Professional Documents
Culture Documents
Volume 2
editors
Alan B. Lumsden MD
Chairman, Department of Cardiovascular Surgery; Medical Director, Methodist DeBakey
Heart & Vascular Center, The Methodist Hospital, Houston, Texas; Professor of Cardiovascular
Surgery, Weill Cornell Medical College, New York, New York
Minneapolis, Minnesota
www.cardiotextpublishing.com
Comments, inquiries, and requests for bulk sales can be directed to the publisher at:
info@cardiotextpublishing.com.
All rights reserved. No part of this book may be reproduced in any form or by any means without the prior
permission of the publisher.
All trademarks, service marks, and trade names used herein are the property of their respective owners and
are used only to identify the products or services of those owners.
This book is intended for educational purposes and to further general scientific and medical knowledge,
research, and understanding of the conditions and associated treatments discussed herein. This book is not
intended to serve as and should not be relied upon as recommending or promoting any specific diagnosis
or method of treatment for a particular condition or a particular patient. It is the readers responsibility to
determine the proper steps for diagnosis and the proper course of treatment for any condition or patient,
including suitable and appropriate tests, medications, or medical devices to be used for or in conjunction
with any diagnosis or treatment.
Due to ongoing research, discoveries, modifications to medicines, equipment and devices, and changes
in government regulations, the information contained in this book may not reflect the latest standards,
developments, guidelines, regulations, products, or devices in the field. Readers are responsible for keeping
up-to-date with the latest developments and are urged to review the latest instructions and warnings for any
medicine, equipment, or medical device. Readers should consult with a specialist or contact the vendor of
any medicine or medical device where appropriate.
Except for the publishers website associated with this work, the publisher is not affiliated with and does not
sponsor or endorse any websites, organizations, or other sources of information referred to herein.
The publisher and the author specifically disclaim any damage, liability, or loss incurred, directly or
indirectly, from the use or application of any of the contents of this book.
Unless otherwise stated, all figures and tables in this book are used courtesy of the authors.
ISBN: 978-1-935395-22-5
Printed in Canada
16 15 14 13 12 11 1 2 3 4 5 6 7 8 9 10
Contributors ix
Preface xiii
Anthony J. Comerota and Thomas W. Wakefield
Acknowledgments xv
Introduction 1
Mark G. Davies
Index 235
editors
Mark g. Davies MD, PhD, MBA Vice Chairman, Finance and Administration, Department
of Cardiovascular Surgery; Program Director, Vascular Surgery Fellowship and
Integrated Vascular Surgery Residency; Director of Research and Education,
Methodist DeBakey Heart & Vascular Center, The Methodist Hospital, Houston,
Texas; Senior Investigator, The Methodist Hospital Research Institute; Professor of
Cardiovascular Surgery, Weill Cornell Medical College, New York, New York
Contributors
Frank R. Arko III MD Associate Professor, Chief of Endovascular Surgery, Division of
Vascular and Endovascular Surgery, Department of Surgery, University of Texas
Southwestern Medical Center, Dallas, Texas
Gary Burns MBA, RHIA, CIRCC Principal, Medical Asset Management, Inc., Atlanta,
Georgia
Ruth L. Bush MD, MPH Associate Professor of Surgery, Vascular Surgery, Scott & White
Healthcare; Associate Dean for Education, Temple Campus, Texas A&M Health
Science Center College of Medicine, Temple, Texas
ix
Anthony J. Comerota MD, FACS, FACC Director, Jobst Vascular Institute, Toledo, Ohio;
Adjunct Professor of Surgery, University of Michigan, Ann Arbor, Michigan
Bo Eklf MD, PhD Emeritus Professor, Department of Surgery, University of Hawaii; Clinical
Emeritus Professor, Lund University, Sweden
Valerie B. Emery RN, ANP, BC Department of Surgery, Section of Vascular Surgery, Center
for Thoracic Outlet Syndrome, Washington University School of Medicine and Barnes-
Jewish Hospital, St. Louis, Missouri
Joseph P. Hart MD, RVT, FACS Assistant Professor of Surgery and Interventional Radiology;
Chief of Endovascular Surgery, Division of Vascular Surgery, Department of Surgery,
Medical University of South Carolina, Charleston, South Carolina
