetylcholine receptor antibody is spe-
cific to myasthenia gravis and is thus very useful
in diagnosis. The antibody titer is elevated in 90
per cent of patients with generalized myasthenia
gravis and in approximately 70 per cent of patients
with ocular (group 1) disease. Antistriated muscle
antibody is detectable in more than 90 per cent of
patients with thymoma and in approximately 30
per cent of other myasthenia gravis patients. Elec-
tromyography shows a decrementai response in
repetitive nerve stimulation. Administration of ed-
rophonium or neostigmine (Prostigmine) results in
a transient increase in muscle strength. Computed
tomography of the mediastinum is helpful in deter-
mining whether a thymoma is present (9 to 16 per
cent of myasthenic patients). Finally, acetylcholine
receptor antibody titers are elevated in a majority
of patients.
There are five classifications of the clinical sta-
tus of the patient (Table 15-18):
289
1, ocular symp-
toms and signs only, without progression (these
patients are not treated by thymectomy); IIA, gen-
eralized weakness, mild bulbar and skeletal symp-
toms, and IIB, generalized weakness, moderate to
severe bulbar and skeletal symptoms; III, acute
fulminating weakness with severe bulbar involve-
ment; IV, late onset or exacerbation, severe bulbar
symptoms, and severe generalized weakness; and
V, myasthenia gravis with muscular atrophy.
Pharmacology of Myasthenia Gravis
Figure 1540 Schematic diagram of the motor end-plate. In myasthenia gravis, antibodies to the postsynaptic acetylcholine
(Ach) receptors destroy and reduce the number of receptors; hence, the affected muscle fatigues easily. Muscle relaxants have a
much more exaggerated effect because they easily bind the reduced number of receptors. Acetylcholinesterase (Achesterase)
metabolizes Ach. Achesterase inhibitors increase the amount of end-plate Ach and increase Ach-receptor binding and thereby
increase muscle strength.
Approximately 33 per cent of patients with thy-
momas have myasthenia gravis and 10 to 15 per
cent of myasthenia gravis patients have thymomas.
Patients with thymomas tend to be older, have
higher acetylcholine receptor antibody titers, and
have a more severe form of the disease, and the
response to thymectomy is somewhat poorer than
in patients with thymitis.
290
Myasthenia gravis may
Table 15-18 MAIN GROUPS OF ACQUIRED
MYASTHENIA GRAVIS
be associated with thymitis; these patients are usu-
ally younger than 40 years and often have other
autoimmune diseases (Table 15-19), and the re-
sponse to thymectomy is good.
290
The pure ocular
form of the disease responds very poorly to thy-
mectomy and is best treated with steroids (see
following discussion).
The nonoperative therapy of myasthenia gravis
arises from the apparent autoimmune cause of the
known deficiency in acetylcholine receptors. The
mainstay of treatment of patients with myasthenia
Group
Group I
Group IIA,
Group III
Group IV
Group V
Description
Ocular myasthenia gravis (symptoms may
remain persistently confined to the ocular
muscles, particularly when 2 years have
elapsed since the onset)
Mild or moderately severe generalized
myasthenia gravis
Acute severe (fulminating) myasthenia
gravis with respiratory muscle
involvement
Late (chronic) severe disease
Myasthenia gravis with muscular atrophy
Table 15-19 IMMUNE DISORDERS
ASSOCIATED WITH
MYASTHENIA GRAVIS
Rheumatoid arthritis
Hyperthyroidism
Hypothyroidism
Polymyositis
Systemic lupus erythematosus
Pernicious anemia
Sjogren's syndrome
Pemphigus
gravis is use of the oral anticholinesterase pyrido-
stigmine (Mestinon) (Table 15-20). Most patients
tend to prefer pyridostigmine because it has fewer
muscarinic side effects (i.e., sweating, salivation,
abdominal pain, diarrhea, bradycardia) than neo-
stigmine and a more prolonged length of action.
The anticholinesterase maintains the local concen-
tration of acetylcholine at the motor end-plate at a
high level and increases the chance of acetylcho-
line binding to an acetylcholine receptor (Fig. 15-
40). Occasional patients with myasthenia gravis
may not respond to anticholinesterase drug ther-
apy, which may be due to complete absence of
postsynaptic acetylcholine receptors so that even
the increased amount of acetylcholine present has
no place to act. In addition, myasthenics on pro-
longed chronic anticholinesterase drug therapy
may show decreased sensitivity to the drug.
Exacerbation of myasthenia gravis can occur
secondary to emotional and/or surgical stress and/
or adverse drug interactions (antibiotics and antiar-
rhythmia drugs), with the subsequent development
of a myasthenic crisis. During crisis, patients have
decreased responsiveness to anticholinesterase.
This situation must be differentiated from a cholin-
ergic crisis, which is secondary to an anticholin-
esterase overdosage.
291
In both situations, there is
an increase in muscle weakness, which may in-
volve the respiratory musculature and, thereby, ne-
cessitate respiratory support. A small dose of ed-
rophonium (10 mg intravenously) will improve
strength in a patient with myasthenic crisis but will
have little or a negative effect in a patient with
cholinergic crisis. In both myasthenic and cholin-
ergic crisis, it is best to withhold anticholinesterase
medications while providing mechanical suppor-
tive ventilation.
Remission of myasthenia can be anticipated in
80 per cent of patients receiving corticosteroids.
Doses of prednisolone, 30 to 40 mg daily, are usu-
ally recommended. Once remission is achieved, the
dose can be steadily reduced and remission main-
tained in many patients on as small a dose as 10
mg on alternate days. Because the beneficial ef-
fects are usually achieved well within 2 to 3
weeks, corticosteroids must act in some way other
than by immunosuppression. It seems probable
that they protect the acetylcholine receptor from
immunologic attack. Patients taking corticoste-
roids require less anticholinesterase therapy. If the
anticholinesterase drug dose is not reduced, an in-
itial deterioration in the myasthenia may occur in
the first week or 10 days of treatment. It is advis-
able to start corticosteroid treatment with the pa-
tient in the hospital under close supervision.
Relapse is common after the withdrawal of cor-
ticosteroids, and the inevitability of adverse effects
must be appreciated if the dose of prednisolone
cannot be reduced below 10 mg daily. For patients
in whom remission cannot be maintained on 15
mg of prednisolone on alternate days, azathioprine
treatment is instituted (see later discussion).
In treating myasthenia, corticosteroids are used
for those patients who are seriously ill before thy-
mectomy, those who are unsuitable for thymec-
tomy, and those who are insufficiently improved
postoperatively. They are also useful in patients
with ocular myasthenia, who, as a group, respond
poorly to anticholinesterase drugs and to thymec-
tomy.
Azathioprine (2.5 mg/kg/day) is also effective
in reducing anti-acetylcholine receptor antibody
levels and in producing clinical improvement. The
response is slower than that of corticosteroids; im-
provement begins at 6 to 12 weeks and becomes
maximal at a mean of 6 to 15 months. In one study
of 78 patients treated with azathioprine over 11
years, 31 patients were in complete remission, 40
were much improved, and none were worse.
292
A dramatic but short-lived improvement in my-
asthenia can be brought about by plasma ex-
change. It is useful as a short-term measure in
seriously ill patients until other forms of therapy
become effective. However, there is no evidence
that repeated plasma exchange combined with im-
munosuppression confers any greater long-term
benefit than immunosuppression alone.
Thymectomy has become increasingly impor-
tant in the management of myasthenia gravis and
should be offered to all significantly affected pa-
Table 15-20 ANTICHOLINESTERASE MEDICATIONS AVAILABLE IN THE UNITED STATES FOR
THE TREATMENT OF MYASTHENIA GRAVIS
Abbreviations: IV = intravenous; IM = intramuscular.
Drug
Neostigmine
Pyridostigmine
Ambenonium
Trade Name
Prostigmine
Mestinon
Mytelase
Oral Dose
(mg)
15
60
6
Onset/Duration
of Oral Route
15 min/2 hours
30 min/3-6 hours
30 min/4-6 hours
Parenteral Dose
IV
0.5
2
Not available
(mg)
IM
0.5-1.0
3-4
tients who are fit for operation unless they have
purely ocular disease or have minimal symptoms.
Suffice it to say that the thymus contains acetyl-
choline receptors on the myoid cells, and it is
probable that tolerance to these is broken (e.g., by
thymitis) and the circulating antibodies cross react
with acetylcholine receptors on skeletal muscle.
Complete remission or substantial improvement
can be expected in 80 to 90 per cent of patients
without thymomas, although 5 to 8 years may
elapse before the benefits of operation become ap-
parent.
293-296
In patients with thymic tumors, sur-
gical excision is indicated to prevent local spread
(thymomas can be malignant), but the prognosis is
not as good.
Because of the possibility of malignancy, most
centers advocate the most complete resection pos-
sible, including the tumor, the thymus gland,
lymph nodes, and the surrounding fatty tissue as
well.
294
297
To achieve complete resection and rad-
ical dissection, a median longitudinal sternotomy,
a lateral thoracotomy, or a transverse sternotomy
may be used.
297-298
Physical examination should include airway
evaluation and tests of muscle strength and the
ability to cough, chew, and swallow. Vital capac-
ity, peak inspiratory force, and maximum breath-
ing capacity should be measured in every patient.
Those with impaired respiratory function need
complete respiratory function tests performed, in-
cluding analysis of arterial blood gases. A chest
roentgenogram and computed tomogram should be
reviewed to assess the presence and size of the
thymic tumor. Myasthenic patients may also have
associated myocardial degenerative changes,
which make a preoperative electrocardiogram nec-
essary. Thyroid abnormalities may be associated
with myasthenia gravis, and patients need to be
evaluated clinically and by laboratory tests for ev-
idence of hypo- or hyperthyroidism. Nutritional
status should be evaluated by measurement of
serum electrolyte, albumin, globulin, and hemo-
globin levels. Special attention must be given to
serum glucose and electrolyte concentrations in
patients maintained with steroid preparations,
since prolonged therapy may induce fluid and elec-
trolyte disturbances and hyperglycemia or glycos-
uria.
All nutritional deficiencies, cases of dehydra-
tion, electrolyte imbalances, and respiratory tract
infections must be treated preoperatively. Plasma-
pheresis is commonly used in the preoperative p<
riod to optimize the patient's physical status.
At present, it is recommended that anticholine
terase therapy at the regular dose be continue
until the day of surgery. On the day of surgery,
mild cases, none or half the amount of the mornir
dose of pyridostigmine is administered, whereas
severe cases the full morning dose is prescribe
In patients receiving systemic steroids, suppressu
of the pituitary-adrenal axis should be considere
and the regular dose of steroid should be mai
tained throughout the immediate perioperative
riod. Postoperatively, the cortisone administraii
may be tapered from the second to approximate
the fifth postoperative day.
In view of the propensity for stress to cause
myasthenic crisis, and because many of these
tients demonstrate some emotional instability, sp
cial attention must be paid to the psycholo^
preparation of these patients. Premedication is i
dicated, but care should be taken not to depress
already weakened respiratory apparatus.
2. Intraoperative and Postoperative
Considerations
Anesthesia may be induced with thiopental, f
lowed by the administration of a halogenated dn
Thoracic epidural anesthesia with 1.5 per cent
docaine in addition to a light general anesthe
has been used with good success (the epidu
catheter was also used for postoperative pain
lief)
299
Both anesthesia techniques allow avo
ance of muscle relaxants (see following disc
sion). The surgical approach is most commoi
through a median sternotomy (but may be a tra
verse sternotomy), but in patients who have a n
midline thymoma, a lateral thoracotomy may
necessary. If the incision is a median sternotoi
a single-lumen tube may be used, and if the ir
sion is a lateral thoracotomy, a single- or dout
lumen tube may be used.
Muscle relaxation is a special problem in th
patients. Myasthenia gravis patients are resistan
succinylcholine (more than normal is needed; e
1.5-2.0 mg/kg may be needed to produce
onset of good intubating conditions), and succii
choline is associated with an early onset of a pr
II block, which has a prolonged duration.
300
301
'
likely explanation for these abnormal response
that there is a decreased number of acetylcho
receptors at the motor end-plate. Consistent \
this contention is the observation that myasthi
gravis patients in true remission (asymptom
while receiving no therapy) may not demonsl
resistance to succinylcholine.
302
Furthermore, it is not surprising that pati
with myasthenia gravis have a marked sensitivity
to nondepolarizing relaxants, and all nondepolar-
izing muscle relaxants have an unacceptable dura-
tion of action in these patients if administered in
usual doses (e.g., the ED
50
and ED
95
for vecuron-
ium is anywhere from two to five times less than
that for normal patients).
303
'
304
The decrease in
vecuronium requirement is significantly related to
the patient's acetylcholine receptor antibody ti-
ter.
304
Consequently, if relaxation is required, small
doses of nondepolarizing relaxants should be
given.
Atracurium in small doses (5-15 mg)
305
appears
to be the drug of choice because of its short dura-
tion of action and method of elimination by spon-
taneous decomposition (Hoffman elimination).
Vecuronium in small doses (0.01 mg/kg) has been
successfully used.
305
306
No matter what relaxant
was used, or how much, neuromuscular blockade
must be monitored with a nerve stimulator; indeed,
with appropriate monitoring, a succinylcholine-
vecuronium sequence may be used,
306
or vecuron-
ium may be used for intubation as well as for the
procedure itself.
307
Myasthenia gravis patients are much more sen-
sitive, with respect to neuromuscular blockade, to
isoflurane and halothane.
308
The time course of the
neuromuscular blockade after small doses of non-
depolarizing muscle relaxants is similar to that ob-
served after larger doses in nonmyasthenic sub-
jects.
308
At the end of the operation, the residual
effects of the muscle relaxant can be effectively
antagonized by neostigmine.
A scoring system to predict which patients will
require postoperative ventilatory support has been
devised.
309
The components of the scoring system
consist of duration of myasthenia gravis equal to
or greater than 6 years (equals 12 points); other
concomitant respiratory disease present (equals 10
points); pyridostigmine requirement greater than
750 mg/day (equals 8 points); and a vital capacity
less than 2.9 L (equals 4 points). A score of 10
points or more predicts the need for ventilatory
support. The scoring system was derived in pa-
tients undergoing trans-sternal thymectomy with
halogenated drug anesthesia without muscle relax-
ation; the predictive accuracy of this scoring sys-
tem was 80 per cent.
309
However, in patients
undergoing transcervical thymectomy with the
same anesthetic, the predictive accuracy of the
scoring system was only 13 per cent.
310
The varia-
tion in predictability of the scoring system may be
due to differences in stress between the two sur-
gical approaches.
Alternatively, postoperative mechanical ventila-
tion of all patients with myasthenia gravis has also
been recommended.
287
Extubation can be accom-
plished in an unhurried fashion and according to
such objective criteria as vital capacity and peak
inspiratory force. In addition, this approach allows
time to restart the anticholinesterase therapy after
clinical examination of the patient, especially with
regard to bulbar and respiratory function and test-
ing of handgrip strength with a dynamometer. The
aim of the immediate postoperative anticholines-
terase therapy is to maintain adequate spontaneous
respiratory exchange.
If the spontaneous ventilation is adequate or if
mechanical ventilation is instituted, the anticholin-
ergics are not required in the postoperative period.
However, if spontaneous respiratory exchange is
not adequate, half the regular dose of anticholin-
esterases may be administered during the first 3
postoperative days. On the fourth postoperative
day, regular doses of anticholinesterases may be
started. If the patient cannot take oral medication,
parenteral administration of neostigmine (0.5 mg
to 1 mg intramuscularly) can be given every 2 to
3 hours until medication can be absorbed orally, at
which time pyridostigmine can be restarted.
Because the beneficial effects of thymectomy
can be delayed for several weeks to several years
postoperatively, it is necessary to reassess the
doses for a prolonged postoperative period. Pain
relief can be obtained by titrating small doses of
commonly used narcotics to the desired endpoint.
However, one report claimed that lumbar epidural
morphine (14 ml of 0.5 mg/ml solution) provided
superior pain relief compared with intravenous
narcotics.
3
" All patients with myasthenia gravis
should be monitored postoperatively in an inten-
sive care unit setting.
XIII. ONE-LUNG ANESTHESIA IN
MORBIDLY OBESE PATIENTS
Gastric stapling has become an accepted treat-
ment for refractory morbid obesity.
312
313
Although
the procedure is usually performed through an ab-
dominal incision, a transthoracic transdiaphrag-
matic approach with the patient in the right lateral
decubitus position has been described and provides
an improved operative exposure. As with any in-
trathoracic procedure, the transthoracic transdia-
phragmatic surgical exposure can be greatly in-
creased by using one-lung ventilation. Morbidly
obese patients have heavy chest walls that cause a
reduction in FRC below the closing volume of the
lung, low ventilation-perfusion relationships, and
hypoxemia. Consequently, one-lung ventilation in
morbidly obese patients may be thought to be as-
sociated with an increased risk of hypoxemia.
However, in two studies of morbidly obese pa-
tients undergoing one-lung ventilation, no particu-
lar intraoperative or postoperative problems were
encountered.
314
315
The patients were ventilated
with a tidal volume of 15 ml/kg of ideal weight
and an F,0
2
of 1.0; P
a
0
2
ranged from 72 to 230
mm Hg during one-lung ventilation. The most
probable reason good success was obtained during
one-lung ventilation in these morbidly obese pa-
tients was that the lateral decubitus position al-
lowed the panniculus to displace itself on the op-
erating room table, thereby reducing abdominal
pressure against the diaphragm (compared with the
supine position) and allowing increased FRC and
greater tidal diaphragmatic excursion.
316
Postoper-
ative depression of P
a
0
2
and depression of post-
operative pulmonary function test studies were
slightly greater in the one-lung ventilation patients
compared with a similar group of patients
undergoing only an abdominal incision approach.
Thus, it is reasonable to conclude that morbidly
obese patients can tolerate one-lung ventilation/
anesthesia for transthoracic stapling surgery with
safety as opposed to the abdominal approach.
However, with either approach, special attention to
many details is necessary to avoid problems with
positioning (fitting on the operating room table,
avoiding excessive pressure on nerves, elevation
of the upper half of the body to facilitate respira-
tion), monitoring (arterial catheter, access to cen-
tral veins), the airway (acid aspiration, tracheal
intubation, supplementary oxygen), choice of an-
esthesia (fat sequesters inhalation anesthetics and
therefore metabolism is greater; postoperative res-
piratory depression), and postoperative care (posi-
tioning, ventilation, oxygenation, prevention of
thromboembolism, pain control, and motivation/
ambulation).
XIV. THORACIC OUTLET
SYNDROMES
The thoracic outlet syndromes are due to com-
pression of any one or combination of lower bra-
chial plexus (C-8, T-1), upper brachial plexus
(C-5-C-7), the subclavian artery, and the subcla-
vian vein (Table 15-21). The compression may be
caused by the first rib, a cervical rib, an exostosis
of a cervical vertebra, anomalies of the anterior
scalene muscle, and cancer.
1. Benign Causes
In the benign group of causes of the thoracic
outlet syndrome, 70 to 80 per cent will have a
history of trauma and 5 to 10 per cent will have a
cervical rib.
317
318
Compression of the lower bra-
chial plexus causes pain in the supraclavicular and
infraclavicular fossae, the back of the neck and
rhomboid areas, and the axilla and inner arm, with
Table 15-21 THORACIC OUTLET SYNDROMES AND THEIR DIAGNOSIS AND TREATMENT
Thoracic Outlet Syndrome Diagnosis
Treatment (Transaxillary and/or
Supraclavicular Approaches for
Benign Disease)
Compression of the lower brachial
plexus (C-8, Tl)
Compression of the upper brachial
plexus (C-5-C-7)
Compression of subclavian artery
Compression of subclavian vein
Superior sulcus (Pancoast) tumor
causing compression of lower
brachial plexus, subclavian vessels
History and physical (see text), nerve
conduction/function studies; findings
are in the distribution of these nerves
History and physical (see text), nerve
conduction/function studies; findings
are in the distribution of these nerves
History and physical (see text),
arteriography
History and physical (see text),
venography
All of the above and staging procedures
(see Figs. 15-1 and 15-8)
Removal of first rib, cervical rib (i.e.,
bony floor of thoracic outlet), and
anomalous fibromuscular bands
Removal of parts of the anterior scalene
muscle
Removal of first or cervical rib, various
vascular surgery techniques (resection
of aneurysm, endarterectomy)
Removal of first rib, anomalous
fibromuscular bands, possible
fibrinolytic enzymes if thrombosis
present
Irradiation followed by en bloc removal
of apical chest wall (including
posterior portions of first three ribs,
transverse processes of these thoracic
vertebra, intercostal nerves, lower
brachial plexus, stellate ganglion, and
dorsal sympathetic chain), underlying
lung, artery and vein procedures as
dictated by tumor involvement
numbness and tingling radiating through the ulnar
nerve distribution from the axilla to the ring and
small fingers.
317
Symptoms may include dysesthesias of coldness
and a dead feeling, with impaired strength and
dexterity of the hand and fingers. The plexus is
tender, and the 3-min elevated arm stress test
quickly reproduces the symptoms. Treatment usu-
ally consists of removing the bony floor of the
thoracic outlet, the first rib, and the cervical rib if
present, along with the various types of anomalous
fibromuscular bands, which usually provides re-
markably good relief.
Compression of the upper brachial plexus
causes pain in the anterolateral aspect of the neck
and in the region of the brachial plexus itself just
behind the sternocleidomastoid muscle.
317
Pain ra-
diates up to the mandible and ear, posteriorly to
the scapula, anteriorly into the upper chest, later-
ally across the top of the shoulder, and down the
outer aspect of the arm in a C-5, C-6, and C-7
nerve distribution.
Examination shows exquisite tenderness of the
C-5 nerve and upper trunk of the plexus, especially
with the head tilted to the opposite side, and much
less incidence of edema, coldness, paresthesia, and
motor impairment of the hand compared with the
lower plexus pattern. Treatment usually consists of
removing the anomalous encasing anterior scalene
muscle.
Compression of the subclavian artery (almost
always caused by a cervical or first rib) causes
ischemia of the upper extremity and may cause
aneurysm formation and atherosclerotic plaques.
Treatment consists of removal of the offending rib
and appropriate vascular repair. Compression of
the subclavian vein causes congestion and edema
in the arm and hand, and venography will reveal
whether the vessel is thrombosed.
Treatment consists of removal of the offending
rib and/or scalene muscle as well as perhaps fibri-
nolytic therapy. In general, using a combination of
transaxillary first-rib resection, supraclavicular cer-
vical rib resection, or scalenectomy, 70 per cent of
patients will obtain long-lasting relief.
318
The
transthoracic route (anterolateral thoracotomy) has
provided the same or better benefit with less risk
of brachial plexus damage.
3
'
9
With all approaches,
a small percentage of patients may experience recur-
rence because of new scar tissue formation.
317
"
319
2. Superior Sulcus (Pancoast) Tumor
Although a variety of nonbronchogenic tumors
may produce the clinical pattern peculiar to being
located at the thoracic inlet, it is generally accepted
that the most common cause of Pancoast's syn-
drome is a bronchogenic carcinoma arising in or
near the superior pulmonary sulcus and invading
the adjoining extrapulmonic structures by direct
extension. It is the location of the tumor that is
significant in producing the characteristic clinical
pattern and not its pathologic structure or tissue of
origin.
Bronchogenic carcinomas situated in the narrow
confines of the thoracic inlet invade the lymphatics
in the endothoracic fascia and involve, by direct
extension, the lower roots of the brachial plexus,
intercostal nerves, stellate ganglion, sympathetic
chain, and adjacent ribs and vertebrae, producing
severe pain and Horner's syndrome (Pancoast's
syndrome) (Fig. 15-41).
320
The symptoms are
characteristic of the location of the tumor in the
superior pulmonary sulcus or thoracic inlet adja-
cent to the first and second thoracic and eighth
cervical nerve roots, the sympathetic chain, and
stellate ganglion. Initially, there is localized pain
in the shoulder and vertebral border of the scapula.
Later it may extend down the ulnar distribution of
the arm to the elbow (involvement of T-l) and
finally to the ulnar surface of the forearm and
small and ring fingers of the hand (C-8). If the
tumor extends to the sympathetic chain and stellate
ganglion, Horner's syndrome and anhidrosis de-
velop on the same side of the face and upper
extremity.
The complete diagnostic workup and staging
and treatment plan is shown in Figure 15-42.
320
Staging is determined by the location of the lesion
and its metastases. The true superior sulcus tumor
(Pancoast's tumor) is usually T3, which describes
the extension of the tumor through the visceral
pleura into the parietal pleura and the chest wall.
No lymph node metastases (NO) or metastases
only to the hilar nodes (Nl) are the usual criteria
for operation. Mediastinal node metastases (N2)
indicate "inoperability" (except with an N2 para-
tracheal mediastinal node on the ipsilateral side of
the right chest that contains an intranodal metasta-
sis, which may be considered "regional" spread).
MO is the only stage for the operable group be-
cause peripheral metastases (Ml) signal a poor
prognosis and contraindicate surgery.
Previously, superior sulcus tumors were consid-
ered inoperable and were not often successfully
palliated with irradiation alone. Best results seem
to occur when the tumor and the localized adjacent
area, including the superior mediastinal nodes, are
treated preoperatively with 3000 rads given over 2
to 3 weeks.
The purpose of the preoperative irradiation is to
shrink the tumor, to weaken any cells that might
be "spilled" at the time of surgery, and to block
the lymphatics temporarily, which absorb cells
5 8 2 Anesthesia for Special Elective Therapeutic Procedures
Figure 15-41 Composite illustra-
tion of a patient with clinical mani-
festations, chest radiograph of tumor
with superior sulcus location, and
gross pathologic diagnostic demon-
stration of the lung carcinoma invad-
ing the chest wall, brachial plexus,
and sympathetic nerves. (Copyright
1979 CIBA-GEIGY Corporation.
Reproduced with permission from
The CIBA Collection of Medical Il-
lustrations by Frank H. Netter, M.D.
All rights reserved.)
during the operative procedure. This is a reasona-
ble approach considering the close margin of en
bloc dissection in this type of tumor. Preoperative
treatment of greater than 4000 rads may lead to
poor healing after surgery.
An interval of 2 to 4 weeks after radiation ther-
apy allows the radiation to have maximal effect;
resection of the tumor en bloc with the chest wall
is then performed. Typical results of this treatment
plan are a survival rate of 55 per cent at 3 years,
321
27 to 35 per cent at 5 years,
320
321
and palliation of
pain in 72 per cent.
321
There are several anesthetic considerations tha
are unique to surgery for thoracic outlet problems
First, if the surgical approach is extrapleural, on<
needs to be especially watchful for an inadverten
surgical entry into the operative pleural space
causing either tension pneumothorax and/or col
lapse of the lung. Second, the transthoracic ap
proach is greatly facilitated by one-lung ventilatio!
(collapse of the lung on the opposite side). Thirc
if the disease and surgery involve the subclavia
Figure 1542 Superior pulmonary sulcus carcinoma diagnosis (left column), staging (middle column), and treatment (right column).
(CBC = complete blood count; SMA 20 = 20 blood chemistry tests; CT = computed tomography.) (From Urschel HC Jr: Superior
pulmonary sulcus carcinoma. Surg Clin North Am 68:497-509, 1988. Used with permission.)
vessels, large-bore intravenous access needs to be
ensured to deal with major blood loss. Fourth, if
the chest wall is unstable at the end of the proce-
dure, then a period of postoperative mechanical
ventilation should be seriously considered.
REFERENCES
1. Dedhia HV, Leroy L, Jain PR, et al.: Endoscopic laser
therapy for respiratory distress due to obstructive airway
tumors. Crit Care Med 13:464-467, 1985.
2. Gelb AF, Epstein JD: Laser in treatment of lung cancer.
Chest 86:662-666, 1984.
3. Unger M: Bronchoscopic utilization of the Nd:YAG laser
for obstructing lesions of the trachea and bronchi. Surg
Clin North Am 64:931-938, 1984.
4. Parr GVS, Unger M, Trout RG, et al: One hundred neo-
dymium-YAG laser ablations of obstructing tracheal neo-
plasms. Ann Thorac Surg 38:374-381, 1984.
5. George PJM, Clarke G, Tolfree S, Garrett CPO, Hetzel
MR: Changes in regional ventilation and perfusion of the
lung after endoscopic laser treatment. Thorax 45:248-
253, 1990.
6. Gelb AF, Tashkin DP, Epstein JD, Szeftel A, Fairshter R:
Physiologic characteristics of malignant unilateral main-
stem bronchial obstruction. Am Rev Respir Dis 138:
1382-1385, 1988.
7. Mohsenifar , Jasper AC, Koerner SK: Physiologic as-
sessment of lung function in patients undergoing laser
photoresection of tracheobronchial tree tumors. Chest
93:65-69, 1988.
8. Warner ME, Warner MA, Leonard PF: Anesthesia for
neodymium-YAG (Nd-YAG) laser resection of major air-
way obstructing tumors. Anesthesiology 60:230-232,
1984.
9. Jacobson MJ, Lo Cicero J: Endobronchial treatment of
lung carcinoma. Chest 100:837-841, 1991.
10. Goldberg M: Endoscopic laser treatment of bronchogenic
carcinoma. Surg Clin North Am 68:635-644, 1988.
11. Snow JC: Fire hazard during C0
2
-laser microsurgery on
larynx and trachea. Anesth Analg 55:146, 1976.
12. Hirshman CA, Smith J: Indirect ignition of the endotra-
cheal tube during carbon dioxide laser surgery. Arch Oto-
laryngol 106:639, 1980.
13. Burgess GE, LeJeune FE: Endotracheal tube ignition dur-
ing laser surgery. Otolaryngology 5:561, 1979.
14. Patel KF, Hicks JN: Prevention of fire hazards associated
with the use of carbon dioxide lasers. Anesth Analg
60:885, 1981.
15. Brutinel WM, McDougall JC, Cortese DA: Broncho-
scopic therapy with neodymium-yttrium-aluminum-gar-
net (Nd-YAG) laser during intravenous anesthesia. Chest
84:518-521, 1983.
16. Vourch G, Fischler M, Personne C, et al: Anesthetic
management during Nd-YAG laser resection for major
tracheobronchial obstructing tumors. Anesthesiology
61:150-151,1984.
17. Rudow M, Hill AB, Thompson NW, Finch J: Helium-
oxygen mixtures in airway obstruction due to thyroid
carcinoma. Can Anaesth Soc J 33:498-501, 1986.
18. Mizrahi S, Yaari Y, Lugassy JG, Cotev S: Major airway
obstruction relieved by helium/oxygen breathing. Crit
Care Med 14:986-987, 1986.
19. Pashayan AG, Gravenstein JS, Cassisi NJ, McLaughlin
G: The helium protocol for laryngotracheal operations
with CO
:
laser: A retrospective review of 523 cases.
20. Sosis M: Anesthesia for laser surgery. J Voice 3:163-
174, 1989.
21. Sosis MB: On the development of a new laser resistant
endotracheal tube. J Clin Anesth 4:87-88, 1992.
22. Sosis MB: Evaluation of five metallic tapes for protection
of endotracheal tubes during CO, laser surgery. Anesth
Analg 68:392-393, 1989.
23. Green JM, Gonzalez RM, Sonbolian N, Rehkopf P: The
resistance to carbon dioxide laser ignition of a new en-
dotracheal tube: Xomed Laser-Shield II. J Clin Anesth
4:89-92, 1992.
24. Rampil IJ: Anesthetic consideration for laser surgery.
Anesth Analg 74:424^*35, 1992.
25. Lejeune FE, Guice C, LeTard F, Marice H: Heat sink
protection against lasering endotracheal cuffs. Ann Otol
Rhinol Laryngol 9:606-607, 1982.
26. Goldhill DR, Hill AJ, Whitburn RH, Feneck RO, George
PJM, Keeling P: Carboxyhemoglobin concentrations,
pulse oximetry and arterial blood gas tensions during jet
ventilation for Nd:YAG laser bronchoscopy. Br J Anaesth
65:749-753, 1990.
27. Sosis M: Evaluation of a new laser resistant anesthesia
circuit protector, drape and patient eye shield. Anesth
Analg 72:S265, 1991.
28. Brutinel WM, Cortese DA, Edell ES, McDougall JC,
Prakash UBS: Complications of Nd:YAG laser therapy.
Chest 94:902-903, 1988.
29. Cavaliere S, Foccoli , Farina PL: Nd:YAG laser bron-
choscopy: A five year experience with 1,396 applications
in 1,000 patients. Chest 94:15-21, 1988.
30. Arabian A, Spagnolo SV: Laser therapy in patients with
primary lung cancer. Chest 86:519-523, 1984.
31. McDougall JC, Cortese DA: Neodymium-YAG laser
therapy of malignant airway obstruction. Mayo Clin Proc
58:35-39, 1983.
32. McElvein RB, Zorn GL: Indications, results and compli-
cations of bronchoscopic carbon dioxide laser therapy.
Ann Surg 199:522-525, 1984.
33. Chan AL, Tharratt RS, Siefkin AD, Albertson TE, Volz
WG, Allen RP: Nd:YAG laser bronchoscopy: Rigid or
fiberoptic mode? Chest 98:271-275, 1990.
34. Schneider M, Probst R: High frequency jet ventilation via
a telescope for endobronchial laser surgery. Can Anaesth
37:372-376, 1990.
35. Personne C, Colchen A, Leroy M, Vourc'h G, Toty L:
Indications and techniques for endoscopic laser resections
in bronchology. J Thorac Cardiovasc Surg 91:710-715,
1986.
36. Dumas JF, Shapshay S, Bourcereau J, Cavaliere S, Meric
, Garbi , Beamis J: Principles for safety in application
of neodymium-YAG laser in bronchology. Chest 86:163-
168, 1984.
37. McElvein RB: Treatment of malignant tracheobronchial
tree obstruction. Ann Thorac Surg 48:463-464, 1989.
38. George RJM, Garrett CPO, Nixon C, Hetzel MR, Nanson
E, Millard FJC: Laser treatment of tracheobronchial tu-
mors: Local or general anesthesia? Thorax 42:656-660,
1987.
39. Hetzel MR, Smith SGT: Endoscopic palliation of tracheo-
bronchial malignancies. Thorax 46:325-333, 1991.
40. Bloomquist S, Algotsson L, Karlsson SE: Anaesthesia for
resection of tumours in the trachea and central bronchi
using the Nd-YAG laser technique. Acta Anaesthesiol
Scand 34:506-510, 1990.
41. Benumof JL: Another use for the plastic transparent
dressing (letter). Anesthesiology 63:334, 1985.
42. Grant RP, White SA, Brand SC: Modified rigid broncho-
scope for Nd:YAG laser resection of tracheobronchial
obstructing lesions. Anesthesiology 66:575-576, 1987.
43. Edell ES, Cortese DA: Bronchoscopic phototherapy with
hematoporphyin derivative for treatment of localized
bronchogenic carcinoma: A 5-year experience. Mayo
Clin Proc 62:8-14, 1987.
CHAPTER 16
Anesthesia for Esophageal
Surgery
Preoperative Considerations
A.
B.
C.
D.
E.
F.
History
Nutritional Status
Perioperative Regurgitation and
Aspiration
Preoperative Chemotherapy
Diagnostic Workup Logic for
Esophageal Lesions (With Special
Emphasis on Esophageal
Carcinoma)
1. Esophagoscopy
a. Indications
b. Anesthetic Technique
c. Complications
Benign Esophageal Diseases (Often
Requiring Surgical Repair)
1. Hiatal Hernia
2. Benign Esophagorespiratory
Fistula
3. Benign Esophageal Stricture
4. Achalasia
5. Esophageal Rupture and
Perforation
a. Clinical/Diagnostic Features
b. Treatment of Esophageal
Perforations
G. Surgical Approaches/Incisions for
Esophageal Surgery (With
Emphasis on Esophageal
Carcinoma)
II. Intraoperative Anesthetic
Considerations
A. Monitoring
B. Induction of Anesthesia
C. Airway Considerations
1. General Airway Considerations
2. Airway Considerations in Cases
of Esophagorespiratory Fistula
D. Hemodynamic Considerations
E. Esophagogastrointestinal
Considerations
III. Postoperative Considerations
592 Anesthesia for Esophageal Surgery
I. PREOPERATIVE
CONSIDERATIONS
A. History
1
The most common symptom of esophageal dis-
ease is difficulty in swallowing, which is termed
dysphagia. The difficulty may be characterized as
various degrees of obstruction and/or pain during
swallowing. Dysphagia that has been present for
any significant length of time is almost always
accompanied by weight loss, dehydration, prerenal
failure, hypoalbuminemia, decreased plasma on-
cotic pressure, anemia, electrolyte abnormalities,
and depressed immune mechanisms and muscle
power (see Nutritional Status).
Heartburn is also a common symptom and is
due to reflux of gastric contents into the esopha-
gus. Heartburn is often a postprandial symptom
brought on by recumbency, leaning forward,
belching, and severe exercise. In addition, many
patients with esophageal disease (those with a sig-
nificant pouch proximal to an obstruction and
those with hiatal hernia) are prone to chronic re-
gurgitation and aspiration and may have chronic
lung disease (see Perioperative Regurgitation and
Aspiration).
Finally, patients with esophageal cancer are
often treated with anticancer drugs that may cause
considerable organ toxicity; doxorubicin (Adria-
mycin) can cause very severe refractory left ven-
tricular myopathy, bleomycin can cause respira-
tory failure, and mitomycin can cause both
pulmonary and nephrotoxicity (see Chapter 5 and
Preoperative Chemotherapy discussed later).
B. Nutritional Status
Patients with esophageal diseases present with
dysphagia, nausea, and vomiting; poor nutrition
may result, which, in turn, leads to many secon-
dary metabolic and functional changes that have
important anesthetic implications (Table 16-1).
2
Dehydration resulting from poor fluid intake may
be present, rendering the patient more susceptible
to anesthesia-induced hemodynamic suppression.
Dehydration may also cause prerenal failure as
manifested by decreased urine output and in-
creased blood urea nitrogen and creatinine.
Hypoalbuminemia renders the patient more sus-
ceptible to an exaggerated response to drugs that
are normally protein bound (thiopental, muscle re-
laxants, local anesthetics). Decreased plasma on-
cotic pressure will aso make the patient more sus-
ceptible to pulmonary edema. Low levels of
hemoglobin, resulting from poor iron intake or
chronic bleeding from the site of the tumor, de-
Table 16-1 POOR NUTRITION PROBLEMS
CAUSED BY DYSPHAGIA AND
NAUSEA AND VOMITING AND
THEIR ANESTHETIC
IMPLICATIONS
Dysphagia-lnduced
Nutrition Problem Anesthetic Implication
Dehydration
Hypoalbuminemia
Hemodynamic instability, prerenal
failure
Exaggerated response to drugs,
increased susceptibility to
pulmonary edema
Decreased hemoglobin Decreased 0
2
transport, tissue
hypoxia
Hypomagnesemia Arrhythmias, altered neuromuscular
transmission
Malnutrition Depressed immune responses,
sepsis, decreased muscle power,
poor postoperative respiratory
effort
crease oxygen transport potential and often result
in preoperative elevated cardiac output and a right-
shifted oxygen-hemoglobin dissociation curve.
Subsequent decreased cardiac output caused by
anesthesia and left shifting of the oxygen-hemo-
globin dissociation curve resulting from hyperven-
tilation may result in tissue hypoperfusion and hy-
poxia.
Electrolytes may be altered secondary to star-
vation, commonly producing either hypokalemia
or hypomagnesemia. Hypokalemia may be accen-
tuated by unintentional hyperventilation or meta-
bolic alkalosis (as may be caused by citrated
blood, administration of sodium bicarbonate, loss
of gastric acid, diuretics), which may cause cardiac
arrhythmias, especially in a digitalized patient.
Low levels of magnesium may cause alterations in
the electrocardiogram and in neuromuscular trans-
mission.
Malnutrition also suppresses the immune re-
sponse and impedes wound healing; these latter
two abnormalities may increase the risk of wound
infection. Malnutrition also decreases muscle
power, and a poorly nourished patient is more
likely to experience postoperative respiratory com-
plications.
All these metabolic and functional abnormalities
can now be corrected by total parenteral nutrition
before surgery.
3
Consequently, parenteral nutrition
is now frequently used preoperatively, and a typi-
cal daily intravenous nutrition prescription is
shown in Table 16-2.
4
The calorie source for pa-
renterally fed patients is usually a combination of
carbohydrate (dextrose) and fat. For most adult
patients, 40 to 45 calories per kilogram body
weight per day are adequate. Intravenous protein
requirements are usually provided as synthesized
amino acids.
Table 16-2 DAILY INTRAVENOUS NUTRITION
PRESCRIPTION
Source
10% Fat solution
Dextrose 50% (500 g)
Amino acids 8.5%
Multiple essential ions/
minerals and vitamins
Amount
(ml)
500
1000
1000
Nutrition
550 calories
1700 calories
85 g
However, it is important for the anesthesiologist
to realize that parenteral nutrition may pose two
common special problems during anesthesia and
surgery (Table 16-3). First, the risk of intraopera-
tive hypoglycemia is increased in these patients.
5
The islet cells of the pancreas are stimulated dur-
ing the infusion of dextrose-rich solutions, and
these patients may have high levels of endogenous
insulin. If the dextrose-rich infusion is stopped
when the patient arrives in the operating room, and
the islet cells continue to secrete insulin, hypogly-
cemia may develop intraoperatively within 45 to
60 min.
If the hypoglycemic episode is severe and pro-
longed, it may result in delayed awakening or
coma in the postoperative period. This problem
can be avoided either by infusing the parenteral
nutrition solution throughout the intraoperative pe-
riod at half the presurgical flow rate or by substi-
tuting 10 per cent dextrose in water at the same
infusion rate.
2
Either technique prevents the intra-
operative hypoglycemia, but both techniques re-
quire perioperative monitoring of the blood sugar.
Second, the preoperative administration of large
amounts of glucose-rich parenteral nutritional so-
lutions may result in Hpogenesis.
6
Lipogenesis re-
sults in increased carbon dioxide production,
which is removed in the awake patient by a spon-
taneous increase in ventilation. However, when
these patients are paralyzed and mechanically ven-
tilated intraoperatively, the use of a normal minute
ventilation may result in severe hypercarbia.
7
This
can be prevented by hyperventilation, sometimes
in excess of twice the normal minute ventilation.
Postoperatively, the increase in C0
2
production
(caused by lipogenesis) may be a critical factor
inhibiting the weaning of a patient from mechani-
cal ventilatory support. Decreasing the glucose
load or changing to fat emulsions, which causes a
respiratory quotient of 0.7, may be useful in these
circumstances (see chapter 3).
6
C. Perioperative Regurgitation and
Aspiration
Patients with esophageal disease are prone to
regurgitation and aspiration. In very debilitated pa-
tients with poor laryngeal reflexes, regurgitation
may lead to chronic aspiration, atelectasis, pneu-
monia, and chronic lung disease. In many patients,
the aspiration of acid gastric contents may take
place during sleep.
8
9
It is thought that the gastric
reflux may be the cause of asthma in some pa-
tients, and both diseases/problems may be cured
by an antireflux surgical procedure.
10
In patients with an esophageal obstruction, it
may take many hours for the esophagus above the
stricture to empty, and having the patient fast over-
night before surgery does not guarantee an empty
pouch above the stricture or obstruction. Conse-
quently, if there is any question about the proximal
esophageal pouch, it should be emptied by passing
a large nasogastric tube and suctioning before in-
duction of anesthesia. In spite of these attempts,
the pouch may not be completely empty, and it is
reasonable to consider these patients as having a
full stomach.
The gastroesophageal barrier consists of both a
valve (caused by the angle at which the esophagus
enters the stomach) and a sphincter (the lower
esophageal sphincter)." Patients with hiatal hernia
or esophageal carcinoma may have an obliterated
esophagogastric angle and physiologically incom-
petent or dysfunctioning lower esophageal sphinc-
ter (often indicated by history of reflux).
12
They
are also thus prone to regurgitation and aspiration
during induction of anesthesia. With these consid-
erations in mind, intubation should be performed
either in an awake patient with sedation and topi-
cal anesthesia or by rapid-sequence induction.
A nasogastric tube should be available toward
the end of the operation for threading through any
esophageal anastomosis and into the stomach, if
the surgeon wishes. The purpose of the tube is to
prevent gastric distention and to indicate whether
bleeding is continuing from the stomach. It pro-
vides no guarantee against aspiration of gastric
contents into the lungs. Patients without significant
pulmonary disease can be allowed to breathe spon-
taneously after an uncomplicated procedure, but
they should not be extubated until they are alert
and sitting upright because there may be no me-
chanical barrier to reflux and aspiration into the
lungs postoperatively.
Table 16-3 INTRAOPERATIVE PROBLEMS
THAT CAN BE CAUSED BY
PREOPERATIVE TOTAL
PARENTERAL NUTRITION
1. Hypoglycemia (reactive insulin secretion)
2. Hypercarbia (increased C0
2
production because of
lipogenesis)
D. Preoperative Chemotherapy (See
Chapter 5)
Patients with esophageal cancer are often treated
with chemotherapeutic drugs before surgery. The
anticancer effect of these antibiotics is produced
by formation of relatively stable complexes with
deoxyribonucleic acid (DNA), which inhibits
DNA and/or ribonucleic acid (RNA) function and
synthesis.
13
These drugs affect not only the cancer
cells but also rapidly growing normal cells (eryth-
ropoiesis, leukocyte and platelet production, and
gastrointestinal tract lining). Commonly used che-
motherapeutic drugs in carcinoma of the esopha-
gus belong to the antibiotic group, which includes
doxorubicin, bleomycin, and mitomycin C.
Toxicity secondary to doxorubicin includes se-
vere cardiomyopathy, seen in 1.8 per cent of pa-
tients treated. When cardiomyopathy develops, it
has been shown to be irreversible in 60 per cent of
the patients, with death occurring within 3 weeks
of the onset of the symptoms.
14
The left ventricular
failure that occurs with doxorubicin is refractory
to inotropic drugs. Electrocardiogram abnormali-
ties are also part of doxorubicin toxicity, but they
resolve 1 to 2 months after cessation of therapy.
15
Bleomycin's action is similar to the effect of
radiation, and bleomycin and radiation may act
synergistically during simultaneous therapy.
16
Pul-
monary toxicity is the most life-threatening, drug-
limiting effect, reported in 15 to 25 per cent of
patients.
17
-
I8
Predisposing factors include age
greater than 20 years, dose greater than 400 units,
underlying pulmonary disease, and prior radiation
therapy. Signs and symptoms of pulmonary toxic-
ity are cough, dyspnea, and basal rales. The dis-
ease may manifest itself with minimal radiologic
changes and normal resting P
a
0
2
, or it may pro-
gress to severe hypoxemia at rest, with radiologic
changes similar to severe adult respiratory distress
syndrome. A controversial factor that may predis-
pose patients to pulmonary toxicity is the admin-
istration of oxygen in high concentrations.
1920
Mitomycin C is also highly toxic and can cause
pulmonary fibrosis and nephrotoxicity.
21
Thus, pa-
tients receiving mitomycin C should have their
pulmonary and renal status fully evaluated in the
preoperative period.
E. Diagnostic Workup Logic for
Esophageal Lesions
22
(With Special
Emphasis on Esophageal Carcinoma)
There are numerous esophageal lesions and dis-
orders. In order of decreasing frequency of occur-
rence, they are tumors (squamous cell carcinomas
and adenocarcinomas are most common), hiatal
hernia, benign strictures (ingestion of caustic fluids
causes strictures in the cervical esophagus,
whereas reflux esophagitis causes strictures in the
lower third of the esophagus), foreign bodies, di-
verticula, achalasia, esophagorespiratory tract fis-
tula, traumatic perforations, and various motility
disorders (such as scleroderma).
After routine screening chest X-rays, a barium
swallow (esophagogram) under cinefluoroscopy
should be performed (Fig. 16-1). The only con-
traindication to obtaining an esophagogram is
when a fistula to the trachea or bronchus is sus-
pected. The barium swallow in a very high per-
centage of cases gives a strong indication of diag-
nosis. If a structural lesion is suspected, the
esophagogram should be followed by esophagos-
copy.
Esophagoscopy should be performed with a
flexible fiberoptic instrument; however, when dif-
ficulty is encountered while obtaining an adequate
biopsy specimen, a rigid esophagoscope may pro-
vide improved conditions for collecting a biopsy.
Rigid esophagoscopy should be avoided as the first
diagnostic test whenever possible because of the
risk of inadvertent perforation of an esophageal
tumor. Less often, the esophagogram shows that
the symptoms are not due to a structural obstruc-
tion, and manometric motility studies, pH studies,
and provocation studies (Bernstein test [repro-
duces heartburn and pain by dripping a gastric
acid-like solution through a nasogastric tube 30 tc
35 cm from the nose, which is the distal esopha-
gus] and the Tensilon test [reproduces chest pair
by causing increased esophageal contractions]) are
indicated. If the sequence of esophagogram anc
esophagoscopy has not made possible diagnosis
thoracotomy rarely will be required for diagnosis.
If the diagnosis of carcinoma of the esophagu;
is being entertained, strong consideration shouk
be given to examination of the mediastinum (me
diastinoscopy or computed tomographic scanninj
and nuclear magnetic resonance imaging; how
ever, the value of computed tomography and nu
clear magnetic resonance imaging has been sen
ously questioned)
23
because 60 to 80 per cent o
esophageal tumors may involve the mediastinum
by the time surgery is performed
24-27
(see discus
sion of surgical approach to esophagectomy).
Bronchoscopy should be performed in patient
with upper or middle one third esophageal lesions
because nearly one sixth of these lesions will hav
tracheobronchial tree involvement or vocal cor
paralysis.
27
The staging of carcinoma of the esophagus i
similar to that of the lung (Table 164). Based o
preoperative staging, 60 per cent of these lesior
are considered to be potentially resectable (Fij
16-2). Unfortunately, and perhaps because of th
Figure 16-1 Preoperative evalua-
tion logic of esophageal lesions. See
text for full explanation.
limitations of mediastinal examination,
23
final stag-
ing is often done at the time of surgery and is most
often upgraded by the intraoperative and postop-
erative (histologic) findings.
27
The vast majority of
esophageal carcinoma is squamous cell, and a
small percentage is adenocarcinoma and small-cell
carcinoma. All esophageal cancers have a large
male preponderance (approximately 3:1).
Nonstructural lesions may be initially treated
with antireflux (Table 16-5) or antimotility (Table
16-6) measures. The distinction between these two
categories (reflux vs. motility disorder) is impor-
tant because some of the treatments are contrain-
dicated for the other disease (compare Table 16-5
with Table 16-6).
The physiologic assessment of overall cardio-
respiratory function of the patient requiring esoph-
ageal surgery is similar to that of a patient with
lung cancer. As with resection of lung carcinoma,
age per se is not a contraindication to esophagec-
tomy (operative mortality of 13 per cent in patients
older than 70 years compares favorably with the
overall operative mortality of 10-20 per cent [see
later discussion]). As with lung carcinoma, resec-
tion is the best hope of cure for patients with
esophageal carcinoma and is frequently the best
form of palliation.
28
1. Esophagoscopy
a. INDICATIONS
The major indication for esophagoscopy is dem-
onstration of an esophageal lesion following con-
trast studies that needs either etiologic or anatomic
clarification.
29
Such lesions include all strictures,
intraluminal filling defects, mucosal abnormalities,
and upper gastrointestinal bleeding. In most of
these situations, biopsy specimens and cytologic
brushings will also need to be obtained. Candi-
dates for esophagoscopy include patients who
have symptoms of difficulty in swallowing and
those with esophageal reflux. Finally, therapeutic
esophagoscopy is required for removal of foreign
bodies, dilatation of strictures, placement of plastic
prosthesis across a malignant stricture, the injec-
tion of sclerosing agents into esophageal varices,
and coagulation of bleeding lesions.
Table 16-7 compares the advantages and dis-
advantages of fiberoptic esophagoscopy and open-
tube rigid esophagoscopy.
29
Generally, fiberoptic
esophagoscopy lends itself to greater patient com-
fort, the ability to examine the entire upper gastroin-
testinal tract, and greater safety. Consequently, flex-
ible fiberoptic esophagoscopy is usually carried
out under local anesthesia. Disadvantages of the
Table 16-4 UICC/AJCC TNM STAGING SYSTEM FOR CARCINOMA OF THE ESOPHAGUS't
*From Muller JM, Erasmi H, Stelzner M, et al: Surgical therapy of oesophageal carcinoma. Br J Surg 77:845-857, 1990. Used
with permission.
tThe evaluation of the primary tumor (T) no longer considers anatomic location, length of tumor, or percentage of wall
circumference involved. Only depth of esophageal wall invasion is assessed. The evaluation of regional lymph nodes (N) has also
been simplified. Anatomic site and side of involvement are no longer considered. Whether a regional lymph node contains metastatic
carcinoma is the only staging criterion. Sites of distant metastasis (M) have been redefined to include celiac lymph nodes.
fiberoptic instrument are the inability to obtain
deep biopsy specimens and inadequate instrumen-
tation for foreign body removal.
Specific indications for the use of open-tube
esophagoscopy include removal of a foreign body,
dilation of strictures, and determination of the
source of massive esophageal bleeding. Although
open-tube esophagoscopy can be carried out under
Figure 16-2 Survival rate after resection according to stage (Union Internationale Contre Cancer; 1978) of the esophageal
carcinoma. Stage 1 (): - 739; Stage II ( ): = 1592; Stage III (): = 2995; Stage IV (): - 1392. (From Muller JM,
Erasmi H, Stelzner M, et al: Surgical therapy of oesophageal carcinoma. Br J Surg 77:845-857, 1990. Used with permission.)
J
Anesthesia for Esophageal Surgery 5 9 7
Table 16-5 INITIAL THERAPY FOR
GASTROESOPHAGEAL REFLUX
1. Diet; avoid irritating substances (citrus juices, tomato
products, coffee, spices).
2. Avoid substances that lower the lower esophageal sphincter
pressure (nicotine, fatty foods, alcohol, peppermint,
nitrates, anticholinergics, theophylline, calcium blockers).
3. Avoid nighttime sedatives; arousal from sleep may help to
clear the esophagus.
4. Elevate head of bed.
5. Take antacids.
6. Take H
:
-blockers.
7. Take metoclopramide.
8. Undergo surgery.
topical anesthesia, it is more readily accomplished
under general endotracheal anesthesia.
30
Open-
tube esophagoscopy is to be avoided in the patient
with a large thoracic aneurysm because instrumen-
tation may cause rupture. The procedure is also
contraindicated in patients with acute pharyngitis
because infection may be aggravated. Finally, the
risk of perforation during dilation of malignant
strictures with a rigid esophagoscope has been ap-
proximately 10 per cent.
31
b. ANESTHETIC TECHNIQUE
Topical anesthesia of the oropharynx for fiber-
optic esophagoscopy can be achieved by having
the patient gargle viscous lidocaine followed by a
tetracaine or lidocaine spray. The patient is further
sedated before the examination with small doses
of a sedative and/or narcotic.
General anesthesia is usually required for rigid
esophagoscopy (certainly for children).
32
Atropine
premedication is useful in decreasing secretions
and in preventing a vagal reaction to gastric disten-
sion. Oropharyngeal topical anesthesia (with lido-
caine gargle and spray) is useful in minimizing the
Table 16-6 INITIAL THERAPY FOR
ESOPHAGEAL MOTILITY
DISORDERS OTHER THAN
ACHALASIA
Mild Cases Severe Cases
Nitrates
Nitroglycerin (0.4 mg prn)
Isosorbidc ( 1020 mg qid)
Anticholinergics
Dicyclomine (20 mg tid)
Tranquilizers/antidepressants
Diazepam (5 mg tid)
Doxepin (50 mg hs)
Reassurance
Calcium channel-blocking
agents
Diltiazem (60-90 mg tid)
Nifedipine (l 0-20mg tid)
Esophageal dilatation (50
bougie prn)
Psychological evaluation
Esophagomyotomy (only in
extreme cases)
Abbreviations: prn = as occasion requires; qid = four times
daily; tid = three times daily; hs = at bedtime.
amount of general anesthesia required. Preoxyge-
nation is followed by intravenous induction using
cricoid pressure to control reflux of esophageal
contents. After intravenous induction of anesthesia
and paralysis, a small-sized endotracheal tube
should be inserted to prevent anterior compression
of the esophagus (as a large-sized endotracheal
tube might do) and thereby allow more room for
esophageal instrumentation and passage of the
esophagoscope.
The endotracheal tube should be positioned to
the left side of the patient's mouth so that the
esophagoscope can be introduced through the right
side of the mouth. The connecting tubing should
be fastened over the front of the patient's chest to
allow the surgeon easy access to the mouth. The
patient's eyes should be protected because they are
likely to be concealed by a drape around the head.
Full relaxation should be maintained at least
until the esophagoscope has been passed through
the cricopharyngeal sphincter. If this passage
proves difficult, it may be necessary to deflate the
cuff on the endotracheal tube for a few moments
to provide a little more room for the esophago-
scope.
Although spontaneous ventilation can be per-
mitted once the esophagoscope is through the cri-
copharyngeal sphincter, it is preferable to retain
control of ventilation and maintain paralysis to
prevent any coughing or bucking during the pro-
cedure, which may cause the esophagoscope to
damage or tear the esophagus. Certainly, full re-
laxation should be maintained throughout the pro-
cedure if repeated instrumentation through the cri-
copharyngeal muscle is required. Arrhythmias may
occur as the esophagoscope passes behind the
heart, but they are rarely a source of actual concern
and are usually only transitory.
Recovery of consciousness should be supervised
with the patient on the side, slightly head-down on
a gurney that can be tipped. A postoperative chest
X-ray should be obtained to rule out signs of
esophageal tear, such as subcutaneous emphysema,
pneumoperitoneum, pneumothorax, or pneumome-
diastinum. Emergency thoracotomy for primary
repair or diagnosis may be required if esophageal
rupture is suspected (see Esophageal Rupture Per-
foration below).
C. COMPLICATIONS
The complications of esophagoscopy include
perforation, hemorrhage, and cardiopulmonary
complications.
33 34
The most frequent site of per-
foration is the hypopharynx, which carries a mor-
tality rate ranging from 34 to 84 per cent.
33
"
37
Per-
foration of the esophagus is most likely to occur
in patients with esophageal carcinoma (e.g., 12 per
cent).
38
Biopsy specimens that are taken too deeply
Table 16-7 FIBEROPTIC VERSUS OPEN-TUBE ESOPHAGOSCOPY
Advantages Disadvantages
Patient comfort
Ability to examine entire upper
gastrointestinal tract
General anesthesia not required
Greater safety
Fiberoptic Esophagoscopy
1. Expensive equipment required
2. Greater occurrence of mechanical failure
3. Dilation not possible
4. Smaller biopsies
1. Deeper biopsy possible
2. Greater aspiration capacity
3. Dilation feasible/indicated
4. Foreign body removal
5. Ease of sterilization
Open-Tube Esophagoscopy
1. Patient discomfort
2. Safety
3. Less readily done on outpatient basis
4. Greater risk of perforation
(i.e., full thickness) can also result in perforations.
Hemorrhage may occur from esophageal varices,
lacerations at the gastroesophageal junction in-
duced by forceful retching, and biopsy. Cardiopul-
monary problems are due to aspiration, pneumo-
nia, cardiac arrhythmias, and myocardial
infarction. Septicemia may occur in immunosup-
pressed patients, and prophylactic antibiotics are
indicated in this group.
F. Benign Esophageal Diseases (Often
Requiring Surgical Repair)
The surgical management of complex benign
esophageal disease is challenging and has no easy
solution. Many patients with this condition have
undergone a previous operation. Table 16-8 shows
the wide variety in 63 operations initially per-
formed in 35 patients, and Table 16-9 shows the
wide variety of subsequent reconstructive opera-
tions performed in the same 35 patients in one
institution.
39
1. Hiatal Hernia
Two types of hiatus hernia have been described.
Type I hernias, also called sliding hernias, make
up approximately 90 per cent of esophageal hiatal
hernias. In this type, the esophagogastric junction
and fundus of the stomach have herniated axially
through the esophageal hiatus into the thorax. The
type II, or paraesophageal hiatus hernia, is charac-
terized by portions of the stomach herniating into
the thorax next to the esophagus. In the presence
of a type II hernia, the esophagogastric junction is
still located in the abdomen. The goal of surgical
Table 16-8 COMPLEX BENIGN ESOPHAGEAL
DISEASE: PREVIOUS
OPERATION*
Fundoplication
Nissen (17)
Collis-Nissen (7)
Collis-Belsey (2)
Belsey (2)
Thal-Nissen (2)
Hill (1)
Esophagomyotomy
Diaphragmatic hernia repair
Esophageal exploration and/or repair
Esophageal exclusion
Cardiectomy
Diverticulectomy
Miscellaneous
Total
No.
31
11
5
4
4
3
2
3
63
*From Ellis FH, Gibb SP: Esophageal reconstruction for
complex benign esophageal disease. J Thorac Cardiovasc Surg
Table 16-9 COMPLEX BENIGN ESOPHAGEAL
DISEASE: SUBSEQUENT
"DEFINITIVE" OPERATION*
No.
Operation
Esophagectomy 27
Esophageal exclusion 4
Cardioplasty 2
No resection _2
Total 35
Reconstruction
Acid suppression and alkaline diversion 21
Colon interposition 8
Substernal (5)
Intrathoracic (3)
Esophagogastrostomy 6
Inkwell (3)
End to side (2)
Transhiatal (1)
Total 35
*From Ellis FH, Gibb SP: Esophageal reconstruction for
complex benign esophageal disease. J Thorac Cardiovasc Surg
repair of a sliding hernia is to obtain gastroesoph-
ageal competence. Because restoration of the nor-
mal anatomy is not always successful in prevent-
ing subsequent reflux, several antireflux operations
have been developed, the so-called wraparound
procedures. For example, Nissen fundoplication,
which can be performed via an abdominal or tho-
racic incision, entails wrapping the distal esopha-
gus with the fundus of the stomach.
Barrett's esophagus is the eponym commonly
used to describe the presence of columnar epithe-
lial lining within the distal tubular esophagus. The
change in the lining of the esophagus may be due
to chronic gastric reflux (as may be caused by a
hiatal hernia). Unfortunately, there is a moderately
strong association between the columnar-lined
esophagus and esophageal adenocarcinoma.
2. Benign Esophagorespiratory Fistula
Benign esophagorespiratory fistula is a rare con-
dition and an unusual cause of chronic pulmonary
suppuration in adults. Although the ratio of benign
to malignant esophagorespiratory fistulas is ap-
proximately 1:5, benign esophagorespiratory fistu-
las are of greater surgical interest because of their
potential for total curability.
Benign esophagorespiratory fistulas may be
either congenital or acquired; the latter is more
common. Acquired benign esophagorespiratory
fistulas are most commonly inflammatory or post-
traumatic in origin, and an increasing number of
the latter seem to have been recorded as a compli-
cation of blunt trauma to the chest in automobile
accidents and secondary to cuffed endotracheal or
tracheostomy tubes in patients requiring prolonged
intubation and mechanical ventilation.
The nonspecific nature of presenting symptoms
makes the diagnosis sometimes difficult; the main
classic finding of sudden coughing seconds after
ingestion of fluids or solids (Ono's sign) should
arouse suspicion of the diagnosis. The diagnosis
(esophagotracheal or esophagobronchial commu-
nication) is established by means of contrast ra-
diography. Esophagoscopy and bronchoscopy are
not as accurate as radiography.
40
3. Benign Esophageal Stricture
Chronic reflux of acidic gastric contents leads to
ulceration, inflammation, and, eventually, esopha-
geal stricture. Reflux is the most common cause of
benign stricture formation in the lower esophagus.
The pathologic changes are reversible if the acidic
gastric contents cease contact with the esophageal
mucosa. Surgery may be necessary if medical
treatment (diet, H
2
-blockers, antacids; see Table
16-5) and dilatations are inadequate.
There are two types of surgical repair, both of
which are usually approached via a left thoracoab-
dominal incision. Gastroplasty after esophageal
dilatation interposes the fundus of the stomach be-
tween esophageal mucosa and the acidic milieu of
the stomach. The remaining fundus may be sewn
to the lower esophagus to create a valve-like ef-
fect. The second type of repair is resection of the
stricture and the creation of a thoracic end-to-side
esophagogastrostomy. Vagotomy and antrectomy
are performed to eliminate stomach acidity, and a
Roux-en-Y gastric drainage procedure is per-
formed to prevent alkaline intestinal reflux.
4. Achalasia
Achalasia is a rare motor disorder characterized
by almost complete disruption of the primary peri-
stalsis function of the esophagus. The neurologic
defect that is responsible for the development of
achalasia remains poorly understood. Consistent
pathologic findings include degeneration of gan-
glion cells in the myenteric plexus, occasionally
accompanied by chronic inflammatory cell infil-
trates, and loss of nerves innervating the smooth-
muscle cells of the lower esophageal sphincter.
41
Esophageal manometry remains the best means for
diagnosing achalasia. Initial therapy can include
either pneumatic dilation or esophagomyotomy.
Symptomatic improvement occurs in 71 per cent
of patients after pneumatic dilation, with a risk of
perforation of 1.4 per cent.
41
Surgical procedures for achalasia can be per-
formed through either an abdominal or a thoracic
incision. Nearly all authors favoring an abdominal
approach add an antireflux operation to esophago-
myotomy, whereas many authors advocating a
transthoracic esophagomyotomy believe that an
antireflux wrap is unnecessary. Overall results for
the various surgical approaches used as initial ther-
apy are excellent, with symptomatic improvement
in 89 per cent of patients, a mortality rate of less
than 1 per cent, and development of gastroesoph-
ageal acid reflux in less than 10 per cent.
41
5. Esophageal Rupture and Perforation
A rupture is a bursting injury and may be a
spontaneous event as from uncoordinated vomit-
ing, straining associated with weightlifting, child-
birth, and defecation, or it may be externally
caused as in crush injuries to the chest and abdo-
men. These ruptures are due to a sudden increase
in abdominal pressure with a relaxed lower esoph-
ageal sphincter and an obstructed esophageal inlet,
and the tear is usually located within 2 cm of the
gastroesophageal junction. In contrast to a perfo-
ration, in the presence of a rupture, the stomach
contents enter the mediastinum under high pres-
sure and the patient becomes symptomatic from
mediastinitis much more abruptly (i.e., septic syn-
drome).
Perforations are usually unaccompanied by high
intraluminal pressure and are usually due to iatro-
genic instrumentation (the most common cause)
(the origin of an iatrogenic tear may be either
intraluminal or extraluminal; see Table 16-10),
trauma, or ingestion of a perforating substance/
material. Causes of iatrogenic esophageal perfora-
tion include endoesophageal intubation with any
catheter/instrument (e.g., endotracheal tubes, la-
ryngoscopes, esophageal obturator airways, esoph-
ageal balloon tamponade, and nasogastric tubes).
Table 16-10 lists the causes of esophageal rup-
ture and perforation. The relative frequency of
these causes of esophageal rupture and perforation
from reported series is depicted in Figure 16-3,
42
and the relationship of cause to location is shown
in Figure 16-4.
42
In 60 per cent of patients, there
is esophageal disease (benign stricture [most com-
mon cause], diverticulum, tumor, achalasia) under-
lying the perforation.
42
a. CLINICAL/DIAGNOSTIC FEATURES
There is a sharp rise in mortality rates when
treatment is delayed. The mortality rate for pa-
tients treated in fewer than 24 hours is near 10 per
Table 16-10 CAUSES OF ESOPHAGEAL
RUPTURE AND PERFORATION*
I. Instrumentation/iatrogenic
A. Intraluminal
1. Esophagoscopy
2. Bougienage
3. Pneumatic dilation
4. Sclerosis of esophageal varices
5. Placement of intraesophageal tubes (nasogastric,
Sengstaken-Blakemore, prostheses)
6. Endotracheal tube placement
B. Extraluminal
1. Mediastinoscopy
2. Intraoperative injury
a. Thyroid resection
b. Leiomyoma enucleation
c. Proximal gastric vagotomy
d. Pneumonectomy
3. Radiation therapy
II. Traumatic
A. Blunt
B. Penetrating
C. Ingestion of caustic substance
III. Spontaneous (barogenic)
A. Postemetic
B. Straining (weightlifting, bowel movement, childbirth)
IV. Ingestion of foreign body
V. Tumor
VI. Surrounding infections
*Based on data from Jones and Ginsberg.
42
cent, but it rises sharply to greater than 50 per cent
for those treated later.
43
"*
5
The major reason for
this sharp increase in mortality rate is the rapid
development of necrotizing mediastinitis, com-
bined with the inability to close the perforation
surgically. Because of the propensity to develop
mediastinitis, thoracic perforations have three
times the mortality of cervical perforation,
45
and
spontaneous ruptures are much more lethal than
perforations because of the explosive contamina-
tion of the mediastinum as well as the delay in
diagnosis, because they most often occur unob-
served outside the hospital and may be clinically
silent initially.
Cervical tears may be due to either instrumen-
tation or trauma and occur at the level of the upper
esophageal sphincter (the cricopharyngeal mus-
cle), which is the narrowest point in the esophagus
(and therefore this is the area where foreign bodies
most frequently impact in children). In addition,
the upper esophagus is compressed by the cervical
vertebrae (exacerbated by hyperextension of the
neck) and at the level of the C-5 and C-6 verte-
brae, the posterior esophageal mucosa is covered
only by fascia.
The early signs of cervical esophageal perfora-
tion include neck stiffness and a dull neck ache,
regurgitation of blood material, and the finding of
cervical subcutaneous emphysema. Inflammatory
changes in the neck may not develop for several
hours; signs of systemic sepsis often do not occur
for up to 24 hours.
In contrast, perforations of the thoracic esopha-
gus result directly in mediastinal contamination,
leading to a more rapid development of pneumo-
mediastinum and mediastinitis than after cervical
perforations. The thin mediastinal pleura is usually
ruptured by the inflammatory process, producing
contamination of the pleural space and a pleural
effusion. Gastric contents and fluids are then
drawn into the pleural space by the negative intra-
thoracic pressure, resulting in further inflammation
and fluid sequestration, hypovolemia, and the early
appearance of tachycardia and systemic sepsis.
Chest pain and subcutaneous emphysema are usu-
ally present, and dyspnea is often prominent even
in the absence of pneumothorax.
Intra-abdominal esophageal perforations occur
into the free peritoneal cavity and result in perito-
nitis. A dull, retrosternal ache in association with
epigastric pain radiating to the shoulders is char-
acteristic because of the relationship of the intra-
abdominal esophagus to the diaphragm. Systemic
signs such as tachycardia, tachypnea, and fever
develop early; progression to sepsis and shock oc-
curs within hours.
Although contrast esophagograms are the stan-
Figure 16- 3 04) Cause of 511 cases of esoph-
ageal perforation in recent series and (B) types of
instrumental perforations among those cases.
Other causes of iatrogenic esophageal perforation
include endoesophageal intubation with any cath-
eter/instrument such as endotracheal tubes, laryn-
goscopes, esophageal obturator airways, esopha-
geal balloon tamponade, and nasogastric tubes.
Data are taken from References 49, 50, 51, 72,
112 and 113 of Jones (my Reference 42).
dard diagnostic procedure in cases of suspected
esophageal perforation, it should be noted that the
false-negative rate of these examinations can ex-
ceed 10 per cent. Thus, "negative" studies may
not completely eliminate the possibility of a per-
foration. When positive, contrast studies have the
advantage of demonstrating the level of the perfo-
ration and the presence of extension into the
pleural cavity. In the patient in whom esophageal
perforation is highly suspected clinically but con-
trast esophagograms are negative, flexible esopha-
goscopy and computed tomography can be useful
diagnostic adjuvants.
b. TREATMENT OF ESOPHAGEAL
PERFORATIONS
Once the diagnosis of esophageal perforation is
established, multiple factors must be considered in
selecting the appropriate therapy. The cause of the
perforation, its location, the presence of underlying
esophageal disease, and the interval between per-
foration and diagnosis are critical factors in the
determination of treatment. In addition, the condi-
tion of the esophagus and the extent of soilage or
injury to adjacent organs and tissues as well as the
age and general condition of the patient must also
Figure 16-4 Relationship of
cause to location of esophageal per-
forations in recent series.
be considered. Treatment options for esophageal
perforations are shown in Table 16-11.
The surgical approach is dependent on the lo-
cation of the perforation. Cervical esophageal tears
are best exposed through an incision parallel to the
left sternocleidomastoid muscle and anterior to the
carotid artery and internal jugular vein. Perfora-
tions in the upper two thirds of the thoracic esoph-
agus can be reached through a right posterolateral
thoracotomy in the fourth or fifth intercostal space,
whereas those in the lower third are best ap-
proached through a posterolateral thoracotomy in
the left sixth or seventh interspace. Abdominal
esophageal injuries usually require an upper mid-
line laparotomy. Once the esophagus is exposed,
localization of subtle perforations may require in-
stillation of methylene blue into the esophageal
lumen or the insufflation of air into the esophagus
immersed in saline solution. Figure 16-5 shows
the variety of diversion, drainage, and feeding
Table 16-11 VARIOUS POSSIBLE
TREATMENTS OF ESOPHAGEAL
PERFORATION
1. Operative management
a. Primary closure
b. Primary closure reinforced with tissue (pleural flap,
omentum, intercostal flap, pericardial fat, diaphragm,
gastric wall)
c. Resection (transhiatal or transthoracic)
d. Drainage alone
e. T-tube drainage (creating esophagocutaneous "venting"
fistula)
f. Exclusion and diversion (surgical closure of proximal
and distal esophagus with cervical esophagostomy and
gastrostomy drainage tubes)
g. Intraluminal stent (for carcinoma cases)
2. Nonoperative management
tubes a patient with an esophageal tear may re-
quire.
The following guidelines have been suggested
for selecting nonoperative treatment: (1) clinically
stable patients; (2) instrumental perforations de-
tected before major mediastinal contamination has
occurred or perforations with such a long delay in
diagnosis that the patient has already demonstrated
tolerance for the perforation without the need for
surgery; and (3) esophageal disruption that is well
contained within the mediastinum or a pleural lo-
culus.
Nonoperative management is much more feasi-
ble in children because mediastinal tissues seem
more resistant and esophageal perforations appear
more contained in children than adults.
46
Nonoperative management consists of nothing
by mouth, total parenteral nutrition, broad-spec-
trum antibiotics, and some of the various tubes
shown in Figure 16-5. Final outcome is related to
cause (e.g., spontaneous ruptures have a worse
prognosis compared with iatrogenic instrumental
causes), underlying disease (e.g., underlying car-
cinoma has a much worse prognosis than benign
disease), location (e.g., thoracic and abdominal
tears have a much worse prognosis than a cervical
tear), and type and timing of treatment (late treat-
ment has a much worse prognosis than early treat-
ment).
G. Surgical Approaches/Incisions for
Esophageal Surgery (With Emphasis on
Esophageal Carcinoma)
47
The distribution of esophageal carcinoma be-
tween the upper, middle, and lower thirds of the
esophagus is approximately 17 to 27 per cent, 40
Figure 165 Diagram of esopha-
gogastric-intestinal temporary diver-
sion, drainage, and feeding tubes.
to 54 per cent, and 20 to 30 per cent, respec-
tively.
27
48
~
50
Resectional surgery of the midthora-
cic esophagus is technically very complicated
(modern immediate mortality is high but with a
wide reported range of approximately 3-20 per
cent).