Peter K. Henke MD Leland Ira Doan Professor of Surgery, Section of Vascular Surgery,
University of Michigan, Ann Arbor, Michigan
Brandi Huf PharmD Scott and White Anticoagulation Clinic, Scott and White Memorial
Hospital and Clinics, Temple, Texas
Kaj H. Johansen MD, PhD, FACS Swedish Heart & Vascular Institute, Swedish Medical
Center, Seattle, Washington
Patricia Hightower Lambden MSN, CNS-BC, APN McLennan Community College, Waco,
Texas
Alan H. Matsumoto MD, FSIR, FACR, FAHA Professor and Chair, Department of Radiology
and Medical Imaging, University of Virginia Health System, Charlottesville, Virginia
L. Bernardo Menajovsky MD, MS Associate Professor of Internal Medicine, Scott & White
Healthcare, Temple, Texas
Wael E. A. Saad MBBCh, FSIR Associate Professor of Radiology, Division of Vascular and
Interventional Radiology, University of Virginia Health System, Charlottesville, Virginia
CleAnn Toner PharmD Scott and White Anticoagulation Clinic, Scott and White Memorial
Hospital and Clinics, Temple, Texas
Venous thromboembolism (VTE) includes deep venous thrombosis (DVT) and pulmonary
embolism (PE). VTE is a national health concern, with mortality exceeding that of acute
myocardial infarction and acute stroke. VTE affects over 1,000,000 patients per year, with
over 300,000 deaths per year in the United States. VTE fatalities have remained constant over
the past 30 years, exceeding breast cancer and AIDS combined. The incidence of DVT has
been increasing with the aging of the population. In those 85-89 years old, the incidence is
reported to be as high as 310 people per 100,000 in the population. Additionally, treatment
costs are in billions of dollars per year. The local consequence of DVT, termed postthrombotic
syndrome (PTS), affects between 400,000 and 500,000 patients annually with pain, edema,
pigmentation, and skin ulcerations. It has been reported that after having an iliofemoral DVT
treated with anticoagulation alone, 95% of patients have chronic venous insufficiency, 40%
have venous claudication and nearly all have reduced quality-of-life. Even asymptomatic DVT
has been associated with PTS.
Although Virchows triad of stasis, vessel wall injury, and hypercoagulability has defined
the events that predispose individuals to DVT formation for the past 150 years, today the un-
derstanding of events that occur at the level of the vein wall, including the influence of the
inflammatory response on thrombogenesis, is increasingly becoming recognized, although we
still have much to learn.
The critical nature of venous thromboembolism to our nation and the world has been
recognized, and a Surgeon Generals conference held in Bethesda, Maryland, in May 2006
in conjunction with the National Institutes of Health led to a call to action against VTE in
2008. The fact that death due to pulmonary embolism remains the most common cause of
in-hospital death today places this problem in stark perspective.
xiii
With the current emphasis that has been placed on VTE, the production of this excellent
book which addresses all aspects of venous thromboembolic disease is very timely. In part 1,
many different aspects of venous thromboembolic disease are covered, ranging from patho-
physiology of DVT and PE to diagnosis and treatment of those conditions. Treatment involves
not only pharmacologic agents, but also mechanical and pharmacomechanical approaches.
Operative thrombectomy remains a good option for treating patients with iliofemoral DVT,
and the well-designed Scandanavian trial illustrates the value of a strategy of thrombus remov-
al in patients with iliofemoral DVT. However, VTE is not limited to the lower extremities. In
part 2, upper extremity venous thrombosis related to thoracic outlet (thoracic inlet) syndrome
is discussed. Part 3 involves issues of central venous disease with a major emphasis on dialysis
access. The lifeline of patients with end-stage renal disease is a functional dialysis access. It is
evident that the majority of successful accesses are dependent on good-quality venous outflow.
Part 4 discusses practice management of venous disease.
Venous Thromboembolic Disease, volume 2 of Contemporary Endovascular Manage-
ment, has contributions by leaders in the field, with evidence-based recommendations. We
highly recommend this volume to those who are serious students of VTE and who want to
update themselves on the current status of VTE pathophysiology, diagnosis, and treatment.