25
28
'
45
51-53
The main causes of mortality are
anastomotic leaks and cardiorespiratory complica-
tions.
Surgery for midthoracic esophageal lesions usu-
ally involves an additional proximal cervical inci-
sion (to approach the upper third of the esophagus)
and a distal abdominal incision (to approach the
lower third of the esophagus and/or mobilize the
stomach). The midthoracic esophagus can be ap-
proached through either left-sided or right-sided
thoracotomy. Tumor resection is performed
through the thoracotomy. The stomach is then
pulled up into the chest and anastomosed to the
proximal esophagus. For esophageal tumors in-
volving the stomach, an esophagogastrectomy is
performed, and the jejunum is anastomosed to the
proximal esophagus.
Advantages of a left thoracotomy include the
ability to split open the left diaphragm, mobilize
the stomach, and bring it up into the chest without
having to change the patient's position. A right
thoracotomy has the advantage of good exposure
because the midthoracic portion of the esophagus,
except for its most distal part, is primarily on the
right side of the mediastinum. Also, if the azygos
vein is infiltrated by tumor, it is safer to dissect it
from the right side.
The disadvantages of a right thoracotomy are
that the position of the patient must be changed
after the abdominal part of the operation (per-
formed first) is completed. In addition, after an
extensive procedure in the abdomen and an open-
ing of the chest, the lesion may be found to be
nonresectable. Finally, esophagectomy can be per-
formed without thoracotomy through an abdomi-
nal transhiatal approach (requires only abdominal
and neck incisions and the esophagus is freed by
blunt finger dissection). The purported advantages
of this approach are that it (1) is associated with
decreased cardiopulmonary mobidity and (2)
places the anastomosis in the neck.
The surgical literature is very controversial with
respect to the appropriate surgical approach to
esophageal cancer. The issues involved in the con-
troversy are whether the goal is cure versus pallia-
tion (i.e., whether the resection is complete and
includes lymph nodes; the transthoracic approach
can potentially be curative, whereas the transhiatal
approach is only palliative), ease of taking care of
postoperative anastomotic leaks (transthoracic ap-
proaches usually result in an intrathoracic anasto-
mosis; an intrathoracic anastomotic leak is much
more difficult to take care of than a cervical anas-
tomotic leak and has a 40 per cent mortality),
54
the
length of surgery and the amount of intraoperative
bleeding (transhiatal pull-through is blind and may
be associated with mediastinal bleeding), and post-
operative esophageal motility.
Esophageal surgery involves a number of other
controversies, including questions of delayed ver-
sus immediate reconstruction and the use of colon
or stomach to restore continuity.
49
Without quoting
a large number of references, it is my impression
that the left thoracoabdominal incision is most
commonly used today.
Unfortunately, it is not the operation but rather
the stage and biological behavior of the tumor at
the time of esophagectomy that determine sur-
vival. Only about 10 per cent of patients have
stage I disease (tumor that is superficial and con-
fined to the mucosa and submucosa and therefore
potentially curable by virtually any extirpative
technique).
26
Perhaps in the 10 to 25 per cent of
the patients with stage II disease, a traditional
"cancer operation"transthoracic esophagec-
tomy with regional lymph node dissectionoffers
a chance for better survival than does the transhia-
tal approach, but this has by no means been estab-
lished. The finding of mediastinal lymph node me-
tastases from esophageal carcinoma carries an
ominous prognosis. The 5-year survival for stage
III tumors without lymph node metastases (T3,
NO, MO) was 34 per cent compared with 8 per cent
for patients with lymph node involvement (T3,
Nl-3, MO).
53
Unfortunately, the majority of pa-
tients with esophageal carcinoma have mediastinal
lymph node metastases, and that is the primary
determinant of their survival.
Overall, the 5-year survival rate is 10 to 15 per
cent (Fig. 16-6)
26
but has ranged as high as 54 per
cent.
26
55
56
The preponderance of evidence indi-
cates that combined modality therapy (surgery,
chemotherapy, and radiation) slightly and moder-
ately increases survival (from 5-10 per cent to -
5 per cent) compared with surgery alone.
57-59
Ob-
viously, there is agreement that we need to find a
way to recognize esophageal cancer early. Encour-
aging reports from China, where carcinoma of the
esophagus is endemic (there is a 50-fold greater
incidence of esophageal carcinoma in China com-
pared with the United States), have described mass
screening techniques and early esophageal resec-
tions.
60
The resectability rate was 100 per cent,
hospital mortality was 2.5 per cent, and the 5-year
survival rate was 85.9 per cent.
As with obstruction of the tracheobronchial tree
by lung carcinoma, obstruction of the esophagus
by esophageal carcinoma may be palliated (unob-
structed) by endoscopic laser
61
and photodynamic
therapy.
62
With these therapies, malignant stenoses
have been recanalized or widened in all patients,
and the ability to swallow was improved in about
80 per cent with no procedure-related deaths.
61
62
Other forms of palliation of a totally obstructed
esophagus, but with significant procedure-related
mortality, are by-pass operations (with a hospital
mortality of 20 per cent) and esophageal dilation
and intubation (stenting with a catheter; 30-day
mortality is 10 per cent, with much of it resulting
from perforation
38
).
63
M
With all forms of pallia-
tion (laser therapy, bypass, intubation) median sur-
vival is approximately 4-7 months.
62
II. INTRAOPERATIVE ANESTHETIC
CONSIDERATIONS
A. Monitoring
In addition to all the considerations discussed in
chapter 7, an arterial catheter is indicated if (1)
one-lung ventilation is to be used, (2) blood loss
will be significant, (3) the heart will be manipu-
lated (by retraction or passage of gastrointestinal
tract behind the heart), (4) an esophagorespiratory
fistula exists, and/or (5) chronic lung disease is
present. Central venous access is required for mon-
itoring the effects of manipulating the heart (see
later discussion), preload measurement, blood loss
replacement, and drug administration.
B. Induction of Anesthesia
Because patients presenting for esophageal sur-
gery may be at risk for aspiration, either an awake
100
PATIENTS
56 CONSIDERED
POTENTIALLY
RESECTABLE
44
UNRESECTABLE
49 SURVIVE
THE OPERATION
7 DEATHS FROM
PERIOPERATIVE
COMPLICATIONS
20-30 ALIVE
AT
1 YEAR
10-20
ALIVE AT
5 YEARS
Figure 16-6 The fate of 100 patients with esophageal carcinoma. Statistics are based on treatment outcome from I20l articles
published between 1980 and 1988 and include the use of adjuvant chemotherapy and radiation. (From Muller JM, Erasmi H,
Stelzner M, et al: Surgical therapy of oesophageal carcinoma. Br J Surg 77:845-857, 1990. Used with permission.)
intubation or a rapid-sequence induction with cri-
coid pressure may be indicated. However, rarely,
neck, trachea, or an esophageal foreign body may
prevent the application of cricoid pressure (neces-
sitating an awake intubation).
65
Additionally, a pa-
tient with mediastinal lymphadenopathy may have
tracheal compression and collapse of the airway
with the onset of muscle relaxation. Ventilation
may be possible only by passage of an endotra-
cheal tube beyond the obstruction (see chapter 15,
Mediastinal Mass).
C. Airway Considerations
1. General Airway Considerations
There are several important airway considera-
tions that are generally related to the performance
of esophageal surgery. First, for lower esophago-
gastric resections/procedures via a left thoracoab-
dominal incision or an abdominal incision, it is not
necessary to collapse the left lung using a double-
lumen endobronchial tube. A single-lumen endo-
tracheal tube can be placed, and surgical exposure
can be obtained by gentle retraction of the left
lung.
For esophageal surgery via a standard thoracot-
omy alone, it is usually necessary to place a dou-
ble-lumen endobronchial tube or bronchial blocker
tube to collapse the ipsilateral lung. However, a
patient with relatively normal lungs undergoing
esophageal surgery may be more likely to experi-
ence hypoxemia during one-lung ventilation than
a patient presenting for lung resection. This is be-
cause the patient presenting for lung surgery may
already have limitation of blood flow to the dis-
eased lung and, thus, less ventilation/perfusion
mismatching during one-lung anesthesia. Also,
during lung resection, the surgeon ligates the pul-
monary artery or a branch thereof, which decreases
the shunt.
Second, as large a tracheal tube (either double-
lumen tube, single-lumen tube, Univent bronchial
blocker tube) as possible should be used if the
patient has chronic lung disease in order to effec-
tively suction secretions. Chronic lung disease in
these patients is due to chronic aspiration, and the
secretions are frequently copious. In addition, pa-
tients with chronic aspiration have a greater risk
of intraoperative bronchospasm secondary to in-
creased airway reactivity caused by chronically ir-
ritated airways and may require bronchodilation
^
2
-agonist, adequate anesthesia).
Third, in thoracotomy cases, one-lung ventila-
tion (as created by either a double-lumen tube or
bronchial blocker) will facilitate exposure of the
esophagus. However, when the thoracotomy is fol-
lowed by a cervical anastomosis, it should be re-
membered that a very large tracheal tube or over-
inflated cuff may cause the posterior membrane of
the trachea to bulge into the anterior surface of the
esophagus and interfere with the surgical field.
Fourth, high concentrations of nitrous oxide are
contraindicated with bowel present in the chest
because the resultant bowel distention may cause
respiratory impairment and possible interference
with surgical exposure.
Fifth, patients who have chronic lung disease
and copious secretions should be mechanically
ventilated and extubated only when they demon-
strate an adequate cough. If the patient is to remain
intubated at the end of the procedure in which a
double-lumen endobronchial tube has been used,
reintubation with a single-lumen endobronchial
tube must be done very carefully (see Jet Stylet
Method, chapter 9). If the neck is going to be fixed
in a hyperflexed position (either by mandibular-
manubrium wires or chin to sternal skin sutures)
to avoid/minimize tension on an esophageal
anastomosis,
66
67
then the single-lumen tube should
be in place before fixation of the neck in a hyper-
flexed position (placement of the single-lumen
tube may require neck extension).
Sixth, in nonthoracotomy cases (without a chest
drain), a chest X-ray should always be obtained
immediately postoperatively to exclude a pneu-
mothorax from inadvertent entry into either pleural
cavity. Finally, after extubation, aspiration may
occur because of recurrent laryngeal nerve injury
from cervical dissection.
2. Airway Considerations in Cases of
Esophagorespiratory Fistula
Anesthetizing the patient with an esophagores-
piratory tract fistula carries with it unique consid-
erations, distinguishing it from other types of
esophageal surgery. The level of the fistula should
be identified preoperatively (esophography or en-
doscopy) to determine the most appropriate tra-
cheal tube (double-lumen tube, single-lumen tube,
bronchial blocker) (see later discussion). If the fis-
tula is very large, positive-pressure ventilation
may result in massive abdominal distention (re-
sulting in secondary respiratory insufficiency and
cardiovascular compromise), and loss of inspired
gas (with resultant inadequate ventilation) through
a gastric tube (if present). Thus, spontaneous ven-
tilation should be maintained, if possible.
Awake intubation is, therefore, an appropriate
method of securing the airway, if possible, in the
presence of a very large fistula. If the fistula is in
the mid to proximal trachea, a single-lumen tube
can be used with the tip placed distal to the fistula
so that the cuff of the endotracheal tube may oc-
elude the fistula. If the fistula is in the distal tra-
chea or main-stem bronchus, a double-lumen tube
will be necessary to isolate the fistula. A right-
sided double-lumen tube should be used for a fis-
tula in the distal trachea, left main-stem bronchus,
or left lung, and a left-sided double-lumen tube
should be used for a fistula in the distal trachea,
right main-stem bronchus, or right lung because,
in either case, the endobronchial cuff guarantees
isolation of the fistula (Fig. 16-7).
If the site of the fistula cannot be determined
preoperatively, a right-sided double-lumen tube
should be placed because, statistically, the fistula
is most likely to communicate with the trachea or
left main-stem bronchus. Right-sided ventilation
should be attempted first. Either gastric distention
or loss of delivered tidal volume indicates pres-
ence of a right-sided fistula, and left-lung ventila-
tion should be used instead.
If the patient must be generally anesthetized be-
fore intubation, spontaneous ventilation should be
maintained until gentle (low positive pressure)
ventilation by mask has been shown to provide
effective ventilation. If the stomach distends with
positive-pressure ventilation, then it may need to
be vented (nasogastric tube, gastrostomy) (Fig.
16-8). If venting the stomach results in loss of
tidal volume, then appropriate one-lung ventilation
must be instituted (see prior discussion). The
amount of tidal volume loss is a function of size
of the fistula, compliance of the lungs, peak inspi-
ratory pressure, and whether there is a stomach
tube or not.
The lung on the side of the fistula is likely to
have a decreased compliance; consequently, the
contralateral lung tends to receive most of the tidal
volume. When one-lung ventilation is used and
chronic lung disease in the ventilated lung pre-
vents adequate arterial oxygenation, operative lung
continuous positive airway pressure (5-10 cm
H
2
0) is indicated. When two-lung ventilation is
used after the esophageal portion that contained
the fistula has been excluded (i.e., the creation of
a blind esophageal pouch), the excluded portion
must be drained to protect against proximal or
distal pouch suture line disruption resulting from
distention of the pouch by air delivered through
the fistula during positive-pressure ventilation
(Fig. 16-9). If the pouch is not drained, it should
not be exposed to positive-pressure ventilation
(i.e., use one-lung ventilation, spontaneous venti-
lation, block fistula, and so on).
68,69
In view of these considerations, spontaneous
ventilation at the end of the procedure is desirable.
Isolation of Esophago-Respiratory Tract Fistula
By Double-Lumen Tube Intubation
Fistula in Distal Trachea,
Left Mainstem Bronchus or Lung
Fistula in Distal Trachea,
Right Mainstem Bronchus or Lung
Figure 16-7 Either a right- or left-sided double-lumen tube can be used to isolate a distal tracheal fistula. A left-lung fistula is
isolated by a right-sided double-lumen tube, and a right-lung fistula is isolated by a left-sided double-lumen tube.
Anesthesia for Esophageal Surgery 6 0 7
Figure 16-8 Algorithm for the
management of the airway during the
induction of anesthesia in a patient
with an esophagorespiratory tract fis-
tula.
Surgical Treatment of
Esophago-RespiratoryTract Fistula
Figure 16-9 Exclusion of malig-
nant esophagorespiratory fistula pro-
cedure.
In this circumstance, meticulous suctioning of se-
cretions, avoidance of sedation, and provision of
appropriate analgesia are not only desirable but
necessary. However, the timing of extubation
should be conservative if chronic lung disease and
secretions are present.
D. Hemodynamic Considerations
Two major hemodynamic considerations are
unique to major esophageal surgery. First, trans-
thoracic or transhiatal tumor resection may lead to
surgical compression and/or manipulation of vital
organs (vena cava, heart), as may passing of distal
gastrointestinal tract (stomach, small intestine, co-
lon) retrosternally into the chest. The compression/
manipulation of the central cardiovascular struc-
tures may cause arrhythmias, decreased cardiac
output, hypotension, and perhaps myocardial is-
chemia (in patients with pre-existing coronary ar-
tery disease).
70
7I
Each time the surgeon's hand
enters the thorax, it may be expected that the mean
arterial pressure and cardiac index will decrease
by an average of 46 per cent (Figs. 16-10 and -
1 ), and premature ventricular contractions occur
in 50 per cent of patients.
71
Pulmonary artery occluded pressure signifi-
cantly increases with each intrathoracic manipula-
tion (see Figs. 16-10 and 16-11). All hemody-
namic parameters return to control values once
manipulation is stopped (see Figs. 16-10 and lo-
ll). The simultaneous decrease in mean arterial
pressure and cardiac index are clearly related to
impaired venous return and/or ventricular ejection
from manual compression.
From these findings, it can be assumed that
long-lasting alterations in hemodynamic status
could be detrimental to patients with cardiac dis-
ease. Therefore, careful intraoperative hemody-
namic monitoring is advisable. The obvious re-
sponse to the mechanical/surgical interference of
cardiac function is to terminate the offending stim-
ulus (see Figs. 16-10 and 16-11), but in situations
in which an arrhythmia is prolonged, antiarrhyth-
mia drugs may need to be administered. If the
electrocardiogram indicates ischemia, nitroglyc-
erin and calcium channel-blocking drugs (for cor-
onary artery vasospasm) may need to be adminis-
tered.
Second, the average blood loss associated with
esophagectomy is approximately 2000 ml.
72
The
negative hemodynamic consequences of manipu-
lation of the heart described previously will be
greatly magnified if the patient is concomitantly
hypovolemic. The preload monitoring suggested
HEMODYNAMIC CHANGES OBSERVED DURING MANUAL DISSECTION
Figure 16-10 A typical recording of the hemodynamic events observed during intrathoracic esophageal resection. " I n" repre-
sents the moment the surgeon's hand entered the posterior mediastinum. "Out " represents the moment the surgeon's hand was
removed. Total time elapsed between the two points in and out was 55 sec in this case. (A.P. = aortic pressure; P.C.W.P. =
pulmonary capillary wedge pressure; ECG = electrocardiogram.) (From Yakoubian K, Bougeois B, Marty J, et al: Cardiovascular
responses to manual dissection associated with transhiatal esophageal resection. J Cardiothorac Anesth 4:458461, 1990. Used with
permission.)
ing air to define the esophageal anatomy better.
Air may also be injected through a retrograde gas-
troesophageal catheter. Similarly, to test the integ-
rity of the anastomosis or repair, methylene blue
may need to be injected via a naso- or oroesopha-
geal tube, the lower end of which has been care-
fully positioned by the surgeon in relation to the
anastomosis.
Finally, it may be necessary to pass bougies per
os to dilate strictures and/or define anatomy. Dila-
tions may progressively increase in size to the 50
to 60 French range. The final bougie may be used
as a stent around which to perform a fundoplica-
tion. Whenever a catheter or bougie is passed into
the esophagus and resistance at the esophageal in-
let is felt, which may be due to collapse of the
esophagus by the tracheal tube and/or its cuff, it
may be necessary to pull the trachea anteriorly by
grabbing it at the right and left lateral margins (if
the neck is not sterile) or by transiently deflating
the cuff of the tracheal tube.
III. POSTOPERATIVE
CONSIDERATIONS
Figure 16-11 Changes in mean arterial pressure (MAP),
cardiac index (CI), and pulmonary capillary wedge pressure
(PCWP) during intrathoracic manipulations, -p < .05 versus
before. (From Yakoubian K, Bougeois B, Marty J, et al: Car-
diovascular responses to manual dissection associated with
transhiatal esophageal resection. J Cardiothorac Anesth 4:458-
The major routine postoperative problems con-
sist of obtaining satisfactory gas exchange and,
related to gas exchange, postoperative analgesia.
Postoperative mechanical ventilation is discussed
in chapter 20 and postoperative analgesia in chap-
ter 21. The major nonroutine complication of
esophageal surgery is leakage from an esophageal
to gastrointestinal tract anastomosis. Leakage from
a cervical anastomosis can usually be easily han-
dled by incision and drainage. Leakage from an
intrathoracic anastomosis is the most feared com-
plication of esophageal surgery and carries a 50
per cent mortality rate. The treatment options are
listed in Table 16-11, and Figure 16-5 shows the
variety of diversion, drainage, and feeding tubes
that are often required to deal with an intrathoracic
anastomotic leak.
REFERENCES
previously and serial hemoglobin and hematocrit
determinations allow fairly precise replacement of
blood and fluid losses.
E. Esophagogastrointestinal
Considerations
The anesthesiologist may be asked to insert a
nasogastric tube (the exact distance is usually de-
termined by the surgeon) for the purpose of inject-
1. Hardy JD: Diseases of the esophagus: An overview. In
Hardy JD (ed): Textbook of Surgery. Philadelphia, JB Lip-
pincott Co, chapter 37, 1977.
2. Roa TLK, El-Etr AA: Esophageal and mediastinal surgery.
In Kaplan JA (ed): Thoracic Anesthesia. New York.
Churchill Livingstone, 1983, chapter 14, pp -.
3. Ruberg RL, Dudrick SJ: Intravenous hyperalimentation in
head and neck tumor surgery: Indications and precautions.
Br J Plast Surg 30:151 -153, 1977.
4. Weiss SM: Nutritional aspects of preoperative manage-
ment. Med Clin North Am 71:369-375, 1987.
5. Reinhardt GF, DeOrio AJ, Kaminski MV Jr: Total paren-
teral nutrition. Surg Clin North Am 57:1283-1301, 1977.
6. Askanazi J, Carpentier YA, Elwyn DH, et al: Influence of
CHAPTER 17
Anesthesia for Emergency
Thoracic Surgery
I. Introduction
II. Massive Hemoptysis
B. Surgical Considerations
C. Anesthetic Considerations
a. Prevent Asphyxiation
b. Prevent Contamination of
Normal Lung
c. Correct Hypovolemia
2. Intraoperative Considerations
III. Thoracic Aortic Aneurysms and
Dissections/Disruptions
1. Spontaneous Dissection of
Thoracic Aortic Aneurysms
2. Traumatic Disruption
3. Signs and Symptoms of Any
Type (Traumatic or
Spontaneous) of Dissection
IV. Bronchopleural Fistula
V. Lung Abscesses and Empyema
VI. Chest Trauma
A. Overview of the Management of
the Patient With Extensive Trauma
1. Primary Survey
a. Airway and Ventilation
b. Circulation
c. Neurologic Status
d. Mechanism of Injury
2. Secondary Survey
3. Surgical Priorities
B. Specific Chest (Noncardiac)
Injuries: General Considerations
1. Chest Wall Fractures (Flail
Chest)
2. Pleural Space (Hemothorax,
Pneumothorax) and Pulmonary
Parenchyma (Contusions)
a. Hemothorax
b. Pneumothorax
c. Pulmonary Contusion
3. Tracheobronchial Disruptions
4. Esophageal Disruptions
5. Diaphragmatic Disruptions
C. Specific (Noncardiac) Injuries:
Surgical Considerations
Chest)
a. Hemothorax
b. Pneumothorax
D. Specific Chest (Noncardiac)
Injuries: Anesthetic Considerations
Chest)
a. Hemothorax
b. Pneumothorax
VII. Transvenous Pulmonary
Embolectomy
Considerations
VIII. Emergency Room Thoracotomy in
the Management of Trauma
IX. Removal of Tracheobronchial Tree
Foreign Bodies
612
I. I NTRODUCTI ON Table 17-1 CAUSES OF MASSIVE
HEMOPTYSIS*!
The vast majority of cases requiring emergency
thoracic surgery consist of massive hemoptysis,
thoracic aortic aneurysms and dissections/disrup-
tions, bronchopleural fistula (BPF), lung abscess
and empyema, chest trauma (chest wall fractures,
pulmonary parenchymal contusions, tracheobron-
chial disruption, and esophageal and diaphrag-
matic trauma), and tracheobronchial tree foreign
bodies. Since chest trauma can cause massive he-
moptysis and thoracic aortic disruptions and neces-
sitate emergency room thoracotomy, it is obvious
that this classification is an arbitrary one. How-
ever, since each of these indications for emergency
thoracic surgery commonly occurs independently
of chest trauma, they warrant separate considera-
tion.
II. MASSIVE HEMOPTYSIS
Massive hemoptysis is uncommon, occurring in
less than 0.5 per cent of patients admitted to a
large pulmonary medicine service.
1
It has been
arbitrarily defined, on the basis of the amount of
daily volume of blood expectorated, as 200 ml,
2
more than 300 ml,
3
more than 500 ml,
4
more than
600 ml in 24 to 48 hours,
1
5
"
7
more than 600 ml
within 16 hours,
5
and more than 1000 ml in 24
hours.
8
9
However, others evaluated hemoptysis not in
terms of rate of bleeding but from the standpoint
of its threat to vital functions. Thus, a life-threat-
ening hemoptysis was one that caused acute air-
way obstruction or hypotension severe enough to
require blood transfusion (although, in general, the
greater the rate of bleeding, the greater the mortal-
ity).
1
I0
In all of these reports, massive hemopty-
sis often occurred abruptly, unexpectedly, and
without prodromal symptoms. The differential di-
agnosis of massive hemoptysis is shown in Table
17-1."
I2
On occasion, aspirated and/or swallowed
epistaxis and hematemesis can be mistaken for
hemoptysis.
The large majority of reported cases of massive
hemoptysis have had a chronic infectious cause."
I 3
1 4
This is because chronic inflammation leads to
profuse vascularization of the high-pressure bron-
chial artery system. Subsequently, any erosion or
rupture of enlarged bronchial arteries will result in
massive hemoptysis. Bronchitis is the most com-
mon, active tuberculosis the second most common,
and bronchiectasis the third most common infec-
tion causing massive hemoptysis.
12
~
14
Of patients
I. Infection (45-90%)
Tuberculosis
Bronchiectasis
Bronchitis
Lung abscess
Necrotizing pneumonia
II. Neoplasm (7-19%)
Bronchogenic carcinoma
Metastatic carcinoma
Endobronchial polyp
III. Cardiovascular disease
Mitral stenosis
Pulmonary arteriovenous malformation
Pulmonary embolus
Pulmonary vasculitis
IV. Miscellaneous causes
Pulmonary artery catheterization
Exploratory needling
Cystic fibrosis (5-10%)
Pulmonary contusion, laceration, aneurysm
Reperfusion of pulmonary vasculature after
pulmonary embolectomy and after cardiopulmo-
nary by-pass
*Based on data from Garzon & Gourin.
1
Yeoh et al.,
2
Stern
et al.,
3
Ehrenhaft & Taber,
4
Crocco et al.,
s
Gourin & Garzon,
h
Conlan and Hurwitz,
7
Corey & Hla,* Muthuswamy et al.."
Thorns et al.,
10
Wedel," Bone,
12
Johnston & Reisz,
13
Wedzicha
& Pearson,
14
Panos et al.,
15
Cervenko et al.,'
6
and Rice et al."
tRange of percentage of incidence is given for the major
causes.
with pulmonary tuberculosis, 1 to 5 per cent will
have massive hemoptysis.
9
The majority of the remaining causes of hemop-
tysis are due to bleeding neoplasms. Although the
major cause of spontaneous bleeding with neo-
plasms is the same as with the infectious causes
(erosion into bronchial arteries), in present-day
clinical practice massive hemoptysis from neo-
plasms increasingly occurs during diagnostic fiber-
optic bronchoscopic manipulations of exophytic
airway tumors. Hemoptysis from neoplasm has a
worse prognosis than hemoptysis from infection/
The prognosis is especially poor in carcinoma pa-
tients if there is concomitant fungal infection (36
per cent motility).
15
A mechanism for hemoptysis in both cancer and
fungal infections, in addition to direct invasion of
blood vessels, is distal ischemia and avascular ne-
crosis.
1,5
Massive hemoptysis may also occur when
the low-pressure pulmonary circulation is tran-
sected or ruptured, as may occur with accidental
trauma with sharp or blunt objects and pulmonary
artery catheter trauma. The likelihood of massive
endobronchial hemorrhage following pulmonary
artery catheterization is increased by subsequent
heparinization.
16 I7
Thus, massive hemoptysis may
arise from either the pulmonary or the systemic
circulation.
614 Anesthesia for Emergency Thoracic Surgery
Death from massive hemoptysis usually results
from asphyxiation, rarely from exsanguination.
However, the amount of blood that has actually
been lost may be seriously underestimated based
on the history since at least some of the coughed-
up blood is swallowed. In addition, the accurate
measurement of expectorated blood is frequently
difficult.
In a series of 55 pulmonary resections per-
formed for massive hemoptysis (600 ml/16
hours),
5
there was a mortality of 18 per cent re-
ported; this was markedly better than with conser-
vative treatment, which resulted in a mortality of
75 per cent in patients who bled 600 ml or more
in 16 hours and of 54 per cent in those who bled
600 ml or more in 48 hours.
5,6
However, routine
use of surgery has been debated because other
authors have found that somewhat lesser degrees
of hemoptysis may be successfully managed con-
servatively regardless of the amount of bleeding in
the first 24 hours.
2
l8
l9
The intensive care unit
medical protocol used in one institution is shown
in Table 17-2.
9
Nevertheless, surgery (resection)
is probably indicated in patients who require mul-
tiple transfusions, in those in whom bleeding re-
sults in progressive impairment of pulmonary
function (aspiration should be evaluated by fre-
Table 17-2 INTENSIVE CARE UNIT MEDICAL
PROTOCOL FOR MASSIVE
HEMOPTYSIS*
1. Complete bed rest
2. Postural management
3. Nothing per mouth
4. Establishment of large-bore intravenous catheters
5. Correction of abnormalities of coagulation
6. Oxygen therapy to keep arterial oxygen pressure at 75
mm Hg range
7. Accurate record of volume and rate of hemoptysis
8. Antituberculosis chemotherapyt
9. Broad-spectrum antibioticst
10. Judicious transfusion and fluid therapy
11. Cough suppression (codeine and similar drugs)
12. Stool softeners (to avoid straining)
13. Coordination with thoracic surgeons and angioradiologists
14. Bronchoscopy (see Fig. 17-1)
15. Radionuclide scanning for localization selectively
16. Vasopressin intravenously
*Based on data from Muthuswamy et al.
9
tThis consists of four tuberculocidal drugs, including isoni-
azid, rifampin, streptomycin, and pyrazinamide. This combi-
nation is used because of the high incidence of primary drug
resistance in their population of patients.
$ Third-generation cephalosporins are not used to avoid com-
plicating coagulopathies.
This may be used in those patients who are at high risk for
invasive procedures or who are being prepared for surgery.
quent serial chest roentgenograms), and in those in
whom hemoptysis persists for several days despite
optimum medical management.
19
Contraindica-
tions to surgery include inoperable carcinoma of
the lung, inability to localize the bleeding site, and
presence of severe bilateral pulmonary disease and
systemic disease (debilitation). These patients are
candidates for bronchial artery embolization (Fig.
17-1) (see later).
The chest X-ray may contain strong clues (evi-
dence of tuberculosis or opacifications) as to the
source of bleeding. However, considering that
blood can be, and probably has been, aspirated
into the nonbleeding lung, it cannot be safely as-
sumed that the chest X-ray pathologic findings
correspond to the site of bleeding.
20
Indeed, in
many patients with massive hemoptysis, the chest
X-ray may even be normal.
21
Bronchoscopy during active bleeding is the sin-
gle most important technique for determining the
cause and location of bleeding and should be per-
formed in all patients. The procedure should be
done, if possible, in the operating room so that
immediate resection can be performed (see Fig.
17-1).
22
Most bronchoscopists will use a rigid
bronchoscope because of the much greater suc-
tioning and ventilating capability. However, the
flexible fiberoptic bronchoscope may be used if
there is no active bleeding and/or the site of bleed-
ing is thought to be in the upper lobes. The com-
bination of fiberoptic bronchoscopy through a
rigid bronchoscope may be a safer alternative (see
chapter 15). If localization is still uncertain after
bronchoscopy and if the clinical situation permits,
bronchial and pulmonary arteriography can be
helpful. When bronchoscopy is coupled with se-
lective angiography, a high rate of accurate locali-
zation can be achieved. When bronchoscopy has
caused the bleeding (as in diagnostic procedures),
the site of bleeding is known and the patient can
proceed to definitive therapy (see later discussion)
without delay.
The endoscopist may be able to control bleeding
and the spread of bleeding during bronchoscopy
(see Fig. 17-1). Topical iced saline and vasocon-
strictors can be administered through the broncho-
scope to control bleeding, provided the bleeding is
not so massive as to preclude visualization of the
origin.
7
Bronchodilator treatment should not be ad-
ministered because these may have vasodilator ac-
tions and precipitate renewed bleeding. The spread
of blood from one lung to the other can be con-
trolled by the use of a bronchial blocker (Univent
tube or balloon-tipped Fogarty catheter) in the
main bronchus of the bleeding side or a gauze
packing at the bleeding segment or side.
22-26
Yt-
trium aluminum garnet (YAG) laser has proved
efficacious in the treatment of hemoptysis in pa-
Anesthesia for Emergency Thoracic Surgery- 615
Figure 17-1 Treatment algorithm for massive hemoptysis. The most important function of emergency bronchoscopy is to
establish or to diagnose the cause of bleeding. However, the amount and spread of bleeding can also be controlled during emergency
bronchoscopy (see items under Diagnosis of Cause and Treatment). (PEEP = positive end-expiratory pressure: CPAP = continuous
positive airway pressure.)
tients with lung cancer.
27
28
In a very exciting de-
velopment in pulmonary medicine, two reports de-
scribed complete cessation of bleeding in almost
all patients by selective intrabronchial spraying of
fibrin precursors through a catheter in the suction
port of a fiberoptic bronchoscope without any ad-
verse effect.
29
30
This selective intrabronchial local
coagulative method seems to be the most logical,
physiologic, and effecti ve one of all the conserva-
tive medical therapies. During bronchoscopy, the
endoscopist should frequently restore adequate ox-
ygenation and ventilation by intubating the unin-
volved main-stem bronchus.
In some of the patients with a contraindication
to resectional surgery, control of hemorrhage can
be achieved by selective bronchial artery emboli-
zation using resorbable material.
31
"
34
The major
risk of bronchial artery embolization is spinal cord
injury due to spinal cord embolization via arterial
collaterals to the spinal cord. The presence of
spinal cord collaterals on scout arteriograms is an
absolute contraindication to bronchial, emboliza-
tion. The important causes of initial failure of
bronchial artery embolization are extensive and
bilateral disease, technical failure to catheterize or
achieve a secure catheter position in the bronchial
artery for safe bronchial artery embolization, and
pulmonary artery origin of the bleeding. In fact,
several authors strongly advocated both pulmonary
and bronchial arteriography (and embolization,
when appropriate), especially in patients with tu-
berculosis and other suppurative pulmonary dis-
ease who experience massive hemoptysis.
9
35
Recurrence of hemoptysis after an initially suc-
cessful bronchial artery embolization is not un-
common. The most important causes of recurrence
of bleeding are incomplete embolization, progres-
sion of the native disease, and recanalization of the
embolized vessels. Probably all three factors oper-
ate in each patient to a varying degree. Finally, it
must be realized that bronchial artery embolization
is only a palliative procedure and does not cure
the underlying disease that caused hemoptysis in
the first place; therefore, surgery is often indicated
once the patient has been stabilized by emboliza-
tion (see Fig. 17-1).
9
In one patient with hemorrhage from the left
lung, the combined occlusion of left pulmonary
and bronchial arteries was used to control repeated
hemorrhage.
36
In this patient, a Swan-Ganz cathe-
ter was guided by fluoroscopy into the left main
pulmonary artery while bronchial arteries were se-
lectively embolized. When recurrent hemorrhage
occurred, the balloon in the pulmonary artery was
inflated with 4 ml of saline. The decrease in pul-
monary artery blood flow to the bleeding area was
believed to be instrumental in stopping the second
episode of hemorrhage.
Figure 17-1 summarizes most of the medi-
cal/surgical considerations involving massive he-
moptysis. The site of bleeding must be known, and
in the vast majority of patients, this will require
bronchoscopy. The rigid bronchoscope is preferred
(see chapter 15). Unstable patients must be stabi-
lized. If they ultimately require intubation while
still bleeding, a double-lumen tube is preferred
(see Anesthetic Considerations later). During bron-
choscopy, the surgeon may help to control bleed-
ing by performing regional ice lavage; applying
topical vasoconstrictors, fibrin seal, YAG laser;
and instituting tamponade of the bleeding region.
If flexible fiberoptic bronchoscopy has to be per-
formed at the bedside and major life-threatening
obstructing blood clots have to be removed, a Fo-
garty balloon-tip embolectomy catheter can be
passed via the suction/biopsy channel to a position
beyond the clot.
After balloon inflation, partial withdrawal of the
catheter and the bronchoscope may dislodge the
clot, allowing its subsequent suctioning and re-
moval.
37
Of course, the great danger of this proce-
dure is restarting the pulmonary hemorrhage that
ceased because of the airway clot. If the patient
can withstand surgery and has an operable lesion,
surgery should be performed. If the patient cannot
withstand surgery and/or has an inoperable lesion,
bronchial embolization should be tried.
An approach to the clinical management of mas-
sive hemoptysis in patients with cystic fibrosis has
been recommended. It is similar to the prior algo-
rithm but differs in one key respect: Embolization
is the first choice, and surgery is a second choice
of treatment.
38
The difference is because patients
with cystic fibrosis have generalized lung disease,
and preservation of tissue is a much more crucial
issue. The site of bleeding is first identified by
bronchoscopy, ideally under general anesthesia.
Then selective bronchial arteriography is per-
formed instead of surgery. If there are no collater-
als to the spinal cord, then bronchial embolization
is performed. If collaterals to the spinal cord are
visualized, arterial embolization is abandoned and
pulmonary resection is undertaken within the lim-
its dictated by the patient's overall pulmonary
function.
Hemoptysis has a 37 per cent fatality rate; there-
fore, it behooves all users of pulmonary artery
catheters to know how to manage pulmonary ar-
tery rupture caused by a pulmonary artery cathe-
ter.
39
If the hemoptysis is minor, it usually ceases
after the pulmonary artery catheter is moved to a
more proximal position. If the hemoptysis is still
present, gentle inflation of the balloon with 2 ml
of air may stop blood flow in the pulmonary artery
branch and therefore tamponade the hemorrhage
from a more distal perforation. General anticoagu-
lation should be reversed if present.
Preservation of oxygen transport may require
emergency transfusion, oxygen therapy, or endo-
tracheal intubation. A double-lumen endotracheal
tube or bronchial blocker may facilitate ventilation
and prevent blood aspiration in the unaffected lung
(occasionally in dire circumstances, an endobron-
chial intubation with a single-lumen tube may be
life-saving [see chapter 9]). Artificial ventilation
with positive end-expiratory pressure (PEEP) has
been advocated to decrease the pressure gradient
between damaged vessel and the surrounding lung
parenchyma. With a double-lumen tube, a high
level of selective bleeding lung continuous posi-
tive airway pressure (CPAP) may be used. If all
these previous measures fail, emergency thoracic
surgery would be the last therapeutic possibility.
The most important preoperative priorities are
to prevent asphyxiation, localize site of bleeding,
prevent contamination of normal lung if possible,
and correct hypovolemia. It must be realized that
many of these priorities should be fulfilled simul-
taneously. For example, at the same time a double-
lumen tube is being inserted to prevent asphyxia-
tion by drowning in blood, large-bore intravenous
cannulas must be inserted to begin to rapidly cor-
rect hypovolemia. Antibiotics should be adminis-
tered preoperatively, and antituberculous drugs
should be started in patients with tuberculosis.
a. PREVENT ASPHYXIATION
An increased F,0
2
should be immediately ad-
ministered as continuously as possible (oxygen ad-
ministration may be interrupted by episodes of he-
moptysis). Oxygenation and ventilation should be
monitored, when possible, with arterial blood-gas
determinations. If the site of bleeding is known,
the bleeding lung should be placed in a dependent
position to prevent soiling of the nonbleeding
lung. Patients who cannot cough out the blood
effectively enough to prevent contamination of the
nonbleeding lung must have the lungs separated as
soon as possible. In the large majority of these
patients, this can be done most expeditiously and
effectively with a double-lumen tube; however, in
the unusual case in which the bleeding is intermit-
tent and a fiberoptic bronchoscope that was being
used for diagnostic purposes is positioned near the
site of bleeding, placement of a bronchial blocker
is a viable alternative (see chapter 9). Rarely, in-
sertion of a single-lumen tube down a main-stem
bronchus must suffice as a life-saving measure
when the other technically more difficult proce-
dures cannot be accomplished (see chapter 9).
Ventilatory support (intermittent positive-pressure
breathing and PEEP) must be provided as needed.
Suctioning of the tracheobronchial tree must be
aggressive.
b. PREVENT CONTAMINATION OF NORMAL
LUNG
Coughing may increase bleeding. The advisabil-
ity of using sedatives and cough suppressants is
time dependent.
19
In the unintubated patient, the
ability to cough may be life-saving, and suppres-
sants should be avoided. The patient should be at
strict bed rest in the semi-Fowler's position or with
the radiologically normal lung in a nondependent
position. In the intubated patient, suctioning can
replace the cough mechanism, and suppression of
cough may decrease bleeding. If possible, the ini-
tial intubation should be with a double-lumen tube.
A coagulation profile should be drawn early; if
any abnormalities are noted, they should be cor-
rected. During and after bronchoscopy, the surgeon
internist can control bleeding by iced-saline la-
vage; placement of topical vasoconstrictors, fibrin
seal, YAG laser; placement of a bronchial blocker
or gauze packing; and use of bronchial artery
embolization (see prior discussion).
c. CORRECT HYPOVOLEMIA
As soon as possible, large-bore intravenous can-
nulas should be inserted. The patient's blood
should be typed and screened and cross matched
for adequate amounts of blood products (whole
blood, packed red blood cells, platelets, fresh
frozen plasma). Transfusion should begin if ap-
propriate. Finally, the appropriate monitoring
(e.g., arterial line, central venous line) should be
instituted.
The patient with massive hemoptysis can come
to the operating room without an endotracheal tube
in place, with a single-lumen endotracheal tube in
place, or with a double-lumen endotracheal tube in
place (Fig. 17-2). In deciding what type of airway
the patient needs, it is important to remember sev-
eral general principles. First, the important advan-
tages of a double-lumen tube include separation of
the two lungs in order to prevent the spread of
blood from one lung to the other lung; improved
surgical exposure; the ability to ventilate, provide
CPAP, and sigh the operative lung when desired;
and the ability to safely inspect an open bronchus
during resection. Second, the potential disadvan-
tages of a double-lumen tube include the increased
technical difficulty in intubating a bloody trachea
in a hypoxic patient, and, if a large quantity of
blood has already spread to the contralateral lung,
it may be impossible to adequately ventilate and
oxygenate the patient using only the now-soiled
contralateral lung. Third, if a single-lumen tube is
already placed or is inserted, the lungs can still be
separated by a subsequent endobronchial intuba-
tion (right main-stem bronchus blindly and left
main-stem bronchus with the aid of a fiberoptic
bronchoscope) or by insertion of a bronchial
blocker along the side of the single-lumen tube
into the appropriate main-stem bronchus (see
chapter 9). Fourth, the Univent bronchial blocker
tube may be a good alternative in cases in which
rapid intubation and single-lung blockage are re-
quired.
If the patient with massive hemoptysis is with-
out an indwelling endotracheal tube, preoxygena-
tion should be instituted immediately. Adequate
suctioning must be available. It may be necessary
for the patient to be awake for intubation during
massive, active, spontaneous bleeding in order to
avoid the hazard of trying to visualize a blood-
obscured airway in a paralyzed patient. Intubation
performed in the semiupright position may mini-
mize coughing that results in the presence of blood
in the upper airway and, thereby, may provide a
clearer field of vision.
Figure 17-2 Endotracheal intubation options in massive hemoptysis and the sequence of conversion of one option to another.
Double-lumen tubes are the intubation option of choice in the majority of patients.
If the patient without an indwelling endotra-
cheal tube is to be put to sleep, aspiration precau-
tions (e.g., cricoid pressure) should be utilized.
Because these patients are likely to be hypovo-
lemic, anesthesia should be induced with a small
dose of short-acting barbiturate or ketamine or
with narcotics followed in rapid sequence by re-
laxation. If the larynx can be visualized, insertion
of a double-lumen tube is preferable to insertion
of a single-lumen tube (see General Considerations
presented earlier and Fig. 17-2). It should be re-
membered that if the patient is not actively bleed-
ing but has a blood-filled cavity (i.e., a hemor-
rhagic lobe) this cavity will likely empty its
contents into the dependent, unsoiled lung when
the patient is turned to the lateral decubitus posi-
tion; therefore, this situation is a strong indication
for placement of a double-lumen tube. If a single-
lumen tube is inserted, it might as well be the
Univent bronchial blocker tube, which additionally
allows lung separation. If the Univent bronchial
blocker tube is not available, then serious consid-
eration should be given to converting a standard
single-lumen tube to an endobronchial position or
using a bronchial blocker either inside the lumen
or alongside it before turning the patient to the
lateral decubitus position.
If a single-lumen tube is already in place, the
same conversion considerations to bronchial
blocker and endobronchial position apply (Fig.
17-2). In addition, consideration should be given
to converting the single-lumen tube to a double-
lumen tube (Fig. 17-2). If a double-lumen tube is
in place, it should remain and be utilized (provided
serious bilateral lung contamination has not oc-
curred). In all situations, if active tuberculosis is
present or suspected, contamination precautions
should be utilized.
Once the airway has been established, the pa-
tient must be placed in the lateral decubitus posi-
tion with the bleeding lung in the nondependent
position. Use of this position, of course, empha-
sizes the importance of separating the lungs. Blood
products must be administered according to the
continued loss of blood, updated coagulation pro-
files, and hemodynamic monitoring findings. In-
sertion of an arterial line greatly facilitates moni-
toring of cardiovascular status and allows
repetitive sampling for arterial blood-gas analysis
during the operation. The use of blood warmers is
strongly recommended.
At the end of the case, the endotracheal tube
should be left in place and the patient should be
ventilated mechanically. Most of these patients
will have impaired gas exchange postoperatively
owing to pre-existing lung disease, the probability
that the nonbleeding lung has been soiled by the
recent hemoptysis from the diseased lung, and the
physiologic consequence of having just undergone
a major anesthetic and surgical experience. Coag-
ulation profiles, electrolyte concentrations, and
acid-base status must be monitored in the imme-
diate postoperative period, and abnormalities must
be treated quickly.
III. THORACIC AORTIC
ANEURYSMS AND
DISSECTIONS/DISRUPTIONS
A true thoracic aortic aneurysm is a dilatation
of all three layers of the aortic wall. Dissection of
the thoracic aorta is a propagation of hematoma
between the intima and the adventitia (i.e., in the
media). Aortic dissection is caused by the sudden
development of a tear in the aortic intima, opening
the way for blood to enter the media of the aortic
wall, thereby separating intima from the adventitia
for variable distances along the length of the aorta.
The most important forces acting to propagate the
dissecting hematoma are the steepness of the as-
cending pressure pulse wave (dp/dt) and the aortic
blood pressure. Spontaneous dissections occur in
aneurysms, and traumatic dissections occur with-
out being preceded or accompanied by an aneu-
rysm.
7. Spontaneous Dissection of Thoracic
Aortic Aneurysms
Spontaneously occurring aneurysms are usually
repaired electively, and spontaneous acute dissec-
tion in a spontaneously occurring aneurysm usu-
ally requires emergency surgical repair.
40
Sponta-
neous dissection is approximately four times as
common as a spontaneous rupture.
41
Spontaneous
rupture is usually a fatal event.
There are two classifications of thoracic aortic
dissections, and they are based on anatomic loca-
tion. The DeBakey classification
42
consists of
types I, II, and III (Fig. 17-3). Type I dissection
begins in the ascending aorta and extends for vary-
ing distances past the aortic arch and below the
diaphragm. Type II dissection also starts in the
ascending aorta but ends proximal to the left sub-
clavian artery. Type III aortic dissection generally
starts just distal to the left subclavian artery, ex-
tends for varying distances (type 3A remains
above the diaphragm and type 3B extends below
the diaphragm), and may even include the iliac
arteries. Type I aortic dissection is the most com-
mon, occurring in approximately 70 per cent of all
cases of spontaneous thoracic aortic dissection.
The Stanford classification
43
consists of types A
and (Fig. 17-3). Type A encompasses DeBakey
types I and II, and type consists of DeBakey
type III. The Stanford classification emphasizes
the all-important point that Stanford type A and
DeBakey types I and II require immediate surgery.
Numerous conditions predispose to aneurysmal
dilatation and spontaneous dissection of the tho-
racic aorta. All of these conditions cause degener-
Anesthesia for Emergency Thoracic Surgery 6 1 9
ation of the aortic wall. Hypertension, found in 90
per cent of patients with acute aortic dissection,
44
mechanically weakens the thoracic aortic wall be-
cause of chronic wall stress and shear forces. Sim-
ilarly, the mechanical effect of a jet stream due to
aortic valve stenosis and the proximal hyperten-
sion due to a coarctation of the aorta are associated
with aneurysm of the thoracic aortic arch. Turner's
syndrome, which is associated with coarctation of
the aorta, is also associated with thoracic aortic
aneurysms. Inflammation of the aorta will weaken
the wall, and, not surprisingly, giant cell aortitis
and syphilitic aortitis are associated with thoracic
aortic aneurysms. Marfan's and Ehlers-Danlos
syndromes are associated with thoracic aortic
aneurysms owing to hereditary defects in connec-
tive tissue strength. The hormonal changes of
pregnancy can cause a loss of thoracic aortic wall
structural integrity. Finally, although spontaneous
internal tears are seldom seen through atheroma-
tous lesions, thoracic aortic dissections can occur
around atheromatous plaques following retrograde
catheterization of the central arteries. Patients with
thoracic aortic aneurysms have a high incidence of
other arterial diseases, such as cerebrovascular and
other vessel occlusive disease, abdominal aortic
aneurysms, hypertension, and coronary artery dis-
ease.
2. Traumatic Disruption
Thoracic aortic injury caused by blunt objects
results from differential rates of deceleration of the
fixed aortic arch relative to the more mobile car-
diac chambers and descending thoracic aorta. Con-
sequently, traumatic disruption of the thoracic
aorta usually occurs in the region of the ligamen-
tum arteriosum just distal to the left subclavian
artery. A less common site is just above (cephalad
to) the aortic valve. The tear may involve the in-
tima, intima and media, or the entire aortic wall.
Traumatic dissection and rupture of the thoracic
aorta have been estimated to be associated with 10
to 16 per cent of all automobile accident
fatalities.
43
45
When a traumatic thoracic aortic tear
occurs, 80 to 90 per cent of the patients die at the
scene of the accident.
43
45
46
Among the 10 to 20
per cent of patients who survive an initial trau-
matic thoracic aortic tear, the leak is temporarily
controlled by the adventitia of the aorta. Even
though the adventitia is strong, it is not capable of
resisting the same bursting pressure as was the
intact aorta, and increases in blood pressure above
normal limits must be avoided. Most of these pa-
tients will have associated injuries, and a priority
system must be established; often head and ab-
dominal injuries take precedence. In the group of
patients arriving at the hospital alive, as many as
80 per cent may survive postoperatively.
43 47
3. Signs and Symptoms of Any Type
(Traumatic or Spontaneous) of
Dissection
The most common presenting symptom (94 per
cent of patients)
44
is the sudden onset of severe,
unremitting, tearing, or ripping chest or back pain.
Blood pressure is increased in two thirds of pa-
tients (especially if the dissection involves the
renal arteries), and the patients appear cool,
clammy, and vasoconstricted, but hypotension re-
sulting from rupture of the aorta may be present.
Pseudohypotension may be caused by compromise
of the flow through either or both subclavian arte-
ries, and a difference in blood pressure between
the arms is not uncommon.
In dissections involving the ascending aorta,
aortic valvular insufficiency may be present in two
thirds of the patients (which may cause acute
congestive heart failure) as a result of stretching of
the aortic annulus by the dissecting hematoma.
Other presenting signs and symptoms of thoracic
aortic dissection/disruption result from the ana-
tomic course that the propagating hematoma
takes.
48
As the medial hematoma dissects or as the
intimai flap peels off, partial or complete obstruc-
tion of the lumen of an artery arising from the
aorta can occur, and the signs and symptoms re-
flect the distribution of the obstructed artery. In-
volvement of the coronary arteries causes acute
myocardial infarction and frequently sudden death.
Involvement of the innominate and/or common ca-
rotid artery may cause syncope, confusion, stroke,
or coma. Involvement of the innominate and/or
subclavian artery can cause upper limb gangrene
and paralysis and disparity in the pulses in the
extremities. Involvement of the intercostal and/or
lumbar arteries can cause spinal cord ischemia and
paraplegia. Involvement of the renal, mesenteric,
and common iliac arteries can cause oliguria and
renal vascular hypertension, bowel ischemia (nau-
sea and vomiting, abdominal pain, hematemesis,
melena), lower leg gangrene and paralysis, and
disparity in the pulses in the extremities, respec-
tively.
Through-and-through rupture of aorta (the most
common cause of death) can cause extravasation
of blood into any adjacent region; consequently,
blood may be found in the pericardial sac (most
common), pleural space (second most common),
mediastinum (which can cause superior vena cava
and tracheal compression),
49
retroperitoneum, wall
of the pulmonary trunk and/or main right and left
pulmonary arteries (because of common adventitia
with aorta), lung parenchyma, and esophagus.
Rupture might even occur back into the aortic lu-
men, leading to a "spontaneous cure."
44
50
Since there are tremendous differences in mor-
tality with surgical versus medical treatment of
Figure 173 The classification of thoracic aortic dissections is based on anatomic location. The DeBakey classification consists
of types I, II, and III, and the Stanford classification consists of types A and B. DeBakey types I and II and Stanford type A involve
the ascending aorta and aortic arch, and DeBakey type III and Stanford type originate distal to the left subclavian artery.
proximal aortic lesions
48
5I
52
the indications for
surgery are based on the location of the dissec-
tion/disruption. Stanford type A and complicated
Stanford type lesions require immediate surgery.
Complications of type dissection consist of any
of the following: failure to control hypertension,
continued pain, expanding aortic diameter, devel-
opment of neurologic deficit, evidence of compro-
mise of a major subdiaphragmatic aortic branch
vessel, and development of aortic valvular insuffi-
ciency.
On the basis of widespread experience, a con-
sensus has emerged concerning major principles
and management of suspected acute aortic dissec-
tion.
53
"
58
It is most important to decrease blood
pressure and velocity of ventricular contraction.
During the first few hours after admission, the
patient's blood pressure should be stabilized with
sodium nitroprusside (to systolic blood pressure of
100 to 110 mm Hg). Because sodium nitroprusside
causes reflex tachycardia and an increase in dp/dt,
beta blockade (propanolol, esmolol, labetalol) is
also required to decrease contractility and heart
rate (to 60-70 beats/min). Nifedipine may be use-
ful in patients with asthma. An upright chest radio-
graph with a nasogastric tube in place should be
obtained. The chest radiograph criteria for pro-
ceeding to aortic angiography include the presence
of any of the following: mediastinal widening, the
loss of normal contour of the aortic knob, opacifi-
cation of the aortopulmonary window, depression
of the left main bronchus, deviation of the trachea
or nasogastric tube to the right, and the apical cap
sign.
58
Ratios of mediastinal width to chest width
at the aortic knob greater than 0.28 have been
found to be accurate predictors of thoracic aortic
rupture.
59
If the patient is stable, the patient should be
transferred for aortography.
47
Aortic angiography
has a diagnostic accuracy of 95 to 99 per cent
50
; it
is best performed by retrograde aortic catheteriza-
tion. While the patient is in the aortography suite,
use of pain medications or sedatives should be
restricted. The persistence of pain despite an ade-
quate decrease in arterial blood pressure is a sign
of continuing hematoma formation. Similarly,
depression of mental status is an ominous clinical
finding, which may indicate involvement of the
major extracranial cerebral vessels in the dissec-
tion of the hematoma. Both these valuable and
important signs (pain and depressed mental status)
are masked by overmedication with narcotics or
sedatives. Small doses of a mild hypnotic agent
such as diazepam should provide adequate relief
of patient anxiety. Oxygen should be administered
by face mask. If the patient is hemodynamically
unstable, the patient should be immediately trans-
ferred to the operating room.
On the basis of the aortography results (which
can show the intimai tear, the false channel, com-
pression of true lumen, patency of major vessels
arising from the aorta), the lesion can be classified
as Stanford type A or B. This information, coupled
with the patient's in-hospital course and response
to emergency therapy, dictates whether immediate
surgery (type A or complicated type B; e.g., leak-
ing or ruptured vessel, compromise of a branch so
that an organ is jeopardized
60
) or further stabiliza-
tion (e.g., repair intracranial and/or bleeding ab-
dominal lesions or torn airway) with later surgery
(some type B) or no surgery (uncomplicated type
B) will be required. The operative mortality of
type A dissections is approximately 15 to 20 per
cent.
50
In-hospital survival of stable type treated
either medically or surgically is 80 per cent.
50
The
overall 10-year survival is 40 per cent.
50
Unfortunately, there is approximately a 2 per
cent incidence of false-negative aortography re-
sults in aortic dissection.
61
Transesophageal echo-
cardiography has a diagnostic sensitivity and spec-
ificity near 100 per cent
62
and therefore has been
recommended as the primary bedside test that can
identify patients who need surgery without delay
for any other test. Indeed, there are cases in which
transesophageal echocardiography has detected an
asymptomatic descending aortic intimai tear while
aortography was negative.
63
The diagnostic accu-
racy of computed tomographic scanning and mag-
netic resonance imaging is 90 per cent or
better.
64
65
These noninvasive techniques should be
used when aortography is negative, for serial
measurements, and perhaps as a screen when sus-
picion of aortic disruption is low or diagnosis is
needed immediately at the bedside (i.e., with trans-
esophageal echocardiography).
The surgical approach (Fig. 17-4) is dictated by
several considerations. First, and foremost, is the
site of the lesion. Proximal repairs (Stanford type
A) require cardiopulmonary by-pass and a median
sternotomy (approaches A and of Fig. 17-4).
Distal repairs (Stanford type B) utilize a left tho-
racotomy and usually some form of partial by-pass
(approaches C and D of Fig. 17-4). With partial
by-pass the organs proximal to the clamp area are
perfused by the patient's heart and oxygenated by
the patient's lungs, whereas the organs distal to the
clamped area are perfused and oxygenated by a
pump-oxygenator system.
The second consideration is control of proximal
hypertension during partial by-pass. Approach C
decreases proximal hypertension by inserting a
proximal-distal shunt (Gott, TDMAC-heparin
shunt). The heparin-bonded shunt (systemic hepa-
rinization is not used) is inserted proximally in the
ascending aorta, or apex of the left ventricle, or, if
neither is available, the left subclavian artery and
A. Ascending Aorta B. Aortic Arch
Figure 17- 4 See legend on opposite page
distally into the aorta or femoral artery. Approach
D decreases proximal hypertension by a left ven-
tricular, or left atrial, or femoral vein-femoral ar-
tery partial by-pass. The left ventricular, or left
atrial, or femoral vein drain decreases venous re-
turn to the left side of the heart, thereby limiting
left-sided heart preload, cardiac output, and hyper-
tension proximal to the descending thoracic aortic
clamp. It is logical that the patients who would
benefit the most by partial cardiopulmonary by-
pass, with respect to preservation of cardiac func-
tion, are those with depressed left ventricular func-
tion preoperatively (e.g., ejection fraction less than
20 per cent).
66
Associated injuries may contraindi-
cate the use of heparinization required for ap-
proaches C and D.
The third consideration is minimizing neuro-
logic (spinal cord) damage. Approaches C and D
provide some form of distal perfusion. Although
there has been one report of no paraplegia in 168
patients shunted with a Gott shunt (cross-clamp
time was 37 min),
67
in general, there has been no
difference in the incidence of paraplegia between
these approaches (C and D) and simple aortic cross
clamping (approach of Fig. 17-4). The inci-
dence of paralysis has been as high as 24 per cent
in acute traumatic disruption, as high as 20 to 40
per cent with extensive thoracoabdominal disease,
but as low 0 to 2 per cent in patients with elective
repair of chronic aneurysms.
68-71
To reduce the incidence of neurologic damage
further, other adjunctive methods are being evalu-
ated clinically or experimentally. These efforts in-
clude monitoring of evoked potentials,
72
monitor-
ing cerebrospinal fluid pressure and attempting to
reduce it if elevated during aortic cross clamping,
73
use of intrathecal vasodilators,
74
monitoring distal
arterial pressures, use of intravenous steroids,
75
use
of local hypothermia,
76
and pharmacologic
suppression of spinal cord activity (pentothal, ste-
roids, magnesium sulfate, mannitol).
To date, none of these methods (shunting or
otherwise) have been shown to be a means of
avoiding paraplegia.
70
This is because the most
important common denominators for spinal cord
injury are prolonged cross-clamp time (e.g.,
greater than 30 min), prolonged systemic hypoten-
sion, removal of long segments of aorta, and re-
moval of aortic segments from which a collateral
circulation to the spinal cord has not had time to
develop. The presence of these denominators has
little to do with the type of by-pass.
68
69
77
-
78
Of
these factors, aortic cross-clamp time is probably
the most important.
79
In addition, distal perfusion
of the aorta only provides adequate lumbar spinal
cord protection, but the arrangement of the vascu-
lar anatomy still leaves the lower thoracic spinal
cord at risk from ischemia (vascular resistance up
the anterior spinal artery is 50 times greater than
down the anterior spinal artery).
80
It is possible
that, aside from short aortic cross-clamp time,
scavenging oxygen free radicals during reperfu-
sion (e.g., with recombinant superoxide dismutase)
may prove to be a useful way in limiting ischemia-
induced neurologic damage.
81
The fourth consideration is maintenance of renal
function, and partial by-passes are theoretically
most protective.
82
Pharmacologic methods such as
mannitol and dopamine have not proved to be
helpful.
83
Other factors that may cause renal dys-
function/failure (aside from aortic cross clamping
with or without shunting) are hypotension before
and during operation, amount of operative blood
loss, degree of associated injuries, or a combina-
tion of these factors.
82
The fifth consideration is associated injuries.
For example, systemic heparinization is contrain-
dicated with closed-head injuries and would elim-
inate use of pump by-pass. Simple aortic cross
clamping avoids heparinization and allows for
more prompt management of other injuries owing
to shorter surgical time, there are no cannulation
site complications, and postoperative renal func-
tion is normal if clamping time is less than 30 min.
Use of the TDMAC-heparin-bonded shunt also
avoids the use of heparin.
This disease has an extremely high early mortal-
ity, and the aim of preoperative medical manage-
ment is to limit the extent of dissection or prevent
frank aortic rupture. The time spent in establishing
monitoring should be inversely proportional, and
the speed of induction of anesthesia should be
directly proportional, to the likelihood of further
dissection/rupture and the hemodynamic stability
of the patient. Proper judgment as to the appropri-
Figure 17-4 The repair of the various thoracic aortic dissections requires different surgical approaches. A, Complete cardiopul-
monary by-pass (CPB); cardioplegia for myocardial preservation. B, Cardiopulmonary by-pass for cooling prior to cross clamping,
then cessation of by-pass with profound hypothermia during cross clamping. C, TDMAC-heparin shunt for descending aortic
dissection. D. Left ventricular, atrial, or femoral vein-femoral artery partial by-pass (PB) for descending aortic dissection. E. Simple
aortic cross clamping for descending aortic dissection.
ate balance between preoperative preparation and
patient risk is extremely critical with this disease.
Preoperative anesthetic considerations consist of
correction of hypovolemia, institution of monitor-
ing, drug therapy, and laboratory analysis. The
extent of correction of hypovolemia will depend
on the need for speed of control of the thoracic
aorta. At a minimum, adequate intravenous access
must be ensured by at least two large-bore intra-
venous catheters connected to blood warmers, and
the patient's blood must be typed and cross
matched for adequate amounts of blood products
(10 units of whole blood, platelets, fresh frozen
plasma). Consideration should be given to use of
autotransfusion (i.e., the cell saver).
Monitoring should be instituted as permitted by
the overall hemodynamic status of the patient. If
possible, the V
5
or equivalent leads (CMV
5
or
CB
5
) should be used to monitor myocardial ische-
mia. A right radial (or brachial) intra-arterial cath-
eter should be inserted because the left subclavian
artery and, therefore the left radial artery, might be
compromised by aortic dissection and/or the prox-
imal aortic clamp. If right arm arteries cannot be
cannulated, then insertion of a needle into the
proximal aorta in the operative field can be substi-
tuted during the aortic cross-clamping period. In
addition, the left subclavian artery is more likely
to be compromised by intraoperative clamping
than the innominate artery. A second arterial cath-
eter distal to the cross clamp (e.g., femoral artery,
dorsalis pedis artery) may be useful for measuring
distal perfusion pressure during cross clamping.
84
Urine output should be monitored with a urinary
catheter. If time permits, a pulmonary artery cath-
eter or central venous pressure catheter should be
inserted. If access to the central circulation is ob-
tained, a pulmonary artery catheter is desirable
because of the known discrepancies between cen-
tral venous pressure and the pulmonary artery oc-
cluded pressure in patients with poor ventricular
function (e.g., ejection fraction less than 40 per
cent) (which many patients with this problem may
have).
85
86
Both central venous pressure and pul-
monary artery occluded pressure correlate well
with cerebrospinal fluid pressure (which is the
back pressure for spinal cord perfusion).
87
Trans-
esophageal echocardiography, either in use to
make the diagnosis of aortic disruption or inserted
after the induction of anesthesia, can certainly
monitor on-line, ventricular performance (see prior
discussion and chapter 7).
Before the induction of anesthesia, cerebral
function should be noted simply by briefly talking
with the patient and by observing the pupils and
motor function. Somatic sensory evoked potential
monitoring of spinal function has been used, but
unfortunately it has not had a significant impact on
the prevention of neurologic deficit. Specifically,
localization of critical spinal arteries for reattach-
ment has not been possible, and the incidence of
false-negative results has been 13 per cent, and of
false-positive results, 67 per cent.
88
Drug therapy consists of controlling blood pres-
sure with sodium nitroprusside and heart rate with
propranolol. The sodium nitroprusside should be
titrated to an end point of systemic arterial blood
pressure of 100 to 110 mm Hg, and propranolol
should be titrated to the end point of a heart rate
of 60 to 70 beats/min. Laboratory analysis should
consist of complete blood count, determination of
electrolytes and blood urea nitrogen/creatinine
concentrations, clotting studies (prothrombin time
[PT]/partial thromboplastin time [PTT]/bleeding
time/platelet count), and a 12-lead electrocardio-
gram (EKG). Oxygen should be administered by a
face mask throughout all these procedures.
The important intraoperative considerations are
mentioned in approximately the order that they
arise during the repair of the thoracic aortic aneu-
rysm (Table 17-3). The aim of induction of anes-
thesia is to prevent both hypertension and hypoten-
sion. In patients with acute dissection, the
induction should be smooth and rapid but con-
trolled. In these patients, anesthesia should be in-
duced intravenously with small to moderate doses
of narcotic, ketamine, or thiopental (or some com-
bination of these drugs, depending on the starting
blood pressure and the degree of hypovolemia
thought to be present), followed by profound mus-
Table 17-3 IMPORTANT ANESTHETIC
CONSIDERATIONS FOR
DESCENDING THORACIC AORTIC
ANEURYSMS/DISRUPTIONS
I. Control hemodynamic responses:
A. Nitroprusside (systolic blood pressure 110 to 100 mm
Hg)
B. Propranolol or labetalol (heart rate 60 to 70 beats/min)
II. One-lung ventilation (collapse left lung, ventilate right
lung)
III. Following clamping control proximal hypertension:
A. Nitroprusside,
B. Increase partial by-pass drainage,
C. Decrease intravenous infusion
IV. Prior to and during unclamping avoid hypotension:
A. Infuse volume,
B. Taper nitroprusside,
C. Discontinue by-passes,
D. Administer bicarbonate as indicated,
E. Restore coagulation
cle relaxation, cricoid pressure, and rapid endotra-
cheal intubation. For elective aneurysm repair,
higher induction doses of narcotic are usually
used, and the induction pace can be more leisurely.
Adjuncts for control of hypertension include ad-
ministration of increasing sodium nitroprusside
dose, intravenous lidocaine (Xylocaine) bolus ad-
ministration, intravenous propranolol bolus admin-
istration, further intravenous anesthetic drug ad-
ministration, and the initiation of inhalation
anesthesia. Adjuncts for control of hypotension
consist of volume infusion and vasopressor drug
administration. The goal of the maintenance of
anesthesia is to control systemic blood pressure,
which can be done in a variety of ways (halogen-
ated drug anesthesia, high-dose narcotic anesthesia
with sodium nitroprusside).
Unless there is a strong contraindication, these
cases should be done with a double-lumen tube
and collapse of the left lung and ventilation of the
right lung (with patient in the right lateral decubi-
tus position for left thoracotomy). Ordinarily, a
left-sided double-lumen tube should be used (see
chapter 9), but if the aneurysm distorts the left
main-stem bronchus, a right-sided double-lumen
tube should be used. In fact, a dilated aorta may
impinge on the trachea and carina sufficient to
make ventilation difficult
89
and may indicate the
use of cardiopulmonary by-pass.
90
Examination of a preoperative chest film will
often permit the correct initial decision of which
bronchus to intubate. If it is anticipated that differ-
ential lung ventilation will not be needed, the Uni-
vent tube (or a single-lumen tube with an indepen-
dent bronchial blocker) will provide satisfactory
one- and two-lung ventilation.
There are three important reasons why one-lung
ventilation and double-lumen tube intubation are
particularly advantageous for these cases. First, as
in many other thoracic surgery cases, surgical ex-
posure is greatly improved, and collapse of the
adjacent left lung is particularly important during
the difficult period of initial aortic mobilization.
Second, initial surgical dissection can cause bleed-
ing into the bronchi of the left (nondependent)
lung because the aorta is often adherent to the
adjacent lung tissue. The double-lumen tube pro-
tects the right (dependent) lung, which might oth-
erwise be flooded by this blood traversing the ca-
rina. Third, collapse of the left lung protects the
left lung from damage during periods of heparini-
zation. If the left lung were being ventilated during
this period, considerable retraction would have to
be placed on it to allow for adequate surgical ac-
cess. This, in a fully heparinized patient, could
again initiate bleeding in the left lung.
One-lung ventilation, therefore, confers several
benefits to the patient. However, it should be noted
that these patients are likely to have a sodium
nitroprusside infusion, which is a potent inhibitor
of hypoxic pulmonary vasoconstriction in the col-
lapsed left lung. Consequently, use of sodium ni-
troprusside during one-lung ventilation requires
that arterial blood-gas concentrations be monitored
frequently (pulse oximetry is useful in this situa-
tion). If severe hypoxemia occurs, nondependent-
lung CPAP or temporary clamping of the left pul-
monary artery to diminish the shunt flow through
the left lung should be instituted (see chapter 11).
Blood loss may be considerable during the ini-
tial mobilization of the aorta, and adequate volume
replacement is a high priority. Once the section of
the aorta that is diseased has been freed and mo-
bilized, arterial cross clamps will be placed proxi-
mal and distal to the lesion. Prior to placement of
the cross clamps, a sample of unheparinized blood
must be drawn from the patient to preclot the pros-
thetic graft, and the patient must be adequately
anticoagulated with intravenous heparin; the acti-
vated clotting time (ACT) should be approxi-
mately four times the control value.
91
From the
point of view of both possible preclamp hypovo-
lemia and postclamp hypoperfusion, the well-
being of the kidneys must be monitored closely by
the urine output. If oliguria or anuria is present, an
osmotic diuretic should be administered.
Placement of an aortic cross clamp causes an
increase in left ventricular afterload. If the ascend-
ing aorta or aortic arch is cross clamped, the heart
must be fibrillated and cardiopulmonary by-pass
must be used (Fig. 17-4, approach A or B). When
the descending aorta is cross clamped, structures
that are proximal to the cross clamp are still per-
fused by the heart, and structures distal to the cross
clamp are either unperfused or perfused via a shunt
or partial by-pass pump (see Fig. 17-4, approaches
C, D, and E). However, because the heart is now
pumping into a markedly reduced vascular bed,
and vasoconstrictor substances are released by aor-
tic cross clamping,
92
hypertension proximal to the
clamp occurs, which is dangerous because the
proximal hypertension may cause the heart to be-
come ischemic and/or excessively distended. In
this context, pulmonary artery catheter monitoring
(pulmonary artery occluded pressure, giant CV
waves, and so on) can be very useful.
85
Conse-
quently, proximal hypertension must be controlled
by infusing vasodilator drug, moderately increas-
ing the inhaled halogenated drug concentration,
93
increasing the drainage of the heart by the partial
by-pass (see Fig. 17-4, approach D), and/or de-
creasing the intravenous volume infusion (let the
systemic blood pressure drift downward).
Structures distal to the cross clamp, especially
the kidneys and lower spinal cord, must also re-
ceive blood flow. In this regard, a distal perfusion
pressure of 40 to 70 mm Hg must be maintained.
94
During the period of cross clamping, the heart
should not be depressed by very high doses of
inhaled halogenated drugs.
During the period of cross clamping, intraoper-
ative monitoring of somatosensory evoked poten-
tials as a guide to the state of spinal cord perfusion
has been used. Ideally, an early warning of cord
ischemia may allow for modification of surgical
technique, such as reanastomosis of sacrificed in-
tercostal artery or repositioning of hemostatic
clamps, in an attempt to restore spinal cord perfu-
sion.
95
Unfortunately, the incidence of false-posi-
tive and false-negative results has been high with
somatosensory evoked potential monitoring.
88
Pharmacologic agents used to decrease spinal in-
jury, in addition to sodium nitroprusside, are ste-
roids, pentathol, magnesium sulfate, and mannitol.
The intravascular volume management before
unclamping is critical. Just as the placement of the
cross clamp decreased the arterial space and vas-
cular compliance, unclamping will have the oppo-
site effect. To prepare for this sudden return to
normal vascular size and compliance and to avoid
precipitous hypotension, the anesthesiologist
should begin transfusion to high-normal cardiac
filling pressures as the final vascular anastomosis
is performed. If hypotensive agents have been nec-
essary to treat proximal hypertension, careful drug
infusion tapering should begin. Adequate amounts
of blood and blood products should be available if
further transfusion is necessary once the cross
clamp is removed. Unclamping the descending
thoracic aorta must be accompanied by simultane-
ous discontinuance of the distal perfusion mecha-
nism. With unclamping, bleeding may occur along
fresh suture lines and must be monitored and re-
placed. Metabolic acidosis after unclamping
should be expected in unshunted patients (even
with a mean aortic cross-clamp time of just 32 min
the pH decreases by a mean of 0.2 units)
96
and
should be corrected as indicated by arterial blood-
gas tensions. A small to moderate transient in-
crease in mean pulmonary artery pressure should
be expected (approximately 4-5 mm Hg) within
the first 5 min after abdominal aortic declamping
perhaps because of a sudden decrease in S
v
0
2
or
increase in P
a
C0
2
(causing hypoxic pulmonary
vasoconstriction).
97
Urine output should be monitored closely in
order to detect oliguria early. After aortic un-
clamping, the coagulation status should be man-
aged in the following sequence: reverse heparin
with protamine; check ACT (or equivalent method
to detect residual'heparin effect); give additional
protamine as indicated to return ACT to control
level; perform coagulation tests (laboratory), in-
cluding PT, PTT, and platelet count; and treat ac-
cording to results of clotting studies. After aortic
unclamping, the goal of anesthesia should be to
end the case so that the patient is paralyzed and
adequately sedated. As the patient emerges from
deeper levels of anesthesia, hypertension must be
treated to prevent suture line stress. When the pa- ;
tient is in the supine position, the double-lumen
tube should be converted to a single-lumen tube
and the patient should be ventilated mechanically
until all vital functions are adequate and have sta-
bilized.
IV. BRONCHOPLEURAL FISTULA
A BPF may be caused by the rupture of a lung
abscess, bronchus, bulla, cyst, or parenchymal tis-
sue (as in the case of high levels of PEEP during
mechanical ventilation) into the pleural space; ero-
sion of a bronchus can occur by a carcinoma or
chronic inflammatory disease and by breakdown
of a bronchial suture line following pulmonary
resection.
Other abnormal connections between distal por-
tions of the respiratory airways and other anatomic
structures are abdominobronchial and bronchovas-
cular, and they also usually produce significant
pulmonary compromise. A variety of abnormal
connections between the lung and abdominal or-
gans occur usually as a result of infection, trauma,
or malignancy. These fistulas most often produce
pulmonary rather than visceral symptoms even
though the primary disease initiating the fistula
originates in the abdominal organ. It is postulated
that inflammation, infection, or tumor invasion
within the abdomen causes diaphragmatic and
lung penetration. The pulmonary consequences are
nonspecific and include consolidation, atelectasis,
and hypoxemia. Fistulas between the . lung and
blood vessels or the heart usually present clinically
with life-threatening hemoptysis or air emboliza-
tion. Two examples of such acute and catastrophic
connections include blunt and penetrating wounds
to the lung and rupture of the pulmonary artery
into the airway as caused by a balloon flotation
catheter.
The diagnosis of BPF is usually made clinically.
In early postpneumonectomy patients, the diagno-
sis is based on sudden dyspnea, subcutaneous em-
physema, contralateral deviation of the trachea,
and disappearance of the fluid level on roentgeno-
grams of the chest. In patients after lobectomy,
persistent air leak, purulent drainage, and expec-
toration of purulent material are usually diagnos-
tic. When the fistula appears after removal of the
chest tube, the diagnosis of a BPF is made on the
basis of fever, purulent sputum, and a new air-
fluid level in the pleural cavity on the roentgeno-
gram of the chest. The diagnosis is confirmed by
bronchoscopic examination in most, bronchogra-
phy in a few, and sinograms (of the fistula) in an
occasional patient.
98
Other methods consist of injection of an indica-
tor such as methylene blue into the pleural space
with its recovery from the sputum, accumulation
of radionuclide in the pleural space after inhalation
of xenon or when bronchography shows spillage
of contrast into the vacant hemithorax. In an effort
to eliminate the dreaded complication of postsur-
gical (resection) leak, tissue adhesive has been
prophylactically applied at the time of bronchial
closure."
In postpneumonectomy patients, if the disrup-
tion occurs early, it is possible to resuture the
stump. Late postpneumonectomy bronchial disrup-
tion is usually associated with empyema and has
been managed by conservative drainage, but defin-
itive operative closure is now considered the treat-
ment of choice. Operations for late postpneumo-
nectomy disruption consist of closure with
pedicled flaps of muscle and omentum using a
lateral thoracotomy and staple closure through an
anterior transpericardial approach via median ster-
notomy.
100
In non-postpneumonectomy cases, if the lung
expands to fill the thoracic cavity, the leak can
usually be controlled with chest tube drainage
alone. The proven rationale for thoracostomy suc-
tion is to evacuate the pleural space and seal the
parietal and visceral pleurae against one another
so that the BPF is occluded. With time, most fis-
tulas scar and heal. However, if the fistula is large,
and a large leak through a large persistent pleural
space occurs, it is unlikely that the fistula will
close, and surgical resection is usually necessary.
98
The most commonly performed operations are
staged multiple-rib resection thoracoplasties (to
obliterate the pleural space), resuturing of bron-
chial stump, and suture of intercostal muscle ped-
icle (or pleuras, thymus, pericardium, or omentum)
over the open bronchial stump.
98
If the lung re-
maining in the hemithorax is unable to expand
because of pleural thickening, a decortication pro-
cedure may also be performed. An empyema com-
plicating a bronchopleural fistula should, if possi-
ble, be drained before surgery. This is done under
local anesthesia with the patient in an upright po-
sition to minimize the possibility of contaminating
healthy lung tissue with the cavity's purulent ma-
terial.
Finally, a spontaneous pneumothorax (no prior
lung resection involved) is pathophysiologically
similar to a BPF. A spontaneous pneumothorax is
due to a ruptured pseudocyst: either a bleb or a
bullae. Blebs are subpleural and small, less than 1
cm in diameter.
There are three types of bullae. Type I is a thin-
walled cyst with little communication with the
bronchial system that is basically extrapulmonary
in nature. It is usually a few centimeters in diame-
ter but sometimes grows to a large size: 15 to 25
cm. At this size, such pseudocysts are readily vis-
ible on a plain chest X-ray film and cause com-
pressive bullous emphysema. Surgical bullectomy
for compressive bullous emphysema has been
most successful for type I bullae. A type II pseu-
docyst is a medium-sized bulla and has thick, fi-
brous walls. It is located deep in the lung paren-
chyma. Several type II bullae may be found
together in one lobe. Patients with type II bullae
usually remain asymptomatic, and the bullae are
not visible on plain chest films. When they rup-
ture, causing a spontaneous pneumothorax, how-
ever, surgical intervention is frequently required.
Type III is a large pseudocyst and is usually found
in more than one lobe (diffuse bullous emphy-
sema). They are the most common cause of diffuse
bullous emphysema. Bronchial communications
are abundant. When they rupture, surgical inter-
vention is required.
There are three situations in which definitive
surgical treatment is indicated in treating a spon-
taneous pneumothorax. First, surgery is required
when conventional tube drainage and suction have
been unsuccessful in clearing the pleural space
(approximately 80 per cent of patients with spon-
taneous pneumothorax have cure without recur-
rence with simple chest tube drainage),
101
and
when, in effect, a BPF has formed.
Second, surgical intervention is usually indi-
cated when a second ipsilateral or first contralat-
eral spontaneous pneumothorax occurs. Third,
considering that a spontaneous pneumothorax has
a recurrence rate of 20 per cent, if after the initial
event the patient's lifestyle is such that a recur-
rence might be life-threatening or highly incon-
venient, definitive treatment is indicated.
In younger patients with a simple pneumo-
thorax, pleurectomy is the prophylactic procedure
of choice and results in a low recurrence rate.
102
I03
In older patients with severe, chronic airflow ob-
struction or young patients with cystic fibrosis,
chemical pleurodesis with talc, dextrose, tetracy-
cline, AgN0
3
, or other irritant is safer but also less
effective.
104
There are a number of closed-chest closure
methods of a BPF or a ruptured pseudocyst (Table
17-4).
105
Most of these methods are based on a
Table 17-4 CLOSED-CHEST CLOSURE
METHODS FOR BRONCHIAL
FISTULAS AND RUPTURED
PSEUDOCYST (BLEBS, BULLAE)*
I. Differential lung ventilation (see Table 17-5, 13)
II. Bronchoscopic
1. Nd:YAG and C0
2
laser
2. Lead fishing weights or shot
3. Balloon occlusion (septostomy catheter)
4. Tissue adhesive (bucrylate, Histoacryl)
5. Silver nitrate to directly visualize stump fistulas
6. Fibrin glue
7. Autologous blood instillation to form an obstructive
clot and doxycycline as an irritant
8. Absorbable gelatin sponge (Gelfoam)
III. Chest tube chemical pleurodesis
IV. Thoracoscopic
1. Instillation of talc or other irritant chemical into fistula
and/or pleural space (pleurodesis)
2. Nd:YAG pleurodesis
3. Fibrin glue
4. Tissue glueHistoacryl
5. CO, laser, Nd:YAG laser ] of blebs and
6. Electrocautery J bullae
^Modified from Powner DJ, Bierman MI: Thoracic and ex-
trathoracic bronchial fistulas. Chest 100:480-486, 1991. Used
with permission.
Abbreviation: Nd:YAG = neodymium: yttrium aluminum
garnet.
temporary physical occlusion of the airway until
an inflammatory response to a foreign material
effects a permanent seal. All of the occlusion
methods carry a risk that the obstructed segment
may become infected or the obstructing object
may act as a one-way valve, leading to distal ex-
pansion rupture of other previously normal lung
units. Simultaneous use of a bronchoscope and
thoracoscope combines the advantages of two of
the methods.
Preoperatively, it is useful to estimate the loss
of tidal volume through the BPF. This may be
done in two ways (Fig. 17-5). First, one should
determine whether air bubbles intermittently or
continuously through the chest tube. If air bubbles
intermittently, it means the BPF is small. A small
BPF only allows air to enter the pleural space with
large transpleural pressure swings (deep inhala-
tion, during cough against a partly closed glottis,
or during positive-pressure ventilation). With quiet
breathing, one generally sees air bubbling through
the water-seal chamber when the patient exhales
and intrapleural pressure rises.
In contrast, when one has a large, low-resistance
BPF or bronchial rupture, air may bubble continu-
ously through the water-seal chamber of the chest
tube drainage system. A constant air leak with
quiet respiration suggests that the damage is so
great that ordinary transpleural pressures drive air
into the pleural space during most of the ventila-
tory cycle. The application of pleural suction may
actually increase the transpleural pressure gradient
and increase gas flow from the airway to the drain-
age system. Consequently, the lung will not ex-
pand to contact the chest wall, and closure of such
a large BPF may be difficult.
Second, the size of the BPF may be quantitated
by the difference between inhaled and exhaled
tidal volumes. In the nonintubated patient, this
may be determined with a tight-fitting mask and a
fast-responding spirometer, and in the intubated
patient by direct attachment of the spirometer to
the endotracheal tube. The larger the leak, the
greater the need to isolate the BPF (double-lumen
tube, bronchial blocker).
During the induction of anesthesia, suction on
the chest tube should be discontinued to decrease
the loss of tidal volume with the initiation of pos-
itive-pressure ventilation.
106
The presence of the
chest tube still protects against the development of
pleural air trapping and a tension pneumothorax.
The bubbling through the chest tube usually in-
creases with the initiation of positive-pressure ven-
tilation (even though the suction is discontinued).
Although virtually all surgical procedures, in-
cluding and especially thoracoscopic procedures,
require a double-lumen tube and one-lung ventila-
tion, selective lobar bronchial blockade of a lobe
with a BPF (achieved by placing a fiberoptic bron-
choscope and Fogarty catheter down one of the
lumens of an in-situ double-lumen tube) has been
described.
107
At the conclusion of surgery, the integrity of the
airway may be tested by pressurizing the airway
and looking for bubbling of air in a saline-filled
hemothorax. Nevertheless, whether the chest is
open or closed, the lung should be expanded
slowly to decrease the risk of re-expansion pul-
monary edema.
108
Therapy to decrease gas flow via the fistula in
anticipation of spontaneous closure may be used
when urgent intervention is needed, multiple fis-
tulas are present, or full closure techniques fail.
Such treatment is directed toward reducing the
pressure gradient across the fistula by changing
gas pressure at one of both ends of the BPF. This
is accomplished by either reducing the force cre-
ated by positive airway pressure or altering the
subambient pressure in the pleural space. These
techniques have been extensively reviewed and are
summarized in Table 17-5.
There have been several medical (nonsurgical)
approaches (use of various mechanical ventila-
tion-chest tube drainage systems) to the treatment
Anesthesia for Emergency Thoracic Surgery
QUANTIFICATION OF TIDAL VOLUME LOSS
THROUGH BRONCHOPLEURAL FISTULA
Mj Intermittent or
Continuous
Bubbling = Tidal
Volume
Loss
(2) Spirometer Connected to Mask
(non-intubated patient) or
Endotracheal Tube ( ETT):
Difference between Inhaled (I)
and Exhaled (E) Tidal Volume
= Tidal Volume Loss
Figure 17-5 The loss of tidal volume through a bronchopleural fistula (BPF) may be semiquantitated in two ways: (l) the
amount of bubbling through a chest tube and (2) the difference between inspired and expired volume.
of patients with BPF that are important for the
anesthesiologist and intensive care physician to
know about'
09-
"
8
(see Fig. 12-9). In all of these
approaches, initial adequate conventional positive-
pressure ventilation through a single-lumen endo-
tracheal tube was generally difficult to achieve be-
cause of loss of much of the tidal volume through
the fistula. This, in turn, retarded healing and may
have even made the fistula larger. One solution
was to intubate selectively the bronchus of the
Table 17-5 TREATMENT TO DECREASE GAS
FLOW THROUGH A
BRONCHOPLEURAL FISTULA*
I. Reduce positive airway pressure
1. Minimize tidal volume, positive end-expiratory pressure
(PEEP), expiratory retard, inspiratory flow, and
inspiratory time during mechanical ventilation
2. Emphasize spontaneous breathing modes of ventilation
3. Independent lung ventilation via a double-lumen
endobronchial tube
4. High-frequency jet ventilation
II. Change pleural pressure
1. Increase or decrease chest tube suction
2. Application of PEEP or inspiratory closure valves on
chest tubes
3. Decubitus body position with the fistula dependent
*From Powner DJ, Bierman MI: Thoracic and extrathoracic
bronchial fistulas. Chest 100:480-486, 1991. Used with permis-
sion.
contralateral normal lung (a double-lumen tube is
the best way to do this) and provide ventilation to
only one lung, thereby allowing the fistula to heal
in a relatively quiet environment
109
(see Fig. 12-
9). However, this approach may not be applied to
patients with a BPF who have associated pulmo-
nary pathologic changes and are, therefore, unable
to tolerate the large transpulmonary shunt inherent
in this method of treatment. In addition, this type
of patient may require PEEP during mechanical
ventilation to maintain functional residual capac-
ity, and unilateral PEEP may increase the shunt
through the nonventilated lung.
It is difficult to apply PEEP effectively in the
presence of a large BPF because a constant air
leak is present."
2
"
3
Furthermore, the suction chest
drainage, which is normally present in these pa-
tients, may cause the ventilator to cycle if it is set
in the assist mode."
2
"
3
One solution to the prob-
lem of maintaining PEEP in the presence of a large
BPF has been the application of positive intra-
pleural pressure equal to the end-expiratory pres-
sure desired during mechanical ventilation
110
(see
Fig. 12-9). This method is effective in maintaining
PEEP and in preventing gas leak during exhala-
tion, but it does not prevent an air leak during
positive-pressure inspiration. Alternatively, the in-
sertion of a unidirectional valve into the chest tube
drainage apparatus solves the problems of avoid-
ing an air leak during positive-pressure inspiration
and using PEEP.
111
The unidirectional chest tube
valve is closed by the inspiratory phase of the
ventilator and opens during expiration. A small
tension pneumothorax is created with each me-
chanical inspiration, but it is immediately relieved
when the chest tube valve opens. Deleterious car-
diovascular effects have not been noted with this
technique. However, the chest tube valve must fre-
quently be checked because chest tube drainage
fluids may cause the valve to either stick closed or
remain permanently open.
One report describes a conservative approach to
managing a BPF by decreasing airway pressure
delivered to the diseased lung using differential
lung ventilation.
114
A double-lumen endotracheal
tube was placed in the patient, and independent
volume settings for each lung were achieved by
using two synchronized ventilators. The nondis-
eased lung was ventilated normally (tidal volume
800 ml, PEEP 12 cm H
2
0, rate 10/min), while
tidal volume was kept low in the diseased lung
(tidal volume 150 ml, PEEP 5 cm H
2
0, rate
10/min). Oxygenation and ventilation were im-
proved over that provided by conventional me-
chanical ventilatory support, although the patient
eventually died. The authors stated that, in retro-
spect, they should not have ventilated the diseased
lung at all; rather, a level of CPAP just below the
critical opening pressure of the fistula might have
been more beneficial.
High-frequency ventilation (HFV) has been pro-
posed as treatment for a large BPF (see chapter 12
and Fig. 12-9). The purported advantages of high-
frequency ventilation for BPF included the follow-
ing: (1) There is minimal loss of tidal volume
through the BPF; (2) the fistula may heal more
quickly; (3) the airway resistance (or lack of) and
pulmonary compliance have minimal influence on
ventilation; (4) low airway pressures result in min-
imal effects on cardiac output; and (5) spontaneous
ventilatory efforts are usually abolished, thereby
lowering oxygen consumption from respiratory
muscles and eliminating the need for paralysis and
excessive sedation. However, the gas-exchange ef-
fect of HFV in patients with BPF has not been
predictable, and in some patients
119
and in well-
controlled animal studies
120
HFV has increased gas
leak through the BPF.
The first report of the successful treatment of a
BPF with high-frequency jet ventilation appeared
in 1980.'
I5
This article described a patient in whom
a BPF occurred along with an empyema several
months after a right upper lobectomy. The patient
was ventilated (rate 115/min) for several weeks
until the fistula closed and weaning was possible.
The use of dual-mode independent lung ventilation
utilizing both conventional and high-frequency jet
ventilation for the intraoperative repair o a BPF
has also been reported."
6
This article described a
patient who was managed during thoracoplasty
with a double-lumen endotracheal tube; the non-
diseased lung was ventilated in a conventional
manner (tidal volume 500 ml, rate 9/min), while
the lung with the BPF received high-frequency jet
ventilation (rate 150/min). This method was con-
tinued for 2 hours following successful repair of
the BPF until extubation was accomplished. Fi-
nally, high-frequency jet ventilation has been used
in the nonoperative management of a patient with
bilateral BPF."
7
This report described a patient
who required high-frequency jet ventilation for al-
most 2 months; at the time of extubation, small
fistulas were still present but eventually healed.
Thus, although the eventual role of high-frequency
jet ventilation for treating BPF is not fully estab-
lished,"
8
it is emerging as the best nonsurgical
method to treat this condition when mechanical
ventilatory support is required and the proper
equipment and personnel experienced with its use
are available.
No matter which ventilation method and chest
tube drainage system are used, the entire approach
should be evaluated by continuously measuring
the volume (flow) of gas passing from the chest
tube by inserting a sterile "flow-tube" sensor into
the chest tube drainage system.
106
The usual meth-
ods of visually assessing the amount of bubbling
in the water-sealed chamber or subtracting the
measured exhaled lung volume from the present
inspired tidal breaths, although useful for quick
preoperative assessment (see prior discussion and
Fig. 17-5), are only intermittent semiquantitative
measures of lost volume, which are especially in-
accurate if the patient is breathing spontaneously.
In addition, flow through the fistula may fre-
quently be changed during the patient's clinical
course. Continuously measuring fistula flow al-
lows on-line titration of tidal volume, PEEP, inspi-
ratory occlusion of the chest tube, and balancing
airway/chest tube PEEP.
106
In addition, the re-
sponse time of the flowmeter is short enough to
separate fistula flow during the inspiratory phase
of ventilation from that of exhalation in most pa-
tients.
Gas recovered from the fistula may have partic-
ipated in physiologic gas exchange, as evidenced
by an increased amount of carbon dioxide in gas
leaving the chest. The magnitude of such gas ex-
change was dramatically demonstrated by a patient
who had a long-term thoracostomy window and,
with mouth and nose occluded, could meet his
respiratory needs by ventilating only through the
bronchocutaneous fistula.
106
Obviously from the prior discussion, and for the
preoperative nonintubated patient, the ability to
deliver positive-pressure ventilation adequately to
a patient with a BPF and chest tube drainage, or
to a patient with a bronchopleurocutaneous fistula,
must be carefully considered preoperatively. If a
fistula is obviously small, chronic, and uninfected,
a standard endotracheal tube can likely be used
safely. When in doubt, positive-pressure ventila-
tion can be tested, and, if found inadequate, the
standard endotracheal tube can be replaced with a
double-lumen endotracheal tube. If, after the chest
is opened, an excessive leak is encountered when
using a standard endotracheal tube, ventilation can
be improved by lung packing and manual control
of the air leak.
121
For large fistulas or fistulas of unknown size,
and for those in which an associated abscess or
empyema is known or suspected to be present,
intraoperative use of a double-lumen endotracheal
tube has the great triple benefits of permitting pos-
itive-pressure ventilation of the normal lung with-
out loss of the minute ventilation through the fis-
tula, avoiding the great hazard of contamination of
the uninfected lung with infected material when
tion,
122
"
124
and facilitating surgical exposure of the
operative lung. Indeed, in one series of 22 patients
undergoing operations for BPF following pulmo-
nary resection for tuberculosis or tuberculous em-
pyema, management with a single-lumen endotra-
cheal tube, despite intubation in the head-up
position and frequent suctioning, resulted in exten-
sive contamination of normal lung in two pa-
tients.
125
In one patient, the operation had to be
terminated, and in the other emergency bronchos-
copy had to be performed. The use of a double-
lumen endotracheal tube effectively isolates the
leaking and perhaps infected lung cavity from the
normal lung. For patients who cannot have a dou-
ble-lumen tube (e.g., small children, patients who
cannot tolerate being taken off the ventilator, those
with anatomic difficulties, postpneumonectomy
patients,
126
) bronchial blockade and endobronchial
intubation are less satisfactory alternatives.
V. LUNG ABSCESSES AND
EMPYEMA
Aspiration secondary to alcoholic stupor has
classically been reported as the factor that most
commonly precipitates lung abscess. Empyema is
the accumulation of pus in the pleural cavity. Any
lung abscess may produce empyema. In descend-
ing order of frequency, causes of empyema are
postpneumonic (abscess), postoperative (any tho-
racic surgery), thoracic trauma (infection of resid-
ual clotted blood after a hemothorax that was
treated by chest tube placement), septicemia, and
iatrogenesis (entry into the pleural space by needle,
scope, and so on, especially if there is a concurrent
intra-abdominal injury or infection).
127-129
Both a
lung abscess and an empyema may erode a bron-
chus and cause a BPF (see Chest Trauma and Fig.
17-8).
The precise mechanism by which empyema de-
velops is still obscure, but the following scenario
seems likely.
130
Inflammation of the visceral pleura
occurs as a result of pulmonary inflammation. This
encourages an increase in the production of pleural
fluid; in time, an expanding pleural effusion,
which may be sterile, develops. Unfortunately,
pleural fluid is an excellent culture medium, and
organisms reaching the pleural space from the lung
probably contribute the greatest number of cases
of empyema. However, any organisms iatrogeni-
cally introduced into the pleural space by unsterile
tapping are likely to flourish.
Continued irritation of the pleura not only re-
sults in the production of pleural fluid, which may
or may not be infected, but also edema and thick-
ening of the pleura itself. Fibrin within the pleural
fluid forms a coagulum. Coalescence of this coag-
ulum will form septations and hence loculation of
pus within the pleural space, the presence of which
will make drainage of the pleural space more dif-
ficult to accomplish. In addition, partial or com-
plete collapse of the lung may be caused by pres-
sure from the accumulating fluid. Thickened
visceral pleura will then trap the lung in this posi-
tion.
The symptoms of lung abscess and empyema
are similar and include cough, purulent sputum
production, fever, chest pain, and dyspnea. Many
patients have some form of systemic illness such
as rheumatoid disease, diabetus, chronic alcohol-
ism, advanced malignancy, chronic malnutrition
and debility, renal failure, drug abuse, or acquired
immunodeficiency syndrome.
128
I31
The most frequent physical findings include de-
creased breath sounds, rales, tachypnea, rhonchi.
dullness on percussion, and tachycardia.
129 I31
The
diagnosis of thoracic empyema is most often first
strongly suspected by the finding of a pleural ef-
fusion on chest roentgenogram and is confirmed
by diagnostic pleural aspiration from which either
frank pus was obtained or organisms were grown.
As with all abscesses, drainage of the pus will
dramatically alleviate symptoms and signs and
speed the patient's recovery. Table 17-6 lists all
the drainage methods in increasing order of inva-
Table 17-6 METHODS FOR DRAINING AN
EMPYEMA IN ORDER OF
INCREASING INVASIVENESS*
1. Needle aspiration.
2. Tube drainage, either with a fine bore percutaneous tube
placed accurately with radiographic or ultrasound control,
or a large bore tube (26-30 French gauge). Absolute
control over the insertion of any of these tubes is
mandatory if subdiaphragmatic placement with injury to
the abdominal organs is to be avoided.
3. Tube drainage plus irrigation of the pleural space plus
instillation of uro- or streptokinase.
4. Thoracoscopy, through which adhesions and septae may be
broken down, and irrigation through a tube placed at the
time of surgery.
5. Thoracotomy and decortication with excision of the
empyema.
6. Rib resection and insertion of an empyema drainage tube
or the creation of an Eloesser flap.
7. Thoracoplasty.
8. Resection of lung.
*From Wells FC: Empyema thoracis: What is the role of
surgery? Respir Med 84:97-99, 1990. Used with permission.
siveness. The choice of treatment should be tai-
lored to the stage of development that the em-
pyema has reached (Table 17-7). Various
combinations of the therapies listed in Table 17-6
may be used for the different classes of empyema
in Table 17-7.
Obviously, the success rate for a given therapy
will depend on the class of empyema. For exam-
ple, aspiration or closed-chest tube drainage can-
not help a loculated class IN empyema and multi-
ple reports have urged surgical aggressiveness
when confronting class II and class III empye-
mas.
128 129
132
-
134
For one dramatic example, the
mortality rate for patients with class II and class
III empyemas managed completely nonoperatively
has ranged from 58 to 100 per cent, whereas it has
been 16 to 19 per cent for those receiving opera-
tive drainage.
128 129,132
Of course, many nonopera-
tive treatment failures have become operative
treatment successes.
128
129
132
-
134
The overall mor-
tality for all classes of empyema has ranged from
5 to 19 per cent.
128
129132
-
135
In patients with lung abscess and empyema,
some of the preoperative findings may be very
unusual. First, the ipsilateral lung may be col-
lapsed. Second, because the infectious process
may diminish hypoxic pulmonary vasoconstriction
and hypoxemia may be profound. Third, the me-
diastinal structures may be shifted toward the dis-
eased side. Fourth, the mediastinal shift may cause
changes on the EKG.
If a surgical procedure is going to be done under
general anesthesia in a patient with either a lung
abscess or an empyema, then double-lumen tube
intubation is absolutely indicated to prevent con-
tamination of the uninfected lung by the infected
Table 17-7 A CLASSIFICATION AND TREATMENT PLAN OF EMPYEMA THORACIS
1
Class of
Empyema Description of Class Treatment of Class
Class I:
Class II:
Class III:
Post pneumonic pleural effusion
bacteria may/or may not be
isolated; usually no pus
no loculi within the pleural space
pleural fluid usually of low
viscosity
Classic uniloculatef empyema
pus in the pleural space
positive cultures of pleural
aspirate
absence of multiple locations
pus may be very thick and
viscous with sediment
Complicated empyema
multiple loculations
grossly thickened pleura
pus within empyema cavity which
may be sterile, but is usually very
viscous
trapped lung
Simple aspiration (thoracentesis)
Percutaneous catheter drainage to underwater seal
1. Tube thoracostomy for continuous drainage (may
need multiple tubes)
2. Irrigation with saline, antibiotics, enzymes
3. Thoracoscope to break empyema down to one space
(break loculations) plus 2 above
1. Thoracoscope as above
2. Thoracotomy for
a. decortication and breakdown of loculations
b. excision of empyema cavity
c. rib resection for extensive and dependent
drainage
d. closure of bronchopleural fistula
e. lung resection
3. Thoracoplasty
*From Wells FC: Empyema thoracis: What is the role of surgery? Respir Med 84:97-99, 1990. Used with permission.
tThere may be fibrinous strands criss-crossing the pleural space but discrete loculi are not present.
Table 17-8 CAUSE OF INJURIES LEADING TO
ACCIDENTAL DEATH IN THE
UNITED STATES (PER CENT) IN
1985*
Motor vehicle accidents 48
Suicide 29
Homicide 22
Miscellaneous 1
*From LoCicero J, Mattox KL: Epidemiology of chest
trauma. Surg Clin North Am 69:15-19, 1989. Used with per-
mission.
lung. In addition, if an empyema is going to be
treated with thoracoscopy, collapse of the diseased
lung greatly aids access to the empyema.
136
I37
The
seal of the endobronchial cuff should be tight and
can be quantitated by the bubble-under-water tech-
nique described in chapter 9. The position of the
tube should be determined by fiberoptic bronchos-
copy. Both of these maneuvers (checking of tube
by fiberoptic bronchoscopy and determination of
pressure level of cuff seal) should be done before
tion. As in many thoracic surgery cases, one-lung
ventilation facilitates surgical exposure (although
the surgeon will usually also be intent on distin-
guishing fibrous adhesions and in expanding the
diseased lung), and lung separation allows differ-
ential lung ventilation, as might be needed with
the presence of a BPF.
During the surgical procedure, the diseased side
should be suctioned frequently (fiberoptically if
necessary). Whenever possible, but particularly at
the end of the procedure, the diseased lung should
be fully expanded manually. When the lung has
been fully expanded, the surgeon should check
carefully for the presence of a BPF. The pleural
cavity may be irrigated with antibiotics at the end
of the procedure, and one must be mindful of the
neuromuscular blockade effects of antibiotics.
VI. CHEST TRAUMA
A. Overview of the Management of the
Patient With Extensive Trauma
Motor vehicle accidents are the most common
cause of accidental death, followed by suicide and
Table 17-9 FREQUENCY OF VARIOUS
INJURIES (PER CENT) IN MOTOR
VEHICLE ACCIDENTS*
Extremities 34
Head and neck 32
Chest 25
Abdomen 15
*From LoCicero J, Mattox KL: Epidemiology of chest
trauma. Surg Clin North Am 69:15-19, 1989.
homicide (Table 17-8).
138
In a motor vehicle acci-
dent, multiple organs may be injured (Table 17-
9).
138
Extensive trauma has been defined as an
injury severity score (ISS) greater than 25 (Table
17-10).
139
The principles of management in all cases of
extensive trauma are simultaneous assessment and
resuscitation, complete physical examination and
diagnostic studies if the patient becomes hemody-
namically stable, and life-saving surgery (the tim-
ing of surgery depends on the patient's stability).
Resuscitation has two components: the primary
survey and initial resuscitation, and the secondary
survey and continuing resuscitation (Figs. 17-6
and 17-7). All patients undergo the primary survey
of airway, breathing, circulation, and neurologic sta-
tus. Only patients who become hemodynamically
stable will progress to the secondary survey, which
focuses on a complete physical examination that
directs further diagnostic studies. The great major-
ity of patients who remain hemodynamically un-
stable require operative intervention immediately.
1. Primary Survey
The goals of the primary survey are, in order, to
establish a patent airway and adequate ventilation,
maintain the circulation (including cardiac func-
tion and intravascular volume), assess the global
neurologic status, and determine the mechanism of
injury.
a. AIRWAY AND VENTILATION
Patients with extensive trauma who are uncon-
scious or in shock benefit from immediate endotra-
cheal intubation. To prevent injury to the spinal
cord, the cervical spine must not be excessively
flexed or extended during intubation. Oral endotra-
cheal intubation is successful in the majority of
injured patients. On rare occasions, bleeding, neck
deformity, airway laceration, or edema from max-
illofacial injuries necessitates cricothyroidotomy
or tracheostomy.
b. CIRCULATION
A patient who is in shock with flat neck veins
is assumed to have hypovolemic shock until
proved otherwise (see later discussion). Indeed, in
one large series of acute blunt thoracic trauma
patients (n = 303), the cause of all but one intra-
operative deaths (n = 15) was exsanguination.
141
If the neck veins are distended and the blood pres-
sure is low, there are five possibilities: (1) myocar-
dial contusion, (2) myocardial infarction, (3) ten-
sion pneumothorax, (4) air embolism, and (5)
pericardial tamponade. Myocardial contusion is an
increasingly recognized cause of arrhythmias and
cardiac failure in patients with trauma. Indeed, the
Table 17-10 THE INJURY SEVERITY SCORE*
The injury severity score is based on the abbreviated injury score (AIS), a score on a numerical scale ranging from l (indicating
minor injury) to 6 (virtually unsurvivable injury). An AIS is assigned to each of six regions of the body: head or neck, face, chest,
abdomen or pelvic contents, extremities, and body surface. The injury severity score is defined as the sum of the squares of the
AISs for the three most severely injured body regions. The procedure can easily be seen in the following example:
Body Region Injury AIS AIS
2
Head
Face
Chest
Abdomen
Extremities
Epidural hematoma
Ear laceration
Rib fractures
Ruptured spleen
Fractured femur
16
1
4
9
9
In this hypothetical case, the sum of the square of the three highest AISs (those for the most severely injured body regions) is
16 + 9 + 9, or 34. The ISS is thus 34. The AIS for flail chest is 4, moderate pulmonary contusion 3, simple pneumothorax 3.
*From Trunkey D: Initial treatment of patients with extensive trauma. Engl J Med 324:1259-1263, 1991. Used with permission
incidence can be as high as 70 per cent when it is
searched for with thallium 201 scintigraphy.
142
The
clinical diagnosis is made by EKG changes consis-
tent with acute injury, increased creatine kinase-
MB greater than 5 per cent of the total creatine
phosphokinase (CPK), or an abnormal echocardio-
gram consistent with acute injury such as hypoki-
nesis, pericardial effusion, valvular injury, apical
thrombus, or wall thickening with edema and/or
hemorrhage.
143
Myocardial infarction from coronary occlusion
is not uncommon in elderly trauma patients. Ten
sion pneumothorax should be ranked high in thi
physician's differential diagnosis of shock witl
distended neck veins because it is the easiest life
threatening injury to treat in the emergency depart
ment (needle the chest anteriorly and apically). /
simple tube thoracostomy is the definitive methoi
of management.
Air embolism has come to be appreciated as aj
important complication in injured patients (esti
mated as high as 4 per cent of patients with majc
Figure 17-7 The algorithm for management of thoracic shotgun wounds is a specific application of Figure 17-6. (O.R. =
operating room; IV = intravenous.) (From Walker ML, Poindexter JM, Stovall I: Principles of management of shotgun wounds.
Surg Gynecol Obstet 170:97-105. 1990. Used with permission.)
trauma).
139 I44
It consists of air in the systemic
circulation and is caused by a bronchopulmonary
fistula near the hilum of the lung (pulmonary veins
are in close proximity to their bronchi only in the
hilar region; see chapter 2).
144
Not surprisingly,
therefore, airway bleeding is a common finding in
patients with air embolism.
144
A high index of sus-
picion is necessary to make a diagnosis of air
embolism. Any patient with chest trauma (espe-
cially penetrating) who has no obvious head injury
but has focal or lateralizing neurologic signs may
have air bubbles occluding the cerebral circulation.
Theoretically, the observation of air in the retinal
vessels on funduscopic examination will confirm
the presence of cerebral air embolism, but in real-
ity the diagnosis is most often made only at the
time of thoracotomy by direct visualization of air
in the coronary vessels or by aspiration of air from
the left ventricle. Any intubated patient on posi-
tive-pressure ventilation who has a sudden cardio-
vascular collapse should be presumed to have
either tension pneumothorax or air embolism in
the coronary circulation. The definitive treatment
is immediate thoracotomy in the steep head-down
position, clamping of the hilum of the injured lung
to prevent further embolism, expansion of the in-
travascular volume, and introducing proximal aor-
tic hypertension (manual occlusion of aorta) to
increase coronary blood flow. Open cardiac mas-
sage, intravenous administration of epinephrine,
and venting of the left side of the heart and the
aorta with a needle to remove residual air may be
required. The pulmonary injury is definitively
treated by oversewing the laceration or resecting
the offending lobe.
Pericardial tamponade is most commonly en-
countered in patients with penetrating injuries to
the torso. The diagnosis is often obvious. The pa-
tient has distended neck veins with poor peripheral
perfusion; a few have pulsus paradoxus. Ultraso-
nography and pericardiocentesis are occasionally
useful. The proper treatment is immediate thora-
cotomy, preferably in the operating room, al-
though thoracotomy in the emergency room can
be life-saving.
If the primary cause of shock is blood loss, the
four steps to be taken are to (1) gain access to the
circulation, (2) obtain a blood sample from the
patient, (3) determine where the volume loss is
occurring, and (4) give fluids for resuscitation. The
fastest and most reliable way to gain access to the
circulation is by a surgical cut down on the saphe-
nous vein at the ankle. Incision in this anatomi-
cally constant vein allows the physician to achieve
access quickly with a large-diameter tube. Alter-
natively, the resuscitating physician can perform
Table 17-11 FLOW RATES OF CATHETERS
(ml/min) THROUGH 6 FEET OF
TUBING WITH PERFUSION
PRESSURE OF 300 mm Hg*
Catheter
16
14
8F
IV
g central line
g IV
introducer
tubing
Crystalloid
158
387
492
486
Whole Blood
60
200
265
248
*Reprinted with permission from Millikan JS, Cain T, Hills-
brough J: Rapid volume replacement for hypovolemic shock:
A comparison of techniques and equipment. J Trauma
24(5):428-431, 1984.
Abbreviation: IV = intravenous.
bilateral percutaneous femoral vein cannulation
with a large-bore catheter or an 8 French intro-
ducer, more commonly used for passing a pulmo-
nary artery catheter. Some experienced physicians
prefer to gain access through the subclavian or
internal jugular vein. Table 17-11 shows the pres-
surized flow rates from various sized catheters.
As soon as the first intravenous catheter has
been established, base-line blood work should be
undertaken; it should include the hematocrit, toxi-
cologic screening, blood typing and cross match-
ing, and a screening battery of laboratory tests if
the patient is elderly or has premorbid conditions.
Blood-gas determinations should be performed
early in the course of resuscitation. Next, it must
be determined where occult blood loss is occur-
ring. Three sites of hidden blood loss are the
pleural cavities (a possibility that can be elimi-
nated by rapid chest radiography), the thigh, and
the abdomen, including the retroperitoneum and
pelvis. In most cases, a diagnostic peritoneal la-
vage will be performed. Table 17-12 shows the
interpretation of the results, which has approxi-
mately an 85 per cent success rate in detecting all
abdominal injuries and close to 100 per cent suc-
cess in intra-abdominal hemorrhage. Common
sense dictates that if the patient's chest film is
normal and the femur is not fractured, the patient
who remains in shock must be suspected of having
Table 17-12 BLUNT TRAUMA:
INTERPRETATION OF
PERITONEAL LAVAGE
Parameter Negative Indeterminate Positive
RBCs <50,000/1 50-100,000/ >100,000/1
WBCs <100/! 100-500/1 >500/1
Amylase <75 U/dl* 75-175 U/dl > 175 U/dl
Bile None Present
Abbreviations: RBCs = red blood cells; WBCs = white
blood cells.
ongoing hemorrhage in the abdomen or pelvis.
Most such patients require immediate laparotomy
if death from hemorrhage is to be averted.
The fourth step for the resuscitating physician is
to administer fluids for resuscitation, beginning
with balanced salt solution and adding type-spe-
cific whole blood as soon as possible. For the
patient who is exsanguinating, the order of blood
administration, in order of increasing preference,
is type O (e.g., the type has not yet been deter-
mined), type specific, and finally in the case of a
stable patient, it is prudent to wait for typed and
cross-matched blood.
c. NEUROLOGIC STATUS
The next task during the primary survey is to
assess neurologic status and to initiate diagnostic
and therapeutic procedures. The simplest evalua-
tion is alertness and response to verbal and painful
stimuli. The complete Glasgow Coma Scale is
shown in Table 17-13. It should be remembered
that apnea occurs with cerebral and cord lesions
above C-3, and intercostal paralysis and diaphrag-
matic breathing occurs with cord lesions above
C-6.
A decreasing level of consciousness is the single
most reliable indication that the patient has a seri-
ous head injury or secondary insult (usually hy-
poxic or hypotensive) to the brain. An improving
neurologic status reassures the physician that re-
suscitation is improving cerebral blood flow. Neu-
rologic deterioration is strong presumptive evi-
dence of either a mass lesion or important
neurologic injury. Computed tomographic scan-
ning of the head is the preferred technique for
diagnosing head injury and should be performed
as soon as possible.
Table 17-13 GLASGOW COMA SCALE*
I. Eye opening
Spontaneous 4
To voice 3
To pain 2
None l
II. Verbal response
Oriented 5
Confused 4
Inappropriate words 3
Incomprehensible words 2
None l
III. Motor response
Obeys commands 6
Purposeful movements (pain) 5
Withdraws (pain) 4
Flexion (pain) 3
Extension (pain) 2
None 1
Total Glasgow Coma Scale points = I + II + III.
d. MECHANISM OF INJURY
The primary survey must take into account the
mechanism of injury. There are distinct patterns
of injury according to whether the patient has
suffered blunt injury (high velocity, low velocity,
or crush injury; Table 17-14) or penetrating
trauma.
2. Secondary Survey
If the patient becomes hemodynamically stable
during the initial phase of assessment and resusci-
tation, it is then appropriate to perform diagnostic
studies as indicated to determine surgical priori-
ties.
3. Surgical Priorities (Fig. 17-6)
The priorities for surgical intervention are
straightforward. A mass lesion in the cranial vault
must always be treated immediately. Injuries to the
torso are next in line for surgical repair, but they
can be treated at the same time as head injuries.
As stated previously, the surgeon can easily deter-
mine from a chest film whether there is hemor-
rhage into the chest. Obviously, impalement inju-
ries to the thorax (the imbedded piece should
always remain in situ from the field) must go di-
rectly to the operating room.
147
A patient in shock
who has a normal chest film and no other obvious
source of blood loss must have hemorrhage within
the peritoneal cavity or the retroperitoneum. Such
a patient requires immediate laparotomy to control
the hemorrhage. Peripheral vascular injuries are
assigned the next priority for surgery, and ortho-
pedic injuries the last priority.
B. Specific Chest (Noncardiac) Injuries:
General Considerations
Blunt thoracic injury is usually associated with
other injuries to the head, abdomen, or limbs, and
it is the associated injuries that often determine the
prognosis for the patient. Associated head injury is
often severe, and severe head injury is the single
most common cause of death in patients with chest
injury. The presence of associated abdominal in-
juries makes respiratory failure and the need for
mechanical ventilation more likely. Orthopedic in-
juries cause patient discomfort and require immo-
bility, which makes management of the chest in-
jury technically difficult.
Overall, less than 15 per cent of all patients with
chest trauma will require a thoracotomy. The man-
agement of those patients near death from chest
trauma requiring immediate thoracotomy upon ar-
rival at the hospital is covered under Emergency
Room Thoracotomy in the Management of
Trauma. Patients requiring less urgent thoracot-
omy after chest trauma may have a variety of
thoracic injuries. Figure 17-8 shows the variety of
primary and secondary thoracic injuries, and the
inter-relationships of the injuries, as a traumatic
force penetrates inward. Most of the initial primary
injuries can result in major chronic secondary
complications.
1. Chest Wall Fractures (Flail Chest)
Flail chest may be defined as an abnormal
movement of the chest wall occurring as a result
of fractures of two or more ribs in two places on
the same side. If a scapula or sternum is fractured,
high energy transfer has occurred to the thorax,
and one must suspect that deeper structures have
Table 17-14 MECHANISMS AND PATTERNS OF BLUNT THORACIC TRAUMA*
Type of
Impact Chest Wall Injuries
Possible Thoracic
Visceral Injuries Common Associated Injuries
High Velocity
(Deceleration)
Chest wall often intact or fractured
sternum or bilateral rib fractures
with anterior flail (caused by
steering wheel)
Ruptured aorta
Cardiac contusion
Major airways injury
Ruptured diaphragm
Head and faciomaxillary injuries
Fractured cervical spine
Lacerated liver or spleen
Long bone fractures
Low Velocity
(Direct Blow)
Lateral-unilateral fractured ribs
Anterior-fractured sternum
Pulmonary contusion
Cardiac contusion
Lacerated liver or spleen if ribs
6-12 are fractured
Crush
Anteroposterior crushbilateral rib
fractures with or without anterior
flail
Ruptured bronchus
Cardiac contusion
Fractured thoracic spine
Lacerated liver or spleen
Lateral crushipsilateral fractures
with or without flail
Possible contralateral fractures
Pulmonary contusion Lacerated liver or spleen
'From Westaby S, Brayley N: Thoracic traumaI. Br Med J 300:1639-1643, 1990. Used with permission.
been injured. With a flail chest, each rib section
between the two fracture sites essentially floats
free. With spontaneous inspiration and the devel-
opment of a more negative intrathoracic pressure,
the injured segment moves inward upon the lung
(paradoxic chest wall movement), preventing lung
expansion and impairing oxygenation. The reverse
occurs during a spontaneous exhalation. The un-
derlying lung is frequently contused, which greatly
contributes to the poor gas exchange.
148
This injury
most commonly involves the anterolateral aspect
of the chest. The posterior wall is heavily fortified
with muscle and the scapula and, therefore, is
rarely involved. The thoracic cage in older age
groups is more calcified and brittle and, therefore,
more susceptible to flail and other serious injuries.
In contrast, the thoracic cage in pediatric patients
is extremely elastic and resilient and affords much
greater protection to underlying structures.
An opening in the chest wall allows atmospheric
air to enter the pleura, permitting the intrapleural
pressure to rise toward atmospheric pressure and
producing pneumothorax and collapse of the lung.
The result is the familiar "sucking wound," with
air entering the pleural space during inspiration
and exiting during exhalation.
Laceration of intercostal vessels may produce
hemothorax; bleeding may continue until thoracot-
omy becomes necessary. When the wound is para-
sternal, the internal mammary vessels may be in-
jured. Bleeding from the intercostal or internal
mammary arteries, being under systemic pressure,
is not tamponaded by hemothorax and usually
continues until shock or frank pulmonary insuffi-
ciency occurs. Surgical ligation is usually required.
The major concerns with chest wall and sternal
injuries are correction of inadequate gas exchange
resulting from flail chest, the provision of adequate
analgesia (to decrease splinting), and the diagnosis
of possible underlying injury. Indications for insti-
tution of mechanical ventilation are severe hypox-
emia, severe hypercarbia, clinically obvious exces-
sive work of breathing (usually present with major
flail and paradoxic chest wall movement), and se-
vere associated injuries (e.g., parenchymal contu-
sion and hemothorax). Associated injuries, espe-
cially parenchymal contusions, are likely to be
present with sternal fractures and fractures of the
upper ribs because these fractures require a large
force in order to occur. For example, with sternal
injuries, one should always suspect rupture of the
trachea, rupture of major arteries (Fig. 17-9), and,
Figure 17-9 Rupture of the trachea and avulsion of the
innominate artery by a fractured manubrium. (From Pate JW:
Tracheobronchial and esophageal injuries. Surg Clin North Am
69:111. 1989. Used with permission.)
especially, myocardial injury, because the heart is
compressed between the sternum and the verte-
brae.
149
Some patients can be effectively treated
without intubation and mechanical ventilation if
they are provided with adequate analgesia (see An-
esthetic Considerations and chapter 21); in addi-
tion to increasing lung volume (by decreasing
splinting), adequate analgesia decreases the respi-
ratory rate, which, in turn, makes the overall flail
of the chest much less pronounced.
150
I51
In summary, the approach to flail chest injuries
should be flexible.
151 152
In a prospective study of
36 patients with this injury,
151
only 30 per cent
required mechanical ventilation, even though 69
per cent developed pneumonia and about 30 per
cent developed the adult respiratory distress syn-
drome. In the ventilated patients, the mean dura-
tion of ventilation was 10.5 days. The overall mor-
tality was approximately 10 per cent. Depending
upon the series, the mortality may vary between 6
and 50 percent.
153 154
a. HEMOTHORAX
155 l56
The hemithorax is a potential space that can
easily accommodate 30 to 40 per cent of a pa-
tient's blood volume (greater than 2000 ml in a
70-kg adult). Consequently, patients with a mas-
sive hemothorax have severe hypotension.
157
In
addition, as blood progressively accumulates in the
pleural space, the underlying lung is compressed.
The chest is dull to percussion, the breath sounds
are diminished, and a shift in the mediastinum
toward the uninvolved hemithorax (displaced api-
cal cardiac impulse and trachea) may compress the
uninvolved lung; the compression of the contralat-
eral lung will exacerbate the ventilatory impair-
ment. Thus, these patients also can have severe
hypoxemia.
157
b. PNEUMOTHORAX
Pneumothorax occurs secondary to blunt or pen-
etrating trauma to the chest wall. With penetrating
trauma the mechanism of air escape is obvious;
the most common cause of penetrating pneumo-
thorax is iatrogenic and is due to subclavian punc-
ture, which has an overall incidence of pneumo-
thorax of 2 per cent.
158
Penetrating knife wounds
and gunshot wounds have close to a 100 per cent
incidence of pneumothorax. With blunt trauma the
mechanism of air escape may be tearing of the
lung by the edge of a rib fracture or development
of an alveolar bursting force (with a closed glottis)
during the trauma. With a pneumothorax, ipsilat-
eral breath sounds are decreased, and the chest is
tympanic to percussion.
A tension pneumothorax occurs when air enters
the pleural cavity during inspiration but, owing to
a ball-valve action, cannot escape during exhala-
tion. A tension pneumothorax can cause circula-
tory embarrassment because of decreased venous
return and mediastinal shift and causes respiratory
embarrassment via compression of ipsilateral lung
by direct tension and contralateral lung by medias-
tinal shift.
An open pneumothorax communicates with the
environment and has been referred to as a "suck-
ing" chest wall defect. The sucking refers to air
being drawn into the hemithorax from the environ-
ment during a spontaneous inspiration; the phys-
iology of the situation is identical to the paradoxic
respiration and mediastinal flap described in chap-
ter 4.
Pneumomediastinum and pneumopericardium in
the adult trauma patient are often incidental find-
ings of air that has tracked along the bronchial or
vascular sheaths of the thorax into the medias-
tinum and pericardium.
159
Occassionally, this can
be a grave sign of a ruptured bronchus or esopha-
gus, uncontrolled pneumothorax, or ventilator
barotrauma. Air within the pericardial space can
develop into cardiac tamponade.
c. PULMONARY CONTUSION
Pulmonary contusions occur either with pene-
trating lung injury or as the result of rapid decel-
eration forces (blunt injury), which cause the lung
to forcibly impact upon the chest wall.
155, 156
Rib
fractures occur in approximately 50 per cent of the
cases of pulmonary contusion. Deceleration forces
cause rupture of alveoli and vessels with hemor-
rhage, and increased endothelial and epithelial
permeability causes interstitial and intra-alveolar
edema. The edema phase develops progressively
with time (hours) and with fluid administration and
is accompanied by progressive chest X-ray
changes and a decrease in P
a
0
2
. Consequently, the
initial chest roentgenogram is often not a good
indicator of the severity or extent of the contusion.
However, with computed tomographic scanning,
the lesion can be seen immediately.
160
Tracheobronchial disruptions should be sus-
pected in any patient with either penetrating or
blunt trauma to the neck or chest, especially if
accompanied by subcutaneous or mediastinal em-
physema (escape of air into adjacent tissues), he-
moptysis (bleeding within the bronchus), pneumo-
thorax (if no chest tube is in place), and/or BPFs
(escape of air into the pleural space and out the
chest tube, if present). It should be noted that a
lung that is lacerated by a knife may have many
small transected bronchioles and behave like a
Bpp iei although the trachea and major bronchi
may be involved at any level (especially with pen-
etrating trauma), greater than 80 per cent of inju-
ries with blunt objects are within 2.5 cm of the
carina (see later discussion).
162-165
Penetrating
trauma to the neck may cause combined tracheo-
esophageal injuries.
166
Tracheobronchial disruptions due to blunt ob-
jects require a large force and are, therefore, often
associated with trauma to adjacent structures.
167
In
fact, in one series of tracheobronchial disruptions
due to blunt trauma, there was an average of 3.8
injuries per patient.
167
The major late complica-
tions of tracheobronchial disruption consist of
BPF, empyema, and mediastinitis. Proximal tra-
cheobronchial tree injuries, which cause dramatic
clinical symptoms, are often recognized and
treated earlier and, therefore, have a better out-
come than distal injuries, which involve a smaller
volume of air, go undetected longer, and have
higher incidence of major late complications. Flex-
ible fiberoptic (or rigid) bronchoscopy should be
performed if the diagnosis of tracheobronchial tree
disruption is being considered.
168
There are three reasons why tracheobronchial
disruptions anatomically localize around the ca-
rina.
169
First, anterior chest compression causes
rapid lateral movement of the lungs (pulls the
lungs apart), which creates a shearing force on the
carina. Second, an increase in intrathoracic pres-
sure with the glottis closed causes the greatest wall
tension (and, therefore, risk of tear) to occur in the
airways with the largest diameter (Laplace's rela-
tionship - X R, in which = tension, =
transmural pressure difference, and R = radius of
curvature). Third, the fixation of the trachea above
(cricoid cartilage) and below the carina predis-
poses the more mobile carina to shear forces.
4. Esophageal Disruptions^
5 l56 17

(see Chapter 16)
Esophageal ruptures occur from internal trauma
(after medical instrumentation of any type, inges-
tion of penetrating or corrosive material); as a re-
sult of external trauma (blunt and penetrating);
spontaneously (postemetic); and as part of the nat-
ural history of a pre-existing esophageal lesion
(e.g., tumor and caustic burns) (see chapter 16).
Unrecognized and untreated traumatic esopha-
geal injury has an extremely high late mortality
(20 per cent) due to mediastinitis, empyema, and
sepsis.
171
172
In a recent series, all patients operated
on within 24 hours of perforation survived,
whereas there was a 33 per cent mortality when an
operative delay of more than 24 hours occurred.
173
Unfortunately, esophageal injury may go unrecog-
nized for hours to days in a multiply traumatized
patient when attention is not focused on the esoph-
agus. The diagnosis should be suspected in pa-
tients who have chest pain, dysphagia, hemateme-
sis, cervical or mediastinal emphysema, and fever
(especially after an endoscopic procedure, removal
of a foreign body, or trauma). Pain is the most
striking and consistent symptom.
The chest X-ray is often suggestive by revealing
subcutaneous emphysema, pneumothorax (rupture
of mediastinal pleura), pleural effusion (also
caused by tear of the mediastinal pleural, resulting
in esophageal contents irritating the pleural space),
pneumoperitoneum, and retropharyngeal swelling.
However, chest X-ray findings take several hours
to develop and depend on the site of perforation.
174
If a chest tube has been inserted because of asso-
ciated injuries, continuous bubbling may reflect an
air leak from the esophagus, and drainage of par-
ticulate matter should raise suspicion of an esoph-
ageal tear. The diagnosis can be confirmed with
meglumine diatrizoate (Gastrografin) roentgeno-
graphically (esophagogram). Esophagoscopy is
frequently unnecessary. If the tear is diagnosed
early (less than 6 hours), primary repair is accom-
plished, often with good results.
Diaphragmatic injuries can occur secondary to
either penetrating or nonpenetrating (blunt)
trauma.
175-177
Penetrating trauma as high as the
fourth intercostal space can penetrate the dia-
phragm (at the domes) and cause intra-abdominal
injury. Alternatively, stab wounds of the abdomen
(subcostal) can penetrate the diaphragm (at the
lower circumference border) and cause intratho-
racic damage. Extreme forces are necessary to
cause diaphragmatic rupture with blunt trauma;
therefore, there is a high incidence of associated
injuries and deaths with blunt diaphragmatic
trauma (median ISS = 34).
I78
Seventy-five to 95
per cent of diaphragmatic tears after blunt trauma
occur on the left side because of the protection
afforded the right side by the liver.
There are several important pathophysiologic
consequences of diaphragmatic tears. First, if the
disruption is extensive, diaphragmatic movements
will be ineffectual and the thoracic cage will be-
have as a flail chest. Second, abdominal pressure
is greater than thoracic pressure; therefore, there is
always the potential for abdominal viscera to work
their way into the thorax through the diaphrag-
matic defects. Small diaphragmatic tears are more
likely to capture and strangulate abdominal organs.
The organs that most commonly herniate (in de-
creasing frequency) are stomach, small bowel,
spleen, omentum, liver, and kidney.
I78
l79
Third,
herniation of abdominal contents into the thorax
can cause lung compression, mediastinal shift, and
decreased venous return.
C. Specific (Noncardiac) Chest Injuries:
Surgical Considerations
The treatment varies with the severity of the
injury, ranging from simple supportive therapy
such as oxygen enrichment, physical therapy, and
pain management to CPAP by mask to full venti-
latory support. Although mechanical ventilation is
obviously required for patients in whom acute or
subacute respiratory failure develops, most pa-
tients can be managed by 5 cm H
2
0 CPAP via
mask.'
48
CPAP by mask is effective because in
many patients it is the underlying contused lung
that is primarily responsible for the respiratory dif-
ficulty. Control of pain by epidural narcotics is an
important component of effective treatment of this
situation. Other much less frequently used treat-
ments consist of external stabilization of the flail
segment by traction on the injured segment'
80
and
stabilization of the chest wall by intramedullary
pinning of fractured ribs'
8
'; indications for these
procedures are gross instability of a large segment
of the chest wall and unremitting pain caused by
the movement of the rib fracture.'
82
a. HEMOTHORAX
About 500 ml of blood must accumulate in the
chest cavity before it is detectable radiologically.
Upright chest films are preferable, if possible.
Tube thoracostomy allows immediate confirmation
of the diagnosis. Hypovolemia from blood loss is
the most common presenting problem in patients
with significant hemothorax. Therefore, the im-
mediate treatment should be directed toward re-
storing blood volume.
Thoracostomy tube insertion (usually in the
sixth intercostal space in the midaxillary line)
alone is the only surgical treatment required in
more than 80 per cent of patients with hemotho-
rax.'
83
In the majority of cases, the source of bleed-
ing is the pulmonary vessels, which normally have
low perfusion pressures. Although chest tube
drainage is the first step in the treatment of hemo-
thorax, one must remember that chest tube drain-
age can occasionally release a major vessel tam-
ponade and cause rapid exsanguination. Patients
who have responded adequately to volume re-
placement, have less than 150 ml/hour thoracos-
tomy tube output after initial drainage, or have re-
expanded the injured lung on chest X-ray without
reaccumulation of hemothorax do not require
emergency thoracotomy. In contrast, bleeding
from systemic vessels is usually much more per-
sistent and voluminous and usually requires thora-
cotomy (for blood loss greater than 300 ml/hour
for 4 hours).
Sometimes residual clot remains in the thorax
following tube thoracostomy drainage of a hemo-
thorax. The residual clotted blood remaining in the
pleural cavity has the potential for causing em-
pyema or pleural fibrosis with lung entrapment.
Consequently, some authors recommended early
thoracotomy to evacuate the residual blood and
decorticate any early fibrosis process.
127
b. PNEUMOTHORAX
Pneumothoraces of less than 20 per cent are
usually not detectable clinically. Patients with
larger pneumothoraces (20 to 40 per cent) usually
complain of chest pain, which is accentuated by
deep breathing. With even larger pneumothoraces
(40 to 60 per cent), cyanosis may be evident and
the trachea may be deviated. The presence of rib
fractures or tissue emphysema should suggest the
diagnosis. Radiologic examination is the best di-
agnostic aid available, and all films should be
taken during expiration.
A simple closed pneumothorax (greater than 10
per cent on chest X-ray) requires only a chest tube;
further surgical intervention is not required unless
there are underlying injuries. Minor-sized penetrat-
ing defects can be sealed with Vaseline gauze,
dressing, and tape and with chest tube evacuation
of the pleural space. A suspected tension pneumo-
thorax (by clinical signs or chest X-ray) requires
immediate decompression; this should first be
done by insertion of a needle in the second inter-
costal space in the midclavicular line. A tension
pneumothorax is a dire emergency that may get
worse, and valuable time should not be wasted
seeking radiologic confirmation. If the diagnosis
was correct, rapid improvement will be observed.
The needle should be followed by chest tube inser-
tion. Large-sized open defects often require de-
bridement and primary closure.
It is especially important to decompress even a
very small pneumothorax if the patient is going to
be transported by air. Most commercial aircraft are
pressurized to 5000 to 6000 feet. At these alti-
tudes, Boyle's law (the volume of a gas is in-
versely proportional to pressure) becomes impor-
tant because gases will expand 20 to 25 per cent
at this altitude compared with their volume at sea
level. Decompression of existing pneumothoraces
is extremely important. Consideration should be
given to the effect of trapped gases in other areas
of the body (e.g., pass a nasogastric tube) or in
medical devices.
c. PULMONARY CONTUSION
There are no special surgical considerations for
pulmonary contusions (they are not resected). The
contusion must be watched for the possible devel-
opment of infection and abscess formation within
the contusion.
A persistent pneumothorax and air leak after
tube thoracostomy and the presence of subcuta-
neous and mediastinal emphysema are suggestive
of a tracheobronchial tree disruption. The few
cases that can be treated nonsurgically include
small distal tears with minimal air leak, those in a
major bronchus that are less than one third of the
circumference and have no air leak, and small tra-
cheal wounds with good apposition of the margins.
Small- to moderate-sized high cervical tracheal
wounds may be treated by endotracheal intubation,
with the cuff of the endotracheal tube placed be-
low the wound site.
184
Tracheostomy is indicated
when there is extensive injury to the cervical tra-
chea and larynx or when endotracheal intubation
cannot be performed.
Surgery is indicated for most tracheobronchial
disruptions. Surgery should be preceded, if possi-
ble, by diagnostic bronchoscopy to identify the site
and severity of lesion. The surgical approach is a
right thoracotomy for right-sided and tracheal le-
sions and a left-sided thoracotomy for left-sided
lesions. The procedures consist of primary suture
repair if technically feasible and, if not, lung resec-
tion.
167
The long-term sequela of these injuries is
airway stenosis.
(see Chapter 16)
All traumatic (caused by external causes), mi-
totic, and spontaneous esophageal tears must be
treated surgically as soon as they are diagnosed;
any delay results in a sharp increase in mortal-
ity. Because there are false-negative results both
with esophagography and esophagoscopy, surgical
exploration is sometimes undertaken empirically
just on the basis of a very high index of sus-
picion.
174
185
Surgical procedures range greatly in aggressive-
ness and consist of primary repair and drainage
(adjacent wall tissue normal), pure drainage pro-
cedure (adjacent wall tissue injured), esophageal
exclusion, and esophagogastrectomies
173
(see
chapter 16). Tears of the upper and middle thirds
of the esophagus are repaired through a right tho-
racotomy, and tears of the lower third are repaired
through a left thoracotomy (see chapter 16). Un-
fortunately, extensive esophageal injuries often
leak after repair and result in the late complica-
tions of tracheoesophageal fistula, mediastinal ab-
scess, wound infection, and carotid artery blow-
out.
166
Patients with a small, well-contained perforation
due to a medical instrument (internal trauma) can
be treated nonoperatively with antibiotics and in-
travenous alimentation.
173
Nonoperative manage-
ment is successful in these cases because the per-
forations are small, and frequently patients
undergoing medical instrumentation have esopha-
geal disease with periesophageal inflammation and
fibrosis, which limit the spread of the mediastinal
contamination.
Patients with massive herniation of intra-ab-
dominal contents into the thorax with obvious em-
barrassment of pulmonary function ("tension en-
terothorax"
186
) require immediate operative
intervention. If the diaphragmatic injury is not life-
threatening, repair can be undertaken after diag-
nostic confirmation by chest X-ray (observation of
elevated diaphragm, hollow viscus in the chest
["loculated pneumothorax"], obscured diaphrag-
matic shadow, and a pneumothorax or hemothorax
when the entry wound is abdominal), diagnostic
pneumoperitoneum (which causes a pneumo-
thorax), and hemodynamic stabilization. Similarly,
abdominal findings with a chest entry wound in-
dicate a diaphragmatic tear.
Finally, if a chest tube is in place, appearance
of peritoneal lavage fluid in the chest tube drainage
and a nasogastric tube or radiocontrast material
that curls back up into the chest after entering the
abdomen are virtually diagnostic of a diaphrag-
matic tear. Nevertheless, in both acute and sub-
acute situations, diaphragmatic tears are still most
often discovered at the time of surgical exploration
of associated injuries.
178
I86
However, undue delay
is associated with an increased incidence of organ
strangulation and perforation. In fact, patients with
a chronic diaphragmatic tear most often have
symptoms of intestinal obstruction.
The surgical approach is most often abdominal
(the source of associated injuries and bleeding
(which is the reason a body cavity is being ex-
plored) is usually in the abdomen
178
l86
but may be
thoracic, or both, depending on the nature of the
associated or concomitant injuries.
178 ISf
^
188
D. Specific Chest (Noncardiac) Injuries:
Anesthetic Considerations
The anesthetic management of patients with
thoracic trauma is complicated by the fact that
there are often multiple, life-threatening, interre-
lated problems. In one large series of acute, blunt,
thoracic trauma patients, three major prob-
lems/concerns were very apparent.'
41
First, vir-
tually all intraoperative deaths are due to exsan-
guination. Second, many of the injuries indicated
that the patients should be intubated awake (e.g.,
there was a possible cervical spine injury, facial
trauma, severe shock). Third, increased airway
pressure was a common intraoperative problem
and, in order of decreasing frequency, the in-
creased airway pressure was due to hemorrhage
into the tracheobronchial tree, pulmonary edema,
intraoperative pneumothorax, lung collapse, and
aspiration.
In the vast majority of these patients, chest tubes
should be inserted before the initiation of positive-
pressure ventilation. Finally, patients who have a
myocardial contusion are at high risk for life-
threatening arrhythmias and hemodynamic insta-
bility unrelated to bleeding (6 per cent inci-
dence).
189
The anesthesiologist must do many things very
quickly when beginning to administer an anes-
thetic to a patient with extensive thoracic trauma,
and these tasks must be done according to priority:
(1) ventilation; (2) measurement of blood pressure;
(3) attachment of an EKG; (4) sorting out intrave-
nous catheters and finding one good one for drugs;
(5) inserting blood warmers in the intravenous
catheter lines; (6) getting blood into the room,
verifying patient identity, and beginning transfu-
sion; (7) starting an arterial catheter; (8) drawing
blood gases and a hematocrit; (9) titrating the an-
esthetic if possible; (10) checking temperature;
(11) measuring urine output; and (12) inserting a
central venus or pulmonary arterial catheter. This
last procedure should be done only after initial
resuscitation has been completed or when time
permits.
The choice between crystalloid and colloid in-
fusion is discussed at length in chapter 13. Suc-
cessful fluid resuscitation is indicated by a de-
creasing pulse rate (below 100 beats/min), a pulse
pressure of greater than 30 mm Hg, adequate urine
output (greater than 1 ml/kg/hour), minimal meta-
bolic acidosis, and the lack of a large respiratory
swing in arterial pressure. Other intraoperative
concerns in extensively traumatized patients are
persistent hypotension in spite of apparently ade-
quate volume replacement (which may be due to
an occult blood loss, air embolism, pericardial
tamponade, myocardial contusion, head injury,
and so on), hypothermia, acidosis, and coagulopa-
thy (see chapter 13).

Patients with chest wall injuries that require sur-
gical repair also require general anesthesia, intu-
bation, and mechanical ventilation. Analgesia for
continued mechanical ventilation is most effec-
tively provided by continuous lumbar or thoracic
epidural narcotics and can be a major determinant
of morbidity and mortality (see chapter 21).
a. HEMOTHORAX
Intravascular volume should be replaced
through large-bore intravenous lines. Preopera-
tively, the potential for sudden massive blood loss
with chest tube insertion should be remembered.
Intraoperatively, the use of an autotransfuser
should be strongly considered. Acute respiratory
failure may occur prior to surgery and may require
intubation and mechanical ventilation. A double-
lumen tube should be considered if there is a large
concomitant air leak out the chest tube (raising the
possibility of a tracheobronchial tree disruption) or
if there is hemoptysis or a significant amount of
blood in the airway. Other specific anesthetic con-
siderations depend upon associated injuries and the
specific site of bleeding (which may not be diag-
nosed until later; for example, following artrio-
graphie demonstration of an aortic disruption).
b. PNEUMOTHORAX
The only surgical response to a pneumothorax
that requires general anesthesia is debridement and
primary closure of a large open pneumothorax. Of
course, if a pneumothorax is associated with other
injuries, general anesthesia will likely be required
for the treatment of the other injuries. In all cases
involving a potential pneumothorax, the anesthe-
siologist must consider the possibility of convert-
ing a small, untreated, simple, closed pneumo-
thorax to a large tension pneumothorax during the
induction of anesthesia and the initiation of inter-
mittent positive-pressure breathing. If chest tubes
have been inserted, they must be monitored for
continued function (i.e., if they fail to function
properly, tension pneumothorax can still occur).
The signs of the development of a tension pneu-
mothorax consist of decreased breath sounds, de-
creased compliance, deteriorating arterial blood-
gas values, tracheal deviation, and cardiovascular
collapse. As soon as a tension pneumothorax is
suspected, a large-bore needle should be inserted
into the pleural cavity, allowing air to drain freely.
Nitrous oxide should be avoided if a pneumo-
thorax is a possibility.
190
C. PULMONARY CONTUSION
It is important to remember that the edema
phase of a pulmonary contusion may coincide with
the intraoperative period when fluid is necessarily
administered. The edema phase will be accompa-
nied by a progressive decrease in P
a
0
2
and com-
pliance and should be treated with PEEP and per-
haps fluid restriction and diuretics. The type of
fluid infused (colloid versus crystalloid) is not a
critical matter since the permeability characteris-
tics of the involved areas are grossly deranged and
the area will become edematous regardless of the
type of fluid infused.
Whether the patient is intubated awake or anes-
thetized, with or without spontaneous ventilation,
using a single-lumen tube versus a double-lumen
tube, inspection of the trachea with a fiberoptic
bronchoscope and passage of the endotracheal
tube over the direct vision fiberoptic bronchoscope
guide are good ideas. In cases of complete tracheal
transection, blind' endotracheal intubation (nonvi-
sualization of the trachea per se) may be hazardous
because it may be difficult to slide by the tear and
intubate the distal trachea in this manner. In addi-
tion, partial tracheal injuries can be converted to
full tears by such an intubation.
Separation of the lungs is often required and is
often life-saving.
161
Because positive-pressure ven-
tilation may convert a simple mucosal tear to a
major BPF or pneumothorax, consideration should
be given to maintaining spontaneous ventilation
during the induction of anesthesia, endotracheal
intubation, and the maintenance of anesthesia if a
single-lumen tube is used. Double-lumen tubes
should be used for injuries located at or below the
carina. Alternatively, small catheters can be passed
below injuries near the carina for HFV and high-
flow apneic ventilation (see chapter 12). Injuries
above the carina are best handled by passing sin-
gle-lumen or double-lumen tubes past the injury
and ventilating one (or both) of the lungs. With
high cervical tracheal transections, it is advisable
to use a fiberoptic bronchoscope as a stent across
the transection. It should be remembered that tra-
cheostomy under local anesthesia may be the saf-
est way to establish the airway in severely trau-
matized patients, and that gentle positive-pressure
ventilation will be necessary when the chest is
opened. The anesthesiologist should have sterile
endotracheal tubes of various sizes available for
bronchial placement from within the chest during
airway repair.
(see Chapter 16)
Anesthesia for esophageal surgery involves res-
piratory, hemodynamic, and gastrointestinal con-
siderations (see chapter 16). Most important, sur-
gical exposure can be greatly facilitated by double-
lumen tube insertion and use of one-lung ventila-
tion. In patients with esophageal disruption, fur-
ther esophageal instrumentation, such as insertion
of esophageal stethoscope and nasogastric tube, is
contraindicated. At the end of the case, the surgeon
may gently guide a nasogastric tube or esophageal
stent past the injured area.
Patients, whether having laparotomy or thora-
cotomy as the initial approach, may be prepared
and draped into the "corkscrew" position to allow
exploration of the second body cavity if appropri-
ate.
191
The pelvis lies flat on the operating table,
but the thorax and shoulders are rotated, and the
uppermost arm is brought across the table on a
support. Optimal access to both the chest and ab-
domen is afforded by appropriate lateral tilt of the
operating table.
If the injury is approached by thoracotomy, then
surgical exposure can be greatly facilitated by dou-
ble-lumen tube insertion and use of one-lung ven-
tilation. Decompression of the stomach may signif-
icantly improve the hemodynamic and ventilatory
status. The approach to weaning and extubation
should be conservative in view of the fact that
chest and diaphragm trauma can induce a marked
decrease in diaphragmatic displacement on the in-
jured side.
192
VII. TRANSVENOUS PULMONARY
EMBOLECTOMY
Greenfield and Stewart classified patients suffer-
ing from pulmonary embolism into two groups
based on physiologic parameters (Table 17-15).
193
The first consists of class I, II and III patients who
can be managed medically. The second group in-
cludes classes IV and V who have either a massive
pulmonary embolus or a pulmonary embolus as-
sociated with pulmonary or myocardial disease
194
and demonstrate a very high morbidity and mor-
tality (50 per cent dead in 30 min, 70 per cent in 1
hour, 85 per cent by 6 hours).
195
196
Several methods of treatment for acute pulmo-
nary embolism have been described, including
heparin therapy, streptokinase, and surgery (open
pulmonary embolectomy). In class IV and V pa-
tients, medical therapy has little to offer because
of the associated high early mortality. Those pa-
tients who are treated in the acute situation with
open embolectomy continue to demonstrate a sig-
nificant morbidity and mortality (approaching 50
per cent).
197
As a result, the technique of transve-
nous pulmonary embolectomy has been advocated
by Stewart and Greenfield.
198
Indications for transvenous pulmonary embolec-
tomy are similar to those for open pulmonary em-
bolectomy (i.e., class IV and V patients with a
strong clinical suspicion for pulmonary embo-
lism). These patients, if identified at an early stage,
could be saved by rapid pulmonary embolectomy.
The technique of transvenous pulmonary embo-
lectomy has been described by Stewart and Green-
field.
198
The catheter is steerable, measuring 100
cm in length, with a suction cup at the end. It is
inserted through the femoral venotomy to the right
atrium and ventricle to the pulmonary outflow
tract. After artriographie localization of the em-
bolus, the distal cup is placed in close proximity
to the embolus, and suction is applied. The em-
bolus is then extracted via the femoral venotomy
along with the catheter. Frequently, multiple at-
tempts are necessary to remove enough clot to
achieve hemodynamic stability, as evidenced by
recovery from shock and a significant improve-
ment in the P
a
0
2
.
Complications of the procedure include ventric-
ular arrhythmia, inability to extract the embolus,
perforation of the pulmonary artery,
198
and fatal
hemorrhage
197
after reperfusion of the lungs.
At the completion of the pulmonary embolec-
tomy, a Greenfield filter is inserted through the
femoral venotomy. The filter is inserted to protect
against further embolization in an already hemo-
dynamically compromised patient. The filter is in-
serted via the femoral or jugular route over a
guidewire using fluoroscopy. It is positioned in the
inferior vena cava at the level of the third lumbar
vertebra. Systemic anticoagulation should be con-
tinued after the conclusion of the procedure be-
cause the filter protects against further emboliza-
tion but does not treat the initiating problem (i.e.,
deep venous thrombosis). Complications of Green-
Table 17-15 CLASSIFICATION OF PULMONARY THROMBOEMBOLISM
1
*From Stewart JP, Greenfield LJ: Transvenous vena cava filtration and pulmonary embolectomy. Surg Clin North Am 62:411-
430, 1982. With permission of the publisher.
Abbreviations: CVP = central venous pressure; PA = mean pulmonary artery pressure; BP = blood pressure; CO = cardiac
output.
field filters include recurrent embolism, retroperi-
toneal hemorrhage, recurrent thrombophlebitis, pe-
ripheral edema, and filter migration.
1
Appropriate monitoring should include arterial
catheter placement for the second-to-second obser-
vation of blood pressure and to obtain arterial
blood gases as indicated. A large-bore peripheral
or central intravenous catheter is indicated for
rapid infusion of volume and/or vasoactive drugs.
Pulmonary artery pressure monitoring is desirable,
and the values are obtained from the angiographic
catheter. A pulmonary artery catheter should be
placed subsequently through the internal jugular
vein (because the insertion of a Greenfield filter is
via the femoral vein) as early as possible after the
embolectomy both to determine pulmonary artery
pressures and to aid in evaluation of hemodynamic
status.