The editors would like to thank the following individuals for their kind contributions to the
completion of this book: Yvette Whittier and Daynene Vykoukal, PhD, at The Methodist Hos-
pital for organizational assistance, author relations, primary editing, and manuscript assembly;
Steve Korn, Mike Crouchet, Caitlin Crouchet, and Carol Syverson at Cardiotext Publishing
for their roles in series development, content management, and technical support; and Ann
Delgehausen, Beth Wright, and Zan Ceeley at Trio Bookworks for their diligent execution of
the publication and their contribution to the series designs and concepts. Without the time
and effort of the contributors, no book can address its goals and we gratefully acknowledge the
time and effort that each of the authors devoted to this project.
Mark G. Davies
Alan B. Lumsden
xv
PTS because the valves and vessel wall are system vein, conservative measures are ade-
damaged as the thrombus is cleared by the quate. For those lesions that encroach on a
body. Most recent guidelines recommend deep vein, recent investigations suggest that
catheter-directed thrombolysis or combined anticoagulation may be more effective than
pharmacomechanical thrombectomy for ligation in preventing DVT and PE. Venous
proximal DVT in the mobile and functional injuries are similarly underreported, and
patient with no absolute contraindications. current recommendations include repair of
Effective lysis of these DVTs can reduce or bypass of injuries to the major proximal
postthrombotic symptoms and improve veins. If repair is not safe or possible, ligation
quality of life after acute iliofemoral DVT. should be performed.
Inferior vena cava filters continue to have
a role among patients with contraindica-
tions to, complications of, or failure of an- reference
ticoagulation. The incidence of superficial 1. Meissner MH, Wakefield TW, Ascher E,
venous thrombophlebitis is underreported Caprini JA, Comerota AJ, Eklof B, et al. Acute
and is considered to occur in approximate- venous disease: venous thrombosis and venous
ly 125,000 patients per year in the United trauma. J Vasc Surg. 2007;46 (suppl) S:
States. For lesions not encroaching on a deep 25S-53S.
Venous
Thromboembolic
Disease
T he occurrence of deep ve-
nous thrombosis (DVT) and pulmonary embolism continues be a
significant problem in clinical practice. guidelines for prevention of
DVT have been updated by the American College of Chest Physi-
cians (ACCP) and DVT prophylaxis remains a high-priority qual-
ity measure for hospitals and practitioners. Prevention of DVT and
pulmonary embolism can save lives and decrease morbidity. As
we refine our knowledge of the biology and progression of DVT,
we will be able to develop better anticoagulants and alternative
strategies to control DVT propagation. While the efforts to prevent
DVT continue, the ability to clear existing DVTs has become more
refined. While use of pharmacological lysis is a mainstay, introduc-
tion of mechanical declotting and ultrasound generating devices
and combined pharmacomechanical techniques have accelerated
treatment strategies and decreased lysis-related complications. In
certain patients with limb-threatening DVT, open surgical embolec-
tomy remains a viable and necessary option in the carefully selected
patient. There has been a surge in the prophylactic placement of
overview of Therapy
for Venous Thromboembolic Disease
eduardo ramacciotti and Thomas W. Wakefield
Once the diagnosis is confirmed, son with UFH, the small molecular weight
immediate systemic anticoagulation should of LMWH means that it has a more predict-
be started, unless contraindicated. A thera- able anticoagulant effect, can be given sub-
peutic dose is required, due to a 4- to 6-fold cutaneously, causes less platelet activation
risk of recurrence with insufficient antico- and bleeding, and is associated with a lower
agulation in the first 24 hours.3 Adequate rate of heparin-induced thrombocytopenia.7
anticoagulation has been shown to prevent In addition, LMWH may be administered
the development of fatal PE both during the subcutaneously in a weight-based manner
initial treatment and after treatment is com- (q.d. or b.i.d.), and in many instances, may
plete.4 However, the recurrence rates are still be administered in the outpatient setting.
high. It is expected that one-third of properly These compounds do not require monitor-
treated patients will present recurrent DVT ing except in certain circumstances such
after an 8-year follow-up period.5 as renal failure and morbid obesity, and
Historically, intravenous unfraction- during pregnancy.8 However, its use in the
ated heparin (UFH) was the initial standard outpatient settings usually requires a team
therapy. Due to large individual variation in approach and a coordinated effort of mul-
heparin effect response, frequent adjustment tiple healthcare providers. There is also lim-
of the dose is required and the treatment is ited evidence that LMWH may decrease the
maintained for an average of 5 days. Patients incidence of postthrombotic syndrome.
usually have to be hospitalized to undergo There are different LMWHs available
the frequent monitoring and dose adjust- for the treatment of VTE. Since they are bio-
ments necessary to maintain therapeutic logic preparations, they are not interchange-
plasma concentrations of UFH. Monitor- able. Moreover, the regimen and doses for
ing of activated partial thromboplastin time the compounds differ, as listed in Table 1.1.