Monitoring should also include an EKG, which
should be able to give a lead II reading. In more
than one third of patients with pulmonary embo-
lism, the EKG is normal, whereas other patients
will show various transient abnormalities of the
ST segment and wave in the precordial leads.'
99
However, tachycardia, both supraventricular and
ventricular, and bradycardia often result from pul-
monary embolus manipulation and therefore a lead
II or transesophageal lead is best elected.
Anesthetic management requires a motionless
patient. If paralysis were used alone, awareness
might result; therefore, some form of sedation or
general anesthesia without cardiovascular depres-
sion is desirable. Pulmonary embolectomy can be
associated with a significant blood loss. Large
amounts of blood loss may not be readily evident
because of inaccessibility to the patient's lower
extremity. This patient will generally be on an
anticoagulant or thrombolytic agent and therefore
may bleed profusely from the cut-down site.
Bleeding secondary to heparin can be treated with
protamine, and bleeding secondary to streptoki-
nase may respond to treatment with epsilon ami-
nocaproic acid.
There appear to be two main periods when
large-volume infusions are necessary. The first is
when the embolus is removed, with reperfusion
and its associated complications. The second is
when the Greenfield filter is being inserted, be-
cause of the surgical incision and bleeding in an
anticoagulated patient.
VIII. EMERGENCY ROOM
THORACOTOMY IN THE
MANAGEMENT OF TRAUMA
The use of emergency room thoracotomy in the
management of trauma has increased greatly in the
last decade. Thoracotomy in the emergency room
is used when the patient's condition is thought to
be so dire that it precludes taking the time neces-
sary to transfer the patient to the operating theater.
Emergency room thoracotomy has been performed
by surgeons as well as emergency room physicians
in a wide variety of clinical conditions. As experi-
ence with the procedure accumulates, it has be-
come apparent that the mode of injury (blunt ver-
sus penetrating trauma), site and extent of injury,
the condition of the patient at the scene of injury
(e.g., presence of vital signs, reactivity of pupils),
and the rapidity of transport are the most important
determinants of the outcome. The procedure is
most useful in penetrating lung injuries (the lung
hilum can be clamped so as to control lung bleed-
ing and prevent pulmonary venous air embolism),
is progressively less useful in penetrating cardiac
injuries (pericardial tamponade can be relieved,
cardiac bleeding can be controlled with digital
pressure and internal defibrillation, and cardiac
massage can be performed) and penetrating ab-
dominal wounds (thoracic aorta can be occluded
to provide proximal control of massive abdominal
arterial bleeding), and is least successful in blunt
thoracoabdominal injuries
200
(Table 17-16).
Although penetrating lung injuries continue to
have the best results (survival) with emergency
thoracotomy, the results of the most recent large
series are not nearly as impressive as those in
Table 17-16, perhaps because patient selection
was very liberal.
201
If the patient is alive at the
scene of injury, the ultimate survival of critically
injured patients will be increased by minimizing
the use of time-wasting field stabilization measures
(e.g., starting many intravenous lines
202
and apply-
ing military antishock trousers
203
) and by trans-
porting the patient as rapidly as possible to the
hospital.
204
Lack of ventricular activity or a palpa-
ble pulse and the absence of reactive pupils at the
scene of injury carry an almost uniformly fatal
prognosis.
201
Patients who are clinically dead at the
scene of injury and who remain so throughout
transport do not benefit from heroic emergency
room thoracotomy.
201,205
In penetrating injuries of the thorax, the lung is
the most frequently injured organ.
206
The majority
of these injuries may be treated with just tube
thoracostomy
206
(see Chest Trauma). However,
Table 17-16 RESULTS OF IMMEDIATE EMERGENCY ROOM THORACOTOMY*
Number of
Number of Cases Thoracotomies Mortality
Type of Injury (Number of Reports) (%) (%)
Penetrating lung 1699 321 9
(7) (19)
Penetrating cardiac 324 All 71
(14)
Penetrating abdominal 194 All 95
(6)
Blunt thoracoabdominal 252 All 96
(7)
*Based on data from Bodai et al.
20
"
with penetrating thoracic battle injuries, the num-
ber of patients requiring thoracotomy is signifi-
cantly increased.
207
Still, of those who require tho-
racotomy for penetrating injury, only a small
fraction ever require thoracotomy in the emer-
gency room. These patients are the ones who ar-
rive in extremis, and thoracotomy performed in the
emergency room may be life-saving
200
2()8
(Table
17-16).
The prehospital mortality of penetrating cardiac
wounds is high (38 to 83 per cent).
200
208
Patients
with penetrating cardiac injuries who survive the
initial injury but show signs of acute decompensa-
tion on or just before arrival in the emergency
room may also benefit from emergency room tho-
racotomy. The majority of these survivors have
cardiac tamponade (as has been demonstrated by
previous immediate thoracotomy experience); sim-
ple pericardiocentesis is recommended only to
transiently improve hemodynamic function, allow-
ing the necessary time for transfer to the operating
room. The more aggressive approach of perform-
ing thoracotomy in the emergency room has re-
sulted in a significant decrease in the mortality due
to this injury
209
210
(Table 17-16).
Thoracotomy in the emergency room has been
used for injured patients with arterial abdominal
bleeding. The rationale for this practice is that
proximal control of the bleeding by thoracic aortic
cross clamping prior to release of the tamponading
effects of blood in the intact abdominal cavity
would increase survival following abdominal de-
compression. However, experience with patients
with penetrating abdominal injuries subjected to
emergency room thoracotomy discloses an ex-
tremely low survival rate
200
208
(see Table 17-16).
The difference between the relatively high success
rate with lung injunes compared with that of ab-
dominal injuries can probably be explained by two
factors. First, emergency thoracotomy and aortic
occlusion will not substantially affect the rate or
volume of bleeding from major abdominal venous
injuries. Such venous injuries include penetrating
injuries to the liver, the vena cava, and major por-
tal venous structures. The second factor is that
many of these patients with penetrating abdominal
injury did not have demonstrable vital signs before
the performance of thoracotomy. The thoracotomy
was undertaken as a "last-ditch," heroic attempt
in the emergency room to save the patient. Thus,
the trauma victim who has suffered a cardiac arrest
from abdominal hemorrhage has a slim chance of
recovery and, in most instances, is not a candidate
for emergency room thoracotomy.
200
208
Only pa-
tients in whom there is a strong suspicion of major
intra-abdominal arterial injury, and who still have
signs of life, are candidates to survive emergency
room thoracotomy. The alive but severely hypo-
volemic patient with penetrating abdominal injury
and massively distended abdomen should go im-
mediately to the operating room, where the sur-
geon may elect to perform a left anterior thoracot-
omy with aortic cross clamping in a thorough
manner before opening the abdomen.
2
"
Patients with blunt injury severe enough to ne-
cessitate emergency room thoracotomy also dem-
onstrate uniformly poor results
200
206 2()8
(see Table
17-16). The reasons for these dismal results prob-
ably include widespread damage, venous bleeding,
and much higher incidence of head trauma.
212
A left anterolateral thoracotomy is most often
performed for aortic cross clamping and left hilar
clamping. The pericardial sac should be inspected
and, if the sac is bulging or dark blue, or if there
are no signs of underlying cardiac activity, it
should be immediately opened. Care should be
taken during the opening of the pericardial sac to
avoid injury to the coronary arteries. Once the
pericardial sac has been opened, clots should be
removed, bleeding should be controlled by digital
pressure, and. if there is no cardiac activity, inter-
nal defibrillation and cardiac massage should be
instituted. Finally, a left thoracotomy provides
good access for descending thoracic aortic occlu-
sion to provide proximal control of massive ab-
dominal arterial injuries. A right lateral thoracot-
omy will be performed only for right-sided lesions
(i.e., suspected right hilar damage).
The patient must be intubated immediately and
ventilated with 100 per cent oxygen. If a double-
lumen tube is not available or cannot be easily
inserted by the available personnel or if a single-
lumen tube is in situ, selective ventilation of the
right lung may be reliably achieved by advancing
the single-lumen tube in 1-cm steps until left lung
breath sounds disappear (mean standard devia-
tion of length of single-lumen tube required to
achieve a right main-stem bronchial intubation is
30 1.1 cm).
213
Unfortunately, the right upper
lobe may be expected to be obstructed nearly 100
per cent of the time by this technique.
Multiple large-bore intravenous catheters must
be inserted, and intravascular volume repletion
should be begun immediately. As soon as a periph-
eral pulse can be palpated, an arterial line should
be inserted for pressure measurement and blood
sampling. As intravascular volume is being re-
pleted, a central venous catheter should be inserted
to measure central venous pressure. Paralysis
should be instituted if the patient is moving at all.
No, or extremely small doses of, intravenous an-
esthetics should be administered. A Foley catheter
should be inserted as soon as possible. The pa-
tient's blood should be typed and cross matched
for appropriate number of units of blood. Use of
the autotransfuser should be considered.
IX. REMOVAL OF
FOREIGN BODIES
The majority of foreign bodies are inhaled by
children younger than 3 years.
214
The lack of mo-
lars at an early age enhances potential aspiration
(e.g., molars grind food into small particles that
are easily swallowed with the saliva stream).
214
In
addition, adults who have an acutely (alcoholism)
or chronically (dementia) depressed sensorium
also are at risk for foreign body inhalation. Acute
total upper airway obstruction is the immediate
hazard, but far more commonly the foreign body
is inhaled deeper into the tracheobronchial tree,
where it impacts and initiates a local inflammatory
reaction. The subsequent local inflammatory reac-
tion holds the foreign body more securely in place
(perhaps preventing removal).
215
Without removal,
the foreign body ultimately causes distal collapse
and infection.
Organic material is more dangerous than inor-
ganic material because it swells after inhalation
and may fragment, even before attempts are made
to remove it. Peanuts are particularly liable to
swell and fragment, and they liberate an irritant oil
that causes severe local inflammation. Sweets are
also a special problem because they dissolve in the
tracheobronchial secretions, forming a viscous hy-
pertonic solution that predisposes to obstruction
and collapse. Sharp foreign bodies, such as bones,
needles, pins, and glass may also be difficult to
remove if they stick into the wall of the tracheo-
bronchial tree. In one large series of inhaled for-
eign bodies, 66 per cent were peanuts, 16 per cent
were vegetable derivatives, and 17 per cent were
inert objects.
214
The majority of foreign bodies
lodge in the right lung because of the disposition
of the right main-stem bronchus relative to the
trachea, but approximately 20 to 44 per cent can
be found on the left side.
214-216
The history remains the most reliable indicator
of aspiration of a foreign body. Paroxysmal cough
and wheeze are the common symptoms in chil-
dren, but dyspnea, stridor, fever, and vomiting also
occur. Superglottic foreign bodies often cause gur-
gling, drooling, and inspiratory stridor; tracheal
foreign bodies cause both inspiratory and expira-
tory stridor; and subcarinal foreign bodies cause
wheezing. On auscultation, a very high percentage
of patients with subcarinal foreign bodies will
have decreased breaths sounds and a wheeze in the
affected area
217
and often the wheeze is general-
ized.
218
In fact, administration of theophylline, epi-
nephrine, and steroids will decrease wheezing in
the uninvolved lung but not in the involved lung;
this differential lung response has been used as a
diagnostic test in cases in which the differential
diagnosis is between bronchial foreign body and
asthma.
218
Table 17-17 shows the differential diagnosis
(based on history and physical examination) of the
typical presentations of the major causes of airway
obstruction in children. A chest infection that fails
to clear in spite of antibiotic treatment in an oth-
erwise healthy child warrants chest radiography
and diagnostic bronchoscopy to exclude the pres-
ence of an inhaled foreign body. This is particu-
larly true in patients with recurrent one-sided in-
fection and/or persistent unilateral wheezing
649
without prior evidence of underlying allergic di-
athesis. Unfortunately, a specific history of inha-
lation is unavailable in approximately 20 per cent
of all patients with this problem; these patients
consist of some children, adults after alcoholic ex-
cess, adults with dementia, and patients under gen-
eral anesthesia having dental surgery without en-
dotracheal intubation.
216
In approximately 70 to 80 per cent of the pa-
tients, the initial chest roentgenogram will be pos-
itive and show the foreign body, distal atelectasis,
air trapping (foreign body acts as a one-way ball
valve in a bronchus), or mediastinal shift.
214-216
In
those with initially normal chest radiographs, ap-
proximately half will develop very suggestive
"positive" X-ray findings (air trapping, atelec-
tasis, infiltrates, the foreign body itself) before the
removal of the foreign body. However, the foreign
body will not be directly visible most of the time.
Thus, approximately one eighth of patients will
come to bronchoscopy on the basis of the history
and physical examination alone and have a normal
chest roentgenogram. If fluoroscopy is added to
routine radiographic techniques, a much higher
yield of positive findings on initial radiologic ex-
amination is obtained.
216
Positive fluoroscopic
findings include a mediastinal shift away from the
affected side consistent with air trapping on expi-
ration (i.e., the foreign body acts as a one-way ball
valve).
214
In a small percentage of patients, the foreign
body may change location between the time of
admission to the hospital and bronchoscopy.
215
Be-
cause foreign bodies may wander, it is essential to
obtain a chest X-ray just before bronchoscopy.
215
In summary, a history compatible with foreign
body aspiration dictates diagnostic and therapeutic
endoscopy with or without radiologic confirma-
tion. To decrease the approximately one fourth
failure rate of plain films in the first 24 hours,
fluoroscopy must be strongly considered as an in-
itial diagnostic technique in foreign body evalua-
tion. In many instances, a 24-hour interval, a
safety zone, can be observed prior to endoscopy.
The safety zone ensures adequate gastric emptying
and unhurried and thorough preparation for the
procedure. During this period of time, postural
drainage, chest percussion and vibration, and ad-
ministration of bronchial dilators (including corti-
costeroids) may be used; in an occasional patient
these measures may result in a spontaneous extru-
sion of the foreign body. However, these maneu-
vers should not be regarded as a substitution for
endoscopic removal
219
and, on rare occasions, have
caused increased airway obstruction and cardiac
arrest.
22a
-
222
If a foreign body is not readily removed at the
first endoscopic procedure, chest physical therapy
and bronchodilatation should be resumed. Vegeta-
ble foreign bodies, including nuts, show potential
for self-extrusion, particularly postbronchos-
copy.
215
A spontaneously extruded foreign body
always has the potential for causing a major (prox-
imal) obstruction of the tracheobronchial tree,
causing hypoxia.
2
'
5
In an effort to avoid thoracot-
omy, repeated endoscopy should be used when
clinically indicated. However, bronchotomy, or
even lobectomy, may be necessary if the foreign
body becomes deeply embedded in a mass of in-
flammatory tissue or erodes through the wall of a
bronchus.
215
As discussed under Laser Resection of Major
Airway Obstructing Tumors (chapter 15), a rigid
bronchoscope was considered preferable to a fiber-
optic bronchoscope because it allowed better con-
trol of secretions and blood, permitted better re-
moval of larger pieces of necrotic material, and
provided a better view. Similarly, foreign bodies
to the tracheobronchial tree should be removed
with a rigid bronchoscope because it allows pas-
sage of much larger instruments and foreign bod-
ies, and the open-ended rigid instrument also
makes ventilation easier in small children.
215
223,224
In fact, it may be impossible for a small patient to
breathe around a fiberoptic bronchoscope because
even the narrowest fiberoptic bronchoscope (1.8-
mm outside diameter) takes up most of the area of
the trachea. Furthermore, and of special relevance
to small patients, rigid bronchoscopes range in
nominal size from 2.5 to nearly 10 mm. The nom-
inal size of a rigid bronchoscope refers to the
smallest internal diameter through which instru-
ments may be passed, not to the outer diameter,
which may be several millimeters greater. For ex-
ample, the 3.5-mm flexible pediatric bronchoscope
will pass through a 3.5-mm Storz rigid broncho-
scope.
A large foreign body cannot be removed
through a small bronchoscope and must, therefore,
be grasped with forceps and extracted while the
bronchoscope is withdrawn simultaneously. Re-
peated instrumentation may be necessary, and
even in skilled hands this will traumatize the vocal
cords and upper respiratory tract. An alternative
solution that has been reported is to pass a Fogarty
embolectomy catheter beyond the foreign body;
the balloon is inflated and the catheter is with-
drawn so that the foreign body is trapped against
the tip of the bronchoscope and the entire assem-
bly is removed together.
217,225
If thoracotomy for bronchotomy is required, the
usual method of localizing the foreign body is by
palpating the object through the firm bronchial
wall. However, the palpation may not be positive
or precise. During thoracotomy, a fiberoptic bron-
choscope can be a useful guide for the rapid and
precise determination of the site of bronchot-
omy.
223
The fiberoptic bronchoscope finds the ob-
ject and shines its light on it.
223
Dimming the lights
of the operating room is helpful in order to better
visualize the light at the end of the bronchoscope.
A small well-placed bronchotomy will predictably
result in lower morbidity. This technique should
be considered for all patients undergoing open re-
moval of a foreign body or bronchial tumor in
whom the precise placement of the bronchotomy
is in question.
(Table 17-18)
If a child has nondistressed respiration and
crying does not cause an embarrassment of respi-
ration, the child can be heavily premedicated intra-
muscularly and brought to the operating room for
a smooth inhalation induction. If the child has
respiratory distress and/or crying causes wheezing
and/or cyanosis, premedication should be with-
held. In this type of patient, anesthesia should be
induced with intramuscular ketamine in the oper-
ating room. Although an intramuscular injection
may cause crying, it is only for a short period of
time, especially if the child is held and comforted.
Since bronchoscopy involves intense airway stim-
ulation, atropine or glycopyrrolate should be in-
cluded in the intramuscular injection to decrease
vagal reactions and secretions. As soon as the
child appears sedated, the induction of anesthesia
can be continued by inhalation, and intravenous
access can be established as quickly as possible.
When adequately anesthetized, the patient
undergoes a laryngoscopy, and the larynx and tra-
cheobronchial tree are sprayed with 4 per cent
lidocaine. If the response to laryngoscopy and
spraying is minimal, the airway may be released
to the endoscopist. The eyes are taped shut, and
soft pads are placed over the globes. The patient is
then positioned for endoscopy. Although succinyl-
choline is kept available, we prefer spontaneous
respiration until at least the nature and location of
the foreign body are known; avoidance of relax-
ants requires the anesthesiologist to achieve ade-
quate levels of anesthesia so that the insertion of
the bronchoscope does not cause laryngospasm,
bronchospasm, and chest wall rigidity as a result
of energetic active exhalation. Spontaneous venti-
lation may also be protective of not forcing the
Table 17-18 IMPORTANT ANESTHETIC
CONSIDERATIONS WITH
FOREIGN BODIES
1. Minimize preanesthetic crying-induced respiratory distress
2. Administer atropine prior to passage of bronchoscope
3. Provide depth of anesthesia to prevent reaction to
bronchoscope
4. Administer lidocaine to decrease reaction to bronchoscope
5. Try to maintain spontaneous ventilation
6. Paralyze if excessive reaction to bronchoscope (straining,
bucking, laryngospasm, bronchospasm)
7. Beware of mechanical problems when the foreign body is
being removed (lodge in trachea, fragment)
8. Consider endotracheal intubation and/or administration of
steroids postoperatively if bronchoscope was passed several
times
9. Some cooperative adults can undergo bronchoscopy under
local anesthesia
foreign body deeper into the bronchial tree.
217
In-
travenous lidocaine may diminish any coughing or
straining reaction to the insertion of the rigid bron-
choscope. However, the inability to quickly re-
verse any of these aforementioned problems at any
time during the case may require the use of relax-
ants (small succinylcholine bolus, succinylcholine
drip, atracurium, vecuronium).
Because these cases may end suddenly, the use
of nondepolarizing relaxants may be problematic.
Controlled ventilation without relaxants is an ac-
ceptable middle ground. On occasion, major air-
way obstruction may be relieved by adding helium
to the inspired gas.
226
During the actual bronchos-
copy, one overhead spotlight is directed onto the
anesthesia machine, and another is directed onto
the patient's feet to assess perfusion and color
without interfering with the surgeon's vision.
Following passage of the rigid bronchoscope
and during removal of the foreign body, the anes-
thesiologist should be aware of several important
mechanical problems related to the physical char-
acteristics of the foreign body (softness, smooth-
ness, size).
227
First, beans and nuts can fragment
and obstruct both main-stem bronchi. If this is
thought to be a possibility, removal of the foreign
body with the patient in the lateral decubitus posi-
tion (with the foreign body in a dependent posi-
tion) should be considered; if this is not practical,
a thoracic surgeon and thoracotomy instruments
need to be available. Second, plastic and smooth
inert objects (beads) may be hard to grasp, and
during removal they can slip out of the grasp of
the forceps and block larger and more proximal
airways. If a large airway does become occluded,
the object should be pushed back to its previous
location to re-establish ventilation and reassess the
situation. Emphasis is necessary on previous loca-
tion because inflammation in new areas of the lung
may occur quickly following relocation. Third, if
vocal cord motion during extraction causes signif-
icant obstruction, succinylcholine should be ad-
ministered. Fourth, the foreign body may some-
times be larger than the bronchoscope, and the
bronchoscope and the foreign body may need to
be removed at the same time. Fifth, infection may
already be well established, and vigilance in suc-
tioning infected secretions is important; rarely, an
abscess may be present distal to the foreign body.
In all of these situations, if the bronchoscope is
withdrawn and the foreign body is not seen, the
pharynx should be quickly inspected. If the object
is not visible, the bronchoscope should be quickly
reintroduced to make certain that the object is not
occluding the trachea because the narrowing of the
airway at the glottis makes this a most probable
location. The advantage of spontaneous respiration
becomes manifest during this situation because tra-
cheal obstruction is instantly obvious (tracheal tug,
intercostal retractions). If the trachea is occluded,
the object is quickly pushed back to its previous
location to re-establish ventilation, and the situa-
tion is reassessed.
If positive-pressure ventilation is used, it should
be possible to maintain adequate gas exchange
without excessive inflation pressures, even when a
small bronchoscope is in use. However, extreme
care must be taken to avoid overinflation of the
lungs when either an instrument or the foreign
body is blocking the bronchoscope, and ventilation
may need to be interrupted from time to time to
prevent the foreign body or its fragments from
being blown deeper into the respiratory tract when
they are loose within the bronchial lumen.
It should be remembered that some cooperative
adult patients can have the foreign body removed
under local anesthesia.
215
The patient should be supervised closely until
fully conscious. Laryngeal stridor is relatively
common, particularly after a difficult and pro-
longed procedure (with many passages of the
bronchoscope through the vocal cords) in a small
child. Consequently, use of steroids and broncho-
dilators in the postoperative period is common. In
addition, after a traumatic endoscopy, nasotracheal
intubation is indicated with a tube 0.5 mm smaller
than indicated by age; this is left in place until
chest X-ray and "air leak" both confirm that
swelling is decreased. Chest physiotherapy should
be used in patients who have postoperative atelec-
tasis and in patients with pneumonia who require
antibiotics. The patient is then extubated and may
be placed in a croup tent for up to 24 hours, if
necessary. Racemic epinephrine may also be use-
ful after extubation.
228
Patients who are extubated
in the operating room should breathe humidified
oxygen by face mask or croup tent postopera-
tively.
REFERENCES
1. Garzon AA, Gourin A: Surgical management of massive
hemoptysis. Ann Surg 187:267-271, 1978.
2. Yeoh CB, Hubaytar RT, Ford JM, et al: Treatment of
massive hemorrhage in pulmonary tuberculosis. J Thorac
3. Stern RC, Wood RE, Boat TF, et al: Treatment and prog-
nosis of massive hemoptysis in cystic fibrosis. Am Rev
Respir Dis 117:825-828, 1978.
4. Ehrenhaft JL, Taber RE: Management of massive hemop-
tysis not due to pulmonary tuberculosis or neoplasm. J
5. Crocco JA, Rooney JJ, Fankushen DS, et al: Massive
hemoptysis. Arch Intern Med 121:495-^98, 1968.
6. Gourin A, Garzon AA: Operative treatment of massive
hemoptysis. Ann Thorac Surg 18:52-60, 1974.
7. Conlan AA, Hurwitz SS: Management of massive hae-
moptysis with the rigid bronchoscope and cold saline
lavage. Thorax 35:901-904, 1980.
CHAPTER 18
Anesthesia for Pediatric
Thoracic Surgery
I. Introduction
II. Special Problems Related to
Premature and Newborn Infants
A. Persistence of the Fetal Circulation
B. Respiratory Distress Syndrome
C. Retinopathy of Prematurity
D. Periodic Breathing and Apnea
E. Thermoregulation
F. Vitamin, Caloric, Electrolyte, and
Fluid Requirements
G. Airway Anatomy
III. Congenital Diaphragmatic Hernia
IV. Esophageal Atresia and
Tracheoesophageal Fistula
V. Ligation of Patent Ductus
Arteriosus in Premature Infants
VI. Vascular Rings
VII. Congenital Parenchymal Lesions:
Lobar Emphysema, Cysts,
Sequestrations, and Cystic
Adenomatoid Malformations
1. Congenital Lobar Emphysema
2. Congenital Bronchogenic Cysts
3. Pulmonary Sequestration
4. Cystic Adenomatoid
Malformations
Malformations
Malformations
VIII. Thoracic Surgical Procedures
Requiring One-Lung Ventilation
A. Bronchial Blocker
B. Main-Stem Bronchial Intubation
C. Bronchial Blockade Plus Main-
Stem Bronchial Intubation
D. Prone Positioning
Diagnostic Bronchoscopy
Bronchography
Asphyxiating Thoracic Dystrophy
(Jeune's Syndrome)
Considerations
IX.
X.
XI.
The Fetal Circulation*
I. INTRODUCTION
This chapter has two major sections. The first
section highlights the important physiologic and
anatomic problems of premature infants and neo-
nates that have special relevance to anesthesia for
thoracic surgery (as opposed to a broader discus-
sion of all anesthesia problems for all pediatric
surgery). These problems consist of persistence of
the fetal circulation; the respiratory distress syn-
drome; retrolental fibroplasia; periodic breathing
and apnea; impaired thermoregulation; precarious
fluid, electrolyte, and caloric balance; and altered
airway anatomy. The second section discusses the
major pediatric thoracic surgical cases, which con-
sist of congenital diaphragmatic hernia, tracheo-
esophageal fistula, vascular rings, congenital lobar
emphysema, conditions necessitating bronchos-
copy, and various pulmonary parenchymal prob-
lems requiring lung separation. Anesthesia for re-
section of mediastinal tumors, thymectomy for
myasthenia gravis, and removal of foreign bodies
are procedures that also frequently involve pediat-
ric patients but are discussed in chapters 15 and
17.
II. SPECIAL PROBLEMS RELATED
TO PREMATURE AND NEWBORN
INFANTS
The unique anesthetic problems with the pedi-
atric patient occur primarily in the neonatal period
and with the premature infant. These patients may
have persistence of the fetal circulation, the respi-
ratory distress syndrome, greatly increased sensi-
tivity to increased arterial oxygen tension, periodic
breathing and apneic spells, difficulty in regulating
temperature and the internal chemical and caloric
status, and an airway anatomy different from that
of the adult.
A. Persistence of the Fetal Circulation
In the fetal circulation well-oxygenated blood
flows from the umbilical vein through the fetal
liver and the ductus venosus to the inferior vena
cava. The mixing of well-oxygenated umbilical
vein blood with poorly oxygenated inferior vena
cava blood results in an inferior vena cava oxygen
tension of about 35 to 40 mm Hg (Fig. 18-1).
Inferior vena cava blood then flows into the right
atrium. Since fetal right atrial pressure is greater
than left atrial pressure, the "trap door" flap of
the foramen ovale in the atrial septum remains
constantly open prior to birth; thus, the foramen
ovale directs most of the relatively oxygen-rich
inferior vena cava blood across to the left atrium.
The relatively well-oxygenated blood that enters
the left atrium from the right atrium goes into the
left ventricle, is then ejected into the aorta, and
perfuses the coronary arteries, the aortic arch, and
the brain. Reversal of the right-left atrial pressure
gradient after birth (caused by a marked decrease
in pulmonary vascular resistance after initiation of
breathing; Figure 18-2) allows functional closure
of the foramen ovale, even though it may not ana-
tomically close for many years.
Venous blood from the superior vena cava with
a relatively low oxygen tension enters the right
atrium, mixes minimally with blood from the in-
ferior vena cava, and crosses the tricuspid valve
into the right ventricle. Right ventricular ejection
of this low-oxygen-tension blood into the pulmo-
nary circulation causes pulmonary vasoconstric-
tion and elevation of pulmonary vascular resis-
tance. The blood in the pulmonary artery
subsequently follows the path of least resistance
and passes from the pulmonary artery into the
aorta via the ductus arteriosus to perfuse the lower
portion of the systemic vascular bed. A small
amount of pulmonary circulation blood (10 per
cent) mixes with bronchial artery blood to drain
directly into the left side of the heart.
Persistence of the fetal circulation means that
right-to-left shunting occurs through a patent duc-
tus arteriosus and/or patent foramen ovale. Any-
thing that increases pulmonary vascular resistance
(hypoxia, hypercarbia, acidosis, stress, vasoactive
drugs) and/or decreases systemic vascular resis-
tance (inhalation anesthetics, vasodilators) will in-
crease shunting through the ductus arteriosus.
1
Hy-
poxemia, in particular, may also cause ductal
dilatation. In addition, anything that increases pul-
monary vascular resistance will increase right ven-
tricular and right atrial pressures and increase
Figure 18-1 This schematic diagram shows the main central blood flow pathways in the fetal circulation. Oxygen-rich blood is
delivered from the mother to the fetus via the umbilical vein. The oxygen-rich blood in the umbilical vein mixes with the oxygen-
poor blood draining the lower fetal body in the inferior vena cava and increases the inferior vena cava oxygen tension up to 35 to
40 mm Hg. This now relatively oxygen-rich inferior vena cava blood flow streamlines through the right atrium into the foramen
ovale into the left side of the circulation, where it perfuses the coronary arteries, brain, and upper body. These areas are drained by
the superior vena cava, which contains oxygen-poor blood. Superior vena cava blood flow streamlines through the right atrium into
the right ventricle, which then ejects the oxygen-poor blood into the pulmonary artery. The oxygen-poor blood causes pulmonary
vasoconstriction, which directs the blood flow through the ductus arteriosus to the descending aorta to perfuse the lower body as
well as to return to the mother via the umbilical artery. The dashed arrows represent minor blood flow pathways.
660 Anesthesia for Pediatric Thoracic Surgery
Figure 18-2 Schematic diagram of the phases of the transitional circulation and their relative time course. Note that the time
line is discontinuous. The period before birth represents minutes to hours. In utero, pulmonary vascular resistance (PVR) slowly
decreases in the days to weeks before birth and thus is not represented in this diagram. The immediate phase is due to the initiation
of breathing (mechanical lung expansion and introduction of oxygen into the alveolar space), the fast phase may be related to
prostacyclin production (as well as perhaps bradykinin and angiotensin as well as decreasing levels of leukotrienes),' and the final
phase may be the result of remodeling of the pulmonary circulation. (From Clarke WR: The transitional circulation: Physiology and
anesthetic implications. J Clin Anesth 2:196, 1990. Used with permission.)
right-to-left shunting through the patent foramen
ovale. The shunting through the ductus arteriosus
and foramen ovale results in a vicious cycle of
increased hypoxemia, pulmonary vasoconstriction,
pulmonary vascular resistance, shunting, and hy-
poxemia. Patients with a persistent fetal circulation
who often require surgery and anesthesia are those
undergoing ligation of the ductus arteriosus, repair
of diaphragmatic hernia, and resection of cystic
malformations.
2
It should be remembered that the
foramen ovale is a potential site of a paradoxic air
embolus should air enter the venous circulation,
and for this reason it is essential to exclude all air
from intravenous fluid administration sets.
The ductus arteriosus enters the aorta just distal
to the left subclavian artery; however, there is
enough mixing of ductal and left subclavian blood
flow to make the left arm arterial oxygen tension
significantly less than the preductal arterial oxygen
tension. Consequently, preductal arterial oxygen
tension must be obtained from the right radial ar-
tery and postductal arterial oxygen tension can be
obtained from the umbilical artery. Simultaneous
analysis of the pre- and postductal arterial oxygen
tensions indicates the degree of ductal shunting.
When the preductal arterial oxygen tension is at
least 15 mm Hg higher than the postductal value,
significant ductal shunting is present.
3
Indeed, a
right-to-left shunt of 20 per cent is usually consid-
ered a "normal" amount of ductal shunting.
4
When the arterial oxygen tension in the preductal
samples is below the value predicted for a 20 per
cent shunt (see Fig. 3-33), preductal shunting
(through the lungs and/or patent foramen ovale) is
present. If the amount of preductal shunting is
severe, it becomes impossible to detect ductal
shunting.
Successful treatment of pulmonary vasocon-
striction and ductal and foramen ovale shunting
involves avoidance of the events that can increase
pulmonary vascular resistance (acidosis and hy-
poxemia: Fig. 18-3)
5
and/or decrease systemic
vascular resistance. Metabolic acidosis is treated
with sodium bicarbonate infusions. Endotracheal
tube suctioning is minimized to avoid even tran-
sient changes in the alveolar and arterial oxygen
partial pressure. The most consistently successful
therapeutic modality is the alkalosis achieved with
hyperventilation (if it can be achieved; VC/V may
be so high that the P
a
C0
2
cannot be lowered less
than 40 to 50 mm Hg).
6 7
Frequently, a threshold
effect is observed such that the arterial oxygen
tension does not rise until arterial pH reaches 7.55
to 7.60.
7
Generally, ventilatory rates of 60 to 120
Figure 18-3 Change in pulmonary vascular resistance
(PVR) with oxygenation and hydrogen ion concentration.
There is a sharp increase in PVR when P
a
0
2
falls below 6.7
kPa (50 mm Hg). Note also the synergistic effect when hypox-
emia is accompanied by acidosis. (Reproduced by permission
of Rudolph AM, Yuan S: Response of the pulmonary vascula-
ture to hypoxia and H
+
ion concentration changes. J Clin Invest
45:399-411, 1966.)
breaths/min are necessary to achieve this level of
hyperventilation.
Table 18-1 shows the methodology and com-
mon pitfalls in mechanically ventilating patients
with persistent fetal circulation.
2
To achieve the
goals in Table 18-1 (high respiratory rate and air-
way pressures), the anesthesiologist must manually
ventilate the patient in the operating room because
few ventilators compatible with anesthesia ma-
chines are capable of producing the required rates.
Once the proper level of ventilation is determined
( S
P
0
2
/ PETC0
2
) , it must be maintained. The temp-
tation is to back off a bit from high pressures and
rates to avoid damaging the lung. This temptation
must be resisted because such attempts can cause
marked lability of the pulmonary vasculature.
Small increases in P
a
C0
2
or decreases in P
a
O
;
or
pH can lead to sudden increases in pulmonary
vascular resistance, resulting in renewed right-to-
left shunting, with worsening of the arterial blood
gases and a rapid downward spiral. Mechanical
ventilation of the patient is considerably aided by
paralysis.
A number of pharmacologic pulmonary vasodi-
lators have been tried, including morphine, pred-
nisolone, chlorpromazine, tolazoline, bradykinin.
and acetylcholine.
3
8
No agent has consistently im-
proved pulmonary hypertension, although tolazo-
line has been most frequently used. If systemic
hypotension occurs, volume repletion (may require
as much as 50-75 ml/kg for a 2-hour diaphrag-
matic hernia repair) and/or dopamine infusion is
indicated because the reduction in systemic pres-
sure increases ductal shunting.
The systemic blood pressure is probably the best
indicator for volume replacement because the cen-
tral venous pressure may not be very useful in the
presence of very high intrathoracic pressure. Pros-
taglandin E, (PGE,), which is a pulmonary and
ductal dilator, has been used to decrease pulmo-
nary vascular resistance in patients with persistent
fetal circulation (the ductus is thought to be al-
ready fully patent in these patients and cannot di-
late any more) and to dilate the ductus arteriosus
in patients with ductal blood flow-dependent con-
genital heart defects (e.g., pulmonary atresia).
9
Figure 18-4 is a flow chart for a suggested
stepwise administration of therapy for patients
with persistent fetal circulation.
2
On the basis of
these considerations, a reasonable anesthetic for
these patients would consist of mainly narcotics,
paralysis, low isoflurane concentration to control
Table 18-1 GUIDELINES FOR OPERATION OF MECHANICAL VENTILATOR IN
PERSISTENT PULMONARY HYPERTENSION OF THE NEONATES*
Factor Goal Common Error
Inspiratory pressure
Ventilator rate
Positive end-expiratory
pressure
Inspired 0
:
concentration
Inspiratory-expiratory ratio
Level necessary to decrease Pco
2
Rapid rates (up to 150 breaths per
minute) to decrease Pco
2
, with
lowest inspiratory pressure
Low values unless pulmonary
parenchymal disease exists
P
a
0
2
lOOmm Hg
Short inspiratory time to prevent air
trapping at rapid rates
Inadequate inspiratory pressure (no movement of
chest wall or increasing Pco
2
); reading on hand
manometer may differ from that on ventilator
Use of ventilator with maximum rate below level
needed to decrease Pco
2
; may require hand
ventilation to achieve possible high rates
Failure to try different levels that improve P
a
CO
;
as
patient condition changes
Weaning too rapidly
Inadequate flow rates to achieve short inspiratory
time
* Adapted from Clarke WR: The transitional circulation: Physiology and anesthetic implications. J Clin Anesth 2:192-211, 1990.
Used with permission.
Figure 184 Suggested sequential implementation of
therapy for infants with known or suspected persistent
pulmonary hypertension of the newborn. Details are
given in the text. (E.C.M.O. = extracorporeal mem-
brane oxygenation.) (From Clarke WR: The transitional
circulation: Physiology and anesthetic implications. J
Clin Anesth 2:192, 1990. Used with permission.)
response to surgical stimulation, F,0
2
of 1.0, no
nitrous oxide, and liberal fluid administration.
The role of patent ductus arteriosus ligation in
the infant with severe persistent fetal circulation is
unclear. The patent ductus arteriosus, although a
major source of right-to-left shunting and, there-
fore, of systemic hypoxemia and acidosis, is also
a relief valve for the pressure-overloaded right
ventricle. Ligation of the patent ductus arteriosus
might cause acute right ventricular failure (and
subsequently left ventricular failure because of in-
terventricular septal shifting; see Fig. 185)
10
and,
therefore, should not be performed on a routine
basis.
8
Ligation of the ductus arteriosus is most
commonly performed under local anesthesia, mod-
erate levels of narcosis, and neuromuscular block-
ade in the neonatal intensive care unit, which elim-
inates the precarious transportation of these very
ill and tiny infants to the operating room.
B. Respiratory Distress Syndrome
The respiratory distress syndrome is a disease
of atelectasis in premature infants caused by 2
failure of the type II alveolar cells to secrete 2
Figure 18-5 Cardiovascular ab-
normalities associated with the de-
velopment of pulmonary hyperten-
sion culminating in decreased CBF
and CO. (RVEDP = right ventricu-
lar end-diastolic pressure; RVEDV
= right ventricular end-diastolic vol-
ume; CBF = coronary blood flow;
CO = cardiac output; LVEDV -
left ventricular end-diastolic volume;
LVEDP = left ventricular end-dia-
stolic pressure; RV = right ventri-
cle.) (From Burrows FA, Klinck JR.
Rabinovitch M, Bohn DJ: Pulmonary
hypertension in children: Periopera-
tive management. Can Anaesth Soc J
33:606, 1986. Used with permis-
sion.)
normal amount of surfactant. Surfactant acts to
reduce surface tension in the alveolus, allowing
for expansion during inhalation and preventing
collapse during exhalation (see Fig. 3-19). When
surfactant is absent or reduced in quantity, the
respiratory distress syndrome occurs with atelec-
tasis and ventilation-perfusion abnormalities. The
presence of respiratory distress syndrome may
cause increased pulmonary vascular resistance. In-
creased pulmonary vascular resistance may cause
persistence of the fetal circulation.
The more severe the respiratory distress syn-
drome, the higher the pulmonary vascular resis-
tance is in relation to the systemic vascular resis-
tance. In severe forms of the respiratory distress
syndrome, pulmonary resistance is higher than
systemic resistance, and right-to-left shunting oc-
curs through the ductus arteriosus, resulting in a
lower arterial oxygen tension below the level of
the ductus (Fig. 186). As the respiratory distress
syndrome improves, pulmonary resistance falls be-
low systemic levels, the shunt is reversed, and
congestive heart failure may ensue and warrant
closure of the ductus either pharmacologically or
surgically."
-13
The primary aim of the treatment of the respi-
ratory distress syndrome is to open atelectatic al-
veoli with positive-pressure ventilation and then to
keep the alveoli open using positive end-expira-
tory pressure. Casual removal of the positive end-
expiratory pressure during transport to and in the
operating room may result in profound alveolar
collapse and the development of marked hypoxia,
acidosis, and hypercarbia. Consequently, the man-
agement of ventilation during anesthesia of a child
with respiratory distress syndrome is critical; it is
important to continue to provide the same ventila-
tory support during the operative period as was
provided in the preoperative period. The important
mechanical ventilation concerns are listed in Table
18-1. In many instances, it may be simplest to
bring the infant's own ventilator into the operating
room to ventilate the patient during surgery and to
use intravenous anesthesic agents and muscle re-
laxants.
Surfactant replacement is effective in prevention
and amelioration of the respiratory distress syn-
drome. The essential differences between the ear-
lier unsuccessful studies and more recent studies
that have met with unequivocal success (84 per
cent of patients) relate to the means of administra-
tion and the nature of the surfactants delivered.
14
The best results (increased P
a
0
2
, lung compliance
during spontaneous ventilation,
15
and radiographic
clearing) are achieved when 60 to 120 mg of sur-
factants derived from bovine lungs or human am-
niotic fluid are introduced into the lung through an
endotracheal tube as a bolus of 3 to 5 ml.
14
Compared with matched controls, respiratory
distress patients treated with surfactant have a
lower incidence of death (16 per cent vs. 52 per
cent, < .001), bronchopulmonary dysplasia (16
per cent vs. 31 per cent, < .001), and acute
complications (p < .001).
16
Good results have
been evident when surfactants were instilled up to
12 hours afterbirth.'
4
C. Retinopathy of Prematurity
Retinal damage due to inhaling an increased
partial pressure of oxygen is unique to the prema-
Figure 18-6 The lack of surfactant in the respiratory distress syndrome (RDS) causes atelectasis and low ventilation-perfusion
regions. The atelectatic and low ventilation-perfusion regions cause an increase in pulmonary vascular resistance (PVR), which
directs pulmonary artery blood flow through the ductus arteriosus, thereby increasing right-to-left (R > L) shunting. The ductus
arteriosus enters the descending thoracic aorta just distal to the left subclavian artery. Blood flow to the retina is preductal, and
oxygen tension in preductal blood flow must be measured by a right-arm arterial blood sample. Postductal oxygen tension can be
measured from an umbilical artery blood sample.
ture newborn. The premature infant's retinal arte-
ries react to increased partial pressure of oxygen
with vasoconstriction, which induces injury, neo-
vascularization, and retrolental fibroplasia that
causes retinal detachment and destruction.
17
18
Pre-
mature infants are at risk and remain at risk for the
retinopathy of prematurity until their retinal vas-
culature completes its growth at approximately 35
to 40 weeks of age after conception. If gestational
age cannot be determined accurately, a birth
weight of less than 1500 g seems to be the best
indicator of risk for the retinopathy of prematur-
ity.
19
Although there is no correlation between
P
a
0
2
levels and blindness risk,
20
21
there is an as-
sociation between the incidence and severity of
retinopathy of prematurity and the duration of ex-
posure to P
tc
0
2
levels of 80 mm Hg or higher.
22
However, in recent years, it has become clear
that factors other than high inspired oxygen con-
centration may be involved in the pathogenesis of
the retinopathy of prematurity.
18
20
For example,
the retinopathy of prematurity occurs in some in-
fants who have never been treated with supple-
mental oxygen, the disease does not occur in all
premature infants treated with high oxygen con-
centration, and the incidence of prematurity ap-
pears to be rising at a time when there has been
pronounced improvement in the technical methods
of oxygen delivery and monitoring.
18
22
In fact, it
does not appear possible to prevent the disease
with current (as of 1992) methods of monitor-
ing.
18
22
It may be that the apparent increase in
incidence may be due to increased survival of very
small infants (less than 1 kg). Some other factors
may be the absolute birth weight (risk is strongly
inversely proportional to birth weight, indicating
that this is a disease primarily caused by prematur-
ity, with high inspired oxygen concentrations a
secondary factor), maternal diabetes, frequent ap-
neic spells, bronchopulmonary dysplasia, degree
of illness, need for blood transfusions, and sep-
sis.
23-25
The anesthesiologist's responsibility is to main-
tain the arterial and/or transcutaneous oxygen
tension between 60 and 80 mm Hg.
26
In patients
without a ductus arteriosus, the arterial and trans-
cutaneous oxygen tension can be measured any-
where. In patients with a ductus arteriosus, the
retinal arteries are supplied by pre-ductus arterio-
sus blood flow (Fig. 18-6). Consequently, in pa-
tients with a ductus arteriosus, the arterial oxygen
tension must be measured from the right radial
artery and the transcutaneous oxygen tension must
be measured from the right arm. The arterial
and/or transcutaneous oxygen tension can be con-
trolled by using nitrous oxide as a diluting gas or
by using an air-oxygen blender to control the in-
spired oxygen concentration. Careful control of the
inspired oxygen concentration and retinal artery
oxygen tension is particularly important and diffi-
cult during thoracic surgery in which the patient
may be either hypoxic or hyperoxic and may vary
between these two states quickly, depending on
whether or not the operative lung is compressed
(retracted) or ventilated.
D. Periodic Breathing and Apnea
Premature infants often have short periods of
apnea (10 sec) that are not associated with hypox-
emia or bradycardia. Long periods of apnea (30
sec) that are associated with hypoxemia and brady-
cardia in the premature infant are distinctly abnor-
mal and can be lethal (sudden infant death syn-
drome [SIDS]).
27
Indeed, irregular and periodic
breathing is commonly seen in infants who have
survived a "sudden infant death" episode (aborted
SIDS), suggesting that immaturity of medullary
and/or peripheral control of breathing (loss of re-
sponsiveness to hypoxemia and hypercarbia) is an
important cause of this syndrome.
28
The risk of
SIDS may be increased by in situ exposure to
smoke.
29
Theophylline and caffeine have been
shown to increase the central chemoreceptor ven-
tilatory response and decrease the number of ap-
neic spells in premature infants.
30
In preparation
for extubation after surgery, the breathing pattern
of premature infants should be carefully observed
for significant periodic breathing, the presence of
which is a contraindication for the removal of the
endotracheal tube. Immediately following extuba-
tion, an oral airway, if tolerated, may be especially
helpful since many infants (60 per cent) may re-
main obligate nose breathers for as long as 2
months of age and may not open their mouths to
breathe even if the nares are totally obstructed for
any reason (e.g., secretions or edema).
31
E. Thermoregulation
The relative surface area of the newborn is one
ninth that of an adult, whereas the weight of a
newborn is approximately 20 times less than that
of an adult. Thus, the ratio of surface area to
weight for infants is approximately two times
greater than for adults. Consequently, an infant
loses heat much more rapidly than an adult.
Hypothermia in premature and newborn infants
is a dangerous condition; there is 90 per cent mor-
tality in newborns whose temperatures are allowed
to decrease below 32C.
32
Figure 18-7 shows the
symptoms and signs of progressive hypothermia
(Wilson WC, personal communication).
33
Ob-
viously, hypothermia can cause a severe metabolic
acidosis and cardiovascular depression below
32C. Oxygen consumption has been shown to
double in the immediate postoperative period if
the child is allowed to emerge from anesthesia
even in a mildly to moderate hypothermic state.
34
Hypothermia below 35C is a relative contraindi-
cation to extubation.
Hypothermia in the newborn and small child
undergoing thoracic surgery is particularly likely
to occur because there is a large surface area ex-
posed to environmental air, allowing high evapo-
rative and convective heat loss. It is far better to
prevent this heat loss rather than to correct it after
it occurs. Accepted techniques to maintain body
temperature include raising the ambient tempera-
ture of the operating room to 30C to 34C (neutral
thermal environment); warming and humidifying
anesthetic gases
35
; using only warm scrub and ir-
rigating solutions; using radiant heat lamps, porta-
ble heated isolettes, and thermal mattresses; and
warming all intravenous fluids (Table 18-2). In
most pediatric thoracic surgery cases, most or all
of these heat loss prevention measures should be
used. Core temperature (either nasopharyngeal,
esophageal, or rectal) should be measured in every
case.
F. Vitamin, Caloric, Electrolyte, and
Fluid Requirements
The newborn may be deficient in vitamin K,
stores upon which the synthesis of coagulation fac-
tors II, VI, IX, and XI is dependent. Therefore,
vitamin K, (0.5 to 1.0 mg) should be given intra-
muscularly or subcutaneously prior to surgery dur-
ing the newborn period, especially if the child has
never been fed. Hypoglycemia is common in the
newborn period and is defined as a blood glucose
level of less than 40 mg per cent during the first
72 hours of life, and less than 50 mg per cent after
72 hours of otherwise normal neonatal life. In low-
Signs and Symptoms of Hypothermia
Normal core temperature
Vasoconstriction begins
Shivering prominent (BMR increased sixfold)
Maximum ventilatory response to pC0
2
changes,
confusion and disorientation (lowest measured
temperature on many thermometers)
Amnesia, lethargy, apathy, cardiac arrhythmias,
bradycardia, prominent shivering stops
Osborn waves regularly seen on ECG
Spontaneous respirations slow
Coma, pupils dilate, tendon reflexes absent,
heart rate half of baseline
Risk of ventricular fibrillation high
Primary vascular paralysis
Venous stasis (blood viscosity 173% normal)
Spontaneous respirations cease
EEG flatline
Figure 18-7 Signs and symptoms of hypothermia. (BMR = basal metabolic rate; ECG = electrocardiogram; EEG = electro-
encephalogram.) (Based on data from Wilson WC [personal communication] and Reuler.
33
)
Table 18-2 TECHNIQUES TO PREVENT
HEAT LOSS IN INFANTS
AND NEONATES
1. Use portable heated isolettes
2. Warm operating room to 30 to 40C
3. Use thermal mattresses
4. Warm intravenous fluids
5. Use radiant heat lamps
6. Warm and humidify anesthetic gases
7. Warm scrub and irrigating solutions
8. Put hat on infant
birth-weight and premature infants, hypoglycemia
is defined as a blood glucose level of less than 30
mg per cent. Infants of diabetic mothers (high fetal
insulin production) and infants experiencing peri-
natal distress of any kind can be hypoglycemic.
The presence of any condition causing stress or a
history of a diabetic mother should make the anes-
thesiologist aware of the possibility of hypoglyce-
mia and the need to determine blood sugar levels
in the perioperative period. Normal glucose re-
quirements for premature and normal newborns
are 4 mg/kg/min.
Neonatal hypocalcemia is related to low para-
thyroid hormone activity and decreased stores of
available calcium.
36
A rapid infusion of 10 mg/kg
of calcium chloride or 30 mg/kg of calcium glu-
conate is usually effective in reversing hypocal-
cemia. Sodium, potassium, and chloride require-
ments are all similar to those of adults at 1 to 3
mEq/kg/24 hours.
Using a solution of 5 per cent dextrose and 0.2
per cent saline with 20 mEq of KC1 added per
liter, maintenance fluids can be approximated by
administration of 4 ml/kg/hour for the first 10 kg
of body weight and an additional 2 ml/kg/hour for
the second 10 kg of body weight.
37
If maintenance
fluids are all that is required in a specific patient,
it is advisable to calculate the preoperative fluid
deficit and correct half of the deficit in the first
hour, with further administration of fluid at 120
per cent of the base-line maintenance rate until the
entire deficit is corrected. Generally, this simple
approach is complicated by the trauma of surgery,
requiring the intraoperative administration of sub-
stantially more fluid.
37
The increased requirement
may vary from 2 ml/kg/hour to 10 ml/kg/hour
above the maintenance level for most surgery
(however, major surgery, like repair of diaphrag-
matic hernia, may require much more; see pre-
vious discussion) and should consist of either a
lactated Ringer's solution or saline. Blood should
be infused if blood loss exceeds 15 per cent of the
blood volume or reduces the hematocrit to less
than 30.
38
G. Airway Anatomy
Anatomically, the child, and especially the in-
fant, has an airway that is different enough from
that of the adult to make intubation more difficult
(Fig. 18-8). The head and occiput are relatively
larger than the rest of the body, which requires the
neck to be flexed on the chest if the face is in a
midline sagittal plane; this is why a newborn lies
with the face turned to one side. However, if the
child's head is put into a good "sniffing" position
(neck flexed on chest and head extended on neck)
by extending the head on the neck, the exposure
of the larynx will be facilitated. Still, during and
after intubation, the large occiput causes the head
to be unstable, with a propensity to roll to one side
or another; this can be remedied by placing the
occiput inside a doughnut-shaped stabilizing
sponge pillow.
There are other differences in the anatomy of
the newborn airway compared with that of the
adult airway. The narrowest portion of the infant's
airway is the subglottic area at the level of the
cricoid cartilage (in the adult the narrowest area is
at the rima glottidis). The functional significance
of the small diameter is enormous; if a neonate's
larynx measures 4 mm at the level of the cricoid
cartilage, a 1-mm reduction in this diameter,
caused by either trauma or infection, will reduce
the overall cross-sectional area of the airway by
approximately 50 per cent. Table 18-3 shows
guidelines for selection of endotracheal tubes in
terms of appropriate internal diameter (i.e., an in-
Figure 188 The head and airway anatomy of the infant differs from that of an adult in the ways that are listed in the figure.
ternal diameter of an uncuffed endotracheal tube
results in a leak around the endotracheal tube at
15-20 cm H
2
0).
39
In a normal newborn, the neck is short in com-
parison to the adult. The larynx is more cephalad
in the child, lying approximately at the level of C-
2-C-3 compared with C-5 in adults. The larynx-
to-carina distance is only 4 cm (range = 2.5-5.0
cm)
40
in the infant, and consequently great care
must be taken to pass the endotracheal tube be-
yond the vocal cords by only 1.5 to 2.0 cm and to
avoid severe head flexion (moves endotracheal
tube caudad by 1 cm), head extension (moves en-
dotracheal tube cephalad by 1 cm), and wide
mouth opening (can move endotracheal tube cau-
dad by 1 cm) to avoid bronchial cannulation or
tracheal decannulation.
40
Table 18-4 shows guide-
lines for the proper depth of insertion of endotra-
cheal tubes as judged from the level of the anterior
gum line.
The infant's epiglottis projects cephalad at 45
degrees or more "from the anterior wall of the lar-
ynx, and it is relatively stiffer, longer, and omega-
shaped than an adult epiglottis. Although this
makes the epiglottis easier to visualize, it also
makes the epiglottis more difficult to displace so
that the laryngeal aperture may be visualized. For
these reasons, a Miller blade is preferable to a
Mackintosh blade for intubation of infants. The
tongue, like the head, is relatively larger in chil-
dren than in adults and may cause obstruction dur-
ing induction of anesthesia and ventilation by
mask and may impair visualization of. the larynx.
Although the triad of more cephalad displacement
of the larynx, increased difficulty in displacing the
epiglottis, and the presence of a large tongue may
make the larynx seem to be more anterior com-
pared with that of the adult, it is not. All of the air
passages (nose, nasopharynx, trachea) are small.
Consequently, they may be easily obstructed, and
at this small size, any obstruction can cause a
catastrophic increase in resistance to airflow (see
cricoid diameter discussed previously). There are
also a number of physiologic differences between
the infant and adult that make the infant more
susceptible to hypoxia (Table 18-5).
41
Finally, there are two primary determinants ol
which main-stem bronchus an endotracheal tube
will enter: namely, the angle of branching from the
trachea (left greater than right) and the side. t<
which a bevel faces (the tip will enter the opposite
main-stem bronchus). First, in infants, the right
main-stem bronchus projects from the trachea at
an angle of 31 5 degrees to the trachea axis (in
the adult this angle is 20-15 degrees), whereas the
left main-stem bronchus projects from the trachea
at an angle of 47 7 degrees (in adult males this
angle is 40 degrees and in adult females this angle
is 50 degrees).
42 4?
Therefore, on the basis of angle
of branching, endotracheal tubes that are inserted
too far into the trachea are more likely to enter the
Table 18-5 NORMAL PULMONARY
FUNCTION IN INFANTS
(MEAN VALUES)*
Variable Newborn Adult
Body weight (kg)
Tidal volume (ml/kg)
Respiratory rate (breaths/min)
Volume expired (ml/kg/min)
Alveolar gas volume (ml/kg/min)
Anatomic dead space (ml/kg)
Physiologic dead space/tidal volume
ratio
0
2
consumption (ml/kg/min)
CO
:
consumption (ml/kg/min)
Calories (kg/hr)
Total lung capacity (ml/kg)
Functional residual capacity (ml/kg)
Thoracic gas volume (ml/kg)
Vital capacity (ml/kg)
Residual volume (ml/kg)
Closing volume (ml/kg)
Closing capacity (ml/kg)
Tidal volume/functional residual
capacity ratio
Functional residual capacity/total
lung capacity ratio
Tracheal length (mm)
Tracheal diameter (mm)
3
6
35
210
130
2.5
0.30
6.4
6.0
2
63
30
36
35
23
12
35
0.20
0.47
20- 50
4
70
6
15
90
60
2.0
0.33
3.5
3.0
1
86
34
34
70
16
7
23
0.20
0.40
120
16
*From Cohen DE: Special problems in pediatric anesthesia.
Int Anesthesiol Clin 23:25-35, 1985. Used with permission.
right main-stem bronchus in both adults and in-
fants (100 per cent incidence in two studies [if the
bevel is facing to the left, as it is when the concav-
ity of the tube is pointing anteriorly]).
42
4?
The
same considerations apply to suction catheters.
Second, the side to which the bevel is facing is
equally important as the angle of branching. With
a left-facing bevel, the tip of the endotracheal tube
is forced to the right of center, making entry into
the right main-stem bronchus certain,
44
whereas
with a right-facing bevel the tip of the tube lies to
the left of the center, and the endotracheal tube
now enters the left main-stem bronchus (Fig. 18-
9). Thus, a 92 per cent success rate in cannulating
the left main-stem bronchus can be achieved with
an endotracheal tube by turning the patient's
face/head to the extreme right and rotating the
endotracheal tube 180 degrees so that the concav-
ity of the endotracheal tube faces posteriorly and
the bevel of the tube faces to the right (see Fig.
18-9).
45
A 100 per cent success rate of endotra-
cheal intubation of the left main-stem bronchus
can be achieved by simply cutting the endotracheal
tube shaft so that a new bevel is fashioned that is
facing to the right.
III. CONGENI TAL DI APHRAGMATI C
HERNIA
Abdominal contents can herniate through a de-
fect in the diaphragm into the thoracic cavity
(Figs. 18-10 and 18-11). The posterolateral defect
of Bochdalek (at the vertebrocostal trigone) ac-
counts for approximately 80 per cent of all dia-
phragmatic lesions,
46
with left-sided lesions eight
times more common than right-sided lesions in
newborns but with right-sided hernias twice as
common as left-sided hernias after the neonatal
The Direction of the Face of an Endotracheal
Tube Bevel and Mainstem Bronchial Cannulation
LEFT FACING BEVEL
Results in Right
Mainstem Bronchial
Cannulation
RIGHT FACING BEVEL
Results in Left
Mainstem Bronchial
Cannulation
Figure 189 The direction of the face of an endotracheal tube (ETT) bevel is a primary determinant of which main-stem
bronchus the ETT will enter. Left panel, A left-facing bevel forces the tip of the ETT to be off center to the right (R) and therefore
enter the right main-stem bronchus, whereas a right-facing bevel forces the tip of the ETT to be off center to the left (L), thereby
entering the left main-stem bronchus (right panel).
period.
47
Fifteen to 20 per cent of diaphragmatic
hernias occur through the esophageal hiatus, and 2
per cent of the hernias occur through the anterior
foramen of Morgagni on either side of the ster-
num. The average incidence of congenital dia-
phragmatic hernia is approximately 1 in 5000.
48
49
Associated anomalies are often present and include
defects of the cardiovascular system (23 per cent),
the gastrointestinal tract (malrotation in 40 per
cent), the genitourinary system, and the central
nervous system.
50
Patients without other congenital
defects have a mortality of 25 per cent.
51
Patients
with congenital heart disease in addition to con-
genital diaphragmatic hernia have a mortality in
excess of 50 per cent.
46-51
The later the presenta-
tion in life, the greater the survival.
52
The major and always-present result of bowel
contents being chronically in the chest during fetal
development is chronic ipsilateral lung compres-
sion, which results in ipsilateral pulmonary hypo-
plasia. Shift of* the mediastinal structures away
from the hernia, causing chronic contralateral lung
compression, may produce contralateral pulmo-
nary hypoplasia. Consequently, significant dia-
phragmatic herniation presents as severe respira-
tory failure with hypoxemia, hypercarbia, and
acidosis. Mediastinal distortion may impede ve-
nous return and cardiac output and cause a meta-
bolic component to the acidosis. The increased
pulmonary vascular resistance of the hypoplastic
lungs, as well as increased pulmonary vasculai
resistance due to hypoxemia, hypercarbia, and aci-
dosis, may cause persistence of the fetal circula
tion with right-to-left shunting of desaturatec
blood at the level of the atrium and/or ductus ar
teriosus, which, if present, greatly compounds th<
degree of hypoxemia.
53
54
Bochdalek hernias pro
duce the most significant degree of respirator
compromise, whereas esophageal hiatus and fora
men of Morgagni hernias are generally small ii
size, do not compromise pulmonary function a
much, and may not be detected in the neonata
period. Patients with Bochdalek hernias are oftei
in extremis and require intubation and mechanics
ventilation before arrival in the operating room.
55
Eventration of the diaphragm is an abnormall;
high or elevated position of the diaphragm am
may be due to either a congenital failure to de
Vena Cava
Anterior Foramen of
Morgagni
Esophagus
Right Posterolateral
Foramen of Bochdalek
Left Posterolateral
Foramen of Bochdalek
Aorta
Figure 18-10 View of the diaphragm from a caudad direction showing potential sites of herniation of abdominal contents in
congenital diaphragmatic hernia.
velop the muscular part of the diaphragm (most
common cause) or paralysis of the diaphragm
caused by phrenic nerve injury at the time of sur-
gery (most commonly a lateral thoracotomy for a
Blalock-Taussig shunt or procedures related to a
patent ductus arteriosus) or trauma during birth.
56
The lack of muscle or muscle function allows the
abdominal contents to billow the diaphragm into
the thorax to a variable degree. The distinction
between a hernia with a sac (which has no mus-
cular elements) and an eventration (which does
have a muscular or fibrous element) may not be
significant in terms of symptoms produced. For
example, cases of slight eventration may be
asymptomatic, but severe cases occur in a fashion
similar to that of left-sided Bochdalek congenital
diaphragmatic hernias. Most eventrations of the
diaphragm should be surgically corrected (dia-
phragmatic tightening procedure), whether they
are symptomatic or not, because even slight,
chronic compression of the lung may cause infec-
tion. Significant eventration associated with other
anomalies/conditions (e.g., congenital heart dis-
ease, malrotation of the bowel) carries a mortality
as great as a diaphragmatic hernia (despite the fact
that the diaphragm has been plicated) and failure
to extubate the patients within 1 week of surgery
is a very ominous sign.
56
The antenatal diagnosis of congenital diaphrag-
matic hernia may be made by ultrasonography;
these lesions are very severe in all respects and the
mortality is 80 per cent.
52
The postnatal diagnosis
of congenital diaphragmatic hernia is usually
straightforward. Approximately 30 per cent of the
cases are associated with polyhydramnios.
57
Cy-
anosis is usually apparent soon after the umbilical
cord is clamped. Signs of respiratory distress are
common and severe and include tachypnea, nasal
flaring, and both sternal and intercostal space in-
spiratory retraction. Frequently, these infants re-
quire endotracheal intubation and mechanical ven-
tilation immediately or soon after birth.
After positive-pressure ventilation, pneumo-
thorax on either side is common. The abdominal
Physiologic Effects of Congenital Left
Posterolateral Diaphragmatic Hernia
Figure 18-11 The physiologic effects of congenital left posterolateral diaphragmatic hernia consist of causing the ipsilateral
lung to be hypoplastic, shift in the mediastinum to the contralateral side, possibly causing a contralateral hypoplastic lung,
persistence of the ductus arteriosus because of the high pulmonary vascular resistance caused by the hypoplastic lung, and possibly
causing a decreased cardiac output as a result of the shifted mediastinum. The numbers indicate the flow of pathologic events.
contour is abnormally scaphoid, whereas the
thorax may be barrel shaped. With a left postero-
lateral diaphragmatic hernia, the mediastinum may
be shifted into the right hemithorax, both right and
left breath sounds are poor, and cardiac sounds
will best be heard in the right chest ("apparent"
dextrocardia). Audible bowel sounds in the chest
may or may not be present (there is no gas in the
bowel at birth; gas formation takes several hours).
A body radiograph usually confirms the diagnosis,
showing intestinal loops in the hemithorax. The
lung on the side of the hernia is compressed into
the hilum, and the mediastinal contents are pushed
to the opposite side. If a nasogastric tube had al-
ready been inserted, the chest radiograph might
also show a decompressed stomach within the tho-
racic cavity. The diagnosis of either hernia or
eventration can be confirmed by fluoroscopy
and/or peritoneography.
56
The surgical tasks are to return the abdominal
contents to the abdominal cavity, correct bowel
malrotation, repair the diaphragmatic defect, and
close the abdomen.
57
Closure of the abdomen may
prove to be difficult because the unstretched ab-
dominal cavity is too small. In this situation, the
surgeon may need to utilize a prosthetic material
such as Gore-Tex to avoid excessive intra-abdom-
inal pressure.
58
If the abdomen is closed under
tension, the chance of recurrence is increased and
the infant will likely require mechanical ventila-
tion for several days until enlargement of the ab-
dominal cavity allows the diaphragm to have its
normal downward excursion. Most surgeons now
use a transabdominal approach for repair of Boch-
dalek hernias since this incision facilitates return
of the bowel to the peritoneal cavity, correction of
intestinal malrotation, repair of the diaphragm un-
der direct vision, and stretching of the abdominal
wall to accommodate the intestines if necessary.
59
A thoracic approach is often used for the operative
repair of Morgagni hernias.
For cases of eventration, the diaphragm may be
plicated through a thoracic (most common) or ab-
dominal approach.
56
Survivors of either congenital
diaphragmatic hernia or eventration (some having
received extracorporeal membrane oxygenation
[ECMO]; see later discussion) have been shown to
have a very high incidence of an extremely ectatic
esophagus with severe gastroesophageal reflux
(probably on a congenital basis), which can be
managed with surgery but may account for the
observed compromised growth.
60
Many of these infants have major cardiopulmo-
nary problems (hypoplastic lungs, shifted medias-
tinum, possible persistent fetal circulation, respi-
ratory distress syndrome, other congenital defects)
and will not tolerate the induction of general an-
esthesia without a guaranteed airway. In addition,
positive-pressure ventilation with a bag-and-mask
system may inflate the stomach and further com-
promise pulmonary function. Consequently, criti-
cally ill neonates with congenital diaphragmatic
hernia should be intubated when awake (Table 18-
6).
46 61-64
Those few infants who are too vigorous
for this approach usually can be intubated without
a relaxant after breathing halothane and oxygen
spontaneously for a few minutes. All patients
should be premedicated with atropine intramuscu-
larly or intravenously at the beginning of induc-
tion.
After intubation, moribund patients are given
minimal intravenous anesthetics, initially 100 per
cent oxygen, and a nondepolarizing muscle relax-
ant is added if the patient moves in response to the
surgical (or any) stimulus. There has been a con-
certed effort by pediatric intensivists and surgeons
to delay surgery until the patient can be stabilized
(e.g., see ECMO later).
52
65
Peripheral perfusion
should be supported with intravenous fluid and
dopamine administration as necessary.
59
Halothane
or narcotic, in addition to relaxant, is given to
those infants who demonstrate cardiovascular sta-
bility; the amount given depends on the clinical
status of the patient. Nitrous oxide should be
avoided since it may distend the gut, further com-
promising lung function prior to hernia reduction
and subsequent abdominal closure.
66
In premature infants at risk of retrolental fibro-
plasia, air or nitrogen is added to lower the arterial
oxygen tension to approximately 80 mm Hg.
26
To
achieve this arterial oxygen tension end point, it is
highly desirable to have an arterial line. It should
Table 18-6 MOST IMPORTANT SPECIFIC
ANESTHETIC CONSIDERATIONS
FOR EACH OF THE MAJOR
PEDIATRIC THORACIC
SURGERY CASES
I. Congenital posterolateral diaphragmatic hernia
A. Awake intubation
B. Avoid nitrous oxide
C. Use small tidal volumes and rapid respiratory rates;
try to induce respiratory alkalosis to decrease PVR
D. Do not attempt to inflate ipsilateral hypoplastic lung
E. High index of suspicion of need for contralateral chest
tube
F. Patient may have high extracellular fluid requirements
G. Postoperative mechanical ventilation
H. Use of ECMO: Understand the physiology of the
circuit, in particular a decrease in venous return to the
pump
II. Esophageal atresia with tracheoesophageal fistula
A. May need tip of endotracheal tube just above carina
(bilateral breath sounds, CO\ analysis of gastrostomy
tube, gastrostomy tube under water seal)
B. Maintain spontaneous ventilation, if possible
C. Gentle positive-pressure ventilation, if necessary
D. Do not use esophageal stethoscope; use left axillary
stethoscope
E. Do not pass nasogastric tube; may need to measure
length of proximal pouch
F. May need to put Fogarty balloon catheter through
gastrostomy into distal esophagus (may need
fiberscope to do this)
G. Postoperative mechanical ventilation
III. Vascular rings
A. May need right main-stem bronchial intubation
B. May need to palpate upper extremity and head and
neck pulses for surgeon
IV. Congenital lobar emphysema and cysts
A. Nitrous oxide contraindicated
B. Positive-pressure ventilation contraindicated until
chest is open
C. May need to separate the lungs
V. Unilateral infections (abscess, empyema)
A. Separate the lungs (bronchial blocker, contralateral
main-stem bronchus intubation)
VI. Bronchography
A. May need to pass small contrast material catheter into
one lung with the aid of a rigid or flexible
bronchoscope or fluoroscopy
B. Discuss with radiologist preferred pattern of
ventilation
C. May need to turn patient to various positions
Abbreviations: PVR = pulmonary vascular resistance;
ECMO = extracorporeal membrane oxygenation.
be remembered that the right radial artery must be
used to estimate pre-ductus arteriosus (retinal ar-
tery) oxygen tension. An umbilical artery catheter
measures post-ductus arteriosus oxygen tension.
Positive-pressure ventilation is administered by
hand with small tidal volumes and rapid rates (60
to 120 breaths/min) in an attempt to keep the mon-
itored inflating airway pressures under 20 to 30 cm
H
2
0.
46
62
63
67
Hyperventilation (preductal P
a
C0
2
= 25 to 30 mm Hg and postductal P
a
C0
2
= 30
to 35 mm Hg) is recommended to lower pulmo-
nary vascular resistance and minimize right-to-left
shunting through the ductus arteriosus.
6
Attempts
to inflate the hypoplastic ipsilateral lung should be
avoided since the pressure required to even mini-
mally expand the hypoplastic lung is higher than
the pressure required to rupture the normal lung.
68
Consequently, it is not surprising to find that pneu-
mothorax occurs more frequently on the side op-
posite the hernia.
62
Any sudden severe deteriora-
tion in heart rate, blood pressure, arterial oxygen
tension, or lung compliance is strongly suggestive
of such an event; time should not be wasted with
radiographic confirmation of a pneumothorax, and
a chest tube should be placed immediately.
69, 70
Indeed, some authors recommended inserting a
contralateral chest tube prophylactically prior to
surgery.
68
71
There are several other important anesthetic
concerns. Metabolic acidosis should be treated
with infusions of sodium bicarbonate (mEq
HC( V = body weight [kg] X base deficit X
0.25).
6
A nasogastric tube should always be passed
to decompress the stomach. Normothermia must
be maintained and hyperoxemia avoided (see Spe-
cial Problems Related to Premature and Newborn
Infants). Since a large extracellular fluid third
space may develop owing to the trauma of han-
dling the bowel, the volume status of these very
sick infants may be borderline, and hypovolemia
may need to be treated vigorously with colloid
and/or blood infusion. Frequently, total fluids ad-
ministered may equal 25 to 50 ml/kg during the
case.
The postoperative period is stormy much more
often than not in these patients (usually after a so-
called stable "honeymoon" period of several
hours) because of an abdomen that is under ten-
sion, the presence of pulmonary hypoplasia, the
presence of associated congenital defects, and the
persistence of the fetal circulation (probably most
important; see later discussion); artificial ventila-
tion is typically required for at least several days.
Consequently, intraoperative narcotics and relax-
ants are not reversed. Attempts should be made to
keep the preductal arterial oxygen tension between
60 and 80 mm Hg and the arterial carbon dioxide
tension between 35 and 40 mm Hg, if possible.
Careful attention should be given to volume status
as well as the nutritional state of the patient. A
gastrostomy is often placed at the time of surgery
so that feeding may take place during recovery.
The hypoplastic ipsilateral lung should not be ex-
pected to fill the hemithorax. The mediastinum
usually returns to the midline slowly.
59
Patients wkh persistent fetal circulation are
often difficult to wean from ventilators. Pulmonary
vasoconstriction, with resultant right-to-left shunt-
ing, can be triggered by minimal changes in in-
spired oxygen concentration, decreases in arteri
oxygen tension, increases in acidosis, or sudde
changes in pulmonary blood volume. Successf
treatment of pulmonary vasoconstriction involvi
avoidance of these events. Although a number
pharmacologic pulmonary vasodilators have be<
tried, no drug has consistently improved pulm
nary hypertension (see Persistence of the Fetal Ci
culation).
3, 7
72
Other major considerations in tl
postoperative period consist of ligation of a pate
ductus arteriosus, poor or absent function of t
ipsilateral diaphragm, recurrence of the hernia (
to 22 per cent, most of which require reopen
tion),
73, 74
risk of pneumothorax (7 to 20 p|
cent),
75
76
and provision of adequate nutrition.
In view of the previously discussed pre-, intri
and postoperative difficulties in managing tl
sickest of these individuals, use of ECMO h
become an accepted means of treating moribui
infants with overwhelming cardiorespiratory fa
ure.
60
Patients who cannot be stabilized with ma
imal conventional therapy are begun on ECM
support. ECMO is discontinued before surgical r
pair if and when relatively low levels of mechai
cal ventilation (peak pressure < 30 cm H
2
0, ra
< 30 breaths/min, F,0
2
< 0.3) result in acceptat
arterial blood gases (Po
2
> 60 mm Hg, Pco
2
< :
mm Hg). Those who do not meet these critej
within 1 week have the diaphragmatic hernia
paired while receiving ECMO.