(aPTT) is crucial during the initial phase of
treatment. Patients who fail to achieve thera-
peutic levels of heparin within 24 hours of
starting treatment have a 15 times higher Prevention of
risk of recurrence than patients who achieve
heparin concentrations within the target Recurrent Venous
therapeutic range, aPPT 1.5 to 2.5 x control.6
In this period, treatment with oral vitamin
Thromboembolism
K antagonist (VKA) is instituted (usually Warfarin (and other VKAs) remains the only
warfarin, 5 mg/day) and an international available oral anticoagulant proven to reduce
normalized ratio (INR) therapeutic (usually the risk of recurrent VTE. Several studies
above 2) is required before stopping hepa- have established that for the vast majority
rin infusion. The usual minimum time for of patients, anticoagulation results in an
heparin initial therapy is 5 days. annual risk of recurrence of less than 2%.9,10
Recently, the treatment of VTE has Because warfarin has a narrow therapeutic
been markedly simplified by the introduc- index, monitoring the INR is necessary. A
tion of low-molecular-weight heparins target INR value of 2 to 3 is appropriate for
(LMWHs). LMWHs are derived from stan- patients with VTE.
dard heparin after exposure to different Warfarin should be started only after
chemical processes. However, in compari- heparinization is therapeutic to prevent war-
LMWH
Enoxaparin SQ 1.0 mg/kg b.i.d. or No
1.5 mg/g q.d.
UFH = unfractionated heparin IV, LMWH= low-molecular-weight heparin. IV= intravenous,SQ= subcutaneous,
aPTT= activated partial thromboplastin time. Taking all of the evidence together, LMWH is now preferred over
standard UFH for the initial treatment of VTE with a level of evidence 1A.
Source: van Dongen CJ, van den Belt AG, Prins MH, Lensing AW. Fixed dose subcutaneous low molecular
weight heparins vs. adjusted dose unfractionated heparin for venous thromboembolism. Cochrane Db Syst Rev.
2004(4):CD001100.
risk of VTE recurrence, and prolonged oral at an increased risk for VTE recurrence.17
anticoagulation is warranted. However, first Moreover, another recent study has demon-
episodes of VTE in patients with heterozy- strated a statistically significant advantage to
gous mutations of FVL/ FII do not carry the resuming Coumadin if the D-dimer assay
same high risk as those in their homozygous is positive compared to remaining off Cou-
counterparts, shortening the required length madin during an average 1.4-year follow-up
of oral anticoagulation. period (odds ratio [OR] 4.26, P = 0.02).18
Another particularly controversial treat- One other factor that impacts the recurrence
ment situation involves patients experienc- rates is the quality of anticoagulation. One
ing unprovoked VTE. These subjects are trial has demonstrated that patients exposed
at substantial risk for recurrent disease after to an INR <1.5 during the first 3 months of
warfarin discontinuation. During 3 years treatment presented higher risk for recurrent
following warfarin discontinuation, approxi- VTE.19
mately 20% of such patients will suffer recur- Decisions about the duration of antico-
rent VTE, particularly on the same vascular agulant therapy are further complicated by
site as the index event.13 The actual length the need to individualize the hazards asso-
of therapy and dose of warfarin best suited ciated with anticoagulation by weighing the
for this particular group are not known. The risks vs. the benefits of anticoagulation. After
PREVENT multicenter trial suggested that, estimating both the risk of thrombosis with-
for idiopathic DVT, low-dose warfarin (INR out warfarin as well as the risk of bleeding
1.52.0) is superior to placebo over a 4-year with treatment, the clinician should con-
follow-up period with a 64% risk reduction sider factors such as adherence history, risk
for recurrent DVT after the completion of of falling, and underlying diseases before
an initial 6 months of standard warfarin making a recommendation about when or
therapy.14 However, a second study has sug- if to discontinue anticoagulation.20 In addi-
gested that full-dose warfarin (INR 23) is tion, if the anticoagulation is not discon-
superior to low-dose warfarin in these same tinued, decreasing risk of bleeding dose
patients, without a difference in bleeding.15 adjustments and using anticoagulation clin-
Taken together, this data suggest that long- ics are emphasized.
term therapy is required for unprovoked Thus, the decisions for discontinua-
VTE patients, with INR targeting 2 to 3. tion of oral anticoagulation should include
When estimating the risk of recurrent thrombosis risk assessment, evaluation of
disease for an individual patient, certain residual thrombus burden, and coagulation
factors have been linked to a higher likeli- system activation (as suggested by D-dimer
hood of future events: an unprovoked ini- measurements). These criteria are given a
tial clot, and if the index event is PE, rather level of evidence of 1A.17-19
than DVT.16 Recently, one additional crite-
rion has been used to determine the length
of anticoagulation. This validated criterion Complications of
involves D-dimer testing obtained 1 month
after warfarin is completed. If the D-dimer
Anticoagulant Therapy
level is elevated above normal, warfarin The most common complication of antico-
should be continued, as this result suggests agulation is bleeding. Increasing the anti-
that the patient is still prothrombotic and coagulant effect of a pharmacological agent
increases the risk of bleeding. By decreasing mediated reaction, and it seems to be safe as
the anticoagulant effect of the compound, a substitute anticoagulant for HIT patients.22
the risk of bleeding decreases, but the risk The use of these alternative agents is given a
of recurrence or even treatment failure 2C and 1C level of evidence.
increases. For UFH, the risk of bleeding dur-
ing the initial 5 days is around 10%. With
the addition of warfarin at an INR between 2
and 3, the annual bleeding risk is 6%. Prolonged
Another complication associated with
heparin use is heparin-induced thrombo-
Anticoagulation
cytopenia (HIT). This immune-mediated in Cancer Patients
disorder occurs in 0.6% to 30.0% of patients
exposed to heparin. Heparin-dependent The safety of LMWH compared with that
antibodies mediated by platelet factor 4 of warfarin has led to a consideration of
(PF4) bind to platelets, activating them. This the long-term use of LMWH as a replace-
activation leads initially to platelet aggrega- ment for oral vitamin K antagonists. Rates
tion and release of thrombogenic micropar- of recanalization have been reported to be
ticles, which can produce thrombosis, both higher in certain venous segments using
arterial and venous. Thrombocytopenia due LMWH vs. traditional oral agents.
to consumption occurs, increasing the risk of Cancer patients are reported to have
bleeding. Both UFH and LMWH have been longer periods out of the therapeutic INR
associated with HIT, although the incidence range compared to patients free of cancer.
is lower with LMWH. Clinically, the patient Additionally, the use of LMWH has been
presents a drop on platelet count greater found to result in improved outcomes in can-
than 50% of baseline, or below 100,000/L cer patients compared with standard heparin
during heparin therapy. Eventually, throm- or LMWH/warfarin therapy when used for 6
bosis occurs.21 The test of choice for diag- months, without differences in major bleed-
nosis is an enzyme-linked immunosorbent ing.23 The use of LMWH in selected can-
assay (ELISA) that detects the antiheparin cer patients for prolonged anticoagulation is
antibody in the patients plasma. Cessation given a 1A level of evidence.
of heparin is mandatory. LMWH should LMWH has also been found to pro-
not be used as substitute, due to a strong vide better DVT prophylaxis compared with
immunogenic cross-reactivity. In addition, placebo for extended 4-week prophylaxis in
warfarin alone is also contraindicated. The patients undergoing abdominal/pelvic can-
initial drop in protein C/S associated with cer surgery, decreasing by half the incidence
beginning VKA treatment in conjunction of postoperative VTE, with similar bleeding
with HIT may produce severe paradoxal rates.24
thrombosis. Direct thrombin inhibitors hiru-
din (lepirudin/Refludan) and argatroban
are the treatments approved by the FDA, Standard Therapy
although other agents such as fondaparinux
have been found to treat this syndrome as
for VTE Algorithm
well.22 Fondaparinux, a synthetic penthas- The algorithm in Figure 1.1 summarizes the
sacharide, does not produce this immune- standard pharmacological therapy for DVT.