The ECMO circuit is shown in Figure 18-12
Briefly, venous blood is drained by gravity frc
the right atrium to a small collapsible reservoir tl
buffers against slight fluctuations in venous retui
Heparin and other fluids are administered into t
reservoir. ECMO blood flow is provided by
occlusive roller pump. The oxygenator is a silice
membrane envelope separating fresh gas frc
blood. Gas exchange occurs by diffusion. Final
the blood is warmed through a heat exchan
before being returned to the aortic arch via a c<
nula in the right common carotid artery. Venti
tion is not required; a Mapleson circuit is used
provide continuous positive airway pressure a
apneic oxygenation with 100 per cent oxygen.
77
All patients receive fentanyl before ECMO
provide sedation and to decrease their pulmoni
vascular resistance. During ECMO, the fentai
infusion is continued and titrated to relieve d
comfort and to minimize movement that could d
lodge the vascular cannulas. Elevations in bla
pressure are managed with antihypertensi
agents. Additional fentanyl and sufentanil are i
ministered at the time of surgery to induce an
thesia and to blunt the cardiovascular response
surgical stimulation. Vecuronium is used for mi
cle relaxation.
The complications of this ECMO circuit are ;
Right Common
Carotid
Artery
Reservoir
Heparin Infusion Pump Right Atrium
Figure 18-12 The extracorporeal membrane oxygenation circuit. See text. (From Truog RD, Schena JA, Hershenson MB. et al:
Repair of congenital diaphragmatic hernia during extracorporeal membrane oxygenation. Anesthesiology 72:750, 1990. Used with
permission.)
lated to decreased venous return to the pump cir-
cuit. This may cause cessation of extracorporeal
circulation, and if pump shutdown does not occur,
air entrainment and air embolism may occur. De-
creased venous return may result from hypovole-
mia, malposition or kinking of the venous catheter,
or surgical retraction that impedes venous flow.
The management of this complication is fluid ad-
ministration and/or correction of the obstruction.
If at any time air is entrained, the ECMO cannulas
must be clamped to prevent an air embolus from
reaching the patient, and conventional ventilation,
which will result in some gas exchange, must be
instituted. The pump should be manually operated
to move the air to an access port where it can be
aspirated. ECMO may then be resumed.
Unlike patients undergoing cardiac surgery dur-
ing cardiopulmonary by-pass, infants receiving
ECMO maintain a variable degree of intrinsic car-
diac output. Interpretation of the patient's condi-
tion during ECMO requires an understanding of
this unique physiology. For example, a decrease in
arterial Po
2
could reflect either an increase in en-
dogenous cardiac output and pulmonary blood
flow with increased shunting through the lungs or
a malfunction of the ECMO circuit. Inadequate
anesthesia can result in arterial desaturation by the
former mechanism (i.e., increased cardiac output
and pulmonary shunting). Proper management in
this case includes additional anesthetic, beta block-
ade, an increase in pump flow, and/or increased
ventilation of the lungs.
IV. ESOPHAGEAL ATRESIA AND
TRACHEOESOPHAGEAL FISTULA
There are many different types and permutations
of esophageal atresia and tracheoesophageal fis-
tula, and these lesions are perhaps best described
by their particular individual anatomies and with-
out reference to use of letters and/or numbers.
Esophageal atresia with distal (to the esophageal
atresia) tracheoesophageal fistula is by far the most
common of these, occurring in 80 to 90 per cent
of cases
78
"
83
(Fig. 18-13A). The distal fistula be-
tween the trachea and esophagus is usually located
one or two tracheal rings above the carina on the
6 7 6 Anesthesia for Pediatric Thoracic Surgery
Figure 18-13 The most common cases of esophageal atresia and tracheoesophageal fistula consist of (A) esophageal atresia with
distal tracheoesophageal fistula (80 to 90 per cent) and (B) esophageal atresia without fistula (7 to 9 per cent).
posterior aspect of the trachea. Seven to 9 per cent
have esophageal atresia without tracheoesophageal
fistula (Fig. 18-135). The remaining types (as
many as 95)
84
are rare. The incidence of esopha-
geal atresia and tracheoesophageal fistula is ap-
proximately 1 in 3000 births.
85
The two most important determinants of sur-
vival in patients with tracheoesophageal fistula are
prematurity and associated congenital anomalies.
86
In the absence of prematurity and severe associ-
ated congenital anomalies, survival currently ap-
proaches 100 per cent.
79,87
88
Prematurity with pul-
monary hypoplasia and pneumonia, which
accounts for at least 25 per cent of the tracheo-
esophageal fistula * population,
88-93
is associated
with a 15 to 90 per cent mortality.
88
92
"
95
The inci-
dence of associated congenital anomalies in tra-
cheoesophageal fistula cases is 30 to 50 per cent,
and in these patients the mortality approaches 50
percent.
78
81
96
97
The cardiovascular anomalies are most trouble-
some and, in order of decreasing incidence of oc-
currence, are ventricular septal defect, patent duc-
tus arteriosus, tetralogy of Fallot, atrial septal
defect, and coarctation of aorta.
81
96
The associated
gastrointestinal anomalies are next most important
with reference to mortality and consist of imper-
forate anus, midgut malrotation, duodenal atresia,
pyloric stenosis, Meckel's diverticulum, and ec-
topic or annular pancreas.
98
Other defects involve
the vertebrae and renal systems and together with
these anomalies constitute the VATER associa-
tion
84
8 5
8 7
(V = vertebral and ventricular septal
defect; A = anal defect; = tracheoesophageal
fistula; = esophageal atresia; and R = renal
dysplasia).
Atresia of the esophagus may be suspected prior
to delivery if polyhydramnios is present." If poly-
hydramnios is present, the diagnosis can be made
shortly after delivery by failing to pass a radi-
opaque nasogastric tube into the infant's stomach
(i.e., the catheter curls up in the proximal pouch).
If the diagnosis is not suspected or made after
delivery, the presence of excessive foamy oral se-
cretions in the first few postnatal hours may sug-
gest the diagnosis.
88
However, in most cases the
diagnosis is usually delayed until coughing, chok-
ing, and regurgitation occur during the first or sub-
sequent feedings. Following the first (or later)
feeding, the respiratory rate increases and respira-
tions become labored (presumably due to pulmo-
nary aspiration). The urgency in making the diag-
nosis is to prevent further pulmonary aspiration. In
addition, with a tracheoesophageal fistula, the ab-
domen usually becomes distended and tympanic
and occasionally the distension can interfere with
breathing.
The diagnosis is confirmed by trying to pass a
radiopaque nasogastric catheter into the stomach.
If the catheter stops abruptly in the proximal
esophageal segment (at a distance of approxi-
mately 10 to 12 cm from the nares) the diagnosis
is virtually ensured.
96
Posterior and lateral chest
radiographs are taken to determine the position of
the nasoesophageal tube tip. The radiograph
should include the entire abdomen so that presence
or absence of air in the stomach and intestines can
be determined. When the upper esophagus is
atretic (tube tip will not pass into the stomach), air
in the stomach is pathognomonic of a fistula be-
tween the trachea or esophagus. Absence of air in
the gastrointestinal tract usually indicates the pres-
ence of esophageal atresia without tracheoesopha-
geal fistula.
Tracheoesophageal fistula without atresia (H
type fistula) may not cause significant symptoms
until adolescence or adulthood.
100
The diagnosis
should be suspected with a history of recurrent
pneumonia.
101
The diagnosis requires bronchos-
copy or bronchography, and occasionally it is not
possible to demonstrate the lesion prior to surgery.
A diagnostic aid is the demonstration of elevated
intragastric oxygen concentration during ventila-
tion with 100 per cent oxygen.
102
The diagnosis
should be considered in any patient who develops
gastric distension while intubated and receiving
positive-pressure ventilation.
100
Large, healthy infants (about half the patients)
may undergo primary repair directly (Fig. 18-
14).
86
In half the infants, respiratory disease has
occurred and gastrostomy is the first procedure to
be done (see Fig. 18-14). In infants with extreme
immaturity or severe lung disease, thoracotomy
may be postponed until improvement in respira-
tory function has occurred and there has been a
gain in weight to more than 1800 g.
79
88 l03
With
the availability of hyperalimentation and continu-
ous upper pouch and gastrostomy suction, anti-
biotics and nursing the patient in a 45-degree up-
right position (e.g., using an infant car seat), both
of these goals are obtainable.
86
Of course, such a
delay is feasible only if the infant's course is not
complicated by repeated aspiration of gastric
contents.
79
I04
The tracheoesophageal fistula may be occluded
by a Fogarty balloon catheter (4 French with out-
side diameter [OD] = 0.95 mm, balloon volume
= 0.75 ml, which when inflated = 5.0 mm OD)
at the time of gastrostomy by using an antegrade
fiberoptic bronchoscope (2.0-mm OD). The fiber-
optic bronchoscope must first be passed through
the endotracheal tube and the adequacy of the ven-
tilation with the fiberoptic bronchoscope inside the
endotracheal tube ensured. The fiberoptic broncho-
scope is then passed through the tracheoesopha-
geal fistula and the gastrostomy incision, and then
the fiberoptic bronchoscope is used to pull the
Fogarty catheter retrograde into the distal esopha-
gus or tracheoesophageal fistula (the fiberoptic
bronchoscope and Fogarty catheter are tied to-
gether with a silk suture). The reader is referred to
the original publication regarding how to untie the
Fogarty embolectomy catheter from the fiberoptic
bronchoscope as well as other details of this com-
plicated, sophisticated procedure.
105
Occlusion of the fistula permits positive-pres-
sure ventilation even in the patient with poor pul-
monary compliance, and it also prevents tracheo-
bronchial aspiration from gastroesophageal reflux.
Of course, the tracheoesophageal fistula can be
divided surgically (without repair of the esopha-
gus) if the Fogarty catheter cannot be properly
located within the distal esophagus or fistula.
105
,06
Patients whose associated anomalies are life-
threatening may require surgery of these anomalies
before esophageal repair.
86
For definitive repair, a right thoracotomy is used
and the dissection is extrapleural. Important sur-
gical tasks include ligation of the fistula between
the trachea and esophagus and esophageal anasto-
mosis (short-gap atresia). If the esophageal seg-
ments cannot be mobilized sufficiently to perform
the anastomosis without tension on the suture line
(long-gap atresia), the upper pouch will be pre-
served, dilated, and stretched until further attempts
at anastomosis are likely to succeed. Esophageal
by-pass operations are being performed with
greatly decreasing frequency.
Depending on the anatomy involved, gastros-
tomy is carried out in the operating room under
Figure 18-14 Therapeutic approach in 100 patients with EA-TEF. All seven hospital deaths occurred in the 32 patients in the ill
baby group. (IRDS = Infantile respiratory distress syndrome; TPN = total parenteral nutrition.) (From Holder TM, Ashcraft KW,
Sharp RJ, Amoury RA: Care of infants with esophageal atresia, tracheoesophageal fistula, and associated anomalies. J Thorac
Cardiovasc Surg 94:828, 1987. Used with permission.)
either general (unusual) or local (usual) anesthesia
(1 per cent lidocaine 2 to 4 mg/kg). The infant
should be maintained in a head-up position to min-
imize the possibility of aspiration, and atropine
(0.01 mg/kg) is administered intravenously on ar-
rival in the operating room. Esophageal stetho-
scopes and temperature probes are avoided for ob-
vious reasons during gastrostomy as well as during
subsequent procedures. Appropriate oxygen is
given as necessary, and careful attention is paid to
maintenance of body temperature and proper fluid,
electrolyte, and glucose administration.
Unfortunately, even patients who have under-
gone a successful repair are often affected by sig-
nificant complications (approximately 50 per
cent).
107
The high incidence of esophageal dysmo-
tility and gastroesophageal flux has been well doc-
umented. Respiratory complications are also com-
mon sources of morbidity, and abnormalities in
lung function several years after repair of esopha-
geal atresia with tracheoesophageal fistula have
been documented. Most commonly, these symp-
toms have been attributed to aspiration resulting
from gastroesophageal reflux.
However, there are many other causes of respi-
ratory complications in patients with repaired
esophageal atresia and a tracheoesophageal fistula,
including esophageal dysmotility, tracheomalacia,
anastomotic stricture, and recurrent or double fis-
tula. Figure 18-15 shows how complicated and
protracted the surgical course of these patients can
be (at least at one respected institution).
107
Premature infants with severe lung disease may
require endotracheal intubation and mechanical
ventilation prior to gastrostomy. Two reports de-
scribed the development of massive life-threaten-
ing air leak out the gastrostomy tube as soon as
the gastrostomy tube was put in.
108
I09
In one case
the leak was successfully dealt with by putting the
gastrostomy tube under a 25-cm H
2
0 seal,
108
and
in the other case it was successfully dealt with by
inflating a Fogarty embolectomy catheter balloon
in the distal esophagus (placed via the gastrostomy
tube with the aid of fluoroscopy).
109
As described
previously, the Fogarty balloon catheter may also
be placed with the aid of a thin fiberoptic broncho-
scope (the fiberoptic bronchoscope is passed ante-
grade through the endotrachea and fistula, and the
Fogarty catheter is then pulled retrograde into the
distal esophagus or fistula).
105
Definitive repair follows the gastrostomy. The
infant should have been premedicated with atro-
pine (.01 mg/kg intravenously) at the time of gas-
trostomy. After suctioning the esosphageal pouch
and gastrostomy, venting the gastrostomy to the
atmosphere, and providing preoxygenation, tra-
cheal intubation is usually done with the infant
awake (some anomalies permit general anesthesia
with positive-pressure ventilation). The appropri-
ate-sized endotracheal tube (usually 3.0 mm) is
passed beyond the vocal cords while breath sounds
are monitored on both sides of the chest.
In patients in whom intermittent positive-pres-
sure ventilation may likely become necessary be-
fore ligation of the fistula (those with known large
fistulas, pulmonary infiltrates and poor compli-
ance, or other congenital anomalies), and assuming
the fistula is above the carina, it may be extremely
important to take the trouble to position the endo-
tracheal tube just above the carina (but below the
Figure 18-15 The algorithm describes the clinical course of 20 patients with successful repair of esophageal atresia and
tracheoesophageal fistula (TEF) with postoperative respiratory symptoms that were initially attributed to gastroesophageal reflux
(GER). Similar complicated algorithms were used for patients whose respiratory symptoms were initially attributed to tracheoma-
lacia, recurrent tracheoesophageal fistula, and esophageal stricture. (From Delius RE, Wheatley MJ, Goran AG: Etiology and
management of respiratory complications after repair of esophageal atresia with tracheoesophageal fistula. Surgery 112:527-532,
fistula) to avoid either a large air leak out the
gastrostomy or gaseous distension of the stomach.
There are three ways to do this (Fig. 18-16).
First, the endotracheal tube (MaGill type without
side hole) (Murphy type has a side hole) is ad-
vanced gently into the right main-stem bronchus
(breath sounds are heard only over the right lung)
and then withdrawn slowly to a position just above
the carina (withdrawal is terminated the moment
breath sounds are heard equally on both sides).
110
The MaGill endotracheal tube is then rotated so
that the bevel faces anteriorly (i.e., the tip of the
tube occludes the fistula). If the endotracheal tube
had a Murphy eye, it could overlie the tracheo-
esophageal fistula, resulting in gaseous distension.
Second, if a gastrostomy tube is in place and the
external end is put under a water seal, then exclu-
sion of the fistula by the tip of the endotracheal
tube will be indicated by the failure of bubbles to
appear in the water seal from the gastrostomy tube
when the trachea is pressurized.
1
"
Third, the gastrostomy tube may be connected
to the sampling catheter of a side-stream type cap-
nogram or dedicated anesthesia gas analyzer; when
the endotracheal tube tip is proximal to the fistula,
the presence of C0
2
and inhaled agents is detected
Three Methods to Assure
Endotracheal Tube (ETT) is Just Distal to
Tracheo-Esophageal Fistula (TEF) but Above Carina
3) Connect
Gastrostomy tube
to Gas Analyzer
Figure 18-16 The schematic shows three physiologic tests to determine when the endotracheal tube (ETT) is above the carina
but below the tracheoesophageal fistula (TEF).
by the analyzer."
2
When the tip is immediately
distal to the fistula, no C0
2
or inhaled agents are
detectable. Thus, during gentle positive-pressure
ventilation via the endotracheal tube, the intermit-
tent analysis of the gases emerging from the gas-
trostomy tube facilitates precise location of the
endotracheal tube. This technique facilitates the
repeated rapid determination of the optimum posi-
tion of the endotracheal tube tip.
Finally, although cases have been described
using selective bronchial intubation and one-lung
ventilation,
113
doubt has been expressed regarding
the wisdom of this approach (blockage of ipsilat-
eral upper lobe, technical difficulties)."
2
Although intubation of the fistula itself is a rare
event (the fistula is on the posterior aspect of the
trachea), rotating the tube so that the bevel faces
posteriorly as the tube is actually being inserted,
as opposed to withdrawing from the right main-
stem bronchus and rotating the tube so that the
bevel faces anteriorly and blocks the fistula (see
previous discussion), may diminish this possibility
even further. Intubation of the fistula itself is rec-
ognized when the endotracheal tube is within the
trachea and positive-pressure ventilation of the
lungs is impossible, while, at the same time, the
stomach becomes distended or the gastrostomy
tube leaks air. Some leakage of air through the
tracheoesophageal fistula may occur even with the
endotracheal tube tip below the level of the fistula,
and the amount of leakage may be large if the
tracheoesophageal fistula is large. Tube position
must be reconfirmed after positioning the child in
the left lateral decubitus position. The best way to
continuously monitor breath sounds during the
case is with a precordial stethoscope in the left
axilla.
It should be realized that, with the endotracheal
tube so close to the carina, intraoperative manipu-
lation of the mediastinum may easily cause the
endotracheal tube to become malpositioned. In-
deed, even when a gastrostomy tube is in place,
approximately 20 per cent of patients will experi-
ence at least one episode of sudden respiratory
embarrassment (defined as P
a
C0
2
> 50 mm Hg or
sudden loss of ventilatory volumes and pres-
sure)."
4
Anesthesia is induced with nitrous oxide, oxy-
gen, and small amounts of halothane with sponta-
neous ventilation. If halothane is not tolerated, a
nitrous oxide, narcotic, or ketamine technique can
be used. If possible, the infant should be allowed
to breathe spontaneously until the fistula is ligated
(Table 18-6). This may be important in some pa-
tients (e.g., those who do not have a gastrostomy
tube) as severe gastric distension can impede ven-
tilation (compress the lung) and circulation (de-
crease venous return), leading to cardiopulmonary
arrest
94
and gastric rupture."
5
However, it should
be realized that in most patients with a gastros-
tomy that is vented to atmospheric pressure the
danger of gastric distension is minimized and it is
usually possible to assist ventilation with gentle
positive pressure. Similarly, positive-pressure ven-
tilation should be cautiously attempted before ad-
ministering muscle relaxants. If gastric distension
occurs with gentle (low) positive-pressure ventila-
tion, the function of the gastrostomy tube and the
position of the endotracheal tube should be
checked, nitrous oxide should be discontinued, and
perhaps a return to spontaneous breathing or the
use of high-frequency ventilation should be con-
sidered."
6
During the procedure, air entry into the depen-
dent lung should be monitored continuously by a
stethoscope secured in the dependent axilla. Since
secretions and obstruction of the endotracheal tube
can be a problem at any time, suction apparatus
and catheters should be readily available for suc-
tioning. Airway obstruction may also result from
surgical manipulation of the trachea; the surgeon
must be informed when this occurs. Occasionally,
the small endotracheal tube can become occluded
by thick secretions, which requires prompt passage
of a suction catheter and, if that fails to clear the
tube, a marked stylet.
The integrity of the fistula repair can be ensured
by the surgeon placing a small amount of saline
into the wound and the anesthesiologist applying
20 cm H
2
0 constant positive-pressure ventilation
to the airway and looking for any leak in the tra-
cheal-suture line. If, after ligation of the fistula,
primary repair of the esophageal atresia is under-
taken with anastomosis of the proximal and distal
segments, the surgeon may ask the anesthesiolo-
gist to pass a small suction catheter into the prox-
imal segment until it can be felt near the end of
the atretic esophageal pouch. When the catheter is
palpable near the end of the pouch, and with the
infant's head in a neutral or slightly flexed posi-
tion, the anesthesiologist can mark on the catheter
with a suture or umbilical tape the level of the
infant's upper alveolar ridge. The catheter is then
withdrawn and carefully measured to determine
the maximum safe distance that catheters should
be inserted for postoperative oropharyngeal suc-
tioning without causing disruption of the esopha-
geal anastomosis. Extension of the neck after re-
pair is avoided for the same reason.
Most infants undergoing repairs for esophageal
atresia and tracheoesophageal fistula are left intu-
bated for several hours to several days after sur-
gery. Weaning from the ventilator begins as soon
as possible after blood-gas values and radiologic
findings indicate improvement in atelectasis or
pneumonia. Feedings are usually started via gas-
trostomy on the third postoperative day. Unless the
esophageal anastomosis is tenuous, oral glucose
feedings are begun on the sixth to seventh postop-
erative day after an esophagogram is performed to
document patency of the esophageal lumen.
1
"
V. LIGATION OF PATENT DUCTUS
ARTERIOSUS IN PREMATURE
INFANTS
The vasoconstrictor response of the ductus arte-
riosus to an increase in the partial pressure of
oxygen in the early neonatal period is proportional
to gestational age; the lower the gestational age,
the greater is the likelihood of the ductus remain-
ing open, and the greater the gestational age, the
greater is the likelihood of ductal closure. Forty
per cent of infants weighing less than 1750 g at
birth will have clinical evidence of patency of the
ductus; whereas, in those weighing less than 1200
g, the incidence rises to 80 per cent."
7
Initial ther-
apy of patent ductus arteriosus is medical and con-
sists of fluid restriction and administration of di-
uretics and, perhaps, digoxin in an attempt to
decrease the effects of volume overload on the
cardiovascular system (see later discussion)."
8
If
this usual medical therapy fails to close the ductus,
then indomethacin, a prostaglandin synthetase in-
hibitor, should be administered in an attempt to
produce pharmacologic closure of the ductus."
8
The earlier indomethacin is administered after a
patent ductus arteriosus has become clinically ap-
parent, the more effective it is in closing the due-
tus.
119
Surgical closure is indicated if the patient
still has significant shunting (see Surgical Consid-
erations following) after the usual medical and in-
domethacin therapies have been tried. However,
routine surgical closure of patent ductus arteriosus
in premature infants was found to be superior to
chemical closure with indomethacin; there was a
15 per cent overall morbidity and a 17 per cent
mortality for operative closure compared with a 51
per cent morbidity and a 40 per cent mortality for
chemical closure.
120
There are two main well-established indications
for ligation of a patent ductus arteriosus in the
premature infant (as the only surgical procedure).
First, ligation is indicated to decrease severe right-
to-left shunting in premature infants with persist-
ent fetal circulation. Second, ligation is indicated
to decrease severe left-to-right shunting (and there-
fore ventricular overload) in patients with improv-
ing respiratory distress syndrome. These cases are
usually done in the neonatal intensive care unit to
avoid the perils of transporting these critically ill,
tiny infants to the operating room. A left or ante-
rior thoracotomy is utilized.
121
The vast majority of these patients will already
be intubated and mechanically ventilated, and
some will have already been treated with nonde-
polarizing muscle relaxants. These infants will al-
most always have monitoring and intravenous
catheters in place. Anesthetic management is usu-
ally a continuance of preoperative supportive
measures; these consist of avoidance of hypoxemia
and hyperoxemia and maintenance of normother-
mia, fluid and electrolyte and caloric balance, and
adequate gas exchange.
If the patients are not paralyzed, they should be.
Anesthesia may be local, or local plus a moderate
dose of narcotic, and if the operation is performed
in the operating room, local anesthesia plus low
concentration of inhalation agent/air/oxygen may
be used. Although blood loss is usually minimal,
it should be remembered that the ductus is friable
and mobilization of the structure in order to ligate
it may result in tearing and sudden massive blood
loss; the anesthesiologist must be in a position to
deal with massive blood loss (must have adequate
vascular access).
VI. VASCULAR RINGS
122
A double aortic arch, a right aortic arch along
with a left-sided ligamentum arteriosum, a left pul-
monary artery sling, and an anomalous origin of
the innominate artery (the "innominate artery
compression syndrome") are vascular abnormali-
ties that can cause a significant tracheobronchial
tree obstruction (usually near the carina) and
esophageal obstruction that requires surgery for
relief. The first two anomalies are true "vascular
rings," in that they form a continuous circle
around the trachea and esophagus and compress
them.
Vascular rings usually cause symptoms in early
childhood. The symptoms are referable to com-
pression of the tracheobronchial tree or esophagus,
or both. Chest computed tomography usually dem-
onstrates these obstructions well. In addition, the
diagnosis can also be established by a chest radio-
graph with barium swallow, the barium-filled
esophagus being indented posteriorly by the vas-
cular structure. Bronchoscopy with or without
bronchography will clearly demonstrate the flat-
tened and obstructed trachea just above the carina.
Operation is performed through a left thoracot-
omy except for correction of the innominate artery
compression syndrome, which is repaired through
a right thoracotomy.
122
The tracheal compression
caused by the vascular ring can be relieved by
division of the ring at the appropriate point; care
is required to preserve the vessels needed for cir-
culation to the head and lower body. Palpation of
appropriate pulses (carotid and radial) will aid this
vascular preservation.
The vascular ring caused by a double aortic arch
is released by dividing the lesser of the two arches,
usually at an atretic area. When both arches are
patent, the lesser is divided between clamps at a
site selected to preserve the brachiocephalic blood
flow (as determined by the carotid and radial
pulses, which are monitored by the anesthesiolo-
gist after the vascular clamps are applied). Correc-
tion of the innominate artery syndrome requires
only that the innominate artery be pulled forward
away from the trachea that it is compressing. The
artery is usually sutured to the posterior perios-
teum of the sternum to hold it away from the
trachea. The results of operation are excellent in
most cases. The vascular ring caused by a right
aortic arch and a left ligamentum are released by
dividing the ductus ligamentum and dissecting the
trachea and esophagus free from adhesive bands.
The pulmonary arterial sling may require the
use of cardiopulmonary by-pass for correction. In
this anomaly, the left pulmonary artery arises from
the right pulmonary artery and passes to the right
of the trachea before crossing back between the
trachea and esophagus (posterior to the trachea and
anterior to the esophagus) to the left lung hilum.
This "sling" thus pulls on and compresses the
trachea and right main-stem bronchus at the ca-
rina.
123
Correction requires division of the left pul-
monary artery, removal of the left pulmonary ar-
tery from behind the trachea, and reanastomosis of
the vessel to the main pulmonary artery just distal
to the pulmonary valve. Alternatively, the stenotic
trachea may be resected first and the reanastomosis
of the trachea performed in such a way so that the
left pulmonary artery passes anterior to the recon-
struction trachea.
124
Fortunately, ventilation usually improves once
the endotracheal tube is in proper position. The
trachea is usually compressed just above the ca-
rina, so the tube must often be inserted more
deeply into the trachea. Occasionally, a direct right
main-stem bronchial intubation is required; in this
situation, a side hole cut near the end of the endo-
tracheal tube will allow ventilation of the left lung
(see Table 18-6). The surgeon will need to par-
tially collapse the left lung to gain exposure of the
vascular structures. Comparison of palpable pulses
in the head and arms may assist the surgeon in
identifying the proper location for division of the
ring (see Table 18-6). Postoperative relief of tra-
cheal obstruction is usually clearly evident and
dramatic. Occasionally, however, substantial tra-
cheomalacia may have been caused by the chronic
presence of the compressing vessel. In these chil-
dren respiratory symptoms may persist postopera-
tively and may become more severe when the
child is agitated since the trachea is more likely to
collapse during forceful inspiration.
VII. CONGENITAL PARENCHYMAL
LESIONS: LOBAR EMPHYSEMA,
CYSTS, SEQUESTRATIONS, AND
CYSTIC ADENOMATOID
MALFORMATIONS
Congenital lobar emphysema is a pathologic ac-
cumulation of air in one lobe of the lung, usually
an upper lobe or the right middle lobe. The cause
of this condition is unknown in over half of the
reported cases.
125-129
In those cases with a defined
cause, extrinsic bronchial compression (vessels or
enlarged lymph nodes) or intrinsic bronchial ob-
struction (bronchial cartilaginous dysplasia, bron-
chial stenosis, or redundant bronchial mucosa) is
most commonly cited. Bronchial cartilaginous
dysplasia is the most common cause (25 per cent)
of all cases, and the abnormality may involve the
main-stem, lobar, or sublobar bronchi; the sof-
tened cartilage collapses on expiration, trapping air
beyond the collapse.
Congenital bronchogenic cysts are derived from
abnormal budding of the primitive tracheobron-
chial tree.
130
When the abnormal bronchial bud-
ding occurs at the level of the carina or first-order
bronchi, the cyst assumes a mediastinal or subca-
rinal location (30 per cent). When the distal tra-
cheobronchial tree is the origin of the abnormal
budding, an intraparenchymal bronchogenic cyst
results (70 per cent). When peripheral cysts are
generalized (multiple lobes, both lungs), the con-
dition is called the "honeycomb" lung and the
infant dies soon after birth.
131
Congenital bronchogenic lung cysts are usually
not hyperinflated but, unfortunately, they may be
more infected as a result of decreased bronchial
communication. In addition, the bronchial com-
munication, when it exists, may cause valvular air
trapping, tension cyst, and cardiorespiratory em-
barrassment. With congenital lung cysts, the re-
maining lung may be somewhat hypoplastic, re-
sembling the situation found in congenital
diaphragmatic hernia. Further, a congenital cyst
may be supplied by an anomalous systemic vessel,
which may represent a form of pulmonary seques-
tration.
Pulmonary sequestration is characterized by the
presence of lung tissue that lacks bronchial com-
munication with the normal tracheobronchial tree
and a blood supply that is derived from a systemic
arterial source.
130
Sequestration can be extralobar
(with its own separate visceral pleura) or intralobar
(incorporated within the normal visceral pleural
investment).
4. Cystic Adenomatoid Malformations
If the bronchial buds and alveolar mesenchyma
fail to join at 16 to 20 weeks of gestation, uncon-
trolled overgrowth of the bronchioles, especially
the terminal branches, occurs and is characteristic
of cystic adenomatoid malformations.
130
The non-
communicating alveolar ducts and air sacs develop
from the surrounding mesenchyma and contribute
the alveolar component to this congenital anomaly.
Cystic adenomatoid malformations are thus not
true hamartomas (the lesion is not composed of an
intermixture of epithelium, cartilage, fat, and other
tissues that are usually called a hamartoma of the
lung). Because the amount of bronchiolar and al-
veolar components can vary within the mass, the
tumors can range from solid masses without cysts
to mixed solid and cystic masses within the lungs.
The disease usually involves only a single lobe.
There is a high incidence of congenital heart dis-
ease in patients with these congenital pulmonary
parenchymal problems.
132
"
135
Although both car-
diac and pulmonary lesions may be severe, pul-
monary hyperinflation or infection is generally the
more disruptive situation, and its correction is usu-
ally required before cardiac surgery can be con-
templated.
136
The clinical presentation of clinical lobar em-
physema is quite variable. The usual age of pres-
entation is from birth to 4 months,
137
although an
occasional child may present after 12 months.
136
The symptoms are similar to those of a pneumo-
thorax, although in congenital lobar emphysema
the air is not extrapleural. The direct compression
of surrounding lung and compression of the con-
tralateral lung by mediastinal shift results in dysp-
nea and cyanosis being the most frequent symp-
toms. Because of the mediastinal shift, there may
also be severe cardiovascular compromise (de-
creased venous return, systemic hypotension).
Findings on a chest radiograph include a unilat-
eral radiolucency from lobar distension, compres-
sion atelectasis of surrounding lung, mediastinal
shift away from the affected side, and a flattening
of the ipsilateral diaphragm. The presence of bron-
chovascular markings may help to differentiate
congenital lobar emphysema from congenital lung
cysts (with which it is most often confused). Con-
genital lobar emphysema is almost exclusively a
disease of the upper lobes or the right middle lobe.
Once the diagnosis is made, lobectomy is con-
sidered mandatory. Conservative therapy such as
needle aspiration, placement of chest tubes, and
administration of oxygen and antibiotics simply
does not work and has been associated with a
mortality rate of greater than 50 per cent, whereas
operative mortality is 3 to 7 per cent.
138
Long-term
follow-up of children who have undergone pul-
monary resection for congenital lobar emphysema
problems in infancy shows minimal physiologic
difference from unaffected individuals.
139
Mediastinal cysts usually become evident clini-
cally because they cause partial respiratory tract
obstruction and distal pulmonary infection. Non-
communicating mediastinal cysts appear as soft
tissue masses on chest roentgenograms. The pres-
ence of air or air fluid levels within a mediastinal
cyst suggests some sort of tracheobronchial (or
enteric) communication. In infants whose medias-
tinal bronchogenic cysts cause partial bronchial
obstruction with distal air trapping, the radio-
graphic picture can be similar to that in congenital
lobar emphysema (tension cyst). Because only half
of these infants with bronchogenic cysts who have
lobar emphysema have mediastinal masses visible
on their initial chest radiographs, the surgeon must
consider the possibility of an unrecognized bron-
chogenic cyst whenever an exploratory thoracot-
omy is undertaken for congenital lobar emphy-
sema.
When the bronchogenic cyst is intraparenchy-
mal, it usually shows air-fluid levels due to a par-
tially obstructed tracheobronchial tree communi-
cation. Infection often supervenes, and cysts may
be indistinguishable from cavitating pneumonia or
lung abscesses.
140
Intraparenchymal bronchogenic
cysts must be resected, and in cases in which there
is superimposed infection of the surrounding nor-
mal lung, antibiotic administration to control the
inflammatory process should precede operative in-
tervention. With congenital lung cysts, anomalous
systemic vessels from the hilum (or other places)
must be identified; failure to do so has resulted in
exsanguination.
37
Extralobar sequestrations are often recognized
by 1 year of age as incidental findings on chest
radiographs. Respiratory insufficiency, associated
cardiovascular or thoracic anomalies, and feeding
problems are the most common symptoms stimu-
lating evaluation. In 90 per cent of extralobar
cases, the sequestration resides in the posterior left
costophrenic sulcus adjacent to the esophagus.
141
The systemic arterial supply must be identified.
Intralobar sequestrations are evident relatively
late compared with extralobar sequestrations. Few
intralobar sequestrations are recognized prior to 1
year of age. The most frequent symptom is persist-
ent or recurrent pneumonia within the same bron-
chopulmonary segment. Angiography is diagnostic
of the pulmonary sequestration because it can de-
fine the anomalous systemic blood supply to the
sequestered segment; the location and number of
systemic vessels perfusing the anomaly and the
venous drainage present must be clearly identified.
Pulmonary sequestrations should be resected. In
patients with intralobar sequestrations complicated
by pneumonia, infection should be controlled prior
to thoracotomy with appropriate antibiotics. Lo-
bectomy is the preferred procedure in most pa-
tients. In both intra- and extralobar forms, a careful
search of the thorax must be undertaken prior to
resection to identify and control the aberrant sys-
temic artery or arteries.
The most common presentation of cystic ade-
nomatoid malformation is progressive respiratory
distress in a newborn infant that is associated with
a complex cystic, solid, partially air-filled pulmo-
nary mass on chest radiograph. Typically, the mass
displaces the mediastinum to the contralateral side,
causing contralateral atelectasis, which intensifies
the respiratory insufficiency. Radiographically, the
abnormality produced can be indistinguishable
from a congenital posterolateral diaphragmatic
hernia.
142
-
I43
Differential diagnosis of this rare con-
dition also includes congenital lobar emphysema,
bronchogenic cyst, staphylococcal pneumonia with
pneumatocele formation, tension pneumothorax,
and a sequestered lobe. Cystic adenomatoid mal-
formations can also cause recurrent infections and
become abscessed because the cystic cavities have
poor bronchopulmonary drainage.
The treatment of cystic adenomatoid malforma-
tions is pulmonary resection, almost always a
lobectomy. The malformation may have an abnor-
mal systemic arterial supply from an infradia-
phragmatic aortic branch. This vascular supply,
similar to that of an intralobar sequestration,
should be routinely searched for.
The anesthetic considerations for these lesions
are dominated by the need to avoid either air trap-
ping caused by positive-pressure ventilation and
the need to separate the lungs if the lesion is in-
fected (abscessed). All of the lesions described
previously may present clinically in adulthood. In-
deed, intralobar sequestrations present almost ex-
clusively in the first to second decade of life as
recurrent pneumonia.
144
Atropine should be administered prior to the
induction of anesthesia. A gentle induction of an-
esthesia is performed with spontaneous ventilation
of a volatile agent and 100 per cent oxygen.
137
Alternatively, intramuscular ketamine (6 to 8
mg/kg) may be used, followed by spontaneous
ventilation of a volatile drug.
137
Struggling must
be avoided to prevent further air trapping. How-
ever, the volatile drug must be administered with
care because, if the mediastinum is distorted, the
volatile drug may cause cardiovascular depression,
which is unresponsive to volume infusion. Nitrous
oxide and positive-pressure ventilation are con-
traindicated because of the danger of further ex-
panding the emphysematous lobe (see Table 18-
6).
66
136
The trachea is intubated without the use of
muscle relaxants, and spontaneous ventilation is
allowed until the chest is opened. Occasionally, it
may be useful to intubate the contralateral main-
stem bronchus in order to prevent further expan-
sion of the lobe as well as facilitate surgical expo-
sure (see Thoracic Surgical Procedures Requiring
One-Lung Ventilation later).
If the patient demonstrates cardiovascular
depression after the induction of anesthesia, it is
helpful if the incision is infiltrated with local an-
esthetic in order to lighten the level of anesthesia.
Once the chest is opened, the emphysematous lobe
usually herniates through the incision, which elim-
inates compression of the ipsilateral normal lung
and mediastinal shift. Occasionally, emergency
thoracostomy may be required during the induc-
tion of anesthesia to let the lobe expand out of the
hemithorax. Controlled ventilation is instituted at
this point and may be facilitated by muscle relax-
ants if desired. High-frequency jet ventilation (rate
= 150 breaths/min, driving pressure = 5 psi) has
been successfully used after the chest has been
opened in an effort to avoid positive-pressure ven-
tilation and air trapping.
145
Postoperatively, the
previously compressed ipsilateral lung may not al-
ways expand immediately.
137
The anesthetic management of congenital bron-
chogenic cysts shares many of the same features
as the management of congenital lobar emphysema
(for example, the need to avoid nitrous oxide and
preserve spontaneous ventilation) (see Table 18-
6). A cyst that develops great tension may need to
be aspirated by needle.
131
The two lungs may need
to be separated if the cyst is infected (see Thoracic
Surgical Procedures Requiring One-Lung Ventila-
tion later). The management of mediastinal cysts
has been considered in chapter 14.
If the sequestration is infected, the lungs may
need to be separated (see Thoracic Surgical Pro-
cedures Requiring One-Lung Ventilation later).
Otherwise, the anesthetic considerations are simi-
lar to those for other types of resections.
If the malformation is abscessed, then separa-
tion of the two lungs becomes extremely important
(see Thoracic Surgical Procedures Requiring One-
Lung Ventilation later). High-frequency positive-
pressure ventilation has also been used intraopera-
tively for these cases.
146
Otherwise, the anesthetic
considerations are similar to those for other types
of resections and include massive hemorrhage if a
systemic arterial supply is not identified.
VIII. THORACIC SURGICAL
PROCEDURES REQUIRING ONE-
LUNG VENTILATION
The indications for one-lung ventilation in pe-
diatric patients are the same as for adults. How-
ever, the most common indication for lung sepa-
ration is the presence of infection in one lung,
which could potentially contaminate the other
lung. The infections consist of lung abscess, em-
pyema, and hydatid cysts. Lung abscess may be
caused by foreign bodies, infection within true
congenital lung cysts, lung sequestration, and hy-
poplastic lobes and may occur following pneumo-
nia. Other indications for lung separation consist
of bronchiectasis and bronchopleural fistula. There
are basically three ways to separate the two lungs
in pediatric patients, and the methods consist of
use of a bronchial blocker such as a Fogarty em-
bolectomy catheter, main-stem bronchial intuba-
tion, and a combination of the two. Finally, use of
the prone position may minimize the risk of cross
contamination. See chapter 11 for the ventilatory
management of one-lung ventilation (use 100 per
cent oxygen, decrease the two-lung tidal volume
by 20 per cent, increase the two-lung respiratory
rate by 20 per cent).
A. Bronchial Blocker
A 3 French gauge Fogarty embolectomy cathe-
ter is a balloon-tipped catheter with a capacity of
0.5 ml. A fully inflated balloon has a length of 10
mm and a width of 7.5 mm. When the patient is
sufficiently deeply anesthetized, the tracheobron-
chial tree can be sprayed with local anesthetic, and
either a rigid bronchoscope (3.2-3.6-mm OD) or a
fiberoptic bronchoscope (1.8 mm or greater
OD),
147
148
which is inside an endotracheal tube, is
passed into the appropriate main-stem bronchus
(Fig. 18-17). Under direct vision, the Fogarty em-
bolectomy catheter can be placed in the appropri-
ate main-stem bronchus, and the Fogarty catheter
is inflated with 0.5 ml of sterile normal saline. If a
rigid bronchoscope is used, the patient can be ven-
tilated via the side arm. The bronchoscope is then
removed, and the patient is ventilated and oxygen-
ated. A single-lumen tube (3.5 to 4.5 mm OD) is
then placed in the trachea. The bronchial blocker
may also be guided into place with fluoroscopy
149
(Fig. 18-17).
Alternatively, and if the patient is large enough,
a single-lumen tube may be placed first and con-
nected to an elbow adapter with a self-sealing dia-
phragm (Fig. 18-17). The pediatric fiberoptic
bronchoscope (1.8 mm OD or greater) is then
passed through the self-sealing diaphragm and
identifies the carina. Then, depending on the size
of the endotracheal tube used, the Fogarty embo-
lectomy catheter can be passed alongside or within
the endotracheal tube (also through the self-sealing
diaphragm) into the proper main-stem bronchial
position (Fig. 18-17).
The position of the single-lumen endotracheal
tube and bronchial blocker should be confirmed
with a chest X-ray. The bronchial blocker can be
readily appreciated on chest radiograph if a radi-
opaque material is instilled into the blocking bal-
loon.
It is important to realize that the bronchia
blocker may back out of the main-stem bronchui
at any time. Slight dislodgment may be indicates
by the appearance of infected secretions. Complei
dislodgment into the carinal area results in inabil
ity to ventilate either of the lungs due to blockagi
of the main airway by the inflated balloon. Unde
these circumstances, the balloon must be immedi
ately deflated to allow ventilation of both lungs. I:
addition, with dislodgment of the bronchia
blocker, the separation of the two lungs is lost; t
prevent bilateral flooding of infected.material, th
surgeon may need to manually seal off the area i
question with either a rubber snare or a bronchi;
clamp. These potentially catastrophic events ma
happen especially after external handling of th
main-stem bronchus in question by the surgeon.
B. Main-Stem Bronchial Intubation
44 45
The main-stem bronchus of the healthy lung
the bronchus that requires intubation. When tl
left lung is diseased and the right main-stem broi
chus is to be intubated, insertion of a single-lum<
tube with a left-facing bevel (tip of endotrache
tube is to the right of center) deep within tl
tracheobronchial tree will result in right main-ste
bronchial intubation in 100 per cent of cases (Fig
Insertion of Bronchial Blocker
Figure 18-17 Insertion of a bronchial blocker may be accomplished by a number of methods depending on the age of the patienr
and the size of endotracheal tube (ET) that can be used.
18-18 and 18-19). The single-lumen endotracheal
tube can then be withdrawn slowly until breath
sounds are heard over all of the right hemithorax.
When the right lung is diseased and the left
main-stem bronchus requires intubation, use of a
fiberoptic bronchoscope, fluoroscopy, a right-fac-
ing bevel on the endotracheal tube (a standard
endotracheal tube may be cut to create a right-
facing bevel),
45
turning the head to the right and
the endotracheal tube 180 degrees so that the nat-
ural concavity of the tube is facing posteriorly,
44
or a curved stylet will be necessary (see Figs. 18-
18 and 18-19). With a fiberoptic bronchoscope or
fluoroscopy,
149
precise positioning is possible. If
one uses (1) the bevel of the endotracheal tube, (2)
the combination of turning the head to the right
and rotating the tube 180 degrees, or (3) a curved
stylet, the tube may need to be withdrawn until
breath sounds are heard over the entire left hemi-
thorax. Following any method of insertion, the
position of the main-stem bronchial endotracheal
tube should be further confirmed by chest radio-
graph. With either main-stem bronchial intubation,
the surgeon can further check the position of the
tube at the time of thoracotomy.
C. Bronchial Blockade Plus Main-Stem
Bronchial Intubation
To more securely guarantee lung separation, a
bronchial blocker can be used in conjunction with
main-stem bronchial intubation.
147
A bronchial
blocker must first be placed under bronchoscopic
(either rigid or fiberoptic, see earlier discussion) or
fluoroscopic control (Fig. 18-17) and the main-
stem bronchial intubation under fiberoptic bron-
choscopic and/or fluoroscopic control (Fig. 18-
18). When the bronchial blocker is used with a
main-stem bronchial intubation, a curved stylet is
relatively contraindicated because, if the single-
lumen endotracheal tube enters the main-stem
bronchus in which the bronchial blocker sits, ven-
tilation will not be possible, and it may cause mal-
positioning of the bronchial blocker. Finally, the
position of the bronchial balloon and main-stem
Mainstem Bronchial Intubation
Figure 1818 There are several ways to accomplish a left main-stem bronchial intubation, whereas a right main-stem bronchial
intubation can almost always be achieved by simply inserting the endotracheal tube (ETT) deep within the tracheobronchial tree.
See Figure 18-19 for more details on using the bevel of the tube and other non-equipment-aided techniques of left main-stem
cannulation.
bronchial ventilating catheter should be con-
firmed with chest radiograph preoperatively and
by the surgeon intraoperatively as described pre-
viously.
IX. DIAGNOSTIC BRONCHOSCOPY
D. Prone Positioning
The prone position may be used in patients who
have lateral and posterior lesions. Following anes-
thesia induction and intubation, the children are
placed in the prone position, with the head slightly
lower than the chest, thereby allowing secretions
and blood to drain forward without spillage into
the contralateral lung. The surgeon may find that
access to the lung is facilitated, ventilation is more
easily provided, and pulmonary collapse has not
been a problem.
Indications for diagnostic bronchoscopy in chil-
dren include determination of a noninfectious
structural cause of stridor, such as laryngotracheal
malacia, laryngeal webs, tracheal masses, vocal
cord paralysis, and postintubation or posttrache-
otomy stenosis (as opposed to an infectious cause
such as laryngotracheitis, epiglottitis, and so on);
identification of the origin of hemoptysis; the in-
vestigation of persistent pneumonia or atelectasis
and the diagnostic evaluation of a suspected tra
cheoesophageal fistula (in combination witl
esophagoscopy) and a mediastinal mass. Thera
peutic bronchoscopy in children is usually done t<
remove aspirated foreign bodies and was discussei
in detail in chapter 17.
Cannulation of a Mainstem Bronchus
with a Single Lumen Tube
Right
Mainstem
Cannulation
Left
Mainstem
Cannulation
Figure 18-19 Turning the face of the bevel of the endotracheal tube to the side opposite the main-stem bronchus to be cannulated
places the tip of the tracheal tube off center toward the side to be cannulated.
45
In addition, rotation of the endotracheal tube 180
degrees so that the natural concavity of the tube faces posteriorly and turning the face/head to the right results in cannulation of the
left main-stem bronchus a high percentage of the time (92 per cent).
44
The surgical management of pediatric patients
for bronchoscopy has improved significantly with
the introduction of various-sized bronchoscopes
that have a miniature lens system, improved illu-
mination, and a standard 15-mm connector for the
anesthesia circuit. The surgeon must select a tele-
scope that allows enough room for ventilation
around the telescope but within the broncho-
scope.
150
For example, a poor choice would be to
put a 4.0-mm OD telescope into a 4.9-mm internal
diameter bronchoscope; in this case, the 2.8-mm
OD telescope should be used.
150
The consequences
of using a telescope that is too large inside a rigid
bronchoscope are inability to ventilate with posi-
tive pressure and inability to exhale with any form
of ventilation, leading to air trapping.
150
With bronchoscopes larger than 3 mm, small
biopsy forceps or grasping forceps can be passed
through the instrument channel while ventilation is
maintained with the telescope in place.
151
The
bronchoscope should probably not be passed
through an area of severe narrowing because of
the possibility of postoperative swelling.
C Anesthetic Considerations
If an intravenous line is not present, anesthesia
can be induced by intramuscular injection of keta-
mine or by inhalation of halothane/nitrous ox-
ide/oxygen. Alternatively, if an intravenous line is
in place, anesthesia can be induced with intrave-
nous thiopental or ketamine and maintained by
inhalation. Atropine (0.02 mg/kg) should be ad-
ministered intravenously at the time of induction
of anesthesia for its effect on secretions and vagal
tone.
When the patient is adequately anesthetized, lar-
yngoscopy is performed and the trachea is sprayed
with lidocaine (3 to 4 mg/kg). Nitrous oxide is
discontinued if previously used, and anesthesia is
continued with halothane and oxygen by mask for
another 2 to 3 min. Lidocaine (1.0 mg/kg) may be
infused during this time. The rigid bronchoscope
is then inserted (may need to be facilitated by the
administration of succinylcholine), and the anes-
thesia circuit is attached to the standard 15-mm
sidearm connector. Since this is almost a closed
system (some gas leakage may occur around the
outside of the bronchoscope), ventilation can be
controlled (see Fig. 14-10). Controlled ventilation
690
Anesthesia for Pediatric Thoracic Surgery
is desirable in most patients because the work of
spontaneously breathing through the small, high-
resistance, pediatric bronchoscopes is great, and
the patients must be well anesthetized and, there-
fore, will have decreased central drive to sponta-
neously breathe. If the gas leak around the bron-
choscope during positive-pressure ventilation is
large, this can usually be remedied by pharyngeal
packing with saline-soaked gauze and/or by gentle
compression of the cricoid cartilage between the
thumb and forefinger against the bronchoscope.
With either controlled or spontaneous ventilation,
if chest wall rigidity, forceful exhalation, or bron-
chospasm occurs during the case (usually the re-
sult of inadequate anesthesia), muscle relaxation
may be needed quickly, but should not be achieved
with multiple small doses of succinylcholine since
severe bradycardia may occur after the second or
third dose. Rather, the first bolus of succinylcho-
line should be followed by a 1 to 2 mg/ml succi-
nylcholine drip to maintain 50 per cent twitch
depression on the nerve stimulator. Alternatively,
mevacurium or atracurium may be used, if appro-
priate, for the anticipated length of the procedure.
Muscle relaxation, of course, necessitates con-
trolled positive-pressure ventilation.
It must be stressed that anesthesia and ventila-
tion must be adequate and suctioning limited to
only short periods of time followed by lung infla-
tion.
152
153
The combination of hypoventilation and
prolonged suctioning, both of which may lead to
hypoxia, and the presence of light anesthesia may
result in cardiac arrhythmias. These arrhythmias
are usually resolved with manual hyperventilation,
adequate oxygenation, and deepened levels of an-
esthesia. If these measures are unsuccessful, lido-
caine (1 mg/kg) is given intravenously to control
cardiac irritability.
Although not commonly performed in the pedi-
atric population, flexible fiberoptic bronchoscopy
can be done with the patient awake or lightly se-
dated.
154
If the patient's trachea is already intu-
bated or general anesthesia is required for the pro-
cedure, the bronchoscope can be passed through
an endotracheal tube. The smallest endotracheal
tube size that can be used with a 3.2-mm and a
1.8-mm bronchoscope is 4.5 and 3.0 mm, respec-
tively. The flexible fiberoptic bronchoscope must
be introduced through a self-sealing diaphragm in
the elbow connector to the endotracheal tube in
order to maintain an airtight closed system.
155
An-
esthesia technique is as described for rigid bron-
choscopy except that ventilation must be con-
trolled because of the increased resistance to
airflow in the space around the bronchoscope but
within the lumen of the endotracheal tube.
There are four ways of performing flexible fi-
beroptic bronchoscopy in very small children, in
whom the resistance to ventilation around the fi-
beroptic bronchoscope but within the endotracheal
tube is too large (Fig. 18-20). First, the flexible
fiberoptic bronchoscope may be passed using the
laryngeal mask airway as the conduit (Fig. 18-
20).
156
In this situation, the space between the
shaft of the laryngeal mask airway (either size 1
or 2) and an appropriately sized fiberoptic bron-
choscope is ordinarily large and permits adequate
ventilation.
Second, the patient may breathe via a mask con-
nected to a bronchoscopy elbow that has a self-
sealing diaphragm (Fig. 18-205).
Third, the fiberoptic bronchoscope may be
passed through an anesthesia mask with self-sealj
ing diaphragm (Fig. 18-20C). In all three of these
methods, the airway can be examined with little
interference with ventilation. In all three methods
the larynx can be viewed (which is not the cas*
when an endotracheal tube serves as the condui
for the fiberoptic bronchoscope) and there is lesi
trauma to the tracheal mucosa.
Fourth, in the rare and unusual situation of se
vere high tracheal and laryngeal obstruction in in
fants, the institution of transtracheal jet ventilatioi
not only allows an unimpeded examination of thj
larynx but may be life-saving
157
irrespective o
whether a fiberoptic or rigid instrument is subse
quently used for the procedure (Fig. 18-20D).
Bronchoscopic or laryngoscopy instrumentatio
of the upper airway always carries with it the po<
sibility of significant laryngeal or subglotti
edema, especially in smaller infants and childrei
Management of this complication includes pos
operative humidification of inspired gases, ad<
quate hydration, intermittent positive-pressui
breathing of racemic epinephrine (0.5 ml of r;
cemic epinephrine diluted with 3.5 ml of norm
saline) every 2 to 3 hours and dexamethasoi
(0.1-0.3 mg/kg intravenously).
158
X. BRONCHOGRAPHY
Bronchography may be necessary to define tl
extent of certain parenchymal diseases of the Iun
such as segmental bronchiectasis. In order to sele
tively place the contrast material into the lung
gion in question, the contrast material must
introduced via a catheter placed in the lung regie
The catheter can be placed in the lung region \
a rigid bronchoscope or, depending on the age
the patient and the size of endotracheal tube tl
can be used, can be passed alongside or within
endotracheal tube and guided into the lung regi
by a flexible fiberoptic bronchoscope within I
endotracheal tube. This procedure is usually da
in the radiology department, and the patient m
Figure 18-20 Four ways of performing flexible fiberoptic bronchoscopy (FOB) without using an endotracheal tube as the conduit
for the fiberoptic bronchoscope. A, Use of the laryngeal mask airway (LMA). B, Use of a bronchoscopy elbow with a self-sealing
diaphragm. C, Use of an anesthesia mask with a special self-sealing fiberoptic diaphragm. D, Use of transtracheal jet ventilation
(TTJV) through a transtracheal intravenous (IV) catheter while the FOB is inserted from above.
need to be turned to several positions in order to
obtain adequate films. Anesthesia for bronchogra-
phy is similar to that for bronchoscopy, using
either an intravenous induction and halo-
thane/oxygen maintenance or an inhalation induc-
tion as described previously. If an endotracheal
tube is used instead of a bronchoscope to pass the
contrast catheter, and the contrast catheter is
passed within the endotracheal tube, the elbow
connector to the endotracheal tube should have a
self-sealing diaphragm through which the contrast
catheter can be passed and a self-sealing positive-
pressure system maintained.
The anesthesiologist should consult with the ra-
diologist as to whether spontaneous or controlled
ventilation is preferred. Spontaneous ventilation
allows the contrast medium to be drawn gradually
into the bronchi, giving dense and even distribu-
tion; whereas forceful positive-pressure ventilation
may disperse the contrast medium too readily into
the alveoli, so that the bronchial tree is poorly
delineated and obscured by densities at the periph-
ery. Perhaps the best compromise is gentle con-
trolled ventilation during short periods of paraly-
sis. Anesthesia should be sufficiently deep to
prevent coughing when the contrast material is
introduced, because coughing may spread the ma-
terial and result in poor-quality radiographs.
The patient is turned to the lateral decubitus
position to favor filling of an upper lobe. A slight
head-up tilt helps filling of both the middle and
lower lobes. After the necessary radiographs are
obtained, as much of the contrast material as pos-
sible is suctioned from the bronchial tree. The en-
dotracheal tube should be left in place until there
is an active cough reflex.
XI. ASPHYXIATING THORACIC
DYSTROPHY (JEUNE'S
SYNDROME)
159
The major feature of asphyxiating thoracic dys-
trophy (Jeune's syndrome) in the neonate is defor-
mity of the thoracic wall, which prevents normal
intercostal respiratory movement and causes res-
piratory insufficiency. The thoracic wall deformi-
ties include high clavicles, narrowed thoracic cage
with horizontal, and short, stubby ribs with a wid-
ened anterior portion.
The most seriously affected patients have under-
lying pulmonary hypoplasia. The majority of pa-
tients die within the first year of life. Persistent
pulmonary hypertension of the neonate as a result
of decreased cross-sectional area of the pulmonary
arterial vasculature may occur in those surviving
the newborn period, and renal and hepatic failure
from diffuse interstitial fibrosis may appear in late
infancy or early childhood. In older patients, pul-
monary hypertension and cor pulmonale as seque-
lae of chronic hypoxemia may also occur. These
patients may require anesthesia for tracheostomy
and thoracoplasty early in life and for renal trans-
plantation later in life.
These infants are prone to profound hemoglobin
desaturation when agitated because of asynchro-
nous rib and abdominal motion. During the intra-
operative period, peak airway pressure should be
maintained as low as possible to minimize both
barotrauma and any adverse affect of increased
airway pressure on the pulmonary arterial bed. In-
fants who have thoracoplasty almost assuredly will
need long-term ventilator support because of an
increased alveolar-arterial oxygen gradient and in-
effective ventilation with persistent hypercarbia.
REFERENCES
1. Heymann MA: Postnatal regulation of the pulmonary cir-
culation: A role for lipid mediators? Am Rev Respir Dis
136:222-224, 1987.
2. Clarke WR: The transitional circulation: Physiology and
anesthetic implications. J Clin Anesth 2:192-211, 1990.
3. Dibbins AW, Wiener ES: Mortality from neonatal dia-
phragmatic hernia. J Pediatr Surg 9:653, 1974.
4. Nelson NM, Prod'Hom LS, Cherry RB, et al: Pulmonary
function in the newborn infant: The alveolar-arterial ox-
ygen gradient. J Appl Physiol 18:534, 1963.
5. Rudolph AM, Yuan S: Response of the pulmonary vas-
culature to hypoxia and H
+
ion concentration changes. J
Clin Invest 45:399-411, 1966.
6. Peckham GJ, Fox WW: Physiologic factors affecting pul-
monary artery pressure in infants with persistent pulmo-
nary hypertension. J Pediatr 93:1005, 1978.
7. Drummond WH, Gregory GA, Heymann MA, Phibbs
RA: The independent effects of hyperventilation, tolazo-
line, and dopamine on infants with persistent pulmonary
hypertension. J Pediatr 98:603, 1981.
8. Collins D, Pomerance JJ, Travis KW, et al: New ap-
proach to congenital posterolateral diaphragmatic hernia.
J Pediatr Surg 12:149, 1977.
9. Graham TP Jr, Atwood GF, Boucek RJ: Pharmacological
dilatation of the ductus arteriosus with prostaglandin E,
in infants with congenital heart disease. South Med J
71:1238-1246, 1978.
10. Burrows FA, Klinck JR, Rabinovitch M, Bohn DJ: Pul-
monary hypertension in children: Perioperative manage-
ment. Can Anaesth Soc J 33:606-628, 1986.
11. Heymann MA, Rudolph AM, Silverman NH: Closure of
the ductus arteriosus in premature infants by inhibition ol
prostaglandin synthesis. Engl J Med 295:530, 1976.
12. Friedman WF: Pharmacologic closure of patent ductus
arteriosus in the premature infant. Engl J Med 295:526,
1976.
13. Nadas AS: Patent ductus revisited (editorial). Engl J
Med 295:563, 1976.
14. Avery ME, Taeusch HW, Fioros J: Surfactant replace
ment. Engl J Med 315:825-826, 1986.
15. Davis JM, Veness-Meehan K, Notter RH, Bhutani VK
Kendig JW, Shapiro DL: Changes in pulmonary mechan
ics after the administration of surfactant to infants wit]
respiratory distress syndrome. Engl J Med 319:476-
479, 1988.
16. Merritt TA, Hallman M, Bloom BT, et al: Prophylactii
treatment of very premature infants with human surfaq
tant. Engl J Med 315:785-790, 1986.
17. Payne JW, Patz A: Current status of retrolental fibropla
sia. The retinopathy of prematurity. Ann Clin Re
11:205-221, 1979.
18. Phelps DL: Retinopathy of prematurity. Engl J Me
326:1078-1080, 1992.
19. Kinsey VE, Arnold HJ, Kalina RE, et al: P
a
0
2
levels an
retrolental fibroplasia: A report of the cooperative stud]
Pediatrics 60:655-668, 1977.
20. Silverman WA: Retinopathy of prematurity: Oxyge
dogma challenged. Arch Dis Child 57:731-733, 1982.
21. Lucey JF: Retrolental fibroplasia may not be preventabl
J R Soc Med 75:496-497, 1982.
22. Flynn JT, Bancalari E, Snyder ES, et al: A cohort stuc
of transcutaneous oxygen tension and the incidence ar
severity of retinopathy of prematurity. Engl J M<
326:1050-1054, 1992.
23. Purohit DM, Ellison C, Zierler S, et al: Risk factors f
retrolental fibroplasia: Experience with 3,025 prematu
infants. Pediatrics 76:339-344, 1985.
24. Phelps DL: Retinopathy of prematurity: An estimate
vision loss in the United States1979. Pediatri
67:924-926, 1981.
25. Gunn TR, Easdown J, Outerbridge EW, et al: Risk facte
in retrolental fibroplasia. Pediatrics 65:1096-1100, 198
26. Aranda JV, Saheb N, Stern L, Avery ME: Arterial oxyg
tension and retinal vasoconstriction in newborn infan
Am J Dis Child 122:189, 1971.
27. Rigatto H, Brady JP: Periodic breathing and apnea in I
preterm infant: I. Evidence for hypoventilation possil
due to central depression. Pediatrics 50:202, 1972.
28. Steinschneider A: Prolonged sleep apnea and the sudd
infant death syndrome: Clinical and laboratory obser
tions. Pediatrics 50:646, 1972.
29. Musk AW, Shean RE: Passive smoking and health. IV
J Aust 148:27-28, 1988.
30. Aranda JV, Turmen T: Methylxanthines in apnea of
maturity. Clin Perinatol 6:87-108, 1979.
31. Miller MJ, Martin RJ, Carlo WA, et al: Oral breathing
newborn infants. Pediatrics 107:465-469, 1985.
32. Oliver TK Jr: Temperature regulation and heat product
in the newborn. Pediatrics 50:646, 1972.
33. Reuler JB: Hypothermia: Pathophysiology, clinical :
tings and management. Ann Intern Med 89:519-5
1978.
34. Roe CF, Santulli TV, Blair CS: Heat loss in infants dui
general anesthesia and operations. J Pediatr Surg 3:1
1966.
35. Bissonette B, Sessler DI, LaFlamme P: Inspired gas
midification prevents intraoperative hypothermia in
fants and children. Anesth Analg 68:S28, 1989.
36. Schedewie HE, Odell WD, Fisher DA, et al: Parat]
mone and perinatal calcium homeostasis. Pediatr
13:1, 1979.
37. Ward CF: Diseases of infants. In Katz J, Benumol
CHAPTER 19
Early Serious Complications
Specifically Related to Thoracic
Surgery
I. Introduction VIII. Right-Sided Heart Failure
II. Herniation of the Heart IX. Myocardial Ischemia/Infarction
III. Pulmonary Torsion X. Arrhythmias
IV. Major Hemorrhage XI. Right-to-Left Shunting Across a
V. Bronchial Disruption Patent Foramen Ovale
A. Acute Ipsilateral Bronchial XII. Massive Postpneumonectomy
Disruption Malignant Pleural Effusion and
B. Acute Contralateral Pneumothorax Chylothorax
C. Chronic Ipsilateral Fistula XIII. Systemic Tumor Embolism
VI. Respiratory Failure XIV. Neural Injuries
A. Atelectasis and Lung Infection XV. Complications of Intrathoracic
B. ARDS Intercostal Nerve Blocks
VII. Unilateral Re-Expansion
(Intraoperative) Pulmonary Edema
696
Early Serious Complications Specifically Related to Thoracic Surgery 6 9 7
I. INTRODUCTION
There are several life-threatening complications
that are specifically related to major thoracic sur-
gery that may occur in the immediate postopera-
tive period and require prompt diagnosis and treat-
ment. These include herniation of the heart
through the pericardium following radical pneu-
monectomy, massive hemorrhage, blowout of a
bronchial stump following pneumonectomy or lo-
bectomy, respiratory insufficiency, right-sided
heart failure, right-to-left shunting across a patent
foramen ovale, arrhythmias, neural injuries, and
total spinal anesthesia following intrathoracic in-
tercostal nerve blocks. The order in which these
complications are discussed is approximately re-
lated to the usual severity of symptoms and threat
to life, with the most life-threatening complica-
tions considered first. The problem of appropriate
fluid infusion and the relationship of fluid infusion
to postpneumonectomy pulmonary edema is con-
sidered in chapter 13, and the complications of
double-lumen tube insertion are discussed in chap-
ter 8.
II. HERNIATION OF THE HEART
An intrapericardial approach may be employed
for radical pneumonectomy in order to facilitate
access to major vessels and allow a wider hilar
dissection.
1
This technique may result in a large
pericardial defect, which the surgeon does not or
is not able to close after the pneumonectomy is
completed. There have been about 50 cases of
herniation of the heart through the pericardial
opening into the empty hemithorax as of 1992.
2
The herniation can be into either the right or the
left hemithorax,
3 4
and herniation into either side
can cause profound cardiopulmonary function im-
pairment; the mortality is 50 per cent.
2
5
~
7
With a right-sided herniation, the entire heart
protrudes through the defect and is displaced com-
pletely into the right hemithorax. The chest roent-
genogram usually shows displacement of the car-
diac silhouette into the right pleural cavity, with
the apex of the heart pointing to the right side and
at a right angle to the mediastinum, and the apex
may touch the right chest wall (Fig. 19-1, 2).
8
9
The X-ray appearance of partial right-sided cardiac
herniation resembles the shape of a snow cone
(Fig. 19-1, 1 and B).
9
In addition, the heart ro-
tates on its mediastinal attachment, causing twist-
ing and distortion of the structures in the medias-
tinal attachment. Twisting of the superior vena
cava causes an acute superior vena cava syndrome
(Fig. 19-1, 2), and twisting of the inferior vena
cava causes acute cardiovascular collapse
10-14
(Fig.
19-2). If the superior vena cava is distorted, a
central venous pressure catheter will demonstrate
unusual bending and curvature on chest X-ray as
well.
2
Distortion of the distal trachea or left main-
stem bronchus may cause wheezing, and obstruc-
tion of the pulmonary veins may cause pulmonary
congestion and edema (Fig. 19-2). The electrocar-
diogram may reflect ischemia with changes in the
S-T segments and waves and shifts in the QRS
axis
2
-
6,15
(see Fig. 19-2).
With left-sided herniation, the ventricular apex
protrudes out of the pericardium, and the atrioven-
tricular groove is constricted by the pericardium
3
(see Fig. 19-2). The extrapericardial myocardium
becomes ischemic and edematous and, if not re-
duced quickly, causes obstruction of ventricular
outflow, ventricular arrhythmia, and myocardial
infarction. Thus, left-sided herniations are much
Figure 19-1 A, Case l. Right pneumonectomy was performed for uncontrolled bleeding incurred from a stab wound. Recovery
room chest radiograph demonstrates globular right border after right pneumonectomy. Air (arrows) in the pericardium outlines the
heart (Al). Complete cardiac herniation with volvulus occurred 90 minutes later (A2). Cardiac apex is now directed to right lateral
chest wall. Superior mediastinum is widened because of superior vena caval obstruction. B, Case 2. Right pneumonectomy was
performed as part of debulking procedure for mediastinal thymoma. Hemispheric bulge in right contour is again noted. Patient was
asymptomatic. Ninety minutes later, complete herniation occurred. (From St. Luke's Hospital, Milwaukee, WI. Used with permis-
sion.)
698
Early Serious Complications Specifically Related to Thoracic Surgery
Herniation of the Heart Post Pneumonectomy
Causes ( m+>) and Effects ( ( =
Obstruction and/or
Distortion of Superior
and/or Inferior Vena
Cava, Pulmonary
Veins, Trachea and
Mainstem Bronchi
Coughing
Constriction of
Atrioventricular Groove
EKG-Dramatic Axis and/or
Ischemic Changes
Negative
Pressure
(Suction)
To Empty
Hemithorax
Figure 19-2 Following a transpericardial radical pneumonectomy, the heart may herniate through a pericardial defect into the
empty hemithorax. Factors that can cause this complication (solid arrows) are placing the empty hemithorax in a dependent position,
which allows gravity to pull the heart through the defect; applying a high level of negative-pressure suction to the empty hemithorax,
which also pulls the heart through the defect; applying a high level of positive-pressure ventilation to the remaining nondependent
lung, which pushes the heart through the defect; and coughing, which also increases nondependent hemithorax pressure. The effects
of herniation of the heart (open arrows) consist of obstruction and/or distortion of the mediastinal attachments (superior and inferior
vena cava, pulmonary veins, and trachea and main-stem bronchi) and constriction of the atrioventricular groove by a tightly fitting
pericardial defect. The electrocardiogram (EKG) will show dramatic axis and/or ischemic changes.
more likely to demonstrate ischemic electrocardio-
graphic patterns with atrial and ventricular arrhyth-
mias, and the changes in the electrocardiogram
may precede clinical symptoms.
2
Constriction of
the atrioventricular groove may cause acute car-
diovascular collapse. The chest roentgenogram
usually shows that the heart has assumed a spheric
shape and has been displaced laterally and at a
right angle to the mediastinum and that the apex is
against the left chest wall.
8
In most reports, herniation of the heart has oc-
curred either immediately after the patient is
turned from the lateral decubitus to the supine
position (75 per cent) or during the first few hours
of postoperative mechanical ventilation.
2,7
l4, l6
In
one case, a late intraoperative change in the pa-
tient's position from supine to a flexed position
precipitated torsion of the great veins.
17
Obviously,
the patient will still be under the care of the anes-
thesiologist in many of these situations.
14
17
Adhe-
sions form between the heart and pericardium by
the third day after surgery, thereby decreasing the
chance of herniation.
1
However, this complication
has occurred as late as several days after surgery.
18
Events that increase intrapleural pressure in the
nonsurgical (ventilated) hemithorax or that de-
crease intrapleural pressure in the surgical (empty)
hemithorax may predispose the patient to cardiac
herniation. Placing the patient with the empty
hemithorax in a dependent position allows the
heart to be pulled by gravity into the empty hemi-
thorax (see Fig. 19-2). Use of high levels of pres-
sure and volume in the remaining lung can push
the heart into the empty hemithorax (see Fig. 19-
2). Similarly, coughing can increase pleural pres-
sure in the remaining lung and thereby promote
displacement of the heart into the empty hemitho-
rax (see Fig. 19-2). Conversely, inadvertently ap-
plying suction to the chest drain in the empty
hemithorax can pull the heart through a pericardial
defect (tension vacuthorax)
7 l0
(see Fig. 19-2).
The heart can also herniate through a small defect
(3 cm) without any of these factors being present.
14
The differential diagnosis consists of myocar-
dial infarction, cardiac tamponade, massive pul-
monary embolus, airway obstruction, and collapse
of the remaining lung. When circulatory signs per-
mit, the diagnosis can be confirmed by portable
chest roentgenography. A diagnosis of cardiac her-
niation almost always requires immediate re-ex-
ploration
5-7
; however, there is one report of hernia-
tion of the heart that caused only roentgenologic
changes and no immediate clinical signs,
13
and
there is only one report in which herniation of the
heart caused only electrocardiographic changes for
several hours.
2
There are conservative measures that may im-
prove cardiopulmonary function before and during
transfer to the operating room (they are the reverse
of the causes of herniation of the heart shown in
Figure 19-2), and they should be instituted as soon
as the diagnosis is entertained. First, the patient
should be positioned so that the ventilated, nonsur-
gical side is in a dependent position and the empty
surgical side is in a nondependent position.
4 I8 l9
Gravity may return the heart and mediastinum to
their normal anatomic positions.
14
Even if the heart
does not re-enter the pericardium, the reposition-
ing may decrease atriocaval kinking and increase
cardiac output. Second, avoidance of high levels
of pressure and volume in the ventilated lung may
allow the return of the heart to the pericardium.
Consequently, the tidal volume should be reduced,
positive end-expiratory pressure (PEEP) should be
removed, and the respiratory rate should be in-
creased slightly. Third, suction to the empty hemi-
thorax should be discontinued. Fourth, the phar-
macologic support of the circulation should be
provided as needed. Fifth, injecting 1 to 2 L of air
into the surgical hemithorax may push the heart
and mediastinum back to normal anatomic posi-
tions and may perhaps return the heart to the peri-
cardium. Success with this latter technique has
been variable.
1618
20-22
III. PULMONARY TORSION
Pulmonary torsion refers to parenchymal rota-
tion on its bronchovascular pedicle and is thought
to be due to increased mobility of a lobe (as may
occur to the remaining lobes in one hemithorax
after a lobectomy). Torsion after an intrathoracic
surgical procedure has been described in the right
middle lobe,
23
24
in the left lower lobe,
25,26
in the
left upper lobe,
27
28
and in an entire lung.
29
Any
patient with atelectasis or an expanding intratho-
racic mass (and perhaps with severe chest pain)
after thoracotomy should have lung torsion in-
cluded in the differential diagnosis (which also
includes intrathoracic bleeding and progressive
atelectasis).
Pulmonary torsion may also occur after blunt
chest trauma, nonpulmonary thoracic procedures
(see Kucich et al.
28
and Berkmen et al.
30
for a
detailed list of references), and pneumothorax (up-
per lobe hangs down over the lower lobe).
30
It is
possible that infiltrates seen in an upper lobe after
re-expansion of a collapsed lung resulting from
a pneumothorax is not re-expansion pulmonary
edema but hemorrhage caused by venous obstruc-
tion. Because torsion compromises the pulmonary
vasculature (both the arteries and veins) as well as
the bronchi, prompt recognition and surgical inter-
vention are required to avoid the attendant morbid-
ity and mortality (i.e., infarction and gangrene).
31
A double-lumen endotracheal tube should be
inserted before surgery if lung torsion is expected
because, during surgical correction of this condi-
tion (either completion pneumonectomy or un-
twisting and stabilization of the lobe to another
lobe or chest wall), massive hemorrhage may oc-
cur into the airways (as a result of the release of
entrapped blood and continuing bleeding from ne-
crotic lung tissue [i.e., vessels and airways]),
drowning the dependent lung and producing se-
vere hypoxia or death.
27
28
Other intraoperative therapeutic modalities that
the anesthesiologist should consider include the
use of intravenous steroids to decrease any reac-
tive pulmonary inflammation, use of PEEP to re-
expand atelectatic lung tissue, and the use of saline
lavage for the removal of any excess blood, which
might form clots in the normal airways, thereby
obstructing good lung parenchyma.
IV. MAJOR HEMORRHAGE
Postoperative bleeding that necessitates emer-
gency thoracotomy may occur in as many as 3 per
cent of all thoracotomy patients.
32
The mortality of
major postoperative bleeding is approximately 23
per cent.
32
There are several potential sources of
major postoperative hemorrhage (Table 19-1), and
these consist of bleeding from divided pulmonary
arteries and veins due to slippage of surgical liga-
tures, diffuse bleeding from raw surfaces, and sys-
700 Early Serious Complications Specifically Related to Thoracic Surgery
Table 19-1 POTENTIAL SITES OF MAJOR
HEMORRHAGE AFTER
THORACOTOMY
1 Slippage of pulmonary artery and/or vein sutures or
ligatures
2. Large raw surfaces
3. Bronchial arteries
4. Intercostal arteries
temic arterial bleeding (bronchial and intercostal
arteries).
Bleeding may be catastrophic if ligatures slip
off a pulmonary artery or vein. Although these
vessels are not highly pressurized, hemorrhage
from these vessels is into a low pressure-high vol-
ume space and, therefore, may be massive. The
risk of postoperative hemorrhage can be greatly
diminished by adequate intraoperative exposure,
and adequate intraoperative exposure may be
greatly aided in most cases by one-lung ventilation
and in a few cases by an intrapericardial approach
to the hilum. In addition, branches of the pulmo-
nary artery or vein should be directly oversewn
with vascular suture material and/or should be
doubly ligated, and then a transfixion suture or
pursestring suture with a nonabsorbable material
should be placed between the double ligation.
32
Hemorrhage from raw surfaces is more common
after pneumonectomy than after lobectomy be-
cause following lobectomy the apposition of the
remaining lobes of lung against the chest wall and
mediastinum may greatly diminish bleeding from
these surfaces. Bleeding from raw pleural surfaces
is especially likely when vascular adhesions be-
tween the visceral and parietal pleura have been
divided. Hypertension is contributory to diffuse
bleeding from raw surfaces.
Postoperative bleeding may be from systemic
arteries, especially the bronchial arteries. Bleeding
from bronchial arteries is encouraged by extensive
mediastinal dissection that damages bronchial and
mediastinal arteries. Bronchial artery bleeding will
also occur when the bronchial artery is not in-
cluded in the closure of a bronchial stump. If the
free bronchial artery goes into spasm, then it may
retract proximally. Subsequently, when the spasm
is released postoperatively, the vessel will hemor-
rhage. In addition, the intercostal arteries may
bleed if there has been inadvertent placement of
periosteal sutures through an intercostal vessel.
33
The drainage from a patent chest tube is an
excellent monitor of the amount of intrathoracic
bleeding (Table 19-2). In addition, the hematocrit
of the usual fluid* in the chest drainage is always
less than 20 per cent, and intrathoracic bleeding
will increase the hematocrit to some higher value
(see Table 19-2). Significant hemorrhage, of
course, will be accompanied by the hemodynamic
signs of hypovolemia, such as tachycardia, hypo-
tension, and decrease in vascular filling pressures
and urine output (see Table 19-2).
Significant hemorrhage cannot always be ruled
out by the absence of chest tube drainage, since
the chest tube may be blocked by clotted blood or
otherwise obstructed. Nevertheless, the patient will
still show signs and symptoms of hypovolemia. In
addition, if the chest tube is not draining well,
there may be signs of significant shift in the me-
diastinum to the opposite side.
In extreme cases exsanguination may occur rap-
idly unless the bleeding is promptly controlled;
indeed, immediate thoracotomy in the recovery
room in order to obtain manual control of the
bleeding site may be life-saving. Once this has
been accomplished, the patient can be returned to
the operating room for closure under aseptic con-
ditions. In all cases, blood, preferably fresh whole
blood, should be administered. If time permits, the
coagulation profile should be obtained, and spe-
cific coagulation factors and/or platelet concen-
trates should be administered, as indicated by the
laboratory findings.
V. BRONCHIAL DISRUPTION
Development of a bronchopleural fistula is a
serious complication of pulmonary resection and
as recently as 1978 carried a mortality of 23 per
cent.
34, 35
The term refers to any communication
between the tracheobronchial tree and the pleural
cavity and can result from dehiscence of the bron-
chial stump after lobectomy or pneumonectomy,
from rupture of an inflammatory lesion or cavity
within the lung into the pleura, or from trauma or
neoplasm erosion. The symptoms and signs are
variable and depend upon the size of the commu-
nication, the presence or absence of a chest drain,
and whether or not there is fluid of any ort within
the pleural space.
A. Acute Ipsilateral Bronchial
Disruption
Gross disruption of a bronchial stump in the
early postoperative period may result from techni-
Table19-2 SIGNS OF POSTOPERATIVE
INTRATHORACIC HEMORRHAGE
1. Volume of blood in chest tube drainage
2. Hematocrit of blood in chest tube drainage
3. Signs of hypovolemia (tachycardia and decreased systemic
blood pressure, systemic pulse pressure, central filling
pressures, and urine output)
4. Signs of tension hemithorax (with nonfunctional chest
tube)
cal error in bronchial closure, along with the use
of high airway pressure for mechanical ventilation.
The escape of gas is usually signaled by massive
bubbling of gas through the chest tube drainage
system. The patients quickly become hypoxemic
and hypercarbic and develop a sympathomimetic
cardiovascular response to these gas-exchange
changes. Immediate re-exploration and revision of
the bronchial closure are required. As soon as
gross bronchial disruption is recognized, the chest
tube should be removed from suction and left to
underwater seal only. If the leak continues to be
large, strong consideration should be given to rein-
sertion of a double-lumen tube to prevent the es-
cape of gas through the disrupted bronchial stump.
The incidence of this postoperative catastrophe can
be diminished intraoperatively (prior to chest clo-
sure) by the simple technique of filling the pleural
space with irrigating solution and having the anes-
thesiologist exert 35 to 40 cm H
2
0 of pressure to
demonstrate an airtight bronchial closure.
If the chest tube is obstructed, a tension pneu-
mothorax with mediastinal shift will occur; the
presence of a chest tube does not always prevent
tension pneumothorax (e.g., the chest tube may be
in a loculated compartment). High inflation pres-
sures will be required for ventilation, and subcu-
taneous emphysema may become manifest (Fig.
19-3). A tension pneumothorax will shift the me-
diastinum, interfere with venous return, and de-
crease the cardiac output and blood pressure (see
Fig. 19-3). The chest roentgenogram is diagnostic,
but if the clinical situation is critical, therapy
should be instituted without taking one. Treatment
consists of re-establishing the patency of the chest
tube or inserting a needle or a trocar and cannula
into the pleural cavity through the second intercos-
tal space in the midclavicular line on the affected
side.
B. Acute Contralateral Pneumothorax
It should be remembered that a tension pneu-
mothorax may occur in the contralateral side if
very high inflation pressures are used to expand a
previously collapsed lung (as commonly occurs
for one-lung ventilation); the tension pneumo-
thorax may become apparent only after closure of
the chest.
36
Other causes of acute postoperative
contralateral pneumothorax relevant to thoracic
surgery include damage to the contralateral pleura
during surgery, puncture of the pleura during in-
sertion of a central venous cannula or thoracic
epidural needle, or damage to the bronchus by use
of an endobronchial catheter.
C. Chronic Ipsilateral Fistula
Development of a chronic bronchopleural fistula
may be expected to occur in 1 to 3 per cent of
pulmonary resection patients.
34
A chronic bron-
chopleural fistula most frequently becomes evident
within the first 2 weeks after operation, may be
large, and may run a fulminating course character-
ized by sepsis, empyema, purulent sputum, and
respiratory insufficiency. In a much smaller num-
ber of patients, bronchopleural fistulas may de-
velop even later, usually are smaller, and usually
occur in an insidious fashion, with malaise, fever,
and the presence of multiple air-fluid levels on
chest roentgenograms. Factors that predispose to
the formation of a bronchopleural fistula after pul-
monary resection are preoperative irradiation, in-
fection, residual neoplasm at the site of the clo-
sure, and presence of a long or avascular stump.
A large bronchopleural fistula that develops sev-
eral days after pneumonectomy occurs dramati-
cally. The onset is often abrupt, with a bout of
coughing and dyspnea precipitated by a change of
posture that allows the operated side to be upper-
most. Thin red-brown fluid that is characteristic of
the contents of the empty hemithorax is both ex-
pectorated and inhaled into the remaining lung.
This fluid is often infected and always irritant, so
that it provokes bronchoconstriction and local in-
flammation, causing severe respiratory distress and
hypoxemia. Signs of circulatory failure are usual
too, probably caused by both hypoxemia and sep-
ticemia. The diagnosis is usually clear on clinical
grounds alone. If confirmation is required, a chest
roentgenogram will show a considerable fall in the
fluid in the pneumonectomy hemithorax (fluid is
replaced by air) and consolidation and collapse in
the remaining lung.
Prompt resuscitation is essential and may need
to include mechanical and differential lung venti-
lation with a double-lumen tube and support of the
circulation. A functioning chest tube must be in-
serted immediately to evacuate the air (to prevent
a tension pneumothorax) and fluid (to prevent fur-
ther contamination of the opposite lung) from the
pneumonectomy cavity. Postural drainage and
chest physiotherapy have no place in the manage-
ment of a large bronchopleural fistula. The patient
should be recumbent, with the normal side in a
nondependent position. The stump must be closed
surgically.
34
At surgery, a double-lumen tube should be in-
serted to separate the two main-stem bronchi. The
separation of the two main-stem bronchi prevents
further contamination of the remaining or unoper-
ated lung and allows positive-pressure ventilation
of the remaining lung without loss of tidal volume
through the fistula. The loss of tidal volume
Signs of Tension Pneumothorax
Figure 193 A tension pneumothorax will cause a mediastinal shift. The mediastinal shift compresses the contralateral lung
(which causes poor compliance and high inflation pressures) and decreases venous return, cardiac output, and systemic blood
pressure. Since both lungs are under pressure, the diaphragms are depressed and flat, and if the tension pneumothorax occurs
postpneumonectomy, the free air in the pleural space will cause a fall in the postpneumonectomy fluid level.
through the fistula can prevent effective ventilation
in the remaining lung in three ways. First, tidal
volume will be lost through the chest tube if the
chest tube is functioning. Second, tension pneu-
mothorax will develop if the chest tube is not
functioning, which will cause mediastinal shift and
compression of the remaining lung. Third, the es-
caping gas will pressurize the fluid in the empty
hemithorax and force it to enter the fistula, which
can cause contamination of the remaining lung.
From the foregoing considerations, it is obvious
that the lumen to the resected side must be oc-
cluded before positive-pressure ventilation is be-
gun. The endobronchial lumen of the double-lu-
men tube should be in the contralateral main-stem
bronchus. Surgical procedures for chronic bron-
chopleural fistula may include open drainage by
rib resection, multiple-rib thoracoplasty, coverage
of the bronchial leak with muscle flaps, and addi-
tional pulmonary resection; many patients require
multiple procedures.
34
If the chronic fistula is small, the symptoms are
unobtrusive at first and consist of fever, tachycar-
dia, and persistent cough with expectoration of
small amounts of sputum, which is almost always
blood tinged. Episodes of coughing at night and
bouts of wheezy dyspnea are particularly sugges-
tive of bronchopleural fistula. The chest roentgen-
ogram at this stage (postpneumonectomy) may
show a decrease in the volume of fluid in the
pleural space and an increase in the amount of air
at the apex, or the fluid level may fail to continue
rising as it normally does.
A small fistula after pneumonectomy may heal
without surgery. Medical management includes
appropriate use of antibiotics, nutrition, chest tube
placement (Table 19-3), selection of the drainage
device (at - 20 cm H
2
0, suction pumps vary in
capacity, from 2 to 35 L/min [Table 19-4]),
38
ven-
tilator selection and use (Table 19-5) (consider
high-frequency ventilation in a patient with a prox-
imal bronchopleural fistula and a normal lung
compliance), and diagnostic and therapeutic bron-
choscopy (balloon occlusion, use of sealants).
37
With respect to therapeutic bronchoscopy, there
are now many reports of sealing all sizes of bron-
Early Serious Complications Specifically Related to Thoracic Surger\ 7 0 3
Table 19-3 CHEST TUBE IN
BRONCHOPLEURAL FISTULAS
(BPF)*
I. Large diameter chest tube for
A. High-flow BPF
B. Drainage of infected pleural space
II. Therapy (ventilated patient)
A. Positive intrapleural pressure (expiratory phase) = to
desired positive end-expiratory pressure level
B. Chest tube occlusion during inspiratory phase
C. Combination of above
D. Application of sclerosing agent to pleural space
(perhaps aided by thoracoscopy)
III. Negative effects of chest tube (ventilated patient)
A. Loss of tidal volume
B. Some of the lost tidal volume has participated in gas
exchange, making control of ventilation more difficult
C. Negative suction pressure may cause ventilator
cycling
*Modified from Baumann MH, Sahn SA: Medical manage-
ment and therapy of bronchopleural fistulas in the mechanically
ventilated patient. Chest 97:721-728, 1990. Used with permis-
sion.
chopleural fistulas with tissue glue or other bioad-
hesives (e.g., clotted blood), and the sealant is
most often applied through a fiberoptic broncho-
scope (see chapter 17).
However, surgery is still sometimes necessary,
34
and a double-lumen tube (with the endobronchial
lumen in the contralateral main-stem bronchus)
should be used to separate the two lungs during
surgery. Surgical procedures are as previously de-
scribed.
VI. RESPIRATORYFAILURE
There are a number of events/consequences of
general anesthesia for thoracic surgery procedures
that predispose a routine patient to postoperative
respiratory failure. Dependent-lung atelectasis/in-
fection may occur in any generally anesthetized
patient. In addition, there are some unusual causes
of atelectasis/edema/infection in both the operative
and nonoperative lung that are specific to the tho-
racic surgery patient. Finally, the development of
Table 19-4 DRAINAGE SYSTEMS*
Maximal Flow (L/min)
System with - 20 cm H
2
0
1. Emerson pump 35.5
2. Pleur-Evac A4000 34.0
3. Thora-Klex 19.7
4. Sentinel seal 2.3
* Adapted from Capps. JS, Tyler ML, Rusch VW, Pierson
DJ: Potential of chest drainage units to evacuate bronchopleural
air leaks. Chest 88:57S, 1985. Used with permission.
Table 19-5 CONVENTIONAL VENTILATION IN
BRONCHOPLEURAL FISTULAS*
I. Decrease fistula flow (reduce airway pressure)
A. Lowest effective tidal volume
B. Least number of mechanical breaths (allow
spontaneous ventilation if possible; use intermittent
mandatory ventilation)
C. Lowest positive end-expiratory pressure and avoid
expiratory retard
D. Shorten inspiratory time
II. Other maneuvers
A. Selective intubation of both lungs (i.e., double-lumen
tube)
B. Differential lung ventilation
C. Put the bronchopleural fistula in a dependent position
*From Baumann MH, Sahn SA: Medical management and
therapy of bronchopleural fistulas in the mechanically venti-
lated patient. Chest 97:721-728, 1990. Used with permission.
lung infection with gram-negative organisms may
cause/contribute to the development of adult res-
piratory distress syndrome (ARDS) and sepsis.
A. Atelectasis and Lung Infection
Acute respiratory insufficiency (within 30 days
of resection) is one of the most common serious
complications after pulmonary resection. Bron-
chial carcinoma resection data from a large tho-
racic surgery service indicate an incidence of ap-
proximately 4 to 8 per cent,
39
40
and in one study,
the patients in whom acute respiratory failure de-
veloped had a mortality of 50 per cent.
39
Mortality
from respiratory failure after right-sided pneumo-
nectomy is greater than after pneumonectomy on
the left side because the remaining functioning
lung is smaller.
39
Chapter 3 discussed the mecha-
nisms of hypoxemia and hypercarbia during anes-
thesia and surgery, and many of these mechanisms
may continue to function in the postoperative pe-
riod. In addition, pulmonary resection per se in-
volves additional new mechanisms of hypoxemia
and hypercarbia.
Most of the mechanisms of hypoxemia during
anesthesia and surgery involve a decrease in the
functional residual capacity below the closing vol-
ume of the lung, causing either low ventilation-
perfusion regions or atelectasis (Fig. 19-4). First,
postoperative pain will cause splinting of the chest
wall, compression of the lungs, and a decrease in
functional residual capacity. Second, splinting will
also prevent coughing and deep breathing, which
will promote retention of secretions, which, in
turn, will decrease the functional residual capacity.
These first two mechanisms of postoperative hy-
poxemia can be minimized by appropriate control
of pain (see chapter 21). Third, the remaining lung
in either hemithorax may be edematous and/or
Major Mechanisms of
Lung Infection and Impaired Pulmonary Function
Specific to Postoperative Thoracic Surgery Patients
Figure 19- 4 The major mechanisms of atelectasis and lung infection that are specific to postoperative thoracic surgery patients.
The inclusion of intrinsic positive end-expiratory pressure (PEEP) is based on Schmidt and Hall.
42
The notion that atelectasis per se
predisposes to infection is thought to be due to altered immune function in the atelectatic area and is based on data from Schmidt
and Hall.
42
The notion that lung infection may be due to aspiration of gram-negative organisms from the nasopharynx and
gastrointestinal tract is based on data from Niederman et al.
44
soiled with blood. The lung that was dependent
during surgery may be edematous and/or may have
aspirated blood (if a double-lumen tube was not
used) owing to gravitational effects (in zones 3
and 4); the lung that was nondependent during
surgery may be edematous and hemorrhagic owing
to surgical compression and trauma and perhaps
re-expansion of previously collapsed lungs (see
next section). One-lung ventilation can minimize
surgical trauma to the nondependent lung, and
lung separation with a double-lumen tube will pre-
vent aspiration of blood into the dependent lung.
All of the remaining lung may be edematous be-
cause of a decreased vascular bed, increased car-
diac output, decreased lymph clearance capability,
and perhaps, but not necessarily, infusion of exces-
sive amounts of fluid (see chapter 13 for a full
discussion of postpneumonectomy pulmonary
edema).
The impact of diminishing the size of the pul-
monary vascular bed should be predictable on the
basis of preoperative pulmonary circulation and
right ventricular testing results (see chapter 5). The
major mechanism of carbon dioxide retention is
fatigue of the respiratory muscles and the dia-
phragm (which has decreased function after pul-
monary resection)
41
in attempting to move lungs
that are now stiff (and may be hyperexpanded in
areas that have intrinsic PEEP),
42
have a high
airway resistance, and have a diminished gas-
exchange surface.
The lung involved in routine postoperative tho-
racic surgery may become infected via several
mechanisms (see Fig. 19-4). First, for all the rea-
sons cited previously, atelectasis may develop; ate-
lectatic lung is susceptible to infection because of
impaired atelectatic lung immunologic function.
43
Second, major stress such as trauma and surgery
may decrease immune function. Third, indwelling
devices, such as endotracheal tubes, which are
connected to a whole host of respiratory devices,
and chest tubes, may cause lung infection. Fourth,
aspiration of gram-negative bacilli from the naso-
pharynx and gastroesophageal area may initi-
ate lung infection.
44
Respiratory insufficiency is
treated by instituting mechanical ventilation, and,
_J
while the patient is being mechanically ventilated
(see chapter 20), the underlying causes of the fail-
ure should be diagnosed and reversed.
B. ARDS
The presence of lung infection may progress to
ARDS and may be the cause of the septic syn-
drome, and the septic syndrome may indepen-
dently cause ARDS (Fig. 19-5). Lung infection
and the septic syndrome may cause ARDS by ac-
tivation of complement and neutrophils. This
mechanism is illustrated in Figure 196.
45 46
ARDS may also be produced by a whole host of
complement- and neutrophil-independent mecha-
nisms (see Fig. 195).
The pathologic hallmark of ARDS is lung
edema despite low or normal hydrostatic pressures
in the pulmonary microvasculature (i.e., permea-
bility edema); the edema causes small, stiff lungs,
which have a high right-to-left shunt. The specific
diagnostic criteria for ARDS and the septic syn-
drome are defined in Tables 19-6
47
and 19-7,
4
*
49
respectively. Despite the increasing ability to de-
fine the mechanisms of ARDS and its occurrence,
the mortality from ARDS has been and still is
approximately 40 per cent.
45-51
The pathologic findings in the lungs of patients
with ARDS are remarkably similar regardless of
the underlying cause.
50
Widespread alveolar and
interstitial edema resulting from damage to the
epithelial and endothelial cell layers is present
within the first 24 hours of clinical symptoms.
Within the next 24 to 48 hours, thickening of the
alveolar walls caused by capillary congestion and
alveolar hemorrhage occurs. Hyaline membranes,
composed of cellular debris, protein, and fibrin
line the respiratory bronchioles and alveolar ducts.
Type I alveolar epithelial cells, which cover 95 per
cent of the alveolar surface, are extensively in-
volved.
Damage may range from slight swelling to total
cell destruction, leading to denudement of the
basement membrane. The pulmonary circulation is
constricted/obstructed both mechanically (endo-
thelial cell edema, thrombi, platelet, fibrin, leuko-
cyte aggregates) as well as actively (hypoxic
pulmonary vasoconstriction [HPV], acidosis, vaso-
constrictor amines, and peptides) (see Fig. 7-
10).
50
52
53
With the very first signs of respiratory distress,
the chest roentgenogram is most frequently normal
Figure 19-6 The complement ac
tivation-neutrophil-dependent adult
respiratory distress syndrome (ARDS^
pathway. (Based on data from Rin
aldo and Rogers
45
and Said anc
Foda.
46
)
and usually remains normal for the first 12 to 24
hours after the initial insult. The first abnormalities
found in the early stages of ARDS are patchy,
diffuse, symmetric, bilateral interstitial and alveo-
lar infiltrates. The patchy zones of alveolar infil-
trates rapidly coalesce to a more massive airspace
consolidation. The cardiac silhouette is not en-
larged, and the costophrenic and cardiophrenic an-
gles are clear. This pattern of chest roentgenogram
abnormalities is most common and constitutes the
"progressive pulmonary edema" type. The lack of
redistribution of blood flow to the upper lung
zones, sparing of the costophrenic angles, absence
Table 19-6 CRITERIA FOR DIAGNOSIS OF
THE ADULT RESPIRATORY
DISTRESS SYNDROME*
1. An appropriate risk factor (e.g., sepsis, aspiration, or
trauma)
2. Severe hypoxemia while breathing increased oxygen
concentrations
3. Increased pulmonary shunt fraction
4. Reduced lung compliance and volume
5. Radiographic evidence of pulmonary edema
6. Previously normal lung
7. No evidence of heart failure (Ppa
0
<18 mm Hg)
*From Shale DJ: The adult respiratory distress syndrome
20 years on (editorial). Thorax 42:641-645, 1987. Used with
permission.
of obvious pleural effusions in the supine positioi
(although it may be present in the upright positior
and is due to the inability of the lymph clearanc
mechanism to keep up with the transudation), an(
a normal heart size and vascular pedicle help t(
differentiate ARDS from cardiogenic pulmonar
edema.
The principles of management are presented ii
Table 19-7 CRITERIA FOR DIAGNOSIS OF
THE SEPSIS SYNDROME*
I. Two of the following findings indicating possible
infection:
A. Temperature >30C or <36C
B. White blood count <3000 or > 12,000 cells/mm
3
or
with 10% immature granulocytes
C. Blood culture positive for a commonly accepted
pathogen
D. Known or strongly suspected source for systemic
infection that has yielded a culture of a known
pathogen
E. Gross pus in a closed space
II. Plus one of the following, indicating a deleterious systemi
effect:
A. Unexplained metabolic acidosis with a base excess of
>5 mmol (mEq)/L
B. Systemic vascular resistance <800 dynes/sec/cm
C. Unexplained hypotension with systolic blood pressure
<90 mm Hg for more than 2 hours
|
*Based on data from Montgomery et al.
48
and Wiles et al.
4
Table 19-8, which outlines the supportive therapy
used in ARDS (see chapter 20 for the details of
the ventilation and respiratory care steps).
50
Al-
though it is potentially possible to pharmacologi-
cally interfere with the various mediators of ARDS
by the use of corticosteroids, arachidonic acid me-
tabolism blockers, antiproteases, antioxidants, or
combinations of these agents, it remains to be seen
whether applying these pharmacologic agents can
effectively prevent or treat ARDS. In view of this
current situation, it must be clearly stated that to-
day the management of established ARDS is em-
piric and supportive.
50
There has been considerable interest in the pos-
sibility that increasing oxygen delivery may help
improve the survival rate among patients with
ARDS, particularly those with sepsis. There is ev-
idence that many patients with ARDS or sepsis
also have occult tissue hypoxemia as manifested
by an increased oxygen consumption when oxygen
delivery is increased.
54
Figure 19-7 shows the the-
oretical oxygen delivery-oxygen consumption re-
lationship for delivery-independent (normal) and
delivery-dependent (sepsis) systems. Improved ox-
ygen delivery in these patients may also improve
organ perfusion and thereby reduce the incidence
of multiorgan failure. However, no clinical studies
have confirmed this hypothesis.
55
The multiple organs that may fail and the crite-
ria for organ failure that may arise during the sup-
portive management of patients with ARDS are
Table 19-8 SUPPORTIVE THERAPY IN ARDS*
I. Mechanical Ventilation
A. Use a volume-cycled ventilator to deliver tidal
volumes of 7 to 12 mL/kg.t Try to avoid high
inflation pressures >60 cm H
2
0
B. Apply positive end-expiratory pressure (PEEP) by
increasing the level slowly in increments to 5 to 15 cm
H
2
0 while monitoring its effects on cardiac output and
arterial blood gases. Try to achieve a P
a
0
2
>65 with
an F,0
2
<0. 4t
PEEP levels of 9 to 12 cm H
2
0 are most commonly
used.
II. Fluid Management
A. Correct anemia with packed red blood cellst
B. Correct hypovolemia with crystalloids!
III. Pharmacologic interventions (as necessary)
A. Dopamine to augment cardiac output if necessaryt
B. Diuretics to reduce preload and hydrostatic forces (if
reduction in cardiac output can be avoided)
C. Antibiotics to treat underlying or complicating
infection(s)
*From Petty TL: Identification and management of the adult
respiratory distress syndrome. Appl Cardiopulm Pathophysiol
3:3341, 1989. Used with permission.
tThese items address the issue of increasing oxygen delivery
([cardiac output] X [C
a
0
2
]) by either increasing cardiac output
or C
a
0
2
(i.e., Hb or P
a
0
2
). See Figure 19-7 for rationale of
increasing oxygen delivery when oxygen consumption is oxy-
gen delivery dependent.
listed in Table 19-9.
51
A typical sequence of
events that culminates in the diagnosis of multior-
gan failure syndrome begins with an episode of
trauma to the body of some sort (surgery, infec-
tion, profound hypotension, flail chest) followed
by resuscitation. ARDS then develops, and me-
chanical ventilation is usually required. Polyuric
renal insufficiency occurs next, characterized by a
rise in blood urea nitrogen and creatinine levels.
The syndrome is fully developed when jaundice
occurs. Between 16 per cent
48
and 40 per cent
51
die of progressive respiratory failure, whereas the
remainder die of multiple-organ failure and sepsis.
VII. UNILATERAL RE-EXPANSION
(INTRAOPERATIVE) PULMONARY
EDEMA
Sudden evacuation of a chronic or subacute
pneumothorax or pulmonary effusion may cause
edema of ipsilateral lung (re-expansion pulmonary
edema).
56
57
Still other reports described an even
more acute form of re-expansion pulmonary
edema associated with lung re-expansion after
only several hours of atelectasis,
58
59
including re-
expansion of the nondependent lung after one-lung
ventilation to facilitate thoracic surgery
60
^
63
and
after correction of an inadvertent main-stem bron-
chial intubation.
58
M
In two cases, the operative
lung was found to be atelectatic on opening the
chest because of external compression; in one
case, the compression was by a mediastinal mass,
60
and in the other case, the compression was by the
liver and colon through a ruptured diaphragm.
61
In
both cases, after 1 to 2 hours of one-lung ventila-
tion, the operative lung was expanded with 20 cm
H
2
0, and 1 hour later copious pulmonary edema
fluid issued from the operative lung bronchus.
Similarly, re-expansion pulmonary edema has oc-
curred after only 2 hours of atelectasis during the
thoracic stage of esophagectomy.
62
Acute unilateral pulmonary edema has also been
described in the operating room at the end of tho-
racotomy and pleurodesis for recurrent pneumo-
thorax from a lung cyst.
63
Pulmonary edema local-
ized to the contralateral noncollapsed lung has
occurred immediately after re-expansion of a de-
liberately collapsed lung (i.e., one-lung ventilation
for thoracic surgery) after transaxillary sympathec-
tomy; it was postulated that the sympathectomy
was protective of the collapsed lung and causative
of edema in the noncollapsed lung.
65
Unilateral
pulmonary edema of the left lung has occurred in
three patients in one report
64
and in one patient in
another report
58
after insertion of an endotracheal
tube in the right main-stem bronchus and then
Delivery dependent and delivery independent
oxygen consumption
Figure 19-7 Relationship be-
tween oxygen delivery and oxygen
consumption in normal patients and
in patients with multiple-organ fail-
ure as a result of sepsis.
withdrawal into the trachea (allowing re-expansion
of the left lung).
Most clinical and experimental observations
support increased pulmonary vascular permeability
as a major factor in the development of reperfusion
pulmonary edema.
66-68
Possible mechanisms of
the increase in pulmonary vascular permeability
include anoxic damage to the capillary endothe-
lium and mechanical damage to the blood vessels
from overstretching during the process of re-
expansion.
56, 68
Furthermore, one study demonstrated a potential
role for free radicals provided by neutrophils in
the increase in pulmonary vascular permeability as
a cause of reperfusion pulmonary edema.
69,70
Free
radicals mediate damage in a variety of pathologic
conditions, including ischemia in organs such as
myocardium, intestine, and brain.
71,72
Reoxygena-
tion of ischemic tissue results in tissue damage.
73
One potential mechanism of this reperfusion injury
is that oxygen radicals lead to lipid peroxidation
and membrane injury. Thus, one-lung ventilation
followed by bilateral lung ventilation may cause
ischemia and reperfusion injury in the non venti-
lated lung and may increase pulmonary vascular
permeability.
All of this clinical literature indicates that the
rate of re-expansion may be more important than
duration of collapse in the development of reper-
fusion pulmonary edema.
56,62
Thus, other mecha-
nisms of re-expansion pulmonary edema include
generation of markedly negative intrathoracic
pressure during re-expansion and increased pul-
monary capillary pressure and flow on lung re-
expansion.
74
Therefore, the most important factor
in preventing re-expansion pulmonary edema is to
re-expand the lung slowly in a gradual fashion.
The treatment of established re-expansion pulmo-
nary edema is mechanical ventilation, PEEP, re-
striction of fluids, and diuretics.
VIII. RIGHT-SIDED HEART FAILURE
(Table 19-10)
Major pulmonary resection results in a decrease
in the cross-sectional area of the pulmonary vas-
culature and an increase in right ventricular after-
load, which in some patients can result in acute
right-sided heart failure.
75
Although patients at risk
for right-sided heart failure after pulmonary resec-
tion can usually be identified preoperatively,
76
right-sided heart failure may also occur if addi-
tional stresses such as infection, the development
of intrinsic PEEP, the application of extrinsic
PEEP, increased pulmonary blood flow, and new
Table 19-9 CRITERIA FOR MULTIPLE-ORGAN
FAILURE*
Organ Definition
Renal
Cardiovascular
Coagulation
Hepatic
Central
nervous
system
Gastrointestinal
Immunologic
Creatinine >3.5 mg/dl or patient
requiring dialysis or urine output <25
ml/h for 24 h
Cardiac index <l.8L/min/m
2
or systolic
blood pressure <90 mm Hg on
pressor drugs or arterial-venous
oxygen difference >7.0 ml 0
2
/dl
Platelet count <60,000/1 and
prothrombin time or partial
thromboplastin time >l . 5 times the
control
Bilirubin >5 mg/dl and prothrombin
time or partial thromboplastin time
> l .5 the control
Glasgow coma score <8 for three days
Pancreatitis causing shock or rupture/
necrotic/ischemic viscus or
gastrointestinal hemorrhage requiring
more than 2 U of packed red blood
cells
Patient receiving prednisone or the
equivalent of 50 mg/day for at least
seven days or prednisone or the
equivalent >20 mg/day for at least
one month or other known
immunosuppressive agent or patient
with AIDS
*From Suchiya MR, Clemmer TP, Elliott CG, et al: The
adult respiratory distress syndrome: A report of survival and
modifying factors. Chest 101:1074-1079, 1992. Used with per-
mission.
Abbreviation: AIDS = acquired immunodeficiency syn-
drome.
active pulmonary vasoconstriction (such as caused
by hypoxia, acidosis, and vasoactive amines and
peptides) are placed on the right side the heart.
Intrinsic PEEP (trapped gas at end-exhalation),
just like extrinsic PEEP, increases intrathoracic
pressure and adds yet another load on the right
ventricle.
42
Certainly, the development of ARDS
will always cause a significant increase in pulmo-
nary vascular resistance,
52
53
and sepsis is regularly
associated with or is causative of both right and
left ventricular dysfunction.
77
In addition, pulmonary hypertension and right
ventricular afterload will be increased by increased
pulmonary capillary and venous pressures, which
may result from fluid overload and/or decreased
left ventricular compliance. Decreased left ventric-
ular compliance may be due to left ventricular
ischemia and/or failure or interventricular septal
shifting (i.e., right ventricular afterload stress may
produce general dilatation of the right ventricle
and cause interference with left ventricular per-
formance through ventricular interaction).
78
79
Right-sided heart failure may also be manifested
by increasing venous hypertension and peripheral
edema, pulsatile liver, hepatojugular reflux,
marked elevation of the jugular venous waveform,
and tricuspid insufficiency.
A pulmonary artery catheter should be used in
any patient undergoing major pulmonary resection
who is believed to be at risk for postoperative
right-sided (or left-sided) cardiac failure; patients
at risk for right-sided heart failure include those
with an inferior wall myocardial infarction, which
often involves the right ventricle, and those with
pre-existing pulmonary hypertension.
The diagnosis of selective right-sided heart fail-
ure is established when the right atrial pressure
exceeds the left atrial pressure (i.e., the pulmonary
artery wedge pressure) in the presence of an ab-
normally low cardiac output. In addition, pulmo-
nary hypertension with a large pulmonary artery
diastolic to wedge pressure gradient will usually
be present, along with the systemic signs of heart
failure (oliguria, decreased mentation, and periph-
eral edema). One should remember that the pres-
ence of significant amounts of intrinsic PEEP
might confound interpretation of the pulmonary
capillary wedge pressure.
42
The treatment of acute right-sided heart failure
follows the same principles used in treating left-
sided failure: Control heart rate, optimize preload
and the inotropic state of the ventricle, and reduce
afterload (see chapter 13). Preload, as measured by
central venous pressure, can be reduced by fluid
restriction, by diuretics, or by a venous vasodilator
such as nitroglycerin. Inotropic drug support
should take into account the response of the pul-
monary vasculature to the drug chosen. Thus, do-
butamine, a drug that tends to reduce pulmonary
vascular resistance, is a reasonable choice in this
setting.
80
In fact, in this setting, a change from
dopamine to dobutamine results in an increase in
Table 19-10 CAUSES OF INCREASED RIGHT
VENTRICULAR AFTERLOAD AND
FAILURE AFTER PULMONARY
RESECTION
1. Resection of pulmonary vascular bed
2. Requirement for increased pulmonary blood flow
(sympathetic nervous system activation caused by pain,
stress, etc.)
3. Development of remaining lung infection, intrinsic PEEP,
ARDS, HPV
4. Right ventricular dilation * interventricular septal shifting
decrease left ventricular compliance pulmonary
hypertension
5. Decrease left ventricular compliance caused by left
ventricular ischemia - pulmonary hypertension
Abbreviations: PEEP = positive end-expiratory pressure;
ARDS = adult respiratory distress syndrome; HPV = hypoxic
pulmonary vasoconstriction.
stroke index and a significant decrease in right
atrial pressure.
81
Vasodilators that may be effective in reducing
pulmonary vascular resistance include nitric oxide,
nitroglycerin, nitroprusside, phentolamine, hydral-
azine, and prostaglandin E,.
82
Of these drugs, nitric
oxide and prostaglandin E, have the greatest pul-
monary specificity (spares the systemic circula-
tion), nitroglycerin and nitroprusside have inter-
mediate pulmonary specificity, and hydralazine the
least (causes only a moderate decrease in pulmo-
nary artery pressure and resistance and a large
decrease in systemic vascular resistance).
83
In addition to these drugs, treatment should in-
clude still other measures aimed at reducing pul-
monary arteriolar constriction. These include oxy-
gen administration to reduce hypoxic pulmonary
vasoconstriction, administration of bronchodilators
to relieve bronchospasm, antibiotics and chest
physical therapy to clear infection (to diminish the
size of the hypoxic compartment), mechanical
ventilation, and normalization of acid-base bal-
ance. Stabilization of cardiovascular function after
initiation of respiratory therapy may be attributed,
in part, to the disappearance of detrimental heart-
lung interaction.
79
IX. MYOCARDIAL ISCHEMIA/
INFARCTION
-
Myocardial infarction and cardiac failure asso-
ciated with noncardiac thoracic operations are im-
portant causes of mortality and serious morbid-
ity.
84-87
Transient ischemic electrocardiographic
changes have been documented in 3.8 per cent of
these patients and myocardial infarction in 1.2 per
cent.
88
The large majority of ischemic episodes
(and arrhythmias) occur on the second and third
postoperative days (Fig. 19-8).
88
This, of course,
corresponds with the time period of maximum
decrement to postoperative pulmonary function. In
fact, with close monitoring, a direct temporal rela-
tionship between decreases in S
p
0
2
and decreases
in ST-segment level, as well as both severity and
duration of the ST-segment decreases, can be ob-
served
89
(Fig. 19-9). Postoperative decreases in
S
p
0
2
are usually accompanied by increases in heart
rate, but the change in heart rate is usually propor-
tionally less than the changes in S
p
0
2
(see, e.g.,
Fig. 19-9).
89
90
Increases in heart rate are less than the concom-
itant decreases in S
p
0
2
postoperatively perhaps be-
cause the heart rate is usually already increased
postoperatively as a result of pain/stress, and the
potential for further increases is therefore reduced.
Nevertheless, decreased S
p
0
2
and increased heart
rate are a potent combination for producing myo-
cardial ischemia/infarction.
The diagnosis of myocardial infarction is made
by history (pattern of pain), evidence of autonomic
nervous system activation in the form of pallor,
sweating and peripheral vasoconstriction, electro-
cardiogram changes and serial cardiac enzymes.
Creatine kinase (CK) activity rises most rapidly
(peak of 6-9 per cent occurs between 6 and 24
hours) and is cleared most rapidly (decreases to
less than 1 per cent by 48 hours). The specific
isoenzyme CK-MB helps to identify cardiac rather
than skeletal muscle damage. Aspartate amino-
transferase activity is intermediate (peaks at 48
hours), and lactic dehydrogenase or alpha-hy-
droxybutyric dehydrogenase activity rises later
(peaks at 72 hours) and is cleared more slowly.
The principal aspects of the management of
acute myocardial infarction are administration of
oxygen, nitroglycerin, analgesia (morphine), and
aspirin, thrombolysis, and treatment of arrhyth-
mias (beta blocker) and heart failure (oxygen, mor-
phine, diuretic, venous and arterial vasodilators,
ionotrope).
X. ARRHYTHMIAS
Supraventricular arrhythmias, primarily sinus
tachycardia and atrial fibrillation and flutter, are
frequent complications after major pulmonary re-
sections, especially pneumonectomy,
91
92
even in
patients without significant pre-existing cardiac
disease.
88, 93
Atrial tachycardias occur in 16 per
cent of patients; atrial fibrillation is preponderant
(64-87 per cent), followed by supraventricular
tachycardia (23 per cent) and atrial flutter (13 per
cent).
88
93
The potential causes of these arrhythmias after
pulmonary resection include retraction and trauma
to the heart (e.g., arrhythmias occur more fre-
quently after intrapericardial dissection
93
and in
patients who have a pneumonectomy vs. a lesser
procedure)
88,93
and distension of the right ventricle
and atrium (caused by increased pulmonary artery
pressure). Arrhythmias occur more frequently in
patients in whom postoperative pulmonary intersti-
tial edema or perihilar infiltrates develop.
93
The
propensity for arrhythmias to occur in a trauma-
tized distended atrium may be compounded by
pre-existing cardiovascular disease, poor gas ex-
change, and sympathetic nervous system stimula-
tion caused by pain.
These arrhythmias may contribute greatly tc
perioperative morbidity and mortality,
8
especially
in the aged.
94
Indeed, patients with recurrent epl
sodes of arrhythmias have a significantly highe:
mortality than those without episodes or with {
Figure 19-8 Distribution of post-
operative cardiac events in 53 pa-
tients by days after operation (From
Von Knorring J, Lepantalo M, Lind-
gren L, Lindfors O: Cardiac arrhyth-
mias and myocardial ischemia after
thoracotomy for lung cancer. Ann
Thorac Surg 53:642-647, 1992.
single episode only (17 per cent vs. 2.4 per cent;
< .Ol).
88
In pneumonectomy patients, the occur-
rence of an arrhythmia is associated with a signif-
icant increase in mortality (25 per cent vs. 7 per
cent; < .Ol).
93
In survivors, the occurrence of an
arrhythmia doubles the length of stay in the inten-
sive care unit.
95
There are no generally accepted preventive reg-
imens for these arrhythmias. Digoxin has been the
only extensively studied drug to prevent atrial ar-
rhythmias after thoracic operations. However, be-
cause of positive as well as negative reported re-
sults, this regimen remains controversial (see
chapter 5).
88
It was demonstrated that intravenous
infusion of the antiarrhythmic drug flecainide ace-
tate for 3 days after thoracotomy prevented or re-
duced atrial and ventricular arrhythmias without
any side effects.
96
The preventive effect of verap-
amil in one study also seems promising.
97
Once the arrhythmia occurs, general treatment
consists of adequate sedation/analgesia and correc-
tion of hypoxia, hypercarbia or hypocarbia, and
right-sided heart failure (see Right-Sided Heart
Failure). Specific pharmacologic antiarrhythmia
treatment is dependent on the particular arrhyth-
mia
98
(see Table 13-19). Sinus tachycardia is
treated with propranolol (0.5 mg every 2 min up
to 5 mg). Paroxysmal atrial tachycardia can be
treated with verapamil or adenosine. Treatment of
atrial fibrillation is with digitalis (0.25-0.5 mg in-
travenously) if the patient is reasonably stable, ver-
apamil (5 mg/dose) if the patient is moderately
Figure 199 Patient No. 2: ST segment level, saturation, and heart rate (HR) plotted during the first night after operation.
Removal of oxygen was associated with severe decreases in S
P
0
2
and periods of ischemia. No ischemic episode occurred while the
patient received oxygen. (From Reeder MK, Muir AD, Foex P, et al: Postoperative myocardial ischaemia: Temporal association
with nocturnal hypoxaemia. Br J Anaesth 67:626-631, 1991. Used with permission.)
Figure 19-10 Among the general population, 20 to 34 per cent have a patent foramen ovale. Normally, the patent foramen ovale
remains functionally closed because left atrial pressure (LAP) exceeds right atrial pressure (RAP). When pulmonary vascular
resistance is increased, such as following a pneumonectomy or lobectomy, right ventricular (RV) and right atrial pressures may be
increased. If right atrial pressure exceeds left atrial pressure, then the foramen ovale opens and permits right-to-left shunting. (LV
= left ventricle; PVR = pulmonary vascular resistance; COPD = chronic obstructive pulmonary disease; CHF = congestive
heart failure; ARDS = adult respiratory distress syndrome; HPV = hypoxic pulmonary vasoconstriction.)
unstable, and externally synchronized cardiover-
sion if the patient is extremely unstable. Treatment
of atrial flutter consists of ouabain (0.1-0.2 mg
intravenously) or verapamil (5-10 mg intrave-
nously) (equal first choices), propranolol (second
choice), and external cardioversion (third choice).
Ventricular arrhythmias, usually premature
ventricular contractions, can be controlled with
intravenously administered lidocaine and/or beta-
blocking drugs. Ventricular fibrillation and tachy-
cardia must be controlled by electric defibrillation,
followed by lidocaine (Xylocaine) and, perhaps,
propranolol and, rarely, bretylium. Temporary
pacemaker insertion should be strongly considered
for patients in whom complete heart block has
developed.
XI. RIGHT-TO-LEFT SHUNTING
ACROSS A PATENT FORAMEN
OVALE
Many adults have a patent foramen ovale; at
autopsy, the incidence of a probe-patent foramen
ovale is highest at 34 per cent during the first three
decades of life and decreases to 20 per cent by the
ninth and tenth decades." There is normally no
right-to-left shunting across the foramen ovale be-
cause left atrial pressure exceeds right atrial pres-
sure, which keeps the one-way flap valve of the
foramen ovale pressed against the foramen ovale,
resulting in a functionally competent seal. If the
right atrial pressure exceeds left atrial pressure (as
might occur during the use of PEEP
100
l01
; in the
course of pulmonary embolization,
102, 103
pulmo-
nary hypertension,
102
high-altitude HPV,
104
chronic
obstructive pulmonary disease,
105
ARDS,
106
pul-
monary valvular stenosis,
100
107
congestive heart
failure,
100
108
right ventricular infarction,
109
or car-
diopulmonary by-pass
102,
" ' "; and during neuro-
surgical procedures complicated by air emboliza-
tion"
2
), the one-way flap valve can open, and
right-to-left shunting can occur through the fora-
men ovale (Fig. 19-10).
In terms of commonly occuring events, right-to-
left shunting at the atrial level may occur with
reaction to the tracheal tube during emergence
from anesthesia
113
and during the application of
PEEP.
101
The mechanism of increase in right atrial
pressure associated with reaction to the tracheal
tube is a decrease in lung volume that causes in-
creased pulmonary vascular resistance, pulmonary
artery pressure, and right atrial pressure.
113
Several reports documented new-onset right-to-
left shunting across a patent foramen ovale or
atrial septal defect after pneumonectomy
114-120
and
lobectomy
121
122
as a cause of otherwise unex-
plained postoperative dyspnea and systemic oxy-
Early Serious Complications Specifically Related to Thoracic Surgery
713
gen desaturation (Figs. 19-10 and 19-11).
120
Mechanisms for the shunting across the patent fo-
ramen ovale include an increase in right atrial
pressure to greater than left atrial pressure, which
is caused by an increased pulmonary vascular re-
sistance (due to the resection of vascular bed and/
or anesthesia- and surgery-induced lung disease)
and/or a shifting in the anatomic relationship of
the inferior vena cava, right atrium, and intra-atrial
septum resulting from the pulmonary resection, so
that the one-way flap valve is functionally less
competent and permits shunting across the fora-
men ovale (see Figs. 19-10 and 19-11).
120
The diagnosis of right-to-left shunting across a
patent foramen ovale should be suspected in any
patient who is hypoxemic (with relatively normal
chest roentgenogram findings) or who has a pro-
gressive PEEP-induced decrease in the arterial ox-
ygenation tension (these latter patients likely have
abnormal chest roentgenogram findings).
101
In the
latter circumstance, PEEP is presumably further
increasing pulmonary vascular resistance and right
ventricular and right atrial pressures, thereby in-
creasing the right-to-left shunt across the foramen.
In extubated patients, the diagnosis should be sus-
pected by otherwise unexplained clinically signifi-
cant dyspnea and hypoxemia. Contrast two-dimen-
sional and transesophageal echocardiography are
excellent, increasingly readily available methods
by which the diagnosis of this condition, if sus-
pected, may be made.
101
106
' "
3
l 2 0
l 2 2
Contrast an-
giography and dye dilution curve analysis are the
best definitive invasive diagnostic methods.
110
Since surgical closure is stressful to the patient
(it requires cardiopulmonary by-pass) and, in the
presence of pulmonary hypertension, may result in
acute right ventricular failure, therapy should al-
ways first be nonsurgical and aimed at decreasing
pulmonary vascular resistance and right ventricu-
lar and right atrial pressures. Thus, nonsurgical
treatment consists of decreasing pulmonary vaso-
constriction by clearing pulmonary infections, ad-
ministering oxygen, reversing acidosis, eliminating
conditions that release pulmonary vasoactive
amines and peptides (such as hypotension and in-
fection), using pulmonary vasodilators, and de-
creasing right-sided heart pressures by preload ma-
nipulation. These nonsurgical treatment modalities
usually permit functional closure of the foramen
ovale in the majority of patients. In one postpneu-
monectomy patient
123
and one patient with a right
ventricular infarction,
109
interatrial shunting was
successfully managed for a time by occluding the
foraminal orifice with a balloon-tip angiography
catheter. Rarely, a patient may undergo surgical
closure, even though he or she may still be on a
mechanical ventilator and critically ill.
106
109
The general care of a patient with right-to-left
intracardiac shunts must be meticulous. Avoidance
of air bubbles in the intravenous lines is manda-
tory. Forceful flushing of poorly infusing venous
catheters may result in systemic embolization.
108
Blood to be transfused should be filtered through
a micropore filter to exclude particulate material.
Since postoperative patients have an increased pro-
pensity to venous stasis, prophylactic anticoagula-
tion should be considered to reduce the possibility
of systemic thromboembolization.
108
XII. MASSIVE
POSTPNEUMONECTOMY
MALIGNANT PLEURAL EFFUSION
AND CHYLOTHORAX
At the end of a pneumonectomy, care should be
taken to ensure that the remaining lung is well
expanded and the mediastinum is near the midline
(see chapter 13). Excessive midline shift can com-
promise venous return to the heart. During the
weeks to months that follow surgery, fluid grad-
ually accumulates and fills the initially air-filled
hemithorax. The mediastinal structures slowly
shift toward the side of the resection, the ipsilateral
diaphragm progressively elevates, and the remain-
ing lung herniates anteriorly across the midline.
An early mediastinal shift toward the remaining
lung is a cause for concern and may indicate ate-
lectasis, postoperative hemorrhage, or infection in
the residual cavity with possible bronchopleural
fistula. A late mediastinal shift toward the remain-
ing lung is unusual, and the possibility of a malig-
nant effusion or chylothorax should be considered.
Malignant pleural effusions and chylothorax ac-
count for substantial morbidity. Treatment of ma-
lignant effusions is directed toward palliation
because life expectancy usually is limited and cu-
rative therapy does not exist. Ordinarily, the most
popular methods are repeated thoracentesis, chest
tube drainage (chemical pleurodesis is not an op-
tion because, with the lung removed, there is no
visceral pleura to oppose the parietal surface and
thereby obliterate the pleural space), and re-explo-
ration.
124
The same approach may be used for chylo-
thorax, but parenteral nutrition may allow a small
thoracic duct to close (low fat intake decreases the
formation of chylous fluid), and re-exploration is
necessary much more often because the large tho-
racic ducts may have been torn.
124
Pleuroperitoneal shunting for postpneumonec-
tomy malignant effusion
125
and chylothorax
126
has
proved to be an effective alternative treatment. A
unidirectional shunt valve with manual pump is
built into the shunt and ensures clearance of fluid
into the abdominal cavity regardless of the prs-
sure gradient between the thorax and abdomen.
The advantage of this remedy is that it can be
placed with the patient under local anesthesia. An
obvious concern is causing symptomatic intra-ab-
dominal tumor implantation, but this complication
has not been reported. Malignant effusions have
been successfully shunted for periods exceeding 1
year.
127
XIII. SYSTEMIC TUMOR EMBOLISM
Tumor emboli to peripheral vessels usually arise
from left atrial myxomas. However, the second
most common source of seeding is bronchogenic
lung carcinoma; primary lung neoplasms account
for 11 of the 29 cases of systemic tumor emboli
reported in the largest series.
128
Tumor invasion of
the pulmonary vein with subsequent dislodgment
during surgical manipulation leads to systemic
embolization. The common femoral artery is most
often the site of arterial tumor embolism,
129
but
unexpected occlusion of any artery after thoracot-
omy for tumor should raise suspicions of tumor
emboli. For example, intraoperative arterial embo-
lization of a bronchogenic tumor has led to occlu-
sion of the axillary artery (which caused immedi-
ate damping of the radial artery trace during
surgery).
130
XIV. NEURAL INJURIES
During radical hilar dissection or excision of
mediastinal tumors, phrenic, vagus, and recurrent
Figure 19-11 Transesophageal
echogram showing a patent foramen
ovale (PFO). Micro air bubbles were
observed to easily cross the PFO.
This patient had normal oxygenation
before right pneumonectomy and be-
came severely hypoxemic after right
pneumonectomy. Surgical closure of
the PFO resulted in a dramatic im-
provement in oxygenation. (RA =
right atrium; LA = left atrium; AO
= aorta.) (From Berry L, Braude S:
Refractory hypoxaemia after pneu-
monectomy: Diagnosis by trans-
esophageal echocardiography. Tho-
rax 47:60-61, 1992. Used with
permission.)
laryngeal nerves may be injured accidentally or
may be excised deliberately. Phrenic nerve injury
causes respiratory embarrassment by a flail chest
effect and also causes elevation of the ipsilateral
hemidiaphragm. The diagnosis should be sus-
pected in patients who have relatively clear chest
roentgenograms, have adequate gas exchange, and
cannot be weaned from the ventilator.
131
The di-
agnosis can be confirmed by paradoxic movements
of the diaphragm on fluoroscopy. Injury to the
vagus nerve causes gastric and intestinal atony,
which usually is not problematic in the first few
postoperative days. Bilateral partial injury of the
recurrent laryngeal nerve causes adductor spasm
of the vocal cords, which, following extubation.
may result in upper airway obstruction. This musl
be diagnosed promptly and treated with immediate
reintubation and possible tracheostomy until the
dysfunction is resolved.
The recurrent laryngeal nerve or phrenic nerve
usually shows return of function within 2 to
months after injury. If vocal cord function doe;
not return after recurrent laryngeal nerve injury
the involved vocal cord may be injected with ste
roids, which often causes a return of function. Ir
rare cases, surgical procedures, including the injec
tion of Teflon, may have to be performed on th<
involved cord to improve its function. If phreni<
nerve paralysis persists and continues to caus<
marked respiratory embarrassment due to paralysii
of the diaphragm, it is possible to implant j
phrenic nerve stimulator
132
or plicate the paralyze<
diaphragm
133
to correct the respiratory insuffi
ciency.
Aside from clamping of the thoracic aorta, then
are three other causes of paraplegia after thoracot
omy. First, damage to the spinal branches of inter-
costal arteries by dissection or diathermy at the
posterior end of a rib will cause spinal cord ische-
mia.
133
This is most likely to occur with damage to
the intercostal arteries of the left lower lobe.
134
Second, surgical dissection may create a commu-
nication between the epidural space and the
pleural cavity. Clotted blood, or blood that can
later clot, can then enter the epidural space and
cause postoperative spinal cord compression and
ischemia.
135
Third, packing hemostatic gauze into the poste-
rior end of a thoracotomy incision to stop intercos-
tal artery bleeding and leaving the packing in place
at the end of the thoracotomy has caused paraple-
gia in three patients; subsequent neurosurgery
showed that the hemostatic gauze disrupted the
dura and was intradural.
136
Presumably, the hemo-
static gauze was forced/propelled by tissue forceps
through an intervertebral foramen, subsequently
swelled when left in situ, and compressed the
spinal cord.
136
There are other less serious neural injuries that
are also specifically related to thoracic surgery.
The brachial plexus is especially vulnerable to
trauma during thoracic surgery and anesthe-
sia.
137
I38
The brachial plexus has a potentially long
superficial course in the axilla between two points
of fixation: the vertebrae above and the axillary
fascia below. If the two points of fixation are sep-
arated widely, stretching of the plexus will occur.
Suspension of the arm from an ether screen in an
excessive cephalad direction (see Fig. 10-1) in a
paralyzed patient without muscle tone, which in-
creases the distance between the two points of
fixation, is the most common cause of stretching
and damage to the brachial plexus. Stretching of
the plexus is also promoted by extreme abduction,
external rotation, and dorsal extension of the arm
on an arm board. The brachial plexus may also be
injured by a malpositioned chest tube (pushed in
too far cephalad).
139
The intercostal nerves are most exposed to in-
jury during intrathoracic surgical procedures. A rib
fracture occurring at the time of thoracotomy can
compress the intercostal nerve, resulting in an in-
tercostal neuritis, which is manifested by radicular
pain in the postoperative period. When excessive
stretching and tension are placed on the intercostal
nerve roots while the chest is being opened, post-
operative intercostal neuralgia may be quite se-
vere. Median sternotomy may cause a brachial
plexopathy,
140,141
and the sternal wires may cause
a parasternal intercostal neuropathy. The latter is
very effectively treated by local blocks.
142
In addi-
tion, the chest tubes can produce a neuroma or
neuritis by compression of an intercostal nerve.
XV. COMPLICATIONS OF
INTRATHORACIC INTERCOSTAL
NERVE BLOCKS
The performance of intrathoracic intercostal
nerve blocks under direct vision just prior to clo-
sure of the chest is a common method of securing
immediate postoperative analgesia. In one large
institution it has been estimated that these blocks
are performed in 10 to 15 per cent of thoracotomy
patients.
143
However, there have been case reports
of total spinal anesthesia following this proce-
dure.
143 144
In these patients, following placement
of intrathoracic intercostal nerve blocks at several
different levels, there was an immediate profound
decrease in blood pressure and pulse, persistent
apnea, failure to regain consciousness, absence of
reflexes, and dilated pupils, all of which are con-
sistent with total spinal anesthesia. In addition, the
temporal sequence of clinical recovery
145
and a
clear-cut sensory level when the patient awakened
also strongly indicate the occurrence of total spinal
anesthesia.
There are several possible mechanisms for this
complication. First, there can be outward prolon-
gation of the subarachnoid space along the nerve
roots. These durai cuffs can extend as far as 8 cm
past the intervertebral foramen.
145
146
It is possible,
then, that the local anesthetic was introduced into
one or more durai cuffs surrounding the intercostal
nerves. A second possibility is that the local anes-
thetic can spread centrally to the spinal fluid via
the perineural spaces. A third possibility is that the
local anesthetic could have been introduced di-
rectly into the subarachnoid space through an in-
tervertebral foramen because of an unnoticed ill-
placed angulation of the needle.
These reports constitute a relative contraindica-
tion to the performance of intercostal nerve blocks
near the spinal cord with local anesthetics when
the patient is under general anesthesia and when it
is not possible to assess accurately the effect of the
nerve blocks. These reports also emphasize the
necessity of including spinal anesthesia in the dif-
ferential diagnosis of postoperative cardiopulmo-
nary and central nervous system depression when
local anesthesia has been used in the vicinity of
the spinal cord during general anesthesia.
REFERENCES
1. Allison PR: Intrapericardial approach to the lung root in
the treatment of bronchial carcinoma by dissection pneu-
monectomy. J Thorac Cardiovasc Surg 15:99, 1946.
2. Baaijens PFJ, Hasenbos AWM, Lacquet LK, Dekhuijzen
PNR: Cardiac herniation after pneumonectomy. Acta An-
aesthesiol Scand 36:842-845, 1992.
3. Yacoub MH, Williams WG, Ahmad A: Strangulation of
the heart following intrapericardial pneumonectomy.
Thorax 23:261-265, 1968.
CHAPTER
20
Mechanical Ventilation and
Weaning
720
I. INTRODUCTION
The clear-cut indications for postoperative me-
chanical ventilation (at least for a few hours) are
numerous and include severe pre-existing lung dis-
ease (see chapter 5), ventilatory depression due to
residual anesthesia or paralysis (for example, the
peak inspiratory force is less than 20 cm H
2
0),
presence of a significant amount of blood in the
airway, massive intraoperative transfusion follow-
ing many pneumonectomies, large air leak follow-
ing any resection (possible bronchopleural fistula),
presence of flail chest, extensive surgical trauma
to the lung remaining in the operative hemithorax,
multiple and severe organ trauma, sepsis, and un-
acceptable arterial blood-gas concentrations (per-
haps due to the factors just listed).
In addition to these obvious indications, contin-
uing intubation and mechanical ventilation facili-
tates a smooth transition from the operating room
to the intensive care unit (provided the patient is
adequately anesthetized/sedated); allows the intra-
operative use of high-dose narcotic anesthesia; en-
ables adequate postoperative analgesia without
concern for depression of respiration; allows a
more aggressive approach to hemodynamic sup-
port (including intravascular volume loading)
without excessive concern for its effects on pul-
monary function; avoids the need for application
of intensive respiratory therapy in exhausted,
stressed patients in the early postoperative period;
allows restoration of functional residual capacity
(FRC) prior to extubation; and facilitates return to
the operating room if there are concerns about
surgical complications. If postoperative mechani-
cal ventilation is planned, even for only a few
hours, the patient should continue to receive nar-
cotics and be paralyzed, and the double-lumen
tube should be changed to a single-lumen tube
(except perhaps when there is a large broncho-
pleural fistula; see Differential Lung Ventilation
later) prior to leaving the operating room.
The use of positive end-expiratory pressure
(PEEP), continuous positive airway pressure
(CPAP), intermittent mandatory ventilation (IMV),
of either a spontaneously breathing or apneic pa-
tient, pressure support ventilation of a sponta-
neously breathing patient, and an increased under-
standing of the toxicity of unnecessarily high
inspired concentrations of oxygen (F,0
2
) constitute
the cornerstones of modern clinical mechanical
ventilation management. This chapter describes, in
a step-by-step manner, the current practice of me-
chanical ventilation, weaning, and extubation.'^
The recommendations for initial ventilator settings
(see later) are based on the assumption that the
patient's lungs are disordered and, therefore, re-
quire a considerable amount of support. In addi-
tion, it is assumed that a volume-limited ventilator
is being used, the patient is adequately sedated,
and the endotracheal tube and chest tubes are func-
tioning properly.
II. INITIAL VENTILATOR SETTINGS:
INTERMITTENT MANDATORY
VENTILATION
IMV means that the ventilator delivers a posi-
tive-pressure breath a mandatory number of times
per minute while the patient does as much sponta-
neous breathing as desired between these mechan-
ical breaths. The IMV mode of ventilation is at
present the most commonly used and preferable
ventilation mode for several reasons.
5
First, be-
cause the patients are allowed to do, and usually
are doing, some spontaneous ventilation, they are
able to generate a more negative pleural pressure,
which better maintains venous return and cardiac
output (compared with simple intermittent posi-
tive-pressure ventilation.
6-9
The descent of the dia-
phragm into the abdomen with spontaneous venti-
lation also increases abdominal pressure, squeezes
the splanchnic circulation, and contributes to ve-
nous return.
10
Second, and consequently, less intravenous fluid
is usually required to maintain systemic pressure
and perfusion. Third, since it is not necessary to
eliminate these spontaneous breaths, sedation re-
quirements are minimized, and paralysis require-
ments are eliminated. Fourth, the patients benefit
psychologically because they know they are
breathing on their own. Fifth, the respiratory mus-
cles remain coordinated because they are exer-
cised. Finally, when the weaning stage is reached,
IMV allows a gradual transition from ventilator
dependence to independence. There are a number
of other mechanical ventilation modes that also
facilitate mechanical ventilation and/or the transi-
tion from mechanical ventilation to weaning, such
as pressure support (PS), mandatory minute venti-
lation (MMV), inverse ratio ventilation (IRV), and
airway pressure release ventilation (APRV), which
have been introduced; these are discussed under
Other New Mechanical Ventilation and Weaning
Modes.
With conventional IMV, it is possible for the
ventilator to begin to deliver a mechanical breath
just as the patient is completing a spontaneous
inhalation. This results in the stacking of a ma-
chine breath on top of a spontaneous breath and
can lead to overdistension of the lungs and high
peak inspiratory pressures. This problem can be
overcome by use of synchronized IMV (SIMV).
With SIMV, the ventilator intermittently senses
the onset of a spontaneous breath a mandatory
722 Mechanical Ventilation and Weaning
number of times per minute and immediately
causes a machine (positive-pressure) breath to oc-
cur. Consequently, the machine breath overrides
the spontaneous breath, and stacking of breaths
cannot take place.
Mechanical ventilation is initiated with a tidal
volume of 12 ml/kg of lean body weight, an IMV
respiratory rate so that the arterial carbon dioxide
tension (P
a
C0
2
) is 40 mm Hg (which usually re-
quires an initial rate of 8 to 12 breaths/min), and
an inspired oxygen fraction (F,0
2
) of 1.0 (Table
20-1). Although patients may breathe sponta-
neously during IMV (which is especially helpful
with regard to lowering pleural pressure and main-
taining cardiac output), IMV should be regarded
as initially providing total and full ventilatory sup-
port (i.e., the IMV breaths, not the spontaneous
breaths, determine the minute ventilation)."
12
One
hundred per cent oxygen should initially be used
in order to update and document the level of right-
to-left shunting and alveolar-arterial oxygen ten-
sion difference while minimizing the risk of unex-
pected hypoxemia. Administration of 100 per cent
oxygen for a short period of time is not toxic (see
later).
A tidal volume of 12 ml/kg is considered large
tidal volume ventilation. Large tidal volume ven-
tilation is desirable because it prevents the devel-
opment of miliary atelectasis or microatelectasis
that is associated with constant small tidal volume
ventilation (for example, 6 ml/kg).
13
The miliary
atelectasis or microatelectasis that is associated
with small tidal volume ventilation results in a
Table 20-1 INITIAL VENTILATOR SETTINGS
1. IMV or SIMV mode
2. Tidal volume 12 ml/kg
3. Respiratory rate so that P
a
C0
2
=
(usually 8 to 12 breaths/min)
4. F,0
2
= 1.0
5. I:E = 1:2 to 3
6. Inspiratory flow rate 30 L/min
40 mm Hg
Abbreviations: IMV = intermittent mandatory ventilation;
SIMV = synchronized IMV; I:E = inspiratiomexpiration ra-
tio.
progressive increase in the alveolar-arterial oxygen
tension difference and a progressive decrease in
compliance. These deleterious changes associated
with small tidal volume ventilation can be reversed
by intermittent sighs. Large tidal volume ventila-
tion, therefore, functions as frequent intermittent
sighs and, thereby, minimizes the development of
atelectasis. However, tidal volumes in excess of 15
ml/kg may cause the production of a permeability
(high-protein) pulmonary edema caused by
stretching of extra- and intra-alveolar vessels
(causing a decreased extra-alveolar vessel peri-
microvascular pressure)
14-16
and alveolar epithe-
lium
17-19
(Fig. 20-1). A small tidal volume should
be used in patients with bullous or cystic disease
of the lung in an effort to minimize peak inspira-
tory pressures and risk of tension pneumothorax.
The same considerations apply to patients with
severe airflow obstruction (i.e., low tidal volume
minimizes air trapping, hyperinflation, baro-
trauma).
20
Figure 20-1 Scanning electron
micrograph of lung capillaries and
epithelium ruptured at high lung vol-
ume. Arrows show edges of epithe-
lial breaks. Photomicrographs sup-
plied by Drs. John B. West and Odile
Mathieu-Costello. (From Parker JC,
Hernandez LA, Peevy KJ: Mecha-
nisms of ventilator-induced lung in-
jury. Crit Care Med 21:131-143,
There is no evidence that the type of inspiratory
flow pattern (square wave, sinusoidal, accelerating,
decelerating) makes any difference in any impor-
tant respiratory or hemodynamic parameter.
5
How-
ever, the physics of lung time constants mandate
that the greater the time interval from gas delivery
to end inspiration, the more complete the distribu-
tion of gas to areas of low ventilation/perfusion.
Thus, an inspiratory hold is usually desirable and,
with a ventilator rate of 10 breaths/min (6-sec fre-
quency) and an inspiratory delivery time of 1 sec,
an inflation hold of up to a 1 sec may be given
while still maintaining an inspiratiomexpiration
(I:E) ratio of 2:4 or 1:2. However, in patients with
severe airflow obstruction (perhaps manifested by
hyperinflation- [diagnosis made by chest X-ray
and/or measurement of intrinsic PEEP] induced
circulatory depression), I:E ratios of 1:3 or 1:4 are
more appropriate.
20
The respiratory rate should be set so that the
patient has a normal arterial carbon dioxide ten-
sion (this rate is usually 8 to 12 breaths/min). Cer-
tainly, hypercarbia should be avoided in the vast
majority of patients since the accompanying aci-
dosis is dangerous, and the hypoventilation neces-
sary to produce hypercarbia may also cause hy-
poxemia. Mild degrees of hypercarbia (P
a
C0
2
40
to 50 mm Hg) are indicated for patients who are
chronic carbon dioxide retainers; mild hypercarbia
helps to avoid changing the arterial and cerebro-
spinal fluid pH and bicarbonate levels from the
values that the patient normally lives with.
Hypocarbia should be avoided for multiple rea-
sons. First, hypocarbia may cause a decrease in
cardiac output via several mechanisms. In order to
produce hypocarbia, an increased amount of posi-
tive-pressure ventilation will have to be used,
which will decrease venous return. Hypocarbia
also causes a withdrawal of sympathetic tone,
which will decrease the inotropic state of the heart.
Hypocarbia and alkalosis may decrease the level
of ionized calcium, which will also decrease the
inotropic state of the heart.
21
Second, hypocarbia will also shift the oxygen-
hemoglobin dissociation curve to the left, which
increases the hemoglobin affinity for oxygen; at
the tissue level, the increased hemoglobin-oxygen
binding impedes the delivery of oxygen to the
tissues.
Third, hypocarbia and alkalosis will increase
oxygen consumption. The increase in oxygen con-
sumption is due to a pH-mediated uncoupling of
oxidation from phosphorylation; the uncoupling
requires more oxygen to be passed along the res-
piratory chain to generate the same amount of
adenosine triphosphate.
22
23
Quantitatively, at a
P
a
C0
2
of 40 mm Hg, oxygen consumption is 250
ml/min, whereas at a P
a
C0
2
of 20 mm Hg, oxygen
consumption is 500 ml/min. Thus, hypocarbia may
increase tissue oxygen demand (as a result of in-
creased oxygen consumption) at a time when it is
not possible to increase tissue oxygen supply (ox-
ygen-hemoglobin dissociation curve is shifted to
the left and cardiac output is decreased).
Fourth, hypocarbia may cause ventilation-per-
fusion abnormalities by inhibiting hypoxic pulmo-
nary vasoconstriction (HPV) and/or by causing
bronchoconstriction and decreased lung compli-
ance. Fifth, hyperventilation with alkalosis may
cause ventricular irritability (perhaps due, in part,
to a concomitant hypokalemia). However, the ben-
efit of hyperventilation with hypocarbia to de-
crease an elevated intracranial pressure in head-
injured patients may supersede the previously cited
physiologic disadvantages. Fast respiratory rates
should be avoided in patients with severe airflow
obstruction (i.e., avoid hyperinflation).
20
If the measured P
a
C0
2
differs from the desired
level, one can easily calculate the amount of
change in minute ventilation needed to produce
the desired change in P
a
C0
2
by using the relation-
ship:
V
E
2
= (P
a
C0
2
'/P
a
C0
2
2
) X V
E
>
where V
E
2
= the desired minute volume, V
E
' =
the actual minute volume, P
a
C0
2
2
= the desired
P
a
C0
2
, and P
a
C0
2
' = the actual P
a
C0
2
. Use of
total minute ventilation, which includes both al-
veolar ventilation and dead space ventilation, in-
stead of using only alveolar minute ventilation is
clinically acceptable since the dead space remains
relatively constant with moderate changes in tidal
volume.
24
A change in either respiratory rate or
tidal volume, or both, can be used to change min-
ute ventilation.
The use of large tidal volume ventilation usually
makes it possible to achieve normocarbia with rel-
atively slow respiratory rates (8 to 12 breaths/min).
This is highly desirable in two ways. First, slow
rates allow a decrease in the (I:E) ratio. A long
expiratory time increases venous return and aug-
ments cardiac output; it facilitates full expiration
and avoids air trapping and obstructive lung dis-
ease. The I:E ratio should be approximately 1:2 or
3 (except when airway resistance is high and an
even longer exhalation period and lower I:E ratio
are required). Second, slow respiratory rates also
allow slower inspiratory flow while maintaining
an appropriately low I:E ratio. The slower the in-
spiratory flow, the more laminar and less turbulent
is the air stream. This decreases airway resistance,
reduces airway trauma, and allows for a more even
distribution of inspired gas to lung regions with
different time constants. The inspiratory flow
should be approximately 30 L/min.
III. GOAL #1:|F,Q
2
<0.51AND P
a
0
2
ACCEPTABLE
A. Why This Goal Is #1 : Oxygen
Toxicity
An inspired oxygen fraction (F,0
2
) greater than
0.5 is toxic to the lungs,
25-27
and the first and com-
pelling goal is to get the inspired oxygen fraction
to less than 0.5 while maintaining an acceptable
P
a
0
2
(Table 20-2). The mechanism of oxygen-
induced lung damage is the local accumulation of
toxic, tissue-reactive free radicals, such as super-
oxide and hydrogen peroxide, which damage the
pulmonary tissues. For this discussion, an accept-
able P
a
0
2
is defined as one that will be above the
steep part of the oxygen-hemoglobin dissociation
curve (see Fig. 3-32), and for all patients this will
be a value greater than 60 mm Hg. A P
a
0
2
value
that is above the steep part of the oxygen-hemo-
globin dissociation curve provides some margin of
arterial oxygenation reserve and is protective
against decreases in P
v
0
2
. In addition, a P
a
0
2
greater than 60 mm Hg will not cause or promote
ventricular arrhythmias. Other aspects of oxygen
toxicity such as carbon dioxide retention and ab-
sorption atelectasis are discussed in chapter 3.
The rate of onset of oxygen-induced pulmonary
damage is proportional both to the inspired tension
of oxygen and to the duration of the exposure. In
normal human volunteers, the first symptom after
6 hours of exposure to 100 per cent oxygen is
substernal distress, which begins as a mild irrita-
tion in the area of the carina and may be accom-
panied by occasional coughing.
28
As exposure con-
tinues, pain becomes more intense, and the urges
to cough and deep breathe also become more in-
tense. When the inspired oxygen tension has been
1.0 for greater than 12 hours, the pain, dyspnea,
and paroxysmal coughing are severe. As toxicity
progresses, these symptoms are accompanied by
objective signs of lung compromise, such as a
decreased vital capacity and lung compliance, and
deterioration in blood-gas concentrations. Finally,
in animals, after days to weeks, acute respiratory
failure occurs, and, unless gas exchange is sup-
ported by mechanical ventilation, death ensues.
Pathologically, in animals, the lesion progresses
from a tracheobronchitis (exposure for 12 hours to
a day), to pulmonary interstitial edema (exposure
for a few days to a week), to pulmonary fibrosis
(exposure for greater than a week).
29
Often these
pathologic changes are indistinguishable from
those of the adult respiratory distress syndrome
(ARDS). Dose-time toxicity curves for humans
indicate that oxygen toxicity does not occur when
the inspired oxygen fraction is less than 0.5 (even
for long periods of time).
B. How to Achieve Goal #1
PEEP is the treatment of choice of arterial hy-
poxemia due to right-to-left shunting through low
ventilation-perfusion and atelectatic lung regions.