DVT
Confirmed
Discontinue Heparin
UFH or LMWH UFH or LMWH UFH or LMWH UFH or LMWH UFH or LMWH
or Fondaparinux or Fondaparinux or Fondaparinux or Fondaparinux or Fondaparinux Day 6
Warfarin (?) mg
Day 1 Day 2 Day 3 Day 4 Day 5
Day 8
Warfarin (?) mg
Day 9
Day 10
Warfarin (?) mg
Day 11
Decision to stop or continue VKA
Figure 1.1. Algorithm for the standard pharmacological therapy for DVT.
include catheters with microsonic (ultra- for <7 days, good functional status, and life
sound emission) capabilities to facilitate expectancy of >1 year) can be treated with
fibrin dissolution, and catheters that utilize operative venous thrombectomy to reduce
the Venturi effect. acute symptoms and postthrombotic mor-
The effectiveness of mechanical throm- bidity, if appropriate expertise and resources
bectomy alone or in combination with are available (grade 2B). If such patients do
pharmacologic thrombolysis was recently not have a high risk of bleeding, CDT is usu-
compared. Mechanical thrombectomy alone ally preferable to operative venous thrombec-
was successful for removing a thrombus that tomy, with a grade 2C recommendation.28
developed intraprocedurally (which is gen-
erally gelatinous and not cross-linked with
fibrin). However, in patients with thrombosis Future Medical
already installed, only 26% of the thrombus
was removed by mechanical thrombectomy.
Treatments for DVT/PE
The addition of a plasminogen activator solu- The evolution of anticoagulant drugs for
tion to the mechanical technique removed VTE from the 1930s to the present day is a
82% of the thrombus.29 series of agents with increasing specificity.
Traditional therapy has been standard intra-
venous unfractionated heparin, which is a
Operative Venous Thrombectomy good drug for preventing fatal PE. However,
Contemporary venous thrombectomy for heparin has a number of potential prob-
iliofemoral venous thrombosis has been lems, including bleeding, the development
shown to be effective in both short- and long- of osteoporosis and alopecia with high doses
term follow-ups. The long-term benefits of over long periods of time, the development of
venous thrombectomy relate to its ability HIT (heparin-induced thrombocytopenia),
to achieve proximal patency and maintain and importantly, the need for IV administra-
distal valve competency. Both outcomes are tion and frequent monitoring. Additionally,
influenced by the initial technical success contaminated heparins were a problem ear-
and the avoidance of recurrent thrombosis. lier in 2008. Low-molecular-weight heparins
Therefore, attention to operative detail in (LMWHs) are an alternative to standard
terms of complete thrombus removal, cor- unfractionated heparin, which has become
recting underlying venous stenoses, and the gold standard for the prophylaxis and
maintaining therapeutic anticoagulation treatment of VTE. LMWHs work more on
postoperatively is crucial. Pooled data from a inhibiting factor Xa than on inhibiting fac-
number of contemporary reports on iliofem- tor IIa (thrombin). Advantages of LMWHs
oral venous thrombectomy demonstrate that include an improved pharmacokinetic pro-
early and long-term patency of the iliofemo- file, a half-life that is not dose-dependent,
ral venous segment is 70% to 80% compared less physiologic antiplatelet activity, more
with 30% of patients treated with anticoagu- constant anti-factor Xa activity, less protein
lation alone.1 C antigen decrease, less complement activa-
The indication for venous thrombec- tion, less inhibition of physiologic platelet
tomy in proximal DVT patients was reviewed aggregation, and importantly the ability for
in the latest (2008) ACCP consensus: subcutaneous administration with no moni-
patients with proximal DVT (eg, symptoms toring necessary (except in renal failure,
pregnancy, or morbid obesity).30 LMWH is an oral direct factor Xa inhibitor that inhib-
a little better than standard unfractionated its free and fibrin-bound factor Xa activity
heparin regarding recurrent venous throm- and thrombotic activity. It has potent anti-
boembolic events, major hemorrhage, and coagulant effects, and it does not directly
even mortality compared to standard hepa- inhibit thrombin but instead inhibits throm-
rin. Even proximal above-knee thrombi are bin generation via inhibition of factor Xa
associated with improvements with LMWH activity. It has a rapid onset of action (within
therapy. In fact, the latest 2008 ACCP guide- 24 hours), high bioavailability (greater
lines give LMWH a 1A recommendation for than 80%), and does not affect agonist-
the initial treatment of VTE.28 induced platelet aggregation and therefore
Newer drugs include fondaparinux, has no direct effects on primary hemostasis.