PEEP increases end-expiratory lung volume
(FRC), thereby preventing closure of airways at
end exhalation, and recruits alveoli and airways
during inspiration (see Figs. 3-27 and 3-28). The
prevention of airway closure and recruitment of
alveoli increases ventilation-perfusion relation-
ships and reverses atelectasis.
1
PEEP may also
facilitate the movement of water from the less
compliant interstitial space between alveolar epi-
thelium and capillary endothelium to the more
compliant peribronchial and hilar interstitial
spaces.
30
This redistribution of interstitial water
improves lung compliance and oxygen diffusion
across the alveolar-capillary membrane.
31
The ben-
eficial effect of the application of PEEP and the
deleterious effect of removal of PEEP on arterial
oxygenation can be observed within minutes.
32,33
Thus, in post-coronary artery by-pass grafting sur-
gery patients, a nitroglycerin-induced increase in
shunt (caused by decreased HPV) is reversed
within minutes when 10 cm H
2
0 of PEEP is ap-
plied because the PEEP converts the atelectatic
shunt units to gas-exchanging units.
34
Table 20-2 MECHANICAL VENTILATION AND WEANING PLAN
Goal
(Temporal
Sequence to
Be Followed) Goal
Achieved
Primarily By
1. JF,0
2
2. | PEEP
3. 1 IMVrate
F,O
2
<0.5
1 F,0
2
<0.5,
1 F,0
2
<0.5,
i PEEP<10cmH
2
O,
P
a
0
2
>60 mm Hg PEEP titration
P
a
0
2
>60 mm Hg Respiratory care regimen
1 PEEP<10cmH
2
O, i IMV <1 breath/min P
a
0
2
>60 mm Hg Patient having adequate
breathing power
Abbreviations: PEEP = positive end-expiratory pressure; IMV = intermittent mandatory ventilation.
Mechanical Ventilation and Weaning 7 2 5
The end point for PEEP application is the low-
est level of PEEP that provides an adequate P
a
0
2
on an F,0
2
of less than 0.5.
35
~
37
Increasing PEEP
beyond this level to obtain optimal values for var-
ious other end points, such as the production of
maximum oxygen transport,
18
maximum static pul-
monary compliance,
38
shunt less than 15 to 20 per
cent,
37
39
40
minimal arterial end-tidal carbon diox-
ide gradient,
41
42
decreased mixed venous oxygen
tension,
43
and minimal F,0
2
,
44
will not be clinically
helpful and may be harmful.
35,36
In addition, most
of these other end points require invasive monitor-
ing, vary independently of the PEEP level (i.e.,
they have multiple determinants and, therefore,
have a fair amount of background variation), and/
or respond sluggishly to changes in lung disease.
The decrease in inspired oxygen concentration
below 50 per cent is achieved by performing a
dose (PEEP)-response (P
a
0
2
) titration. PEEP is
progressively added in 3- to 5-mm Hg increments
(range = 0-20 mm Hg)
45
until the P
a
0
2
is rela-
tively normal for that patient (or at least above 60
mm Hg) with an F,0
2
less than 0.5. PEEP should
be raised in increments of 2.5 to 5.0 cm H
2
0; the
patient should be allowed to stabilize with regard
to respiratory mechanics (peak inspiratory pres-
sure, compliance), hemodynamics (pulse rate and
rhythm, systemic and central filling pressures, and
cardiac and urine output), and gas exchange (arte-
rial blood gases). Usually each PEEP increment
requires 0.5 to 1 hour to complete. Consequently,
the performance of a PEEP-F,0
2
/P
a
0
2
titration
usually takes only several hours.
In many patients there appears to be a certain
critical level of PEEP above which there is sudden
and dramatic improvement in P
a
0
2
. In the vast
majority of patients, levels of 5 to 12 cm H
2
0 are
sufficient for substantial reversal of intrapulmo-
nary shunting; it is seldom necessary to go as high
as 15 cm H
2
0. Of course, as the PEEP titration is
being performed, other respiratory care modalities
(see later), such as suctioning, turning, and admin-
istration of antibiotics, are instituted. The advent
of fiberoptic bronchoscopy has eliminated the need
for use of super-PEEP (PEEP greater than 20 to
25 cm H
2
0) because regions of collapsed lung that
are resistant to opening by PEEP are usually ob-
vious on chest roentgenogram and can be suc-
tioned and lavaged open under direct vision.
As the PEEP requirement increases, necessity
for cardiovascular support and, therefore, monitor-
ing requirements (central venous and pulmonary
capillary wedge pressures and cardiac and urine
output) generally increase.
46
It may be expected
that hypovolemic patients will have a decrease in
cardiac output as soon as PEEP is applied. Nor-
movolemic patients will maintain cardiac output
within normal limits up to a PEEP of 10 cm H
2
0
because compensatory peripheral venoconstriction
maintains venous return.
47
However, normovo-
lemic patients will have a progressive decrease in
cardiac output with a progressive increase in PEEP
greater than 10 cm H
2
0. Postpneumonectomy pa-
tients with an increased pulmonary vascular resis-
tance may be expected to have a greater decrease
in cardiac output at any given level of PEEP com-
pared with patients with normal lungs.
48
Again, in
all patients, use of the IMV mode with some spon-
taneous ventilation is protective of the cardiac out-
put.
The first cardiovascular support response to
PEEP-induced impairment of hemodynamics
should be moderate augmentation of intravascular
volume (i.e., Trendelenberg's position [10-20
tilt] or infusion of 1 liter of crystalloid).
47
49
If fluid
infusion does not restore satisfactory hemody-
namic balance, then inotropic agents should be
used to provide cardiovascular support. In patients
with acute hypoxemic respiratory failure, dobuta-
mine causes an increase in cardiac output with a
decrease in pulmonary capillary wedge pressure
and end-diastolic and end-systolic ventricular vol-
umes (due to decreased afterload), whereas dopa-
mine increases cardiac output, pulmonary capillary
wedge pressure, and end-diastolic and end-systolic
volumes (due to increased afterload).
50
Since even
small increases in pulmonary capillary wedge
pressure may increase the accumulation of pulmo-
nary water in permeability pulmonary edema,
51
"
53
dobutamine appears to be the inotropic drug of
choice.
50
The unloading of the right ventricle by
dobutamine compared with dopamine results in a
higher cardiac output and lower central venous
pressure.
54
C. Mechanisms of PEEP-lnduced
Decreased Cardiac Output
One of the major limiting factors to the appli-
cation of PEEP is a decrease in cardiac output. In
the past several years, multiple mechanisms have
been proposed for the PEEP-induced decrease in
cardiac output; the following mechanisms have
been reasonably well supported by experimental
and clinical observations.
1. Decreased Venous Return
5551
PEEP causes an increase in intrathoracic pres-
sure, which decreases the gradient for blood to
flow into the central intrathoracic veins from the
peripheral veins (Fig. 20-2). Up to a PEEP level
of 10 cm H
2
0 in normovolemic patients and ani-
mals, sympathetically mediated venoconstriction
increases peripheral venous pressure, thereby
Mechanisms of PEEP-lnduced Decreased Cardiac Output
maintaining a normal pressure gradient for venous
return. However, a progressive increase in PEEP
greater than 10 cm H
2
0 causes a progressive de-
crease in cardiac output in normovolemic animals
and patients. The decrease in cardiac output can
usually be offset by fluid infusion only, which also
results in increased peripheral venous pressure but
at the expense of a hypervolemia relative to the
blood volume that existed at 0 cm H
2
0 end-expi-
ratory pressure.
58
Because a sympathetically me-
diated venoconstriction at least partially offsets the
hemodynamic effect of PEEP, it is not surprising
that a significantly greater decrease in cardiac out-
put at any given PEEP level can be caused by
alpha-adrenergic blockade with phenyoxybenza-
mine.
55
Occasionally, infusion of a venoconstrictor
(alpha-adrenergic stimulation) is necessary to re-
store satisfactory hemodynamics.
55
2. Increased Pulmonary Vascular
Resistance and Right Ventricular
Dysfunction
59
Pulmonary vascular resistance is a U-shaped
function of increasing lung volume (see Fig. 3-
15). The trough of the U-shaped total pulmonary
vascular resistance curve occurs at normal FRC.
As lung volume increases above FRC, the small
intra-alveolar vessels are compressed and flattened
(increasing alveolar dead space),
60
and small-ves-
sel pulmonary vascular resistance increases, caus-
ing an increase in total pulmonary vascular resis-
tance. As lung volume decreases below FRC,
large-vessel resistance increases, causing an in-
crease in total pulmonary vascular resistance. The
increase in large-vessel resistance below FRC is
due to HPV constriction (small lungs are hypoxic);
mechanical tortuosity of vessels is not thought to
contribute significantly to the increased resistance.
It can be seen from Figure 3-15 that the effect
of PEEP on pulmonary vascular resistance is de-
pendent upon the initial lung volume at which the
PEEP is applied (i.e., where the patient is on the
U-shaped curve). If PEEP increases lung volume
from a small volume to a normal FRC, pulmonary
vascular resistance may decrease (due to a de-
crease in large-vessel resistance), and cardiac out-
put may increase. If PEEP increases lung volume
from a normal FRC to a large value, pulmonary
vascular resistance may increase greatly (due to an
increase in small-vessel resistance).
The increase in total pulmonary vascular resis-
tance with large lung volumes increases the after-
load on the right ventricle. The right ventricle is a
thinly muscled cavity and can easily dysfunction
(distend and have poor contractility)
61
62
when pul-
monary vascular resistance (afterload) is greatly
increased (Fig. 20-2 through 20-4). For example,
raising mean pulmonary artery pressure by a small
amount (i.e., 20 mm Hg) reduces right ventricular
stroke volume by as much as 30 per cent. Addi-
tional increases in mean pulmonary pressure lead
to overt right ventricular failure.
61
The left ventri-
cle is a better pressure pump, and small increases
(i.e., 20 mm Hg) in systemic pressure reduce left
ventricular stroke volume by less than 5 per cent
61
(see Fig. 20-4). Thus, seemingly small PEEP-in-
duced acute elevations in right ventricular after-
load (20 mm Hg) cause a relatively large decrease
in right ventricular cardiac output (Fig. 20-2B).
3. Leftward Displacement of
Interventricular Septum
49
Under normal conditions (no PEEP, nondis-
tended heart), the pericardium does not affect ven-
tricular function.
63
There is evidence that progres-
sively increasing PEEP (to treat acute respiratory
failure)
64
causes the following sequence of events:
increased pulmonary vascular resistance and right
ventricular afterload; right ventricular free-wall ex-
pansion and distension, which causes complete
filling of the pericardial sac; and, finally, move-
ment of the interventricular septum into the left
ventricle (Fig. 20-2C). Additionally, PEEP-in-
duced right ventricular distension may decrease
coronary artery blood flow
65
and worsen the sever-
ity of any pre-existing right ventricular
ischemia.
66
67
Both events lead to a progressive
negative cycle by causing further right ventricular
distension.
When the right ventricle has filled the pericar-
dial sac, both ventricles must function within the
common and unyielding stiffness of the pericar-
dium, and the only structure that can give (allow
further right ventricular expansion) in response to
continued or increased pressure loading of the
right ventricle is the interventricular septum. Corn-
Figure 20- 2 Mechanisms of PEEP-induced decreased cardiac output consist of decreased venous return (A), increased pulmonary
vascular resistance and right ventricular dysfunction (B), leftward displacement of the interventricular septum (C), decreased
transmural filling pressure (D), neural reflexes (E), and humoral reflexes (F). ( + = positive pressure (above atmospheric); PEEP
= positive end-expiratory pressure; PVR = pulmonary vascular resistance; RV = right ventricle; R = right; L = left; LV = left
ventricle; ZEEP = zero end-expiratory pressure.) See text for full explanation of each of these mechanisms.
Figure 203 Increased pulmonary vascular
resistance (PVR) causes a decrease in right ven-
tricular (RV) ejection fraction (EF) and an in-
crease in RV end-systolic volume (ESV) and
end-diastolic volume (EDV). The increase in
ESV and EDV results in right ventricular disten-
sion.
pression by the overlying lungs may also prevent
further right ventricular free-wall expansion. Para-
doxically, right ventricular performance is also im-
paired by the development of very negative intra-
thoracic pressure (as might occur during acute
asthma) because the right ventricular free wall is
held outward by the negative pleural pressure (in-
ward motion is restrained).
68
The right-to-left shift in the position of the in-
terventricular septum with PEEP causes the sep-
tum, which is normally convex toward the right
ventricle, to become increasingly flat and perhaps
concave toward the right ventricle. The displace-
ment of the septum into the left ventricle interferes
with left ventricular filling, compliance, and con-
tractility.
69-73
Therefore, by its restraining influence
on both ventricles, the pericardium enforces dy-
namic ventricular interaction
73
74
through septal
shifting, and left ventricular volume and diastolic
pressure depend on the amount of right ventricular
filling.
64
75
It is apparent that at high PEEP levels
the decreases in cardiac output may be due to
acute biventricular failure.
Vascular Resistance
% Increase
Figure 20- 4 The x-axis (vascular resistance per cent in-
crease) and y-axis (ejection fraction relative decrease) in this
diagram are constructed from data in Rose et al.
34
and refer to
the combinations of pulmonary vascular resistance/right ventri-
cle (RV) and systemic vascular resistance/left ventricle (LV).
The right ventricle experiences a much more precipitous de-
crease in ejection fraction when pulmonary vascular resistance
is increased compared with the decrease in left ventricular
ejection fraction when the systemic vascular resistance is in-
creased.
4. PEEP-lnduced Decreases in
Transmural Ventricular Filling
Pressures
47 76 77
The application of PEEP can cause a decrease
in the transmural ventricular filling pressure (intra-
cavitary pressure relative to pleural pressure).
Transmural ventricular filling pressures can de-
crease owing to differential (partial) transmission
of the PEEP applied at the mouth to the pleural
space and then from the pleural space to the cavi-
ties within the ventricles. Consider the example
shown in Figure 20-2D. Initial conditions (pre-
PEEP) consist of pressure of 0 cm H
2
0 at the
mouth, pleural pressure of - 5 cm H
2
0, and a left
ventricular end-diastolic filling pressure (relative
to atmospheric pressure outside the body) of 5 cm
H
2
0, resulting in a transmural ventricular filling
pressure (relative to pleural pressure) of 10 cm
H
2
0. Following the application of 10 cm H
2
0 of
PEEP at the mouth, pleural pressure increases, but
only by 7 cm H
2
0, to a resultant 2 cm H
2
0. Simi-
larly, intracavitary ventricular filling pressure in-
creases (relative to atmospheric pressure outside
the body), but only by 4 cm H
2
0, to a resultant 9
cm H
2
0. However, the transmural ventricular fill-
ing pressure has now decreased to 7 cm H
2
0.
Thus, the 10 cm H
2
0 of PEEP applied at the
mouth was only partially transmitted to the pleural
space, and, in turn, was only partially transmitted
to the cavities within the ventricles; the differential
transmission of PEEP resulted in a decrease in the
transmural ventricular filling pressure from 10 to 7
cm H
2
0. Decrease in true transmural filling pres-
sures of the ventricles decreases preload and car-
diac output (leftward and downward move along a
ventricular function curve).
5. Neural Reflex Cardiovascular
Depression During Unilateral Lung
Hyperinflation
1
*
In experiments in open-chested dogs, acute uni-
lateral pulmonary hyperinflation caused a decrease
in cardiac output and heart rate.
78
"
80
Since the ex-
perimental design precluded changes in ventricular
preload or afterload, these negative cardiovascular
effects were attributed to a neural reflex (Fig. 20-
2E). There are three known afferent receptors in
the lung, and all can be stimulated by mechani-
cally distending the lung. When stimulated, the
stretch receptors evoke a modest increase in blood
pressure and heart rate, the irritant receptors evoke
no known cardiovascular response, and C-fiber re-
ceptors evoke a marked bradycardia and hypoten-
sion. Thus, the observed cardiovascular effects of
unilateral lung hyperinflation in the dog experi-
ments, namely a marked fall in heart rate and
blood pressure and stroke volume, were likely the
result of stimulation of C-fiber receptors within the
lung itself.
6. Humoral Mediation of PEEP-induced
Decreases in Cardiac Output
79
*
5
The PEEP-induced decrease in cardiac output
may be caused by release of a cardiodepressant
humoral substance or metabolite (Fig. 20-2F). For
example, hyperinflation of in vitro and in vivo
lung can release a variety of biologically active
materials, such as the biogenic amines, polypep-
tides, proteolytic enzymes, and prostaglandins.
Most of these substances can cause depression of
cardiac function. A humoral mechanism of PEEP-
induced decrease in cardiac output is further sug-
gested by cross-circulation experiments in animals,
wherein 15 cm H
2
0 PEEP to one member of a
cross-circulated pair caused a decrease in the car-
diac output of the other member, and by in vitro
experiments that show that plasma from animals
subjected to PEEP contains a negative inotropic
agent, which is not present during zero end-expi-
ratory pressure (Fig. 20-2F).
D. Differential Lung Ventilation
In the vast majority of patients with bilateral
lung disease, the application of PEEP to both of
the lungs increases the FRC of both lungs, thereby
reversing atelectasis and increasing the ventilation-
perfusion ratio of both lungs (see Figs. 3-27, 3-
28, and 20-5). Similarly, in the vast majority of
patients with unilateral lung disease, the applica-
tion of PEEP to both of the lungs increases the
FRC of the diseased lung toward normal while not
excessively increasing the FRC of the normal lung
(Fig. 20-6). However, in some patients with uni-
lateral lung disease, PEEP to both of the lungs
may fail to expand the diseased lung and exces-
sively distend the normal lung. The distension of
the normal lung may selectively increase pulmo-
nary vascular resistance in the nondiseased lung
and shunt blood flow to the diseased lung, thereby
increasing the right-to-left shunt and causing a par-
adoxic decrease in the arterial oxygen tension
86
87
(Fig. 20-7). Similarly, in some patients with bilat-
CONVENTIONAL TWO LUNG PEEP
Figure 205 Application of positive end-expiratory pressure (PEEP) to both lungs through a single-lumen tube in patients with
bilateral lung disease usually results in a reversal of low ventilation to perfusion relationships and atelectasis in both lungs. (ATEL
= atelectasis; | V/Q = low ventilation-perfusion ratio.)
Figure 20- 6 Application of positive end-expiratory pressure (PEEP) to both lungs through a single-lumen tube to patients with
unilateral lung disease usually results in a reversal of low ventilation to perfusion relationships and atelectasis in the one diseased
lung. (ATEL = atelectasis; j V
A
/Q = low ventilation-perfusion ratio.)
eral lung disease, in which the disease is in the
basilar regions (with normal apical regions), PEEP
to both lungs may fail to expand the bases and
may excessively distend apical regions (Fig. 20-
8). The excessive distension of apical regions may
selectively increase pulmonary vascular resistance
in the apical regions and shunt blood flow to the
atelectatic and low ventilation-perfusion ratio bas-
ilar regions, thereby increasing the right-to-left
shunt and causing a paradoxic decrease in the ar-
terial oxygen tension.
88,89
In situations in which PEEP increases the right-
to-left shunt and decreases the arterial oxygen ten-
sion (unilateral or asymmetric lung disease and
bilateral basilar disease), differential lung ventila-
tion and PEEP can result in much more selective
application of PEEP to the diseased area. In the
first instance (unilateral lung disease), the applica-
tion of PEEP to the two lungs (via a double-lumen
tube) in an inverse proportion to their compliances
should result in equal and normal FRCs in both
lungs
90
(Fig. 20-9). In this way a high level of
PEEP is applied to only the diseased area in a
much more economic and efficient manner. Since
the PEEP is just being applied to part of the lung,
the effects on cardiac output are much decreased.
In the second instance (bilateral basilar disease),
reversal of atelectasis and improvement in venti-
SELECTIVE APPLICATION OF PEEP TO NORMOXIC LUNG
Figure 20- 7 Application of positive end-expiratory pressure (PEEP) to both normoxic and hypoxic regions may fail to expand
the hypoxic lung regions. This results in the selective application of PEEP to just the normoxic lung. The selective application of
PEEP to only the normoxic lung can compress the small intra-alveolar vessels in the normoxic lung and increase normoxic lung
pulmonary vascular resistance (PVR). The increase in normoxic lung PVR can cause diversion of blood flow to the hypoxic lung
and effectively decrease hypoxic pulmonary vasoconstriction (HPV) in the hypoxic lung.
Figure 208 The application of positive end-expiratory pressure (PEEP) to both lungs when both lungs contain diseased basilar
regions may result in distension of the apical regions, causing a selective increase in pulmonary vascular resistance in the apical
regions, which results in a diversion of blood flow from the apical regions to the basilar regions. This may occur 20 per cent of the
time that PEEP is applied to this kind of situation.
53 M
The distension of the apices results in an increase in functional residual
capacity (FRC), and the diversion of blood flow to the basilar regions results in an increase in shunt (Q
s
/Q,).
lation-perfusion ratio in both lungs can be accom-
plished by placing the patient in the lateral decu-
bitus position, ventilating each lung separately (via
a double-lumen tube), and applying PEEP to only
the dependent lung
9l9?
(Fig. 20-10). The nonde-
pendent lung will improve its function by experi-
encing a low pulmonary vascular pressure, which
results in a low blood flow and shunt in the non-
dependent lung and a propensity to resorb fluid
rather than transudate fluid in the nondependent
lung. The dependent lung will improve its function
by virtue of the application of selective PEEP to
that lung. Again, the selective application of PEEP
to only one lung minimizes effects on venous re-
turn and preserves the cardiac output.
93
High-frequency positive-pressure ventilation
(with a single-lumen tube) has been used success-
fully in two patients with unilateral lung disease;
the success was probably due to the provision of a
uniform distribution of ventilation at a low airway
Figure 20- 9 Differential lung positive end-expiratory pressure (PEEP) applies PEEP to each of the two lungs in an amount that
is inversely proportional to the compliance (C) of each of the lungs. Differential lung PEEP must be applied via a double-lumen
tube. In this example, the diseased lung has one third the compliance of the more normal lung and, therefore, receives three times
as much PEEP. (ATEL = atelectasis; 1 V
A
/Q = low ventilation-perfusion ratio.)
Figure 20-10 Schematic summarization of the results of the study in Hedenstierna et al.
92
Ventilation-perfusion ratios (V
A
/Q)
may be improved in patients with bilateral dependent basilar lung disease (viewed from either the upright or supine position) by
turning the patient into the lateral decubitus position and ventilating the patient with differential lung ventilation via a double-lumen
tube (DLT). The differential lung ventilation consists of zero end-expiratory pressure (ZEEP) to the nondependent lung and positive
end-expiratory pressure (PEEP) to the dependent lung. The nondependent lung will improve its ventilation to perfusion relationship
by virtue of being in a favorable nondependent position (see Fig. 20-12), whereas the dependent lung will improve its ventilation
to perfusion relationship by virtue of selectively receiving PEEP. This method avoids the problem of distension of the apical regions
causing increased apical pulmonary vascular resistance, resulting in blood flow diversion to basilar regions of the lung (see lower
left-hand corner inset).
pressure.
94
Finally, marked prolongation of the in-
spiratory phase (with a single-lumen tube) (inverse
ratio ventilation) has been used successfully in one
patient with unilateral lung disease.
95
The main criterion for initiation of differential
lung ventilation is when PEEP to all of the lung
causes an increase in shunt and a decrease in the
arterial oxygen tension.
90,9I
Measurement of FRC
cannot be used as a criterion because FRC may be
normal or increasing with the application of PEEP
owing to distention of only the normal areas of
lung (either the one lung or the apical regions),
even though there is continued failure to expand
the diseased areas of lung (either the other lung or
the basilar regions) (see Figs. 20-7 and 20-8). Of
course, the chest roentgenogram must show either
severe unilateral asymmetric lung disease or se-
vere bilateral basilar disease.
Three additional specific indications for initia-
tion of differential lung ventilation, which are in-
dependent of the PEEP-induced shunting issue, are
the presence of a bronchopleural fistula that pre-
vents positive-pressure ventilation because of loss
of tidal volume through the fistula, the presence of
massive refractory unilateral atelectasis,
96
and ex-
cessive air trapping (causing hypotension) in a re-
maining emphysematous lung after single-lung
transplantation.
97
There are several methods for demonstrating a
PEEP-induced increase in shunt with unilateral
lung disease. First, shunt can be measured before
and after the application of PEEP. Second, pul-
monary angiograms, which provide a qualitative
measure of regional lung blood flow, can be ob-
tained via a pulmonary artery catheter injection.
87
Third, lung perfusion scans, which provide a quan-
titative measure of regional blood flow, can be
performed with radioisotopes.
98
The specific management of differential lung
ventilation and PEEP for the patient with unilateral
lung disease follows.
99
Because it may be neces-
sary to determine the position of the double-lumen
tube with precision at any time (e.g., after any
significant patient movement), a fiberoptic bron-
choscope should be at the bedside. The risk of the
double-lumen tube becoming malpositioned will
be greatly minimized if coughing, bucking, and
head movement are eliminated by profound paral-
ysis. The most diseased lung should be placed in a
nondependent position. The compliance of each
lung should then be measured separately. The
amount of PEEP applied to each lung should ini
tially be inversely proportional to the compliance
of each lung. For example, if one lung has one
third the compliance of the other lung, then it
would be reasonable to initially apply 15 cm H
2
0
to the diseased lung and 5 cm H
2
0 to the nondis-
eased lung. The tidal volume to each lung should
be equal; equal tidal volumes will, of course, result
in different peak inspiratory pressures. The appli-
cation of PEEP inversely proportional to lung
compliance and use of equal tidal volumes have
most often resulted in equal FRCs and blood flows
to each lung. Poorer results have been obtained
with ventilator settings resulting in equal end-tidal
carbon dioxide concentrations or equal airway
pressures.
100
It was originally thought that the two lungs
would have to be ventilated synchronously be-
cause asynchronous differential lung ventilation
would cause mediastinal movements that would
impair cardiovascular performance. Synchronized
independent ventilation can be achieved utilizing
two MA-1 ventilators (Puritan-Bennett) synchro-
nized with an IMV controller (Healthdyne); the
output of the controller is attached directly to the
"manual" switch of each ventilator.
96
An impulse
directed from the IMV controller therefore sends a
signal simultaneously to both ventilators via the
manual circuitry, and synchronized breaths are de-
livered to the patient. Setting the cycle time on the
IMV controller thus results in a synchronized fir-
ing of the ventilators (e.g., 10 sec = respiratory
rate of 6; 5 sec = respiratory rate of 12). As a
precaution, a back-up respiratory rate lower than
the IMV controller is should be set on each of the
ventilators to ensure protection against possible
failure of the IMV controller.
However, experimental and clinical experience
has not supported the hypothesis that the two lungs
need to be ventilated synchronously.
101
102
In fact,
some studies suggested that asynchronous or alter-
nating differential lung ventilation actually im-
proves compliance of each lung because one lung
does not have to compete with the other during
inspiration for space within the thorax.
101
I03
When
the two lungs have achieved an equal compliance
and the chest roentgenogram has improved, the
double-lumen tube should be switched to a single-
lumen tube, and treatment should proceed to goals
2 and 3 (later).
The specific management of differential lung
ventilation and PEEP for the patient with bilateral
lung disease is similar to that for the patient with
unilateral lung disease, except for a few important
differences. First, the nondependent lung should
receive none or a relatively low level of PEEP,
and the dependent lung should receive a relatively
high level of PEEP (which is the opposite from
differential lung ventilation of the patient with uni-
lateral lung disease in the lateral decubitus posi-
tion). Second, the tidal volume to each lung should
be proportional to the assumed perfusion to each
lung.
91
Third, the patient should be turned from
side to side every 2 to 3 hours so that each lung
may be in a nondependent position for a time.
Fourth, the need for fiberoptic bronchoscopy to
check the position of the double-lumen tube is
increased because of the much more frequent pa-
tient movement.
IV. GOAL #2: F,Q
2
<0.5,
PEEP<10cm H
2
Q, ?|AND
P
a
0
2
ACCEPTABLE
A. Why This Goal Is #2: Match
Ventilation to Perfusion
If patients require more than 10 cm H
2
0 of
PEEP to have a P
a
0
2
greater than 60 mm Hg with
an F,0
2
of less than 0.5, then the ventilation-per-
fusion mismatch is so great that it is illogical to
make patients spontaneously ventilate with such
disordered lungs. In other words, lungs that have
F,0
2
and PEEP requirements greater than 10 cm
H
2
0 and 0.5, respectively, are inefficient to the
point that the work of breathing is likely to be
excessive, and the patients will fatigue and fail.
When the PEEP requirement is less than 10 cm
H
2
0, the F,0
2
is less than 0.5, and the P
a
O
:
is
adequate, the lungs are fulfilling their basic gas-
exchange function in a reasonably efficient man-
ner, and it now makes sense to ask the patients to
do some of the work of breathing on their own.
The second goal is to get the PEEP requirement
to be less than 10 cm H
2
0. Of course, it is assumed
that what was achieved before (F,0
2
<0. 5, accept-
able P
a
0
2
) is retained. This second goal is achieved
primarily by instituting an intensive and aggressive
respiratory regimen (Fig. 20-11 and Table 20-3),
which has four components to it. The first compo-
nent is removing secretions. This may be accom-
plished by coughing routines, tracheal/main-stem
bronchi suctioning, fiberoptic bronchoscopy for
areas of the lung that do not clear with tracheal/
main-stem bronchi suctioning, and chest physical
therapy (percussion and vibration). All of these
secretion-removal techniques can be aided by hav-
ing the patient assume appropriate postures. Fi-
nally, frequent turning will prevent any particular
Figure 2011 Goal #2 is to decrease the positive end-expiratory pressure (PEEP) level to less than 10 cm of water while
maintaining an F,0
2
of less than 0.5 so that the P
a
0
2
is adequate. This is primarily achieved by the application of an intensive and
aggressive respiratory care regimen. (F-O-B = fiberoptic bronchoscope.)
region of the lung from being dependent for a
majority of the time (see below).
Second, infection must be appropriately diag-
nosed and treated. This means that culture and
sensitivity studies must be based on material that
is appropriately peripheral and truly representative
of an infected area. Appropriately peripheral ma-
terial from an infected area may be obtained by a
protected brush specimen and/or bronchoalveolar
lavage with a fiberoptic bronchoscope. Unfortu-
nately, central sputum and tracheal aspirations are
often inadequate for accurate diagnosis.
Empiric antibiotic therapy may be started while
awaiting the results of culture and sensitivity stud-
ies. The patient with hospital-acquired pneumonia
can present in one of three settings. The first and
most common is the patient with known airway
colonization with gram-negative rods, typically a
ventilated patient, and where Staphylococcus au-
reus is not suspected. Empiric antibiotic coverage
is most appropriate with one of the third-genera-
tion cephalosporins: cefotaxime, ceftizoxime, cef-
triaxone, or ceftazidime.
104
The second setting is in the patient with known
airway colonization with Pseudomonas, Serratia,
or Enterobacter. Early empiric therapy should
probably involve two drugs in this situation. The
third-generation cephalosporin of choice would be
ceftazidime. In the patient with normal renal func-
tion, this should be combined with an aminogly-
coside.
104
The third setting is the suspected S. aureus or
anaerobic infection in the hospitalized patient. Ti-
carcillin-clavulanate or ampicillin-sulbactam are
reasonable first-line drugs until an etiologic diag-
nosis can be established.
104
Third, the airways should be dilated. The com-
monly used bronchodilators are p
2
-agonists, anti-
cholinergics, aminophylline (also increases dia-
phragmatic contractility; see later), and steroids.
The bronchodilator that worked best preopera-
tively will probably be the most efficacious drug
during the postoperative mechanical ventilation.
105
Fourth, there are several other general/systemic
maneuvers that may benefit the patient from a res-
piratory care point of view. Humidification of the
inspired gas will prevent mucous plugging. Incen-
tive spirometry will increase lung volume. Diuret-
ics and fluid restriction will decrease lung water
(furosemide may decrease shunt by diuresis, de-
Table 20-3 AGGRESSIVE AND INTENSIVE
RESPIRATORY CARE REGIMEN
I. Removing Secretions
1. Coughing routines
2. Tracheal suctioning
3. Fiberoptic bronchoscopy
4. Chest percussion and vibration
5. All of the above aided by posture
6. Turning frequently
II. Diagnosing and Treating Infections
1. Protected brush
specimen
2. Bronchoalveolar
lavage
3. Antibiotics according to culture and sensitivity
III. Dilating the Airways
1. Bronchodilators (B
2
-agonist, anticholinergics,
aminophylline)
2. Steroids
Accurate
culture and sensitivity
IV. Other General/Systemic Maneuvers
1. Humidification
2. Incentive spirometry
3. Diuretics and fluid restrictions
4. Inotropics
5. Aminophylline to increase diaphragmatic contractility
6. Inhalation of 60% helium
creasing central blood volume and pressures, as
well as direct ventilation/perfusion matching ef-
fect).
106
Inotropic drugs may decrease pulmonary
congestion
107
(see Fig. 19-11). Dopamine has the
advantage of increasing diaphragmatic strength by
increasing diaphragmatic blood flow; presumably
this will also be the case with other inotropic
drugs.
108
In this context, dobutamine also de-
creases pulmonary vascular resistance. Amino-
phylline, in addition to having a bronchodilating
effect, may increase diaphragmatic contractility.
Inhalation of 60 per cent helium will decrease air-
way resistance; Table 20^4 shows the relative in-
crease in flow rate that may be expected by various
concentrations of helium.
109
Although there are no hard data to prove that
any one of these respiratory care treatments is ben-
eficial when used alone, there is a widespread con-
sensus that, when used together and in the context
of the other respiratory care necessary in the post-
operative management of respiratory failure, they
produce a clearing of atelectasis, eradication of
infection, and a decrease in the PEEP require-
ments.
, , 0
, , ,
Other potentially useful treatments include inhi-
bition of cyclo-oxygenase pathway of metabolism
(with indomethacin, administration of almitrine) to
increase HPV,"
2
"
3
infusion of fibronectin (opso-
nizing protein in inflammation [e.g., binds bacte-
ria] and tissue repair [e.g., binds fibrin, collagen]),
decontamination of the gastrointestinal tract,
114
and
infusion of prostaglandin, (to decrease pulmonary
vascular resistance)."
5
"
6
The only way patients can remove secretions
from the tracheobronchial tree by themselves is by
coughing, and effective voluntary coughing is the
most efficient method of mobilizing and removing
secretions.
117
However, many patients cannot or
will not cough (due to depression by drugs, pain,
or disease), and the secretions must be mechani-
cally removed. Consequently, tracheal suctioning
is often necessary and certainly mandatory when-
ever secretions are heard in the tracheobronchial
tree by auscultation. The patients should be preox-
ygenated, and the maximum duration of suctioning
should be 10 sec (S
p
0
2
may decrease because of
apnea, decrease in lung volume, and/or decrease in
F,0
2
and/or causation of bronchospasm). The de-
crease in S
p
0
2
during suctioning may be greatly
minimized or eliminated by passing the suction
catheter through a self-sealing diaphragm in a
bronchoscopy elbow and continuing mechanical
ventilation or by insufflating oxygen at a very high
flow rate (e.g., 12 L/min).
118
Various curved- or angle-tip catheters that have
guide marks on them can improve the accuracy of
cannulation of the main-stem bronchi.
119-123
When
a curved-tip catheter is not available, one can be
easily constructed using a cigarette lighter; the
lighter can heat a straight catheter so it can be
Table 20-4 PHYSICAL PROPERTIES OF RESPIRATORY GASES*
Oxygen
Air
Helium
He-0
2
He-0
2
He-0
2
%/%
100
100
100
20/80
40/60
80/20
Density
1.429
1.293
0.179
1.178
0.678
0.429
VDensity
1.182
1.135
0.423
1.085
0.823
0.655
Viscosity
211.4
188.5
201.75
209.5
207.5
203.6
Kinematic
Viscosity
147.9
145.8
1127.1
177.8
306.1
474.6
Relative
Flow Ratef
0.96
1.00
2.68
1.04
1.38
2.06
*From Gluck EH, Onorate DJ. Castriolta R: Helium-oxygen mixtures in intubated patients with status asthmaticus and respiratory
acidosis. Chest 98:693-698, 1990. Used with permission.
tAir assigned arbitrary value of 1.00. Viscosity measured in micropoise.
molded and cured into a permanent curved posi-
tion (the length of the curved portion should be 3
cm).
124
The guide marks can be made using a felt-
tip pen.
122
To use the curved-tip suction catheter
properly, the tip of endotracheal tube should be 3
to 5 cm above the carina and the head in a mid-
position.
123
The curved-tip suction catheter can re-
sult in a success rate of 89 to 97 per cent for left
bronchial catheterization and 97 to 100 per cent
for right bronchial catheterization.
120-123
Even the
upper lobes may be selectively suctioned by using
a J-shape-tip catheter.
121
Thus, it is possible to
suction almost any part of the tracheobronchial
tree in a blind fashion.
Flexible fiberoptic bronchoscopy can be used
for visually directed endobronchial suctioning and
for obtaining reliable material for culture and
sensitivity.
124125
Certainly, flexible fiberoptic bron-
choscopy should be used whenever roentgenolog-
ically observable lobar obstruction
126
or unilateral
hyperinflation (check-valve mechanism permitting
gas flow only during inspiration when the airways
are larger)
127
is resistant to other standard medical
therapy.
126
A balloon-tip fiberoptic bronchoscope
permits application of positive pressure to specific
regions of the lung in addition to enabling suction
and lavage.
128
A small caliber balloon-tip catheter
has been passed under fluoroscopic guidance to
postoperative atelectatic areas and used to expand
these areas by oxygen injection.
129
Physically trying to shake the secretions loose
from their alveolar and bronchial attachments by
vibration and percussion, in conjunction with pos-
tural gravity drainage of the lungs, may increase
the removal of secretions. External pressure may
be applied to specific chest regions during exhala-
tion in order to increase the expiratory flow of gas
and may also increase the removal of secretions.
Chest physical therapy appears to be especially
beneficial for patients who are acutely ill, have
large volumes of central secretions, and cough
ineffectively.
130,131
Chest physical therapy is also
advantageous in patients with lobar atelectasis.
132
In one study, chest physical therapy of normally
recovering postlobectomy patients was as effective
in preventing the development of postlobectomy
atelectasis as immediate routine fiberoptic bron-
choscopy.
133
Chest physical therapy will not be
beneficial in patients who do not have airway se-
cretions or whose secretions are peripheral.
131
It
should be realized that chest physical therapy in
acutely ill patients may infrequently be associated
with bronchoconstriction and hypoxemia.
130,131
Dependent lung experiences higher pulmonary
vascular pressures, and nondependent lung experi-
ences lower pulmonary vascular pressures. The
level of pulmonary vascular pressure has two im-
portant consequences for diseased lungs (Fig. 20-
12). First, dependent lung has a propensity to tran-
sudate fluid, whereas nondependent lung has a
propensity to absorb fluid. Consequently, if dis-
eased lung is placed in a dependent position, it
may heal more slowly or actually become more
diseased. Conversely, if diseased lung is placed in
a nondependent position, it may heal more
quickly. Second, dependent lung has a higher
blood flow, and nondependent lung has a lower
flow. Consequently, if diseased lung is placed in a
dependent position, it will have a higher shunt
because it is mechanically forced to have a higher
blood flow.
134
Conversely, if diseased lung is
placed in a nondependent position, it will have a
lower shunt because it is mechanically forced to
have a lower blood flow.
134
Table 20-5 shows a
small series of patients with unilateral lung disease
demonstrating the typical arterial oxygenation ef-
fect of changes in the lateral decubitus position.
135
It is apparent from these considerations that dis-
eased lung should be in a nondependent position
for as much time as possible.
136-139
Frequent turn-
ing of the patient will allow all parts of the lung to
be placed in a nondependent position for their pro-
portional share of the time. Use of the Rotabed, a
motor-driven bed, greatly facilitates the turning of
patients by automatically turning the patients 180
degrees side to side every 3 min. This amount of
frequent turning has been shown to decrease the
incidence of atelectasis and pneumonia in a group
of ventilated patients compared with a control
group in conventional bed (p < .01) in two differ-
ent intensive care unit studies.
140, 141
Maintaining
the patient (particularly if obese) in a 45-degree
upright position during the time between lateral
decubitus positions minimizes the reduction of
FRC associated with posture and also reduces the
work of breathing (see goal #3).
Use of the respiratory care regimen outlined
previously usually allows the physician, within a
few days, to decrease PEEP to less than 10 cm
H
2
0 with continued maintenance of an adequate
P
a
0
2
with a low F,0
2
. PEEP should be removed in
the same way it was added (i.e., in 3- to 5-mm Hg
increments and titrated against S
p
0
2
and P
a
0
2
).
45
Once the PEEP requirement is less than 10 cm
H
2
0, no special attempt is made to reduce it fur-
ther. Maintenance of some PEEP is desirable in
terms of maintaining normal FRC and lung com-
pliance and providing a subjective sensation of
good lung expansion. The achievement of an
2
less than 0.5 and a PEEP less than 10 cm H
2
0
with an acceptable P
a
0
2
means that the blood and
gas within the lungs is matched well enough so
that the lungs perform their basic gas-exchange
function in a reasonably efficient manner. The pa-
tient is now a candidate for weaning from mechan-
ical ventilation.
POSITION TREATMENT OF DISEASED LUNG
A. Why This Goal Is #3: Conversion
From Intermittent Mandatory to
Spontaneous Ventilation
Goals 1 and 2 were concerned with avoiding
oxygen toxicity and matching ventilation to perfu-
sion. Once ventilation-perfusion matching is
achieved, it is reasonable to begin converting a
patient who, in theory, required full ventilatory
support (100 per cent dependent on the ventilator)
to one who requires no ventilatory support (no
dependence on the ventilator). The use of IMV
weaning allows a gradual transition from 100 to 0
per cent ventilator dependence. During the transi-
tion, IMV is providing a progressively decreasing
amount of partial ventilatory support.
12
Figure 20-12 The position of the patient can be used to treat lung disease. When diseased lung is placed in a nondependent
position, it experiences a low pulmonary artery pressure (PAP; (see chapter 3). The low pulmonary artery pressure means that blood
flow to the region will be passively decreased, which will cause a decrease in shunt (Q
s
/Q,) through the diseased lung. In addition,
the low pulmonary artery pressure causes the nondependent lung to have a propensity to resorb fluid and, therefore, heal more
quickly. Conversely, when diseased lung is placed in a dependent position, it experiences a high pulmonary artery pressure. The
high pulmonary artery pressure means that blood flow to the region will be passively increased, which will cause an increase in QJ
Q, through the diseased lung. In addition, the high pulmonary artery pressure causes a propensity to transudate fluid, resulting in
more slowly healing lung or even a progressive deterioration in the lung.
7 3 8 Mechanical Ventilation and Weaning
The third goal is to retain what was achieved
before ( FA less than 0.5, PEEP less than 10 cm
H
2
0, and acceptable P
a
0
2
) and to reduce the IMV
rate to less than 1 breath/min. This weaning proc-
ess is accomplished by a progressive decrease in
the IMV rate. The ability to remove a patient from
ventilatory support (i.e., decrease the IMV rate) is
determined by the balance between the patient's
respiratory muscle capacity and respiratory load
(i.e., when capacity significantly exceeds load,
success will be achieved). The rapidity by which
the IMV rate can be reduced initially is dictated
by and is directly proportional to the vital capacity
(VC) and peak inspiratory force (PIF) (Table 20-
6). Both of these breathing power indices can be
rapidly and easily measured at the bedside with a
spirometer (VC) and a dry-gas manometer (PIF).
The VC is the volume exhaled forcefully after a
maximal inspiration. The minimally adequate VC
is 10 to 15 ml/kg (normal range = 50 to 70 ml/
kg), and the minimally adequate PIF is - 20 to
- 25 cm H
2
0 (normal range = - 80 to - 100 cm
H
2
0) (Table 20-6). The VC maneuver requires
cooperation, adequate lung compliance, and mus-
cle strength. Because the VC requires patient co-
operation, it can give some indication of how
much cooperation with breathing treatments will
occur after extubation. When a patient cannot co-
operate for a true VC measurement, an approxi-
mation can be made by measuring the inspiratory
volume after several occluded inspiratory efforts.
The peak inspiratory force is the negative pres-
Table 20-6
Test
BEDSIDE PHYSIOLOGIC
PARAMETERS CONVENTIONALLY
AND COMMONLY USED TO
PREDICT SUCCESS IN WEANING
FROM VENTILATORY SUPPORT
AND WEAKNESS OF EACH TEST
Problem with the Test
1. Tidal volume >5 ml/kg
2. Vital capacity > 10-15
ml/kg
3. Respiratory rate <25-35
breaths/min
4. Minute ventilation >10
L/min
5. Maximum voluntary
ventilation >2 X minute
ventilation
6. Maximum inspiratory
pressure more negative
than 20-25 cm H
2
0
7. Respiratory rate/tidal
volume
1. Ignores determining
inspiratory reserve
2. Requires cooperation
3. Multifactorial
4. Multifactorial
5. Requires cooperation
6. Ignores respiratory
mechanics
7. Combination of above,
but the ratio defines rapid,
shallow breathing
sure generated at the airway when inspiration is
artifically obstructed. This is a good indication of
thoracic muscle strength. It may be more reliable
than the VC in the sense that it does not depend
on patient motivation and may even be performed
in the presence of coma. However, it provides little
indication of subsequent voluntary activity, and,
more importantly, the peak inspiratory force does
not take into account the lung and chest wall me-
chanics and therein lies its main weakness as a
predictor of weaning ability.
142
A poor VC (depen-
dent on cooperation, lung compliance, and muscle
strength) coupled with a good peak inspiratory
force (dependent on muscle strength) means that
the patient is either poorly cooperating and/or has
poor lung compliance.
Other simple indices of readiness for decreasing
the IMV rate are endogenous respiratory rate less
than 25 to 35 breaths/min, endogenous minute
ventilation greater than 10 L/min, maximal volun-
tary ventilation greater than two times the minute
ventilation, and tidal volume greater than 5 ml/kg
(see Table 20-6). With a respiratory rate greater
than 35 breaths/min, expiration usually become
active rather than passive, and the work of breath-
ing increases greatly.
It should be noted that in patients who have a
fresh incision and multiple indwelling catheters
(nasogastric, urinary, and endotracheal tubes, in-
travenous and arterial catheters), the pain input is
high, and spontaneous respiratory rates are often
between 20 and 25 breaths/min at the beginning
of, and during, weaning. A new index of weaning
ability quantitates rapid, shallow breathing as the
ratio of respiratory frequency to tidal volume and
appears to be a very promising predictor; the
threshold between successful and unsuccessful
weaning attempts was less than 105 breaths/
min/L.
143
The work of breathing, calculated as the differ-
ence in oxygen consumption between spontaneous
and mechanical ventilation and expressed as a per-
centage of oxygen consumption during mechanical
ventilation (percentage of change in oxygen con-
sumption), is another, but sophisticated, bedside
measurement for ability to wean.
144, 145
In one
study, the average percentage of change in oxygen
consumption (work of breathing) for eight normal
healthy volunteers was 3.7 1.8 per cent,
whereas it was 7.7 8.8 per cent for patients
recovering from ventilatory failure who were
weaned successfully from respiratory support
(weaning and extubation judgments were based on
the indices of weaning capability shown in Table
20-6). In patients who were not weaned success-
fully, the average percentage of change in oxygen
consumption measured was 24.7 12.3 per
cent.
144
In another study, the change in oxygen con-
sumption between mechanical and spontaneous
ventilation of less than 15 per cent had a sensitiv-
ity of 100 per cent and a specificity of 80 per cent
in predicting successful weaning.
145
Unfortunately,
no single indicator has been uniformly successful
from time to time in the same patient, from patient
to patient, or from study to study in determining
between those who can and cannot successfully
breathe on their own.
142
146
l47
Obviously, several indices together may have a
greater predictive power and the patient who meets
two or more of these criteria stands a better chance
of being separated from the ventilator than the
patient who meets only one criteria.
142
148
Overall,
one may expect a 4 per cent extubation failure rate
(reintubation rate) in a surgical intensive care
unit.
149
Before the IMV (or any type) weaning process
is actually begun, it is wise to pay attention to
several safety factors (Table 20-7). First, the
weaning trial should begin the morning after an
overnight rest period. Second, the tracheobronchial
tree should be suctioned and the patient allowed to
settle down. Third, the P
a
C0
2
level should be near
the patient's normal level to avoid sudden hypo-
ventilation at the beginning of the IMV trial.
Fourth, baseline values for respiratory and hemo-
dynamic parameters should be obtained. Fifth, the
Fj0
2
should be increased by 0.1 to avoid arterial
desaturation during the shallow breathing, which
is common in the initial stage of weaning. Sixth,
the patient should be sitting up or in the semi-
Fowler's position for better diaphragmatic action.
The IMV rate may be decreased in 1-breath/min
increments in patients who achieve only the mini-
mal values shown in Table 20-6. Values much
greater than these (e.g., more positive VC and
more negative PIF) mean that the IMV rate can be
decreased more rapidly, and values much below
these minimal ones (e.g., less positive VC and less
negative PIF) mean that the IMV rate must be
decreased much more slowly. When the decrease
in IMV rate is begun, the head of the bed should
be elevated as much as possible to reduce the
pressure of the abdominal contents on the dia-
Table 20-7 PREWEANING SAFETY
MANEUVERS
1. Begin in the morning after an overnight rest.
2. Suction tracheobronchial tree.
3. Have P
a
C0
2
near normal for the patient.
4. Obtain base line values for respiratory and hemodynamic
parameters.
5. Increase F,0
2
by 0.1.
6. Have patient sitting up as much as possible.
phragm and to allow the patient to reestablish con-
tact with the environment.
As the IMV rate is decreased, the patient is
monitored by spontaneous respiratory rate, S
p
O
:
,
PETC0
2
, VC, PIF, P
a
0
2
, and P
a
C0
2
. The use of the
noninvasive monitors, S
p
0
2
and PETC0
2
, will sig-
nificantly decrease the need for measurement of
P
a
0
2
and P
a
C0
2
.
150
151
When the rate of the IMV
withdrawal is appropriate and is being tolerated,
the patient's spontaneous respiratory rate remains
constant or only slightly increases, the VC and the
PIF remain constant or improve, and the arterial
blood-gas concentrations do not deteriorate. Mild
diaphoresis is not uncommon and should not cause
concern unless accompanied by other abnormali-
ties.
Conversely, when the IMV rate is too rapid and
is not being tolerated, the spontaneous respiratory
rate increases greatly (rapid, shallow breathing is
usually the first and most important indication of
intolerance);
152
the VC and PIF decrease; the S
p
O
:
,
PETC0
2
, and arterial gas values deteriorate; and
the patient becomes hypertensive and tachycardiac
and may have arrhythmias.
A highly stressed system is also indicated by
recession of the suprasternal notch, retraction in
the intercostal spaces, accessory muscle recruit-
ment, and abdominal paradoxical motion. If any of
these very negative findings occur, the patient
should be returned to mechanical ventilation. Most
clinicians provide periods of relatively high IMV
settings (10-14/min), especially at night (to pro-
vide some rest for the respiratory muscles [as is
necessary for all other muscles]) and limit the
weaning attempts to daylight hours.
153
The inability of a patient who has met the first
two goals to tolerate the IMV withdrawal is usu-
ally because either the total clinical status requires
full ventilatory support (e.g., the patient has very
high intrinsic PEEP and is hyperinflated, has he-
modynamic instability or heart failure [causing
pulmonary congestion, decreased compliance, and
poor blood flow to the diaphragm] is malnour-
ished, has hypophosphatemia or hypomagnesemia,
or is sleep deprived or the cumulative effects of
sedation have lasted longer than anticipated or the
patient has a neuromuscular problem) or the wean-
ing system that is being used results in an intoler-
able increase in the work of breathing.
Increased work of breathing through external
apparatus during spontaneous breathing is best in-
dicated by negative fluctuations in inspiratory
pressure greater than 2 cm H
2
0, which is usually
indicative of inadequate inspiratory flow. In this
context, aminophylline may increase diaphrag-
matic contractility and significantly contribute to a
successful wean. Patients who fail to wean or who
take a long time to wean should have each of these
problems addressed and should be started on an
intensive nutritional regimen.
154
Malnutrition may
lead to loss of muscle mass, decreased energy-rich
phosphates, and detrained muscles.
155
However,
large glucose feedings increase carbon dioxide
production (Respiratory Quotient > 1.0 [all car-
bohydrate], RQ = 0.7 [all fat]), and it is important
to substitute fat for glucose in terms of calories as
much as possible.
146
155
When P
a
0
2
is adequate, with the F,0
2
less than
0.5 (goal 1), the PEEP is less than 10 cm H
2
0
(goal 2), VC greater than 15 ml/kg, PIF greater
than - 20 cm H
2
0, IMV less than or equal to 1
breath/min, spontaneous respiratory less than 20 to
30/min, minute ventilation greater than 10 L/min,
maximum voluntary ventilation greater than two
times the minute ventilation, and P
a
C0
2
approxi-
mately 40 mm Hg (goal 3); when no other major
organ systems are in acute major failure or are
unstable; and when the chest roentgenogram find-
ings are reasonably equivalent to the premorbid
findings or are rapidly improving and no new
changes have appeared (such as infiltrates or pneu-
mothorax), the patient is ready for extubation. It
should be emphasized that, throughout this entire
mechanical ventilation and weaning process, anal-
gesics and sedatives must be administered and ti-
trated so that the patient is comfortable, is not
bucking on the endotracheal tube, and is without
active expiratory efforts (see Fig. 3-44).
There are several other methods of weaning (in
addition to decreasing the SIMV rate) (see section
VII. A. and B.). These consist of progressively
decreasing the level of pressure support, increasing
periods of time on a T-piece, and increasing the
trigger threshold of the assist-control mode. There
is no evidence that any one of the techniques of
partial support is inherently better than the others
or more effective at shortening the duration of
mechanical support.
5, 142
146
156
However, the T-
piece method is still very commonly used. With
the T-piece method, with or without CPAP, pa-
tients are periodically disconnected from the ven-
tilator and allowed to breathe spontaneously via a
T-piece adapter attached to a constant flow of hu-
midified, oxygen-enriched air. Adding CPAP
(5-10 cm H
2
0) will prevent decreases in FRC that
might occur with a patient off the ventilator but
with the glottis by-passed by the endotracheal
tube. The flow by mode of CPAP can greatly de-
crease the inspiratory work of breathing.
157
The frequency and duration of T-piece trials
vary according to the patient's tolerance. For pa-
tients with marginal respiratory reserve, two or
three trials per day *are recommended as a starting
point, with an initial duration of 5 to 30 min. The
initial goal of T-piece weaning is for the patient to
spend 6 consecutive hours on the T-piece. All of
the safety factors described for IMV weaning ap-
ply to T-piece weaning.
For the patient who has repeatedly failed wean-
ing attempts but is adequately nourished and is
able to tolerate at least 1 hour of spontaneous
breathing, respiratory muscle endurance may be
improved by using inspiratory muscle-resistive
training (IRT).
158
This is accomplished by alternat-
ing long periods of respiratory muscle rest on me-
chanical ventilation with brief periods of fatiguing
respiratory effort (5-30 min) provided by requir-
ing the patient to inspire against an adjustable in-
spiratory resistor. Two to three daily IRT sessions
are provided, and the length of time and the sever-
ity of the added load are increased as the respira-
tory muscle endurance improves, but never for
longer than 30 min. The method is based on the
principles of endurance training of skeletal muscle:
that acute fatigue and rest periods are required for
effective training but that chronic fatigue is harm-
ful.
Assuming appropriate respiratory care, most
thoracic surgical patients requiring extremely
lengthy "weaning" programs lack adequate re-
serves because of chronic lung or heart disease.
Often inordinate amounts of time, energy, and
emotions are spent on getting the patient off the
ventilator for several days and discharged from the
intensive care unit, only to have them return in
need of ventilatory support without an identifiable
acute insult (4 per cent incidence in a surgical
intensive care unit).
149
Although all patients de-
serve reasonable attempts to be free of the venti-
lator, it must be accepted that some patients will
need ventilatory support for a prolonged time.
Quality of life is best served by early recogni-
tion and acceptance of this circumstance and prep-
aration of the patient and family for tracheostomy
and long-term ventilator support. The purpose of
the tracheostomy is to facilitate the nursing care of
the patient and to enhance patient comfort and
ability to communicate and appearance to the fam-
ily. The timing of a tracheotomy should be driven
by these issues, and it should be performed when
the patient is medically stable.
VI. SUMMARY OF MECHANICAL
VENTILATION AND WEANING
PROCEDURE
The logic of this approach is that the patient's
lungs are first required to have the ventilation and
perfusion matched well enough, as indicated by an
adequate or reasonable P
a
0
2
, with an F,0
2
less than
0.5 and a PEEP level of less than 10 cm H
2
0, so
that the lungs function as an efficient gas-exchange
organ. Next, the patient's lungs are required to
sustain this efficient gas exchange without respi-
rator aid as indicated by an adequate or reasonable
VC, PIF, spontaneous respiratory rate, and P
a
C0
2
on a low IMV. Finally, the process requires that
no new complicating factor has occurred as indi-
cated by lack of new chest radiographic findings,
by resolution of previous pathologic chest radio-
graphic findings, and by the absence of any signif-
icant disorder or complication in any of the other
major organ systems.
VII. OTHER NEW MECHANICAL
VENTILATION AND WEANING
MODES
In the past few years, a number of new mechan-
ical ventilation modes have been introduced. The
new modes are an outcome of microprocessor
technology that enables the ventilator to monitor
breath-by-breath flow, pressure, volume, and res-
piratory mechanics. The ability to sense and trans-
duce these variables accurately, and on-line, al-
lows the interaction between patient and ventilator
to be much more sophisticated than ever before.
The search for new approaches to ventilatory sup-
port for ARDS patients has evolved, in part, from
a general frustration with our inability to reduce
the mortality from this syndrome. However, al-
though all modes of positive-pressure ventilation
work reasonably well, no one form of positive-
pressure ventilation has been shown to be superior
to other forms.
159
Nevertheless, there is a great
need to perform prospective, randomized studies
that compare a specific new mode of ventilation
with conventional therapy (rather than publishing
retrospective and feasibility studies).
160
I61
A. Pressure-Support Ventilation
PS ventilation is a ventilatory mode that pro-
vides additional inspiratory gas flow through a de-
mand valve in response to a spontaneous breath.
The ventilator continually monitors the airway
pressure and adjusts flow to maintain the same
pressure (up to 30 cm H
2
0 above the PEEP level)
throughout the "plateau phase." The pressure and
flow are discontinued when the patient's own in-
spiratory flow rate diminishes considerably, the
exact point being determined differently in differ-
ent ventilators (e.g., 25 per cent of peak flow [see
Fig. 20-13] or 5 L/min [see Table 20-8]).'
62
I63
In
other words, the patient's expiratory muscles pre-
vent more gas flow from entering the lungs, and
the expiratory phase is triggered. Thus, the patient
controls the duration of inspiration, the tidal vol-
ume, and the expiratory time.
Because "supported" tidal volumes may vary
according to the patient's inspiratory flow pattern,
which may vary breath to breath, the tidal volume
may vary breath to breath. If there is a leak around
the endotracheal tube cuff (greater than 25 per cent
of the peak flow or 5 L/min), the PS breath may
not terminate, and a constant high level of CPAP
will be applied.
164
PS usually reduces the work of
breathing
162
l63
because the patient has only to trig-
ger the demand valve, while part or all of the
"spontaneous" breath itself is being supplied by
the ventilator. However, minute ventilation may
be reduced or absent with PS ventilation if dy-
namic hyperinflation of the lungs with high intrin-
sic PEEP occurs. If intrinsic PEEP is equal to the
preset peak pressure during PS ventilation, the
ventilator will not ventilate the patient's lungs.
165
When high levels of PS are used, it resembles
the assist-control pressure-limited mode,
166
al-
Figure 2013 Effect of respiratory muscle
activity on breathing pattern during pressure
support ventilation. When the respiratory mus-
cles are active (pressure support assisted venti-
lation, right panel), inspiratory gas flow (V) is
maintained above the flow cycling threshold
(shaded area) for a longer inspiratory time
(T,) than when the respiratory muscles are inac-
tive (simple controlled ventilation, left panel).
Therefore, during pressure support ventilation,
the patient can regulate, through muscular ef-
fort, the duration of T gas flow, and inspired
volume (V). (T
E
= expiratory time.) (From
Hubmayr RD, Abel MD, Render K: Physiologic
approach to mechanical ventilation. Crit Care
Med 18:103-113, 1990.)
Table 20-8
Ventilator
Puritan-Bennett
7200
Bourns-Bear 5
Siemens-Servo
TERMINATING MECHANISMS OF
PRESSURE SUPPORT*
PS
Range
(cm H
2
0)
0-30
0-72
0-100
Hamilton-Veolar 0-50
Mechanisms of
Termination of PS
tV insp <5 LPM or PAWP
> Pps level by 1.5 cm
H
2
0
tV insp <25% V peak insp
or inspiratory time >5 sec
or inspiratory time >80%
of breath cycle
or inspiratory time >3 sec
tPS initiated on all ventilators when the patient inspires and
the drop in airway pressure exceeds the sensitivity setting.
*From Black C, Grover BS: A hazard of pressure support
ventilation. Chest 93:333-335, 1988. Used with permission.
Abbreviations: peak insp = peak inspiratory flow rate; V
insp = inspiratory flow rate; Pps = pressure support pressure;
PAWP = peak airway pressure; LPM = liters per minute.
though the inspiration is not terminated once the
preset pressure is achieved, as in true pressure-
limited devices. The preset variables for PS venti-
lation are the pressure trigger to initiate the breath
(pressure sensitivity) and the PS level. Some pa-
tients find PS to be a more comfortable ventilatory
mode because it reduces the work of breathing and
allows the patient to have more control over peak
inspiratory flow rate, flow waveform, inspiratory
time, and tidal volume.
167
PS may also be used as
a weaning mode by gradually decreasing the PS
level (i.e., from 30 cm H
2
0) as the patient's own
tidal volume improves.
Available data suggest that, in most patients, 30
cm H
2
0 PS provides full support (comparable to 8
to 10 breaths/min IMV/SIMV), and 15 to 20 cm
H
2
0 PS provides partial support (comparable to 4
to 6 breaths/min IMV/SIMV).
168
-
170
Low-level PS
ventilation (5-10 cm H
2
0) is intended to decrease
the inspiratory work of breathing necessary to
overcome mechanical ventilator/endotracheal tube
resistance and intrinsic PEEP; at this level of PS,
the success or failure of weaning is then governed
by the patient's respiratory mechanics rather than
the equipment attached to the airway (Fig. 20-
14).
m
However, the PS level that exactly elimi-
nates the extra work of breathing through equip-
ment apparatus differs substantially among the
patients.
172
The most reliable clinical guidelines for regulat-
ing the level of PS at all ventilation/weaning stages
appears to be the tidal volume-respiratory rate re-
lationship. That level of pressure support resulting
in a respiratory rate less than 20/min has been
shown to reflect significant unloading of the inspi-
ratory ventilatory work.
168-170
A normal ventila-
tory rate with pressure support has been shown to
correlate with tidal volumes of 10 to 12 ml/
, 168-170. 173
Although perfectly practical in most cases, PS
ventilation does not seem to offer any improve-
ment over existing full ventilation or weaning
techniques and does require a reliable endogenous
respiratory drive for ventilation.
5
B. Inverse-Ratio Ventilation
The difference between inverse-ratio mechani-
cal ventilation and conventional mechanical venti-
Figure 20-14 Net additional
work by mechanical respiratory sys-
tem (positive W
aw
) due to endotra-
cheal tube (ETT) and ventilator cir-
cuit with increasing pressure support
for 9-mm, 8-mm, and 7-mm endotra-
cheal tubes at respiratory rate of 20/
min and tidal volume of 0.5 L. (From
Fiastro JF, Habib MP, Quan SF:
Pressure support compensation for
inspiratory work due to endotracheal
tubes and demand continuous
positive airway pressure. Chest
93:499-505, 1988. Used with per-
mission.)
lation is the prolongation of inspiration such that
inspiratory time relative to total time (T,/T
T
) is
greater than 0.5 (up to 80 per cent of the respira-
tory cycle).
174
I75
Inflation of the lungs, therefore,
proceeds at a lower mean inspiratory flow. Like
conventional mechanical ventilation, inverse-ratio
mechanical ventilation may be either volume or
pressure preset, but it is usually pressure set.
161
With pressure-controlled IRV, the preset variables
are peak airway pressure, the I:E ratio, the respi-
ratory rate, and the applied (extrinsic) PEEP. Nei-
ther peak inspiratory flow rate nor flow waveform
are preset variables. A typical flow waveform is
shown in Figure 20-15.
With pressure-controlled IRV, tidal volume is
determined by the patient's respiratory system me-
chanics. At present, there are no established guide-
lines concerning the level of peak airway pressure
to be set when switching from controlled positive-
pressure ventilation to IRV. Two strategies that
have been used are, first, to set the initial peak
airway pressure on IRV equal to one half of that
observed with conventional positive-pressure ven-
tilation and, second, to adjust peak airway pressure
to produce a similar tidal volume as that obtained
during conventional positive-pressure ventila-
tion.
161
The lower prolonged gas flow rate produces a
lower peak airway pressure, but higher mean air-
way pressure, than does conventional mechanical
ventilation. The lower peak airway pressure may
reduce the risk of barotrauma.
176
177
Inverse I:E
ratio of up to 4:1 is thought to achieve better
oxygenation by a straightforward increase in mean
airway pressure for a long period of time, which
results in progressive alveolar recruitment and
stabilization and increased gas exchange. It has
been suggested that inverse-ratio mechanical venti-
lation be the preferred mode of mechanical venti-
lation for patients with ARDS and pulmonary
edema.
176
"
178
Although an I:E ratio of 2:1 is without signifi-
cant hemodynamic effect,
177
a distinct danger of
higher IRV (i.e., greater than 4:1) is hyperinflation
and a decreased cardiac output, especially in pa-
tients who have intrinsic PEEP and require long
exhalation times.
179
Another potential disadvantage
of inverse-ratio mechanical ventilation is the need
for heavy sedation or paralysis to allow the patient
to tolerate the long inspiration.
C. Airway Pressure-Release Ventilation
APRV is a form of pressure-preset ventilation
during which a constant positive pressure is ap-
Figure 20-15 Schematic illustrations of airway pressure (P
a
J, flow (V), and volume (V = inhaled volume) at I:E ratio of 2:1
with pressure-controlled inverse ratio ventilation. Note the decelerating flow and that inhalation begins before completion of
exhalation because of the short expiratory time. Volume is variable at a constant airway pressure (compliance changes). (Insp =
inspiration; exp = expiration; I = inspiratory time; and = expiratory time.) (From Sassoon CSH: Positive pressure ventilation:
Alternate modes. Chest 100:1421-1429, 1991. Used with permission.)
plied to the airway for approximately 80 per cent
of the respiratory cycle.
180
Thereafter, the lungs are
allowed (released) to intermittently deflate (exhale)
to ambient pressure passively. APRV is similar to
pressure-preset inverse-ratio mechanical ventila-
tion except that patients can take a spontaneous
breath during assisted pressure release ventilation
(Fig. 20-16). The cycle is terminated by time be-
cause opening of the release valve, which cycles
the device from inspiration to expiration, is deter-
mined by a timing mechanism.
The preset variables in APRV are the CPAP
level provided by the continuous high-flow sys-
tem, the level to which CPAP is reduced during
release, the frequency, and the duration of the air-
way pressure releases. The duration of the pressure
release is less than 1.5 sec. Appropriate minute
ventilation is adjusted by titrating the release rate
and level of CPAP, and when the patient is ready
for partial support, the release rate is decreased
until the patient is on CPAP only.
The clinical justifications for the use of APRV
are the same as those for inverse-ratio mechanical
ventilation.
181-183
Both APRV and inverse-ratio
mechanical ventilation may result in dynamic hy-
perinflation. The increase in lung volume probably
is primarily responsible for the improvement in
oxygenation.
175
With inverse-ratio mechanical ven-
tilation and APRV, the tidal volume is a function
of the impedance of the respiratory systems; that
is, if the respiratory system becomes stiff, hypo-
ventilation may occur. APRV has the advantage
over inverse-ratio mechanical ventilation in that
paralysis or heavy sedation is unnecessary. Back-
up conventional positive-pressure ventilation is
available (see later).
D. Mandatory Minute Volume and
Minimal Minute or Mandatory
Ventilation
MMV and minimal minute ventilation (MiMv)
are both methods that ensure the delivery of a
preset minute volume either by spontaneous
breathing or by assistance from the ventilator it-
self. Current microprocessor ventilators supply
MMV and MiMv by continuously comparing the
accumulating minute volume with the preset value
and supplementing any deficit by mandatory me-
chanical breaths. The difference between the two
is that inadequate spontaneous ventilation on an
MMV ventilator is supported by independent pos-
itive-pressure breaths, whereas during MiMv the
mechanical assistance is only in the form of pres-
sure support (i.e., the patient has to generate a
spontaneous breath in order to be assisted). There-
fore, if the minute ventilation decreases below a
predetermined level, the ventilator will either in-
crease the level of pressure support, or, in some
ventilators, the back-up option may be switched
automatically to volume-preset mechanical venti-
lation.
The potential benefit of the MiMv and MMV is
Airway Pressure Release
Figure 2016 Airway pressure re-
lease ventilation (APRV) airway
pressure curves: A, APRV without
spontaneous breathing efforts: B,
APRV with spontaneous breaths.
(Reprinted with permission from
Shapiro BA, Vender JS, Peruzzi WT:
Airway pressure for cardiac surgical
patients: State of the art and clinical
controversies. J Cardiothorac Vase
Anaesth 6:735-748, 1992.)
clear; the patient is guaranteed a minimal minute
ventilation regardless of her or his own sponta-
neous effort or drive, and the weaning process
becomes "automatic" (i.e., the patient passes
"from artificial to spontaneous breathing without
any attention to the controls being required").
184
There are, however, no data describing such "au-
tomatic weaning" in groups of patients, probably
because MiMv and MMV have two main disad-
vantages that stem from the characteristics of its
main parameter, namely, the minute volume. First,
it is difficult to estimate the right value of minute
volume because it varies greatly according to the
patient's dead space ventilation and carbon dioxide
level. If the preselected minute volume value is
too high, the patient will remain assisted by the
ventilator, although he or she may be ready for
extubation. A high preset minute volume value
may also render the patient apneic because of a
too low P
a
C0
2
level. If the preselected minute
volume value is too low, then the patient might be
hypoventilated, distressed, and hypercarbic. Sec-
ond, minute volume is composed of both rate and
tidal volume. Indeed, the patient may achieve the
preselected minute volume value by shallow hy-
perventilation, which is not clinically acceptable
and might signify impending respiratory catastro-
phe. Thus, it is difficult to see the practicality of
MiMv and MMV at this stage.
VIM. EXTUBATION OF THE
TRACHEA
Additional criteria for extubation are that the
patient should be alert, cooperative, and eager to
be extubated, and there should be an air leak
around the endotracheal tube when the cuff is de-
flated.
185
When there is an air leak around the
deflated cuff (and other standard criteria of extu-
bation are met), extubation has been successful.
185
When there is no air leak around the deflated cuff,
approximately 2 of 7 patients may require reintu-
bation.
185
The best timing for extubation is often in the
morning after an overnight rest. Before extubation
of the trachea, the pharynx, nose, nasogastric tube
(if present), and endotracheal tube should be suc-
tioned. Suctioning is followed by several positive-
pressure sighs with 100 per cent oxygen. Lido-
caine, 1 mg/kg, can significantly decrease the
incidence of coughing, bucking, and laryngospasm
after extubation of the trachea.
186
Because extuba-
tion can be a stimulating and possibly stressful
procedure, hemodynamic parameters may need to
be monitored closely and significant changes re-
sponded to in patients with coronary artery dis-
ease.
187
Extubation of the trachea should be accom-
plished by inflating the patient's lungs with 30 cm
H
2
0 pressure, holding the pressure constant while
the endotracheal tube cuff is rapidly deflated, and
then immediately pulling the endotracheal tube out
of the trachea (Fig. 20-17). This maneuver ensures
that the first event following extubation is a force-
ful exhalation from total lung capacity (due to the
elastic recoil of the chest wall); the forceful exha-
lation will blow out any material that has lodged
on top of the endotracheal tube cuff and/or vocal
cords that could not be reached by any of the other
suction avenues (Fig. 20-17).
The patient is then asked to deep breathe and is
observed for any laryngospasm or stridor. In a rare
patient, severe laryngospasm may occur, and it
must be treated immediately by the application of
CPAP by mask. If the laryngospasm is not re-
lieved, a small dose of succinylcholine must be
administered, with continued CPAP mask ventila-
tion efforts. If the patient vigorously tries to venti-
late spontaneously during complete laryngospasm,
pulmonary edema may ensue (which will, of
course, require additional treatment). Racemic epi-
nephrine may be able to reverse stridor resulting
from laryngeal edema if the patient's condition
permits the time to administer the drug.
188 189
If
stridor persists, and the patient otherwise appears
stable, consideration should be given to adminis-
tration of 80 per cent helium and 20 per cent
oxygen (see Fig. 20-18).
190
Humidified oxygen with an F,0
2
0.1 greater
than that used while the patient was intubated is
administered by green face mask. If the patient is
stable on a green face mask, the patient may be
switched to nasal prongs. Nasal prongs are most
comfortable and do not interfere with eating, but
they also do not provide an F,0
2
greater than 0.35
with a flow rate of 3 L/min.
191
When a higher F,0
2
is required (which is usually the case), a face mask
(or shield) should be used, which, with a flow rate
of 5 to 10 L/min, can yield an F,0
2
between 0.3
and 0.5. Face masks with a reservoir bag can ap-
proach an F,0
2
of 0.9. CPAP may also be admin-
istered by a tight-fitting face mask. If the clinician
is concerned with limiting the inspired oxygen
concentration to prevent depression of the patient's
hypoxic drive to breathe, then a Venturi mask is
the most logical choice. A series of Ventimasks
are stamped and color coded with their required
oxygen flow rate and nominal output concentration
(24, 28, 35, 40 per cent). Of course, the expected
F,0
2
of all devices is dependent on appropriate
positioning of the device on the patient's face.
Arterial blood-gas concentrations should be meas-
ured at 0.5 to 1 and 4 to 6 hours postextubation,
and a chest radiograph should be obtained 4 to 6
hours postextubation.
Extubation Procedure
Allow Pressure to
Build to 30 cm H
2
0
(Lung - TLC)
First Event
Post Extubation
is a Forceful Exhalation
Figure 20-17 The proper extubation procedure consists of allowing airway pressure to build to 30 cm of water so that the lung
approaches total lung capacity (TLC). Then, simultaneously, the endotracheal tube cuff is deflated and the endotracheal tube is
pulled out. This procedure forces the first postextubation event to be a forceful exhalation (cough) due to the forceful elastic recoil
of the lung from total lung capacity. The forceful exhalation can blow out of the airway any secretions and/or material that was
lodged between the endotracheal tube cuff and the vocal cords or on the vocal cords per se. In the author's experience, this greatly
diminishes the incidence of postextubation laryngospasm, breath holding, and choking-type respiration.
Bronchospasm may develop after extubation. If
so, racemic epinephrine, or isoproterenol, isoetha-
rine, metaproterenol, terbutaline, and salbutamol
may be administered topically to promote bron-
chodilatation (see chapter 13). Racemic epineph-
rine also may decrease mucosal edema secondary
to airway (especially laryngeal) trauma (which
may cause a reactive hyperemia and edema) and
often has a good effect in stridorous patients in the
immediate postextubation period.
Respiratory care measures instituted during the
mechanical ventilation and weaning process are
continued (e.g., drugs, respiratory therapy). In ad-
dition, incentive spirometry, which emphasizes a
long increase in inspiratory transpulmonary pres-
sure and lung volume (the ideal maneuver; Table
20-9), is the most cost-effective respiratory care
measure and should be begun every hour.
140
One
study has shown that, immediately after thoracot-
omy, incentive spirometry causes large increases
in tidal volume by recruiting both abdominal and
rib cage respiratory components, thereby assisting
in expansion of all lung regions.
192
On the third
postoperative day, however, there is a relative in-
crease in the use of the rib cage component during
incentive spirometry. This increase may be due
either to development of diaphragmatic dysfunc-
tion or, more likely, to an increased rib cage mo-
tion from a training effect induced by repeated
incentive spirometry encounters.
On the other hand, a commonly used expiratory
maneuver, namely, self-administered positive ex-
piratory pressure,
193
(see Table 20-9) has been
found to have no beneficial effect in post-thoracot-
omy patients in preventing atelectasis, as evaluated
by changes in blood-gas tensions and chest roent-
genograms.
194
However, if a patient is experienc-
ing significant postoperative pain, then it is unrea-
sonable to expect good cooperation in performing
the incentive spirometry deep-breathing maneuver;
consequently, pain management must be effective
(see chapter 21). The patient should be positioned
upright, should be encouraged to cough, and
should be encouraged to ambulate as soon and as
much as possible.
For patients with troublesome hypoxemia
Figure 20-18 Proposed therapeutic method
using helium-oxygen mixture in the treatment of
postextubation stridor. (From Fleming MD, Weigelt
JA, Brewer V, Mclntire D: Effect of helium and
oxygen on airflow in a narrowed airway. Arch Surg
127:958, 1992. Used with permission.)
(thought to be due to low lung volume), mask
CPAP may be very efficacious.
195
In patients who
have copious sputum production preoperatively
and in whom secretion retention may be expected
to develop postoperatively, a minitracheostomy
may be helpful. Minitracheotomy is a procedure
of percutaneous tracheal cannulation with a small-
bore tube (4 mm in diameter) (Mini Trach II Kit
Portex LTD) placed through the cricothyroid
membrane that provides constant tracheal access
for sputum suction. The minitracheotomy can be
placed prophylactically at the time of extubation
(the kit has a scalpel, introducer, and cannula)
196
or once sputum retention has begun.
197-199
The nasogastric, urinary, and extraperipheral in-
travenous and arterial catheters and pleural tubes
should be removed after 6 to 8 hours of trouble-
free extubation. It should be remembered that
many (approximately 60-70 per cent) of thoracic
surgery patients have apneas (tidal volume less
than 100 ml for greater than 15 sec), and some
(approximately 15 per cent) meet the criteria for
the sleep apnea syndrome (greater than five apnea/
hour) preoperatively; therefore, S
p
0
2
monitoring
Table 20-9 EFFECTS OF POSTOPERATIVE RESPIRATORY MANEUVERS
Maneuver Alveolar Inflating Pressure Time for Alveolar Inflation Inflating Volume
Ideal maneuver
Expiratory maneuvers
CO
:
hyperventilation
IPPB
Sustained maximal
inspiration
High (40-60 cm H
:
0preferably
negative intrathoracic pressure)
Minimal (mostly deflating)
Moderate
Preset (usually 15-20 cm H,0
positive airway pressure)
Maximal (30-50 cm H
2
0 negative
intrathoracic pressure)
Long (5-15 sec)
Minimal
Short (breathing deeper, but faster)
Long (5-10 sec)
Long (5-10 sec)
Largest possible (6-10 times
tidal volume)
Variable (not emphasized)
2-3 times tidal volume
Unknown (2-3 times tidal
volume)
Largest possible (2-6 times
tidal volume)
Abbreviation: IPPB = intermittent positive-pressure breathing.
may be warranted for the first few hours postextu-
bation days.
200
Common causes of late respiratory
failure postextubation (after a careful wean) that
are not predictable on an objective basis are swol-
len and immobile vocal cords leading to ineffec-
tive coughing, sputum retention and chronic aspi-
ration resulting from an incompetent larynx.
IX. COMPLICATIONS OF
MECHANICAL RESPIRATORY
SUPPORT
The complications of mechanical respiratory
support can be divided into those caused by tra-
cheal intubation
201
and those caused by mechanical
ventilation.
202
The incidence of most complications
related to an in situ endotracheal tube and mechan-
ical ventilation increase with duration (Table 20-
10).
203
Although the list of possible complications of
mechanical respiratory support (see Table 20-10)
appears formidable, most are typically infrequent
events, and, more importantly, most are avoidable.
Proper technique and adequate anesthesia should
make the actual insertion of the endotracheal tube
uncomplicated. Aside from injury to the laryngeal
aperture (edema, granuloma, ulceration), which
has a high incidence of occurrence with intubation
greater than 4 days,
204
providing adequate seda-
tion, using the minimum volume of air in the en-
dotracheal tube cuff to effect sealing, frequently
suctioning, and turning the patient will render the
presence of endotracheal tube benign.
The same considerations apply to prolonged
double-lumen tube intubation.
205
However, it is not
possible to know about or eliminate pressure on
the tissues in all areas (upper airway, trachea, as
well as larynx) of endotracheal tube contact (see
prior discussion).
204
The volume of air in the en-
dotracheal tube cuff should be checked for a min-
imum ("just") seal volume periodically (e.g., after
large inspired oxygen concentration changes) and
only after the mouth has been suctioned. The en-
dotracheal tube should be suctioned whenever
breath sounds indicate the presence of tracheal or
bronchial secretions.
Because the incidence of high airway pressures
correlates with the incidence of barotrauma,
206
-
207
most physicians assume that a cause and effect
relationship exists. The significance of this obser-
vation is difficult to ascertain because various
studies do not differentiate the absolute effects of
the peak inspiratory pressure from the underlying
pulmonary disease responsible for the high peak
inspiratory pressure.
The point of greatest resistance and least com-
pliance is the endotracheal tube. Therefore, the
Table 20-10 COMPLICATIONS OF
MECHANICAL RESPIRATORY
SUPPORT
I. Complications of tracheal intubation
201
A. During intubation
1. Trauma to the eyes, cervical spinal column, nose,
teeth, lips, tongue, larynx, and trachea
2. Aspiration (of blood, tooth, laryngoscope bulb,
gastric contents, and partial dentures)
3. Esophageal intubation
4. Endobronchial intubation
5. Reflexes
a. Sympathetic (hypertension, tachycardia,
myocardial ischemia)
b. Vagal (hypotension, bradycardia, cardiac arrest)
B. With endotracheal tube in place
1. Bronchospasm
2. Tracheomalacia
3. Tracheal perforation (by cuff, by tip of the tube),
fistula formation (to any intrathoracic organ)
4. Inadequate secretion removal (possible obstruction
of the tube)
5. Paranasal sinusitis (nasal tube)
6. Trauma to all areas of contact (exacerbated by
patient movement)
7. Excessive inflation of cuff causing tube lumen
narrowing
8. Ruptured cuff
9. Accidental extubation
C. During extubation
1. Trauma to glottis by persistent inflated cuff
2. Aspiration of supraendotracheal tube cuff
secretions
3. Laryngospasm
4. Respiratory obstruction by immediate edema of
any part of the airway previously in contact with
the endotracheal tube
D. After extubation
1. Sore throat, dysphagia
2. Aphonia
3. Paralysis of the hypoglossal and/or lingual nerve
4. Vocal cord paralysis
5. Ulceration, inflammation, infection, edema,
stenosis of any part of the airway previously in
contact with the endotracheal tube
6. Laryngeal granulomas and polyps
7. Laryngotracheal membranes and webs
8. Tracheal stenosis
9. Sinusitis
II. Complications of mechanical ventilation
202
A. Complications attributable to operation of the ventilator
1. Machine failure or disconnection
2. Alarm failure
3. Alarms not turned on
4. Inadequate humidification
5. Excessive inspired oxygen concentration
B. Medical complications of positive-pressure breathing
1. Inadequate positive-pressure breathing and alveolar
hypoventilation
2. Excessive positive-pressure breathing and alveolar
hyperventilation
3. Massive gastric distension, stress ulceration
4. Pneumothorax, pneumomediastinum
5. Air embolism
6. Atelectasis
7. Pneumonia
8. Hypotension and decreased cardiac output
9. Increased ventilatory dead space
10. Antidiuretic hormone secretion, decreased renal
perfusion, and water retention
Bronchovascuiar
Sheath
Arteriole
Venule
Alveolus
Bronchiole
Figure 2019 Cross-sectional diagram of bronchovascuiar bundle surrounded by alveoli in the normal state (left) and after
alveolar rupture (arrows, right) resulting from increased pressure gradient between air space and perivascular interstitium. (Repro-
duced with permission from Maunder RJ, Pierson DJ, Hudson LD: Subcutaneous and mediastinal emphysema. Arch Intern Med
144:1447-1453, 1984.)
peak inspiratory pressure occurs in the endotra-
cheal tube and must be significantly greater than
alveolar pressures, which are mainly determined
by alveolar gas volume and compliance and are
probably the main determinant of barotrauma. Ma-
nipulation of flow rates will significantly affect
peak inspiratory pressure, whereas totally unre-
lated (and even disparate) pressure changes may
occur in the alveoli. It appears that the high peak
inspiratory pressure coincident with barotrauma
reflects the severity of lung disease and has some,
but not a direct, bearing on the incidence of baro-
trauma. Although the mean airway pressure meas-
urement is not a measurement of the mean intra-
thoracic pressure, it is more useful as a reflection
of the alveolar pressure than is peak inspiratory
pressure.
All forms of ventilator-induced barotrauma de-
velop after rupture of an overdistended alveolus
(Figs. 20-19 and 20-20).
208
-
210
When the alveolus
ruptures, air is introduced into the perivascular ad-
ventitia, which results in interstitial emphysema.
Pulmonary interstitial emphysema was identified
in 88 per cent of 15 adult patients with ARDS
evaluated retrospectively.
209
Pneumothorax never preceded the radiographic
development of pulmonary interstitial emphysema.
Gas may then dissect along perivascular sheaths
Figure 20-20 Pulmonary baro-
trauma: potential gas routes. A and
B, Gas in the perivascular adventitia
results in pulmonary interstitial em-
physema (PIE). C, Gas in the me-
diastinum results in pneumomedias-
tinum. D, Gas travels along cervical
fascial planes to produce subcuta-
neous emphysema. E, Gas may es-
cape retroperitoneally and burst into
the peritoneum to cause pneumoper-
itoneum. F, Rupture of the medias-
tinal parietal pleura causes pneumo-
thorax. (Reproduced with permission
from Maunder RJ, Pierson DJ, Hud-
son LD: Subcutaneous and medias-
tinal emphysema. Arch Intern Med
144:1447_1453, 1984.)
into the mediastinum to produce a pneumomedias-
tinum (see Fig. 20-20).
209
21
Accumulated me-
diastinal gas may decompress along cervical fas-
cial planes into the subcutaneous tissue to produce
subcutaneous emphysema.
Gas may also escape retroperitoneally into the
abdomen and eventually burst into the peritoneal
cavity to cause pneumoperitoneum. If mediastinal
pressure rises abruptly or decompression via other
routes is not sufficient to relieve the tension, the
mediastinal parietal pleura may rupture, resulting
in pneumothorax. Ventilator-associated pneumo-
thorax also occurs with rupture of subpleural air
cysts.
There appears to be a significant relationship
between the degree of renal dysfunction and de-
creased intravascular volume because volume re-
placement returns renal function to base line be-
fore cardiac function returns to base line.
211
The
direct renal effects of positive airway pressure can
be minimized with prior intravascular volume ex-
pansion or the administration of low-dose dopa-
mine.
212
A combination of furosemide and dopa-
mine has been shown to have synergistic effects in
patients with renal insufficiency requiring positive
airway pressure therapy.
213
REFERENCES
1. Shapiro A, Cane RD, Harrison RA: Positive end-expi-
ratory pressure therapy in adults with special reference to
acute lung injury: A review of the literature and suggested
clinical correlations. Crit Care Med 12:127-141, 1984.
2. Benumof JL: Update on mechanical ventilation. An epit-
ome. West J Med 128:147, 1978.
3. Weisman IM, Rinaldo JE, Rogers RM, et al: State of the
art: Intermittent mandatory ventilation. Am Rev Respir
Dis 127:641, 1983.
4. Downs JB, Klein EF Jr, Desautels D, et al: Intermittent
mandatory ventilation: A new approach to weaning pa-
tients from mechanical ventilators. Chest 64:331-335,
1973.
5. Shapiro A, Vender JS, Peruzzi WT: Airway pressure for
cardiac surgical patients: State of the art and clinical
controversies. J Cardiothorac Vase Anaesth 6:735-748,
1992.
6. Mathru M, Rao TL, El-Etr AA, et al: Hemodynamic re-
sponse to changes in ventilatory patterns in patients with
normal and poor left ventricular reserve. Crit Care Med
10:423, 1982.
7. Downs JB, Douglas ME, Sanfelippo PM, et al: Ventila-
tory pattern, intrapleural pressure, and cardiac output.
Anesth Analg 56:88, 1977.
8. Shah DM, Newell JAC, Dutton RE, et al: Continuous
positive airway pressure versus positive end-expiratory
pressure in the respiratory distress syndrome. J Thorac
Cardiovasc Surg 75:557, 1977.
9. Venus B, Jacobs HK, Mathru M: Hemodynamic re-
sponses to different modes of mechanical ventilation in
dogs with normaj and acid aspirated lungs. Crit Care Med
8:620, 1980.
10. Permutt S: Circulatory effects of weaning from mechani-
cal ventilation: The importance of transdiaphragmatic
pressure. Anesthesiology 69:157-160, 1988.
11. Shapiro BA, Cane RD: The IMV-AMV controversy: A
plea for clarification and redirection (editorial). Crit Care
Med 12:472^173, 1984.
12. Cane RD, Shapiro B: Mechanical ventilatory support.
JAMA 254:87-92, 1985.
13. Pontoppidan H, Geffin B, Lowenstein E: Acute respira-
tory failure in the adult. Engl J Med 287:143, 690, 799,
1972.
14. Permutt S: Mechanical influences on water accumulation
in the lungs. In Fishman AP, Renkin EM (eds): Pulmo-
nary Edema. Clinical Physiology Series. Bethesda, MD,
American Physiological Society, 1979, pp 175-193.
15. Albert RK, Lakshminarayan S, Kirk W, Butler J: Lung
inflation can cause pulmonary edema in zone I of in situ
dog lungs. J Appl Physiol 49:815-819, 1980.
16. Albert RK, Lakshminarayan S, Huang TW, Buttler J:
Fluids leaks from extra-alveolar vessels in living dog
lungs. J Appl Physiol 44:759-762, 1978.
17. Bowton DL, Kong DL: High tidal volume ventilation
produces increased lung water in oleic acid-injured rabbit
lungs. Crit Care Med 17:908-911, 1989.
18. Dreyfuss D, Soler P, Basset G, Saumon G: High inflation
pressure pulmonary edema: Respective effects of high
airway pressure, high tidal volume, and positive end-
expiratory pressure. Am Rev Respir Dis 137:1159-1164,
1988.
19. Parker JC, Hernandez LA, Peevy KJ: Mechanisms of
ventilator-induced lung injury. Crit Care Med 21 :131-
143, 1993.
20. Tuxen DV, Lane S: The effects of ventilatory pattern on
hyperinflation, airway pressure, and circulation in me-
chanical ventilation of patients with severe air-flow ob-
struction. Am Rev Respir Dis 136:872-879, 1987.
21. Lang RM, Fellner SK, Neumann A, Bushinsky DA, Bo-
row KM: Left ventricular contractility varies directly with
blood ionized calcium. Ann Intern Med 108:524-529,
1988.
22. Patterson RW: Effect of P
a
C0
2
on 0
2
consumption during
cardiopulmonary bypass in man. Anesth Analg 5:269-
273, 1976.
23. Cohen PJ, Alexander SC, Smith TC, et al: Effects of
hypoxia and normocarbia on cerebral blood flow and
metabolism in conscious man. J Appl Physiol 23:183,
1967.
24. Hedley-Whyte J, Pontoppidan H, Morris MJ: The re-
sponse of patients with respiratory failure and cardiopul-
monary disease to different levels of constant volume
ventilation. J Clin Invest 45:1543, 1966.
25. Winter PM, Smith G: The toxicity of oxygen. Anesthe-
siology 37:210, 1972.
26. Lambertsen CJ: Effects of oxygen at high partial pressure.
In Fenn WO, Rahn (eds): Handbook of Physiology,
section 3. Respiration. Vol 2. Baltimore, Williams & Wil-
kinsCo, 1965, pp 1027-1046.
27. Register SD, Downs JB, Stock MC, Kirby RR: Is 50%
oxygen harmful? Crit Care Med 15:598-601, 1987.
28. Montgomery AB, Luce JM, Murray JF: Retrosternal pain
is an early indicator of oxygen toxicity. Am Rev Respir
Dis 139:1548-1550, 1989.
29. Nash G, Blennerhasset JB, Pontoppidan H: Pulmonary
lesions associated with oxygen therapy and artificial ven-
tilation. Engl J Med 276:368, 1967.
30. Pare PD, Warriner B, Baile M, et al: Redistribution of
pulmonary extravascular water with positive end-expira-
tory pressure in canine pulmonary edema. Am Rev Respir
Dis 127:590, 1983.
31. Demling RH, Staub NC, Edmonds LH: Effect of end-
expiratory airway pressure on accumulation of extravas-
cular lung water. J Appl Physiol 38:907-912, 1975.
32. De Campo T, Civetta JM: The effect of short-term dis-
continuation of high-level PEEP in patients with acute
respiratory failure. Crit Care Med 7:47-49, 1979.
CHAPTER 21
Management of Postoperative
Pain
1.
II.
III.
IV.
V.
Introduction
Anatomy and Physiology of Post-
thoracotomy Pain
Systemic Narcotic Administration
Intercostal and Paravertebral Nerve
Block
Interpleural Analgesia
A. Mechanism and Type of Neural
Blockade
B. Technique
VI.
VII.
VIM.
IX.
1. With the Chest Open (at the
End of Thoracotomy)
2. With the Chest Closed
C. Clinical Experience
Epidural Local Anesthetic
Administration
Cryoanalgesia
Transcutaneous Electric Nerve
Stimulation
Epidural Opioids
756
\. INTRODUCTION
The treatment of pain following thoracotomy is
important not only to ensure patient comfort but
also to minimize pulmonary complications by en-
abling patients to breathe normally (without active
exhalation and/or splinting) and deeply (so that
they can cough) and to ambulate. Normal and deep
breathing require stretching of the skin incision,
which is painful. Postoperative patients will nor-
mally try to prevent stretching on the skin incision
by reflexly contracting their expiratory muscles
(splinting), which limits the stretch on the incision
during inspiration, and by actively exhaling, which
rapidly diminishes whatever stretch there is on the
incision. Failing to cough avoids the deep inspira-
tion prior to the forceful exhalation. Splinting, ac-
tive exhalation, and failing to cough promote
retention of secretions, airway closure, and atelec-
tasis.
There are many techniques of providing post-
thoracotomy analgesia (Table 21-1), and all can
provide satisfactory pain relief as well as minimize
pulmonary dysfunction.
1
The systemic administra-
tion of narcotics can achieve the goals of relieving
pain, diminishing splinting, and promoting cough-
ing, but the therapeutic-respiratory depression
window is small. The risk-benefit ratio in admin-
istering narcotics has been greatly decreased by
the advent of patient-controlled analgesia (PCA).
Intercostal nerve blocks and thoracic epidural
analgesia with local anesthetics are higher risk but
still efficacious procedures; they can be an alter-
native to, or used in conjunction with, conven-
tional narcotic treatment. Intrapleural local anes-
thesia is a relatively new but variably successful
technique. Epidural narcotics are in widespread
use and are, therefore, discussed in greatest length
in this chapter.
Cryoanalgesia is a relatively new, safe, effec-
tive, and long-lasting (3 to 4 weeks) pain-manage-
ment technique. Transcutaneous electric nerve
stimulation may have beneficial effects when used
as an adjunct to one of these techniques.
Table 21-1 METHODS OF PROVIDING
POSTOPERATIVE ANALGESIA
Local
Opioids Anesthetics Mechanical
Intramuscular Peripheral block Cryoanalgesia
Intravenous Intercostal TENS
PCA Epidural
Peridural Intrapleural
Abbreviations: PCA = patient-controlled analgesia; TENS
= transcutaneous electric nerve stimulation.
H. ANATOMY AND PHYSIOLOGY OF
POST-THORACOTOMY PA\N
The pain of thoracotomy has been described as
near the top of many iatrogenic causes (i.e.,
"severe").
2
3
The pain is generated from several
sources, including soft tissue and muscle injury
and inflammation, bone and joint trauma, and vis-
ceral damage. A posterolateral thoracotomy usu-
ally involves seven dermatomes. The pain is
exacerbated by movement, especially by the oblig-
atory movement of ventilation.
Most of the nociceptive output generated by
postoperative thoracotomy pain is conducted cen-
trally to the sensorium via three well-recognized
routes: stimuli from the chest wall and most of the
pleura pass along the intercostal nerves; those
from the diaphragmatic pleura use the phrenic
nerve; and those from lung and mediastinum are
carried by the vagus nerve. The role of sympa-
thetic nerves in the conduction of visceral nocicep-
tion is ill defined, although they are recognized as
routes for cardiac-generated pain. Occasionally, a
characteristic neuritic pain from damaged intercos-
tal nerves is discernible.
Clearly, blockade of the intercostal nerves is
relatively easy by many approaches (e.g., intercos-
tal nerve, paravertebral, epidural, cryoanalgesia,
intrapleural block), whereas blockade of the vital
mixed (sensory and motor) nerves (phrenic and
vagus) is difficult and undesirable (loss of dia-
phragm and coughing function, respectively). A
consequence of effective intercostal nerve block is
that it may bring to the fore some of the other
constituents that make up the post-thoracotomy
pain complex. For instance, patients may become
aware of the shoulder tip pain caused by diaphrag-
matic irritation or the deep-seated discomfort
caused by mediastinal irritation and chest drain
pain.
4
These "revealed" pains may sometimes re-
quire other techniques of analgesia.
The biochemical/neural components of the
stress response include elevated levels of catechol-
amines; sympathetic nervous system activation;
hypercoagulability; metabolic shifts to a catabolic.
protein-wasting state; and immunosuppression.
5
Although this stress response may be beneficial
and self-protective in some situations, it is often
counterproductive. For example, the stimulation of
the sympathetic nervous system that accompanies
severe pain leads to tachycardia, hypertension, and
increased peripheral vascular resistance, which in-
creases myocardial oxygen consumption and can
promote ischemia or infarction.
6
Increased sympa-
thetic outflow also decreases intestinal motility and
may promote or prolong postoperative ileus.
Pulmonary function is the function most ob-
viously depressed/compromised by surgically in-
758 Management of Postoperative Pain
duced pain. In patients undergoing abdominal or
thoracic operations, postoperative pain creates
acute restrictive pulmonary disease. The restriction
is caused by pain and is expressed by failure to
cough and splinting. This respiratory abnormality
presents as an increased respiratory rate and de-
creased tidal volumes, vital capacities, forced ex-
piratory volumes, and functional residual capaci-
ties (FRCs) (Fig. 21-1 ).
7
If FRC decreases below
the closing volume, atelectasis and low ventila-
tion-perfusion units and hypoxemia result.
III. SYSTEMIC NARCOTIC
ADMINISTRATION
Systemic narcotic administration has tradition-
ally been the most frequently utilized treatment for
post-thoracotomy pain. Appropriate narcotic use
achieves a balance between adequate patient com-
fort (which is usually indicated by an absence of
active exhalation and splinting) and avoidance of
excessive sedation or respiratory depression. Since
active exhalation and/or splinting and unwilling-
ness to cough can be observed, small doses of
narcotic (e.g., 2 to 4 mg of morphine intrave-
nously) can be administered to postoperative tho-
racotomy patients until it is evident that these
unfavorable respiratory characteristics are mini-
mized. Avoidance of administration of narcotics
beyond these physiologic end points is important
because narcotics can inhibit coughing, can de-
crease the frequency of sighs, and tend to make
breathing regular, all of which promote alveolar
collapse and the development of atelectasis.
8
In addition, the opiates cause a marked desensi-
tization of the central respiratory apparatus to the
stimulating effects of hypercarbia and hypoxia; in-
deed, an increase of 20 per cent in the arterial
carbon dioxide tension may be expected in patients
adequately dosed with narcotic analgesics.
9,10
The
therapeutic range of systemic intermittent analge-
sics, when used as the sole method of pain relief
and as defined previously, is small and requires
considerable physical attention (i.e., titration of
drug to the desired effect).
The amount of drug required for pain relief is
variable from patient to patient;
10
the variation has
been estimated to be four- to sixfold." In patients
undergoing abdominal surgery, a significant in-
verse correlation was found between the concen-
tration of endorphins in the cerebrospinal fluid pre-
operatively and cerebrospinal fluid pethidine
concentration achieved by PCA postoperatively.
12
In other words, patients who had high cerebrospi-
nal fluid levels of endorphins (presumably giving
the patient natural analgesia) had low pethidine
requirements, and patients who had low cerebro-
spinal fluid levels of endorphins (presumably ren-
dering the patient without natural analgesia) had
high pethidine requirements.
12
This strongly sug-
gests that drug dosage requirement differences be-
tween patients are largely determined by pre-exist-
ing and, perhaps, genetic neurochemical factors.
Subsequent differences may be due to the variable
presence of noxious stimuli unrelated to the inci-
sion, such as nasogastric tubes and Foley catheters.
If adjunctive therapy is added, such as rectal
indomethacin,
13
14
intramuscular ketorolac,
15
or in-
tramuscular diclofenac,
16
narcotic requirements
can be greatly reduced and are, therefore, much
less variable from patient to patient. Other factors
that might be responsible for different patient-to-
patient drug dosage requirements are differences
in pain-assessment methodology, patient popula-
tion (such as age, sex, prior drug exposure), pla-
Figure 21-1 Lung volumes and capacities in the perioperative period and the relative magnitude of their change from preopera-
tive to postoperative measurements. (FRC = functional residual capacity; CC = closing capacity; VC = vital capacity; TV =
tidal volume; IRV = inspiratory reserve volume; IC = inspiratory capacity; ERV = expiratory reserve volume; RV = residual
volume.) (Reprinted with permission from Brown DL, Neal JM: Chronic obstructive pulmonary disease and perioperative analgesia.
Probl Anesth 2:422^34, 1988.)
Figure 21- 2 Relationship among
dose interval, analgesic drug concen-
tration, and clinical effects in a com-
parison of patient-controlled analge-
sia (PCA) system and conventional
intramuscular (i.m.) therapy. (From
Lutz LJ, Lamer TJ: Management of
postoperative pain: Review of cur-
rent techniques and methods. Mayo
Clin Proc 65:584-596, 1990. Used
with permission.)
cebo effects, and subcutaneous (inadvertent) ver-
sus intramuscular injection.
The amount of drug required for pain relief
throughout an individual's hospital course de-
creases in an exponential manner.
10
This is because
pain from the wound decays in an exponential
manner. Thus, with intermittent intramuscular ad-
ministration, mean narcotic doses typically reduce
by one half every 1 to 2 days.
17
However, the usual
interdosing interval of 3 to 4 hours gives a fast
decay rate for pain relief, which is a poor match to
the much slower decay rate of wound pain inten-
sity, and periods of pain are likely to result. In
fact, when a parenterally administered analgesic is
given intramuscularly every 3 to 4 hours, the drug
concentration meets or exceeds the minimal anal-
gesic concentration for just 35 per cent of the
interval (Fig. 21-2).
6
Peak concentrations vary as much as three to
five times the minimal analgesic concentration,
and the time needed to attain peak blood levels
may also vary considerably.
6
PCA (on demand)
and constant infusion systems can eliminate pain
resulting from this discrepancy in the two decay
rates (i.e., pain from the operation and from the
peaks and troughs of the intermittent large bolus
of administered drug) (see Fig. 21-2).
PCA allows the patient, within certain limits, to
determine how frequently a small bolus dose of
analgesic drug should be administered. The limits
consist of the actual bolus dose and a minimum
dosing interval (a waiting period, referred to as the
lockout time), which are set by the physician in
charge of the treatment. The purpose of the control
waiting or lockout period is to prevent administra-
tion of a second dose until after the first dose has
exerted its maximal analgesic effect. Table 21-2
shows the range of bolus doses and lockout inter-
vals for commonly used narcotic agonists and ag-
onist-antagonists.
18
Thus, the patient can fine-tune
the level of analgesia at any particular moment.
Extensive experience with PCA has shown that
previously nonaddicted patients will select a criti-
cal drug level that ensures complete alertness, even
Table 21-2 GUIDELINES REGARDING THE
BOLUS DOSAGES AND LOCKOUT
INTERVALS FOR VARIOUS
PARENTERAL ANALGESICS
WHEN USING A PCA SYSTEM*
Drug
Agonists
Fentanyl citrate
Hydromorphone hydrochloride
Meperidine hydrochloride
Methadone hydrochloride
Morphine sulfate
Oxymorphone hydrochloride
Sufentanil citrate
Agonist-Antagonists
Buprenorphine hydrochloride
Nalbuphine hydrochloride
Pentazocine hydrochloride
Bolus
Dose
(mg)
0.02- l.O
0.1-0.5
5-30
0.5-3.0
0.5-3.0
0.2-0.8
0.003-0.015
0.03-0.2
1-5
5-30
Lockout
Interval
(min)
3-10
5-15
5-15
10-20
5-20
5-15
3-10
10-20
5-15
5-15
*From White PF: Use of patient-controlled analgesia for
management of acute pain. JAMA 259:243-247, 1988. Used
with permission.
Abbreviation: PCA = patient-controlled analgesia.
if it is accompanied by a certain amount of resid-
ual pain.
19
This amount of pain is quite well toler-
ated by the patient because it is the patient's
choice that it should not be completely relieved,
and he or she is assured that additional narcotic is
available at a moment's notice if desired. PCA
also confers psychological benefit. Being able to
control at least one important factor in their treat-
ment calms and considerably relieves many pa-
tients.
Continuous narcotic infusion will also provide a
constant analgesic level. By titrating the hourly
dose to the patient's needs, a steady plasma con-
centration of narcotic may be achieved that will
provide adequate analgesia in a safe manner for a
long period of time to a large majority of postop-
erative patients. However, in practice, several
problems remain. First, there are a number of
strong contraindications to the use of PCA (Table
21-3). Second, initial adequate analgesia can be
obtained in about 80 per cent of patients, but in
approximately 20 per cent it may be insufficient or
excessive.
20
Thus, the problem of selecting the ap-
propriate dose the first time still remains. Third,
the patients still need to be monitored for signs of
overdosage, and the respiratory rate must be
checked frequently. However, once the appropriate
hourly dose of narcotic can be determined, the
patients do quite well.
IV. INTERCOSTAL AND
PARAVERTEBRAL NERVE BLOCK
Intercostal and paravertebral nerve blocks with
long-acting anesthetic agents (e.g., bupivacaine)
are sometimes used as a means of treating post-
thoracotomy pain.
21-31
Paravertebral block pro-
duces similar effects and may be thought of as an
extremely posterior intercostal nerve block that
can spread cephalad and caudad a few ipsilateral
dermatomes as well as slightly to the contralateral
side. Indeed, injection into the intercostal space
can easily spread to the paravertebral space.
32
Table 21-3 CONTRAINDICATIONS FOR
PCA USE
1. Very young children
2. Impaired sensorium/mental status
3. History of chronic obstructive pulmonary disease
4. Severe metabolic disorders, such as sepsis or severe fluid
and electrolyte abnormalities
5. Psychological disorders
6. History of narcotic drug abuse
7. Allergy to morphine or prescribed medication
8. Unexpected postoperative course in the ICU
Abbreviations: PCA = patient-controlled analgesia; ICU =
intensive care unit.
A number of studies showed that intercostal and
paravertebral nerve blocks result in decreased
postoperative pain and narcotic requirements,
21-23

27
33
improvements in arterial blood gases
23
24
26
and lung function,
21
"
23
27,28
33
and earlier discharge
from the hospital.
22_24,33
These blocks have been
used successfully in young patients (6-month-old
infants) as well as in older children and adults.
22
These blocks may be placed either intraopera-
tively (at the time of closure) under direct vision
or postoperatively. Direct-vision intraoperative
blocks may be done by needle or by placement of
catheters in the appropriate intercostal grooves for
injection postoperatively when pain occurs.
29-31,34
Obviously, indwelling intercostal groove catheters
can provide longer term and usually more accurate
blocks than one-time direct-vision needle block or
intermittent percutaneous needle blocks.
28 35
One very successful technique of placing a sin-
gle-site continuous intercostal nerve block catheter
consists of peeling the parietal pleura back from
the incision margin to a point posterior to the in-
cision margin, placing the multihole catheter extra-
pleurally (between the parietal pleura and the in-
tercostal spaces/ribs), and reattaching the parietal
pleural to the margin of the incision.
35
x
This com-
plication-free technique results in a 93 per cent
incidence of patients requiring no additional anal-
gesia in the first 24 hours after thoracotomy and
an 82 per cent incidence of no requirement for
other analgesics in the subsequent 4 days.
35
The
site of action of local anesthetic infused in this
manner, as shown by the spread of radiocontrast
material, is primarily in the paravertebral space.
37
Percutaneous insertion of continuous intercos-
tal
38
and paravertebral
39
nerve block catheters has
been described as well. Postoperative percutaneous
intermittent intercostal nerve blockade may be eas-
ily achieved by contacting the appropriate rib sur-
face with a needle and advancing it below the
inferior margin of the rib. Long-acting local anes-
thetics, such as bupivacaine with epinephrine,
should be used for postoperative intercostal and
paravertebral blocks; bupivacaine has an objective
duration of 12 hours and a subjective duration
(analgesia persisting beyond return of sensation'
as long as 36 hours.
6
Intermittent paravertebral nerve block is per-
formed by inserting a 7.5-cm 22-gauge needle 3
cm lateral to the thoracic dorsal spine, two or threi
interspaces caudad to the most cephalad segmen
to be blocked, in a manner that is perpendicular tc
all planes and advanced until the transverse proc
ess or rib is encountered. The needle is thei
"walked off" the obstruction and advanced 1 en
further and the injection made (20-30 ml of O.i
per cent bupivacaine over 3 min).
The administration of intercostal nerve blocks
however, is not without an incidence of complica-
tions. Several reports documented profound hypo-
tension shortly following the intraoperative place-
ment of intercostal nerve blocks.
40-43
In two of
these reports the fall in blood pressure was thought
to be due to paravertebral spread of the local an-
esthetic agent with resultant sympathetic block-
ade.
40
4I
The other two reports documented total
spinal block that was probably due to unrecog-
nized durai puncture or perineural spread of the
local anesthetic.
41
"*
3
Although this complication
can be avoided by administering the intercostal
block at least 8 cm lateral to the intervertebral
foramen,
43 44
some authors advocated discontinu-
ing altogether the placement of intercostal nerve
blocks intraoperatively .
45
Potential problems with this technique, when
performed postoperatively, include the hazards of
pneumothorax,
24
rapid intravascular absorption of
the local anesthetic, discomfort on injection, and
the considerable time and skilled personnel re-
quired to perform intermittent blocks. Many anes-
thesiologists are unwilling to commit themselves
to patient visits every 6 to 8 hours to perform
intermittent single-injection blocks. Although use
of indwelling intercostal groove catheters de-
creases physician time and increases the safety of
the procedure, the catheters tend to migrate, and
the incidence of failed blocks increases.
V. INTERPLEURAL ANALGESIA
A. Mechanism and Type of Neural
Blockade
Interpleural analgesia has been successfully
used for pain relief in children and adults after
cholecystectomy, renal surgery, breast surgery,
46
~
50
and thoracotomy (for references, see later discus-
sion). As would be expected with any effective
form of analgesia, postoperative pulmonary func-
tion is increased and narcotic requirements are de-
creased with this technique.
48
49
51
On interpleural injection, most of the fluid col-
lects at the lowest point of the pleural cavity irre-
spective of the position of the body (e.g., in the
lateral decubitus position the fluid collects against
the mediastinum, in the Trendelenburg position the
fluid collects against the apices of the lung, and in
the supine position the fluid collects in the poste-
rior paravertebral gutter).
52
The mechanism of an-
algesia produced by this technique is most likely
due to diffusion of local anesthetic solution from
the pleural space through the parietal pleura and
the innermost intercostal muscle to reach the inter-
costal space, where blockade of the intercostal
nerves occurs.
53
The blockade of the intercostal nerves can be
demonstrated by decreased amplitude and in-
creased latency on evoked responses.
54
Therefore,
this technique should be analogous to the blockade
of multiple intercostal nerves unilaterally. How-
ever, differences appear to exist between the con-
ventional form of intercostal nerve block and the
interpleural technique. Intercostal nerve blocks are
capable of providing a surgical level of anesthesia,
whereas the interpleural technique provides anal-
gesia but not anesthesia sufficient for the perform-
ance of surgery.
53
Typical peak venous plasma levels of approxi-
mately 2 g/ml bupivacaine occur after inter-
pleural administration of 150 mg, which is well
below toxic levels.
53
Etidocaine
55
and lidocaine
56
may also be safely used for interpleural analgesia.
The increase in local anesthetic concentration in
plasma is minimized by the addition of epineph-
rine (1:200,000)
57
58
and by the fact that inter-
pleural local anesthetics are significantly cleared
by the lung lymph
59
(as well as lost out the chest
tube, if present; see later discussion). Interpleural
morphine does not provide superior analgesia or
improve pulmonary function compared with sys-
temic morphine.
60
If the patient is placed in the Trendelenburg
position, allowing better soaking of the upper tho-
racic (interpleural) cavity, a unilateral upper tho-
racic sympathectomy and Horner's syndrome may
result.
61
It is likely that mid- and lower thoracic
sympathetic denervation occurs routinely with in-
terpleural injection of local anesthetics. This latter
contention is well supported by sophisticated ex-
perimental studies.
62
Nevertheless, hemodynamic
effects of interpleural analgesia are not clinically
significant.
63
No spinal, epidural, or cerebral neural
blockade can be demonstrated with interpleural an-
algesia.
54
B. Technique
1. With the Chest Open (at the End of
Thoracotomy)
At the end of the surgical procedure and before
the closure of the thorax, an epidural-type catheter
with a Luer lock is placed in the pleural space
(paravertebral gutter) by the surgeon under direct
vision. The catheter is introduced percutaneously
through the intercostal space adjacent to the inci-
sion site by a Tuohy needle. The tip of the catheter
is placed posteriorly in the pleural space (paraver-
tebral gutter), and the catheter is secured with one
parietal pleura and one or two skin sutures to pre-
vent migration of the catheter. Use of two strate-
gically placed catheters (posteromedial and lateral)
may result in better analgesia than one catheter.
64
In the vast majority of reports, the patients re-
ceived 20 ml of 0.5 per cent bupivacaine with
1:200,000 epinephrine in the recovery room as
initial pain therapy. Subsequent doses of 15 to 20
ml of 0.375 to 0.5 per cent bupivacaine (0.3 ml/kg
to a maximum of 20 ml) are given up to four times
a day. The chest tube should be clamped for 5 to
15 min after each administration of bupivacaine. If
the chest tube is not clamped, approximately 30 to
40 per cent of any administered dose of local an-
esthetic will be lost via the thoracostomy tube.
64
The patient should be kept in a supine position
during, and for 10 to 15 min after, the injection.
2. With the Chest Closed
In the closed technique, the patient is placed in
the lateral decubitus position. A skin nick is made
with an 18-gauge needle at the appropriate inter-
costal space (fourth-eighth) in the midaxillary
line. An 18 to 16-gauge Tuohy needle should be
passed through the skin, and the stylet should be
removed. A well-lubricated (with saline), easy-
gliding plunger, 10-ml glass epidural syringe con-
taining 4 to 5 ml of air should be attached. The
plunger should not be contaminated with glove
powder, which could lead to sticking of the
plunger. The needle should be walked off the su-
perior border of the rib in a slight cephalad direc-
tion just so it comes off the superior edge of the
rib. The smooth bevel should be directed toward
the parietal pleura, so that the catheter is directed
parallel to the chest wall and not directly into the
lung. The shaft of the needle should be directed
parallel to the posterior path of the rib toward the
spine.
Consequently, the needle enters the pleural
space at a very shallow angle and not vertically.
As the tip of the needle enters the pleural space,
the plunger will be pulled down by the negative
intrapleural pressure. It is a dramatic movement
that happens by itself and therefore it is not a loss-
of-resistance technique. In the spontaneously
breathing patient, pleural puncture should be made
end inspiration, whereas pleural puncture in the
mechanically ventilated patient should be made
during an apneic end-expiratory period.
On successful identification of the pleural space,
the epidural catheter is inserted 6 to 10 cm. The
epidural needle is removed while controlling the
catheter. A loop of catheter should be covered with
sterile plastic dressing. The catheter should be as-
pirated and then injected with the test dose. Bilat-
eral catheters may be inserted for patients who
have undergone major abdominal or thoracoab-
dominal procedures.
65
Unfortunately, some authors have difficulty
identifying the pleural space in this manner (i.e.,
depend on spontaneous movement of the plunger
of a syringe against an air column).
66-68
Indeed,
blind percutaneous insertion of interpleural cathe-
ters in this manner may be associated with a sig-
nificant incidence of pneumothorax as well as in-
traparenchymal placement.
66
Consequently, blind
insertion may be particularly hazardous if followed
by positive-pressure ventilation. Inward movement
of a liquid column, whether as a hanging drop
67
or
as a vertical column of fluid,
68
has been used to
identify the pleural space. A 100 per cent success
rate has been obtained with both moving-liquid-
column techniques.
67
68
In addition, the epidural catheter has been in-
serted into the pleural space via a Tuohy needle
introducer inserted into a chest tube and the cath-
eter advanced until its tip protrudes about 1 cm
through the distal end of the chest tube (Fig. 21-
3).
69
Ideally, the assembly should be positioned
posteriorly within the pleural space and the chest
tube must be clamped during and after injection
for 5 to 15 min. Finally, a clamped chest tube has
been used alone for administration of pleural space
local anesthetics.
56
C. Clinical Experience
Most of the references cited previously
46-69
have
had good results with interpleural analgesia (good
pain relief, increased pulmonary function, and/or
decreased narcotic requirements). However, sev-
eral reports have been only mildly positive
70
or
frankly negative.
71-74
The most outstanding reason
for failure to achieve adequate analgesia after in-
terpleural administration of local anesthetics has
been failure to clamp the chest tube before admin-
istration of the local anesthetic (i.e., the local an-
esthetic was suctioned out the pleural space into
the chest tube).
58,64
70
71
Indeed, when one group
71
changed technique to include clamping of the
chest tube,
58
negative results were converted to
positive results.
Other reasons for failure of the technique are
posterior placement of local anesthetic, which
prevents adequate analgesia for anterior thoracot-
omy
57
; dilution of local anesthetic by pleural effu-
sion/blood/infected fluid; loculation of local anes-
thetic by adhesions/fibrosis/infection; and loss of
local anesthetic through a bronchopleural fistula.
One review comprising 703 cases detailed the
complications of interpleural analgesics.
75
Pneu-
mothorax was the most frequently registered com-
plication (2.0 per cent) followed by signs of sys-
temic toxicity (1.2 per cent; in one patient seizures
were thought to be due to rapid uptake resulting
from the presence of a highly inflamed pleura
76
)
and pleural effusion (0.42 per cent). Horner's syn-
Management of Postoperative Pain 7 6 3
Figure 21-3 The assembly is placed directly during thoracotomy or percutaneously using standard tube thoracostomy techniques.
The tip of the catheter is positioned posteriorly and at the apex. The catheter is firmly attached to the external portion of the chest
tube with tape. (From Baker JW, Tribble CG: Pleural anesthetics given through an epidural catheter secured inside a chest tube.
Ann Thorac Surg 51:138-139, 1991. Used with permission.)
drome, pleural infections, and catheter rupture
have also been reported.
75
VI. EPIDURAL LOCAL ANESTHETIC
ADMINISTRATION
Thoracic epidural anesthesia (analgesia) essen-
tially results in multiple intercostal nerve blocks
after a single injection of a local anesthetic. It is
not surprising, therefore, that this technique has
the same beneficial effects on arterial blood-gas
values, results of pulmonary function tests, and
narcotic requirements as conventional intercostal
nerve blocks do.
77- 80
Potential major complications
of this technique include significant hypotension
due to sympathetic blockade, cardiovascular and/
or central nervous system toxicity due to intravas-
cular injection of local anesthetic agents, headache
due to inadvertent durai puncture (along with total
spinal block if unrecognized), infection related to
need for chronic catheterization, and trauma to the
spinal cord. In addition, epidural analgesia with
local anesthetic requires considerable physician
time and skill.
The continuous infusion of a dilute concentra-
tion of local anesthetic of 0.25 per cent bupiva-
caine in an epidural catheter has been investigated
in an effort to prolong the block, minimize hypo-
tension, and simplify treatment, but this technique
did not provide adequate analgesia and/or was as-
sociated with a high incidence of major complica-
tions.
8183
In addition, the addition of 0.1 per cent
bupivacaine to fentanyl, 10 g/ml, for epidural
infusion post-thoracotomy does not change the
quality of analgesia (visual analog score range 15-
35 mm), the infusion rate (7-9 ml/hour), serum
fentanyl concentration (1-2 ng/ml), postoperative
pulmonary or bowel function, or incidence of side
effects compared to a control group that did not
receive bupivacaine.
84
In contrast, the addition of
0.2 per cent bupivacaine (twice as much) resulted
in a very slight increase in pain relief and P
a
0
2
but
only for the first post-thoracotomy day.
85
Conse-
quently, and in view of the efficacy of cryoanal-
gesia and lumbar epidural narcotics alone (see fol-
lowing discussion), the risks of epidural local
anesthetic administration outweigh the benefits,
and the technique is seldom used today.
VII. CRYOANALGESIA
Extremely long-lasting intercostal nerve block
may be obtained by intercostal nerve freezing
(cryoanalgesia).
86-93
Direct application of an ice
ball to the nerve causes degeneration of nerve ax-
ons without damage to the support structure of the
nerve (the neurolemma), thereby reversibly dis-
rupting nerve activity. Thus, intraneural and peri-
neural connective tissues are preserved, providing
a scaffolding for regenerating capillaries, axons,
and Schwann cells.
94
The area of anesthesia is
along the dermatomes treated. During the next 2
to 3 weeks after nerve freezing, nerve structure
and function begin to recover in parallel. Usually,
from 1 to 3 months after freezing, there is full
restoration of nerve structure and function, without
untoward sequelae (neuritis or neuroma forma-
tion). The numbness that persists during this time
period is not especially bothersome. Young
women, however, have admitted to some distress
during the period of axonal regeneration due to
loss of sensation in the nipple area if the fifth and
higher intercostal nerves have been frozen.
93
Currently used cryoprobes are about the size of
a pen; the core of the instrument has an exit port
that permits rapid expansion of a gas (nitrous ox-
ide). The rapid expansion of nitrous oxide cools a
surrounding metal sheath. Since the metal sheath
is in contact with a fluid, an ice ball (temperature
= - 60C) forms at the tip of the cryoprobe. The
cryoprobe is applied to the intercostal nerves as
far posteriorly as possible from within the chest at
the level of the incision, plus two or three inter-
spaces above and below this level, just prior to
closure of the thoracotomy wound. The nerves are
approached from within the chest since the probe
has to pierce the parietal pleura to reach the inter-
costal nerves in the subcostal groove. This is easily
accomplished by the surgeon's lifting the nerve
out of the groove with a small nerve hook. The
cryoprobe is placed directly on the nerve and acti-
vated so that the center of the resultant ice ball
encapsulates the nerve. During the freeze, nerve
tissue temperature is approximately 20C.
93
If 2
to 3 mm of tissue remain between the tip of the
cryoprobe and the nerve to be blocked, the nerve
will not become adequately frozen and nerve func-
tion might not be totally eliminated. If the pleura
is thickened, local peeling of the pleura is advisa-
ble (Fig. 21-4).
93
Usually two 30-sec freeze cy-
cles, which are separated by a 5-sec thaw period,
are applied to each of the nerves selected. How-
ever, experience with one 30-sec freeze exposure
resulted in no loss of postoperative pain control
with a significantly reduced period of numbness
(from 3.0 months to 1.2 months).
93
The use of cryoanalgesia in this way has effec-
tively reduced postoperative narcotic requirements
and has improved pulmonary function.
86
88_91
Cryoanalgesia patients have had slight to moderate
pain at the end of the first postoperative day, and
only slight pain after that (as compared with pa-
tients receiving intravenous narcotics who experi-
ence moderate to severe pain at 24 hours). Most
of the pain experienced by the cryoanalgesia pa-
tients has not been incisional discomfort but rather
shoulder or arm pain secondary to chest tube irri-
tation of the pleura.
Further advantage of the cryoanalgesia can be
gained if the chest tubes can be placed within and
emerge from an intercostal space whose nerve has
been subjected to cryoanalgesia.
93
In fact, if the
chest tube is placed and emerges from a non-
blocked dermatome, then the patient will experi-
ence discomfort,
95
which, because of the isolated
origin of this pain, will require significant medi-
cation.
96
Once the chest tubes are removed, usually
on the second or third postoperative day, the pain
in the cryoanalgesia patients has become barely
noticeable.
97
98
Consequently, these patients are
more able to cooperate with the postoperative pul-
monary physiotherapy maneuvers described in
chapter 20.
Return of sensation occurs in most patients by
postoperative day 30. Long-term follow-up for 6
months has shown that the procedure is associated
Figure 21- 4 Positioning of cryoprobe. (From Maiwand MO, Makey AR, Rees A: Cryoanalgesia after thoracotomy. Improvemen
of technique and review of 600 cases. J Thorac Cardiovasc Surg 92:291-295, 1986. Used with permission.)
with a certain worrisome incidence of dysesthesias
and intercostal muscle paralysis,
96
but it is impos-
sible to distinguish the cause of these problems
between injury to the nerves and muscle during
surgery (which does occur; e.g., 44 per cent of
patients in one long-term follow-up study had per-
sistent post-thoracotomy pain" from injury as a
result of the cryoprobe).
95
Since cryoanalgesia causes temporary nerve
damage, and the duration of the nerve damage far
exceeds the duration of significant postoperative
pain (even with a single 30-sec freeze, numbness
persisted for a mean of 38 days),
93
cryoanalgesia
cannot be considered a routine post-thoracotomy
pain treatment. Cryoanalgesia may be a treatment
of choice in thoracic pain situations that are ex-
pected to last a long time (e.g., pain from chest
trauma) and significantly limit respiratory func-
tion.
VIII. TRANSCUTANEOUS ELECTRIC
NERVE STIMULATION
Pain transmitted by small, unmyelinated C-
fibers is inhibited by activation of large myelinated
-fibers. Low-voltage, high-frequency (80-Hz)
transcutaneous electric impulses, which are contin-
uous but variable, stimulate the large myelinated
-fibers. By placing the strip electrodes 2 cm from
both sides of the thoracotomy and activating the
power source, some diminution of post-thoracot-
omy pain occurs.
Transcutaneous electric nerve stimulation has
the advantages of low cost, ease of application,
and lack of undesirable side effects. However, al-
though most patients obtain some relief, it is not
complete, and some patients do not experience any
pain relief. In those patients who do obtain anal-
gesia, the duration of pain relief has varied from 1
hour,
100
to 1 day,
28
to 5 days.
87
Transcutaneous
electric nerve stimulation may, however, reduce
narcotic requirements,
101
help to improve pul-
monary function,
101
l 02
and decrease pulmonary
complications.
101
103
At the present time, transcu-
taneous electric nerve stimulation is generally re-
served for adjunctive use with narcotics to relieve
postoperative pain from thoracotomy.
IX. EPIDURAL OPIOIDS
Since the use of epidural morphine for the treat-
ment of pain was first reported, in 1979,
104
this
therapy has rapidly achieved widespread popular-
ity.
105
Treatment of pain after thoracic surgery with
epidural narcotics has several important advan-
tages. First, the unique feature of epidural narcotic
analgesia is that there is no sympathetic blockade
and neither motor nor sensory loss; therefore, it
allows the patient to ambulate without the risk of
orthostatic hypotension or motor incoordination
usually associated with local anesthetics adminis-
tered epidurally or opioids administered parenter-
als. Second, success in relieving pain is usually
potent, prolonged, sedation free, and predictable.
Third, the duration of pain relief is generally much
longer than that which results from using paren-
teral narcotics.
104106
-
111
It appears that epidural and intrathecal opiates
block the presynaptic and postsynaptic neurons in
the substantia gelatinosa of the dorsal horn of the
spinal cord by binding to receptors and blocking
the release of substance P. Epidural opiates gain
access to the spinal cord by passive diffusion
across the dura into the cerebrospinal fluid
107
"
2
and by absorption into arachnoid villi and poste-
rior radicular arteries (which supply the dorsal
horn region).
113
114
Typically, 2 to 4 per cent of an
epidural dose of morphine finds its way into the
cerebrospinal fluid.
115
In support of the contention that epidural
opioids block spinal cord transmission of pain are
the high density of opiate receptors in the substan-
tia gelatinosa of the dorsal horn;
113
116
the high
cerebrospinal fluid: low plasma concentration ratio
(40:100) of all narcotics after epidural and in-
trathecal use;
107
"
7
and duration of action, which
is related to cerebrospinal fluid narcotic concentra-
tion rather than plasma narcotic concentra-
tion.
107,
"
7
Blockade of the substantia gelatinosa
results in a "selective" spinal block of pain con-
duction; does not block other sensations, such as
external pressure, pinprick, temperature, and pro-
prioception; and does not cause a sympathetic or
motor block. Consequently, epidural narcotics
cannot provide complete anesthesia during sur-
gery.
118
"
9
In contrast, epidural local anesthetics
block nerve impulse conduction in the axons of
nerve roots and long tracks in the spinal cord,
resulting in blockade of sympathetic, sensory, and
motor function, and can be used to provide com-
plete anesthesia.
There is a close relationship between the lipid
solubility of epidural narcotics and onset, duration,
and extent of analgesia. High lipid solubility
means that the narcotic leaves the epidural space
and enters the subarachnoid cerebrospinal fluid
space rapidly; readily fixes to spinal opioid recep-
tors, resulting in a rapid onset of action; does not
diffuse widely throughout the epidural and cere-
brospinal fluid space; and therefore results in a
more segmental block (Fig. 21-5). High lipid sol-
ubility also means that the drug is absorbed into
the vascular space from the epidural and cerebro-
spinal fluid spaces more rapidly, resulting in a

Epidural Narcotics
High Lipid Solubility
Figure 21-5 Epidural narcotics may be either of high or low lipid solubility. Narcotics with high lipid solubility diffuse rapidly
into the cerebrospinal fluid (CSF) space and readily fix to spinal cord receptors, resulting in a rapid onset of a segmental block. In
addition, high lipid solubility narcotics diffuse rapidly out of the epidural space into the vascular space, resulting in a shorter
duration of action. Conversely, low lipid solubility narcotics diffuse slowly into the CSF space and do not readily fix to spinal cord
receptors, resulting in a slow onset of a nonsegmental block. In addition, low-solubility narcotics diffuse slowly out of the epidural
space into the vascular space, resulting in a longer duration of action. The open arrows indicate rapid or major pathways, and the
thin, closed arrows indicate slow or minor pathways.
shorter duration of action but higher plasma con-
centrations (see Fig. 21-5). Consequently, contin-
uous infusion techniques are appropriate and nec-
essary if short-acting lipophilic opioids such as
fentanyl and sufentanil are selected. A 1 g/kg
bolus followed by a 1 g/kg/hour infusion of fen-
tanyl results in a progressive rise in plasma fen-
tanyl concentrations, from 0.42 ng/ml at 1 hour to
1.54 ng/ml at 18*hours.
120
A 1.5 g/kg bolus fol-
lowed by a 1.5 g/kg/hour infusion of fentanyl
results in a 2.0 ng/ml steady-state plasma concen-
tration at 9 hours.
121
Because a plasma fentanyl concentration of 1 to
3 ng/ml from PCA infusion provides good anal-
gesia, these plasma levels and the moderately good
post-thoracotomy pain score correlation with the
plasma fentanyl concentration indicate that a large
systemic component may contribute to the post-
thoracotomy analgesic effect of epidural fen-
tanyl.
121-123
However, because intravenous fentanyl
causes more nausea and vomiting
122
and may not
increase pulmonary function as much as epidural
fentanyl,
124
and not all of the analgesic effect of
epidural fentanyl can be accounted for by plasma
J
concentrations or infusion rates of fentanyl,
124
125
most pain practitioners believe that epidural fen-
tanyl is a useful and unique post-thoracotomy pain
technique.
The lipophilic narcotics are sufentanyl, fentanyl,
methadone, and meperidine (in decreasing order of
lipid solubility). When administered in the epi-
dural space at the segmental level of pain in doses
of 50 g, 0.1 mg, 5 mg, and 30 to 100 mg, respec-
tively, they have a relatively short onset of action
(less than 12 min), provide nearly complete pain
relief in 20 to 30 min, and have a relatively short
duration of action of 3 to 10 hours.
109, 126
By contrast, morphine has a more complex mo-
lecular structure and a low lipid solubility. Low
lipid solubility means that the narcotic leaves the
epidural space and enters the sub-arachnoid space
slowly; does not readily fix to the spinal opioid
receptors, which results in a slow onset of action;
diffuses widely throughout cerebrospinal fluid
space; and therefore results in a relatively nonseg-
mental block (see Fig. 21-5). Low lipid solubility
also means that the drug is absorbed into the vas-
cular space from the epidural and cerebrospinal
fluid spaces more slowly, resulting in a longer
duration of action (see Fig. 21-5). With an epi-
dural dose of 5 mg, morphine has a relatively slow
onset of action (15 to 30 min), provides maximal
pain relief in 40 to 60 min, and has a duration of
action often in excess of 12 hours. A 6-mg dose of
epidural morphine creates a peak plasma concen-
tration of 34 ng/ml at 15 min and a peak cerebro-
spinal fluid concentration of more than 1000 ng/
ml at 1 hour.
127
Because the lipophobic morphine spreads
widely in the epidural and cerebrospinal fluid
spaces and causes a nonsegmental block, the level
of epidural injection is not nearly as important for
adequate pain relief (see later) as it is with the
lipophilic narcotics. The wider spread of morphine
is also responsible for an increased incidence of
late respiratory depression compared with the li-
pophilic narcotics. All the narcotics have a positive
epidural dose-response (intensity and duration of
pain relief) relationship."
1128
129
The major advantage of "selective" blockade
of spinal pain with narcotics lies in the avoidance
of the major complications of local anesthetic
blockade or parenteral narcotic administration de-
scribed previously. Contraindications to the use of
both epidural and intrathecal narcotics include el-
evated intracranial pressure, pre-existing periph-
eral neurologic disease, allergy to narcotics, infec-
tion at the site of injection, and perhaps gross
disturbances of coagulation.
The use of intrathecal (subarachnoid) narcotics
for pain relief is greatly limited by the inability to
give repeated injections owing to the risks in-
volved in leaving an indwelling catheter in the
subarachnoid space. In an attempt to partially by-
pass this problem, single intrathecal injections of
morphine have been successfully used preopera-
tive^ or intraoperatively to provide 18 to 24 hours
of postoperative pain relief.
130-132
Intrathecal mor-
phine results in extremely high cerebrospinal fluid
concentrations (intrathecal dose should be 10 times
less than the epidural dose). Subsequent slow (6
hours), passive circulation of cerebrospinal fluid
allows the lipophobic morphine to reach the cis-
terns of the brain and the respiratory center.
133
Thus, the use of intrathecal morphine appears to
be associated with an increased occurrence of se-
vere late respiratory depression (4 to 7 per cent
compared with much less than 1 per cent for epi-
dural administration).
133
Because the epidural route
can provide much longer-lasting analgesia with
less risk, it has generally replaced the intrathecal
route for treating postoperative pain.
The most serious complications of epidural
opioids are early and late respiratory depression.
The late respiratory depression can be life-threat-
ening and is associated with decreased mentation
and confusion. Fortunately, the late respiratory
depression is an infrequent occurrence (0.25 to 0.4
per cent
133
or lower [0.09%]
134
135
incidence) and
is usually gradual in onset, allowing time for di-
agnosis and treatment."
3
I 33
The early respiratory
depression seems to be directly related to the
plasma concentration of narcotic and has occurred
in 30 min for some of the lipophilic narcotics and
in 98 min for lipophobic morphine. Early respira-
tory depression is more likely with the lipid-solu-
ble narcotics (e.g., fentanyl and sufentanyl) be-
cause of their predilection for rapid systemic
absorption. The late respiratory depression has oc-
curred in 6 to 10 hours for epidural morphine and
is thought to be due to rostral spread of the hydro-
philic morphine within the cerebrospinal fluid.
136
It is important to note that late respiratory
depression is a slow-onset and indolent process
and is associated with progressive somnolence.
Therefore, the current practice of using supple-
mental oxygen and monitoring respiratory rate and
degree of somnolence every 30 to 60 min has
become the standard of care. In the absence of
special risk factors (see later discussion), it may be
safely performed on the surgical wards.
137
"
139
Results of quantitative studies of ventilatory re-
sponses to carbon dioxide, such as minute ventila-
tion (V
E
) and airway occlusion pressure, per-
formed before and after epidural morphine
administration are consistent with clinical experi-
ence and have shown 20 to 30 per cent decreases
in V
E
/C0
2
and P
100
/CO
2
slopes.
140
The rapidity of onset, severity, and incidence of
respiratory depression may be increased by other
factors, such as inadvertent durai puncture,
1
"
141
use of thoracic epidural opioids,
135,142
143
increased
dosage and age,
133
position of the patient after in-
jection,
144
raised intrathoracic and intra-abdominal
pressures (e.g., controlled ventilation, coughing,
vomiting), and especially the administration of
concomitant oral, intramuscular, or intravenous
sedative and/or narcotics.
133
Conversely, chronic administration of systemic
opioids protects against the development of respi-
ratory depression. Indeed, there are no reports of
respiratory depression in chronic pain patients pre-
viously made tolerant to opioids.
117
145_147
Nalox-
one, 0.2 mg bolus, 0.2 to 0.3 mg/hour, or 5 to 10
g/kg/hour infusion (made by mixing four am-
pules [1.6 mg] into 250 ml of D
5
W and giving it
at a rate of 30 to 50 ml/hour [0.2-0.32 mg/hour])
dramatically reverses both early and late respira-
tory depression. The intravenous infusion of the
narcotic agonist/antagonist nalbuphine (0.2 mg/kg
bolus plus 0.05 mg/kg/hour) during post-thoracot-
omy epidural morphine analgesia results in signif-
icantly lower P
a
C0
2
(compared with placebo)
without decreasing the quality of analgesia.
148
In the absence of these known risk factors, pa-
tients receiving epidural morphine should be mon-
itored (respiratory rate and somnolence by trained
nursing staff is adequate) for 12 hours and those
receiving fentanyl and sufentanyl for 6 hours.
135
Patients with a respiratory rate of less than 8
breaths/min or who require excessive physical or
verbal stimulation to be responsive are "de-
pressed" and should receive a naloxone drip.
High-risk patients should be nursed in an intensive
care unit, a postoperative unit, or a step-down unit.
Continuous pulse oximetry may provide added
safety.
135
The other side effects of epidural opioids are
urinary retention, pruritus, and nausea and vomit-
ing. These side effects are much more common
but also far less serious than late respiratory
depression. The incidence of these side effects in-
creases with increasing dosage.
149
Thus, since the
relationship between dose and intensity of analge-
sic response plateaus quickly, the risk-benefit ratio
of increasing dosage rises rapidly once effective
pain relief levels have been reached.
107
-
131
The in-
cidence of side effects may be minimized by con-
tinuous infusions of low doses of narcotics into the
epidural space.
The incidence of postoperative patients' requir-
ing bladder catheterization after epidural morphine
has generally ranged around 15 per cent,
134
al-
though considerable variability in this figure has
been reported, and it can be as high as 53 per
cent.
105
15
The urinary retention occurs approxi-
mately 12 hours after beginning epidural mor-
phine,
151
and is probably due to a sympathetically
mediated increase in the tone of the bladder detru-
sor muscle and bladder neck sphincter; the inci-
dence of the complication can be greatly decreased
by administration of phenoxybenzamine.
150
The in-
cidence of pruritus has also been variable and has
been reported to be as low as 1 per cent,"
4
al-
though in most series it is considerably higher.
108
134, 136 j ^g pruritus is usually generalized,
152
but
may be confined to just the areas affected by the
epidural morphine.
153
The time of onset of pruritus
is usually several hours after injection, and there-
fore histamine release is probably not causative.
Nevertheless, the pruritus may occasionally be re-
lieved by promethazine.
153
The mechanism of ac-
tion of pruritus is most likely a generalized defect
in the modulation of cutaneous sensation.
154
The
incidence of nausea and vomiting has ranged from
17 to 50 per cent;
134
149, 155
they occur when mor-
phine reaches the chemoreceptor trigger zone in
the fourth ventricle.
156
The incidence decreases
with repeated dosing;
145
157
presumably the brain
becomes tolerant to these side effects, as it does to
the respiratory depressant effect.
Other factors that may contribute to the wide
range in reported incidences of epidural morphine
side effects are differences in clinical and experi-
mental conditions, such as use of volunteers (in-
creased frequency of complications) rather than
patients, patient age, operative site, injection level,
and dose of drug administered. Nausea and vom-
iting are responsive to droperidol. Fortunately, the
narcotic antagonist naloxone can reverse all of
these side effects. However, repeated injections of
naloxone are sometimes necessary to eliminate re-
curring side effects and suggest that the binding of
opiate to receptor is strong. Repeated use of nal-
oxone can reverse the analgesic effects as well, so
it must be used with some caution/titration if an-
algesia is to be maintained.
Epidural opioids have had a moderately exten-
sive trial in relief of post-thoracotomy pain. The
epidural catheter my be placed before the induc-
tion of general anesthetics and checked for correct
position by using a small dose of local anesthetic.
If it is placed preoperatively, it may be used intra-
operatively to administer epidural fentanyl, which
has been shown to decrease isoflurane require-
ments during thoracotomy.
158
Alternatively, the in-
itial injection of morphine or fentanyl can be made
60 or 10 min, respectively, before the end of the
surgical procedure if the catheter was placed pre-
operatively.
Alternatively, the catheter may be placed post-
operatively before emergence from anesthesia in
the lumbar region while the patient is still in the
lateral decubitus position. If the catheter is placed
at the end of the procedure (in the lumbar region),
the initial injection is made in the recovery room
Table 21-4 EPIDURAL NARCOTICS THAT HAVE BEEN USED FOR POSTOPERATIVE THORACIC
SURGERY PAIN MANAGEMENT: ROUTES, DOSAGES, AND PHARMACODYNAMICS
Site of
Incision
Thoracotomy
Abdominal
Epidural
Route
Thoracic
Lumbar
Lumbar
Drug
Morphine
Meperidine
Methadone
Alfentanil
Fentanyl
Sufentanil
Morphine
Hydromorphone
Meperidine
Methadone
Fentanyl
Sufentanil
Morphine
Methadone
Fentanyl
Lipid*
Solubility
1.4
39
116
126
813
1778
1.4
8
39
116
813
1778
1.4
116
813
Epidural Dose
(Diluent Saline):
Continuous Infusion
2-4 mg (8 ml): 0.1 mg/hour
30-75 mg (10 ml)
5 mg (10 ml)
200 g(10ml )
50-100 g (10 ml): 0.5-1.5 g/kg/hour
20-40 g (10 ml): 5 g/hour
6-8 mg (10-15 ml)
1-2 mg (10-15 ml)
1 mg/kg (10-15 ml)
5-10 mg (10-15 ml)
1-2 g/kg (18 ml): 1-2 g/kg/hour
30-50 g (20 ml)
6-8 mg (10 ml)
5-10 mg (10-15 ml)
1-2 g/kg (10 ml): 1-2 g/kg/hour
Onset
(min)
40-60
10
15
10
10
5
40-60
20
10
15
10
5
40-60
15
10
Duration
(hours)
12-24
4-8
8-10
2
2-3
1-5
12-24
8-12
4-8
8-10
2-3
1-5
12-24
8-10
2-3
*Octanol: Water partition coefficient.
or intensive care unit. Several important clinical
points have emerged from the post-thoracotomy
epidural opioid analgesia experience.
First, although the throacic epidural route has
been used, the procedure has spinal cord risks (pri-
marily durai puncture and spinal cord damage)
159
and has a higher risk of respiratory depress-
ion.
135
142
The lumbar area for catheter insertion
has been found to be just as satisfactory for pain
relief if a slightly higher dose of morphine or fen-
tanyl and adequate diluent volumes are used.
107
"
m. 122.145. .58.160-166
In
children, the caudal route is
satisfactory,
167
and the caudal approach can be
used to place a catheter in the epidural space at the
lumbar or thoracic level.
168
A caudal catheter has
been successfully used (with morphine) post-tho-
racotomy in a patient with a previous lumbar lam-
inectomy.
169
Morphine can be used with a lumbar
epidural injection for the relief of thoracic pain
because it is lipophobic and will therefore remain
in, and have time to spread throughout, the epi-
dural and cerebrospinal fluid spaces. For relief of
thoracic pain, instillation of 6 mg of morphine in
10 to 15 ml of diluent (preservative-free normal
saline) in the lumbar epidural region has been used
successfully.
110161
There has also been a very widespread and suc-
cessful post-thoracotomy analgesia experience
with lumbar epidural fentanyl;
122
158
164
-
166
l70
ade-
quate diluent/distribution saline volumes are more
important with the use of fentanyl via the lumbar
route (compared with morphine). Although studied
to a much smaller degree, sufentanil,
17
'
172
metha-
done,
173
hydromorphone,
174
175
nalbuphine
176
and
meperidine
177
may also be used by either the tho-
racic or the lumbar epidural route. Not surpris-
ingly, the lumbar route requires a greater diluent
volume to mechanically push the narcotic into a
wider distribution because these lipophilic narcot-
ics will bind too quickly to a spinal cord at, and a
few dermatomes above and below, the segmental
level of introduction.
178
Because the lumbar route is safer than the tho-
racic route, and fentanyl has not been associated
with respiratory depression, the lumbar epidural
route with fentanyl is my post-thoracotomy pain
management technique of choice. However, there
remains a wide selection of drugs (and dosages)
for use by the lumbar and thoracic epidural routes
(Table 21-4). All the drugs listed in Table 21-4
can be, and most have been, administered in the
epidural PCA mode, with presumably the same
benefits that accrue from the intravenous PCA.
A single dose of intramuscular ketorolac (Tora-
dol) does not risk respiratory depression and ap-
pears to be a useful initial supplement. Three or 4
hours after the initial morphine or fentanyl bolus,
either fentanyl infusion (1 to 1.5 g/kg/hour) or a
morphine infusion (0.6 to 1 mg/hour) is started at
a rate of 6 to 10 ml/hour and continued overnight.
If, during the course of the infusion, the analgesia
becomes suboptimal, one can rebolus with 2 ml of
fentanyl (100 g) diluted to 10 ml and then in-
crease the infusion rate by 2 ml/hour above the
present setting. Although the practice may be
widespread, there is little objective evidence that
the addition of dilute solutions of local anesthetics
(e.g., 0.1 per cent bupivacaine) to the epidural
narcotics improves the quality of post-thoracotomy
analgesia.
121
Second, breakthrough pain is an issue that all
acute pain services must address. Some patients
experience sore throats, backache, nasogastric tube
irritation, minor headaches, or incomplete surgical
analgesia. Additionally, some patients frequently
need to experience systemic sedation in view of
their recent surgical ordeal. While coadministra-
tion of parenteral narcotics has been associated
with a greater risk of respiratory depression, all
clinical studies provide some protocol for break-
through pain. A common technique is to use one
third to one half of the conventional dose of par-
enteral narcotic in patients who have received neu-
roaxial opiates or low-dose PCA.
Third, epidural morphine
162
and fentanyl
145

179,180 ^
6
^
e e n r e
p
0 r
ted to be effective after tho-
racic trauma. In patients with multiple fractured
ribs, epidural morphine resulted in well over 6
hours of analgesia with each dose. Alternatively,
continuous epidural infusion of fentanyl (1 to 2
g/kg/hou following 2 g/kg bolus) appears to
be a good choice in the thoracic trauma situation
(as well as for post-thoracotomy pain) because it
has a decreased likelihood of causing respiratory
depression. Since there is no sympathetic blockade
with epidural opioids, there is no need to place the
patient supine for top-up doses, and patients with
some degree of hypovolemia may be managed
more safely than with epidural local anesthetic ad-
ministration.
Fourth, in the post-thoracotomy pain relief ex-
perience (summarized in Table 20-1), there have
been relatively few significant side effects.
Fifth, catheters have been left in situ and used
for as long as 5 days without the development of
tolerance.
153
Sixth, pain relief failures have usually been due
to improper catheter localization since no pain re-
lief could be demonstrated following injection of
local anesthetics.
153
Seventh, the expected increase in the mean post-
thoracotomy expiratory flow rate, forced vital ca-
pacity, FRC, and ability to tolerate respiratory care
maneuvers following epidural pain relief have
been demonstrated.
153
Even larger increases in
ventilatory capacity have been demonstrated in the
much larger experience with epidural pain relief
after abdominal surgery.
152
These increases are
generally greater than, or equal to, those obtained
by intravenous narcotics or epidural local anes-
thetic administration.
Finally, epidural narcotics have been used to
treat intractable pain due to thoracic tumors.
117
Less dosage is required and longer duration of
analgesia with each intermittent dose is obtained
for intractable cancer pain compared with acute
postoperative pain.
117
However, although results
have been satisfactory after the first few series of
injections for intractable pain, tolerance has devel-
oped, and dosages have had to be increased owing
to continuous bathing of the spinal cord with a
cerebrospinal fluid concentration of opioid.
145
For
treatment of chronic pain, permanent catheters
have been implanted in the subarachnoid and epi-
dural spaces and connected to reservoirs or perfu-
sion pumps.
145
m
These implants have been found
to be effective, but again tachyphylaxis has been
observed.
145
REFERENCES
1. Miguel R, Hubbell D: Postthoracotomy pain relief: Does
the technique really make a difference? Anesthesiology
75:A1091, 1991.
2. Conacher ID: Pain relief after thoracotomy. Br J Anaesth
65:806-812, 1990.
3. Loan WB, Morrison JD: The incidence and severity of
postoperative pain. Br J Anaesth 39:695-698, 1967.
4. Conacher ID, Paes ML, Jacobson L, Phillips PD, Heavi-

Benumof - Thoracic Anesthesia

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Benumof - Thoracic Anesthesia

Uploaded by

Copyright:

Available Formats

Anesthesia

are dynamically determined

false-negative results most fre-

You might also like