among others. This indirect thrombin It does not require a cofactor such as anti-
inhibitor is a synthetic pentasaccharide that thrombin, it may be safely administrated
is identical to the antithrombin III binding with concomitant agents, and there are no
sequence of heparin. Fondaparinux is useful dosage requirement adjustments needed
for the prophylaxis and treatment of VTE, for gender, age, or extreme body weight.
and it has been FDA-approved for the pro- It has been evaluated in treatment studies
phylaxis of DVT, which may lead to PE in and prophylaxis studies, where it has been
patients undergoing abdominal surgery, found to be quite effective in reducing the
hip fracture surgery, extended prophylaxis risk for VTE, between approximately 30%
for hip fracture surgery, hip replacement and 80%. Apixaban, another factor Xa
surgery, and knee replacement surgery. inhibitor, has also successfully undergone
Additionally, it is indicated for the treat- both treatment and prophylaxis protocols.
ment of acute DVT when administered in Finally, dabigatran etexilate is a new oral
conjunction with warfarin sodium and for direct thrombin inhibitor. This drug has
the treatment of PE when administered in predictable anticoagulant effects, no need
conjunction with warfarin sodium and ini- for monitoring, and binds directly to throm-
tial therapy is given in the hospital. A cousin bin with high affinity and specificity. This
of fondaparinux, Idraparinux, has a 10-day drug was approved for prophylaxis in total
half-life. However, it did not meet the non- hip replacement and total knee replace-
inferiority target against PE in a recent study ment in Western Europe and Canada, but
and was also associated with significant approval will require additional trials in the
intracranial bleeding. Thus, its development United States as testing against LMWH 30
was halted. The drug has been biotinylated mg b.i.d. failed to meet the noninferiority
so that its effects can be reversed by avidin. target. This drug shows a low rate of bleed-
This drug is called SSR126517 and is under- ing comparable to LMWH, demonstrates
going evaluation. elevated liver enzymes in only a small por-
Other new drugs target specifically tion of patients (comparable with LMWH
either factor Xa or IIa and are oral agents. treatment), and offers oral dosing without
These include rivaroxaban, apixaban, and coagulation monitoring.
dabigatran.31 A comparison of these 3 drugs HIT requires special components for
reveals that they all are orally administered, its treatment. Agents available for patients
they have varying bioavailability, and their with HIT include direct thrombin inhibi-
half-lives are not the same. Rivaroxaban is tors lepirudin (Refludan), argatroban, and
11. Hyers TM, Agnelli G, Hull RD, Morris TA, 19. Palareti G, Legnani C, Cosmi B, Guazzaloca
Samama M, Tapson V, et al. Antithrombotic G, Cini M, Mattarozzi S. Poor anticoagulation
therapy for venous thromboembolic disease. quality in the first 3 months after unprovoked
Chest. 2001;119(1 suppl):176S-193S. venous thromboembolism is a risk factor for
12. Schulman S, Rhedin AS, Lindmarker P, long-term recurrence. J Thromb Haemost.
Carlsson A, Larfars G, Nicol P, et al. A 2005;3(5):955-61.
comparison of six weeks with six months of 20. Garcia DA, Spyropoulos AC. Update in the
oral anticoagulant therapy after a first episode treatment of venous thromboembolism.
of venous thromboembolism. Duration of Sem Resp Crit Care M. 2008;29(1):40-6.
Anticoagulation Trial Study Group. N Engl J 21. Alving BM. How I treat heparin-induced
Med. 1995;332(25):1661-5. thrombocytopenia and thrombosis. Blood.
13. Eichinger S, Weltermann A, Minar E, 2003;101(1):31-7.
Stain M, Schonauer V, Schneider B, et al. 22. Kovacs MJ. Successful treatment of
Symptomatic pulmonary embolism and the heparin induced thrombocytopenia (HIT)
risk of recurrent venous thromboembolism. with fondaparinux. Thromb Haemost.
Arch Intern Med. 2004;164(1):92-6. 2005;93(5):999-1000.
14. Ridker PM, Goldhaber SZ, Danielson 23. Lee AY, Levine MN, Baker RI, Bowden
E, Rosenberg Y, Eby CS, Deitcher SR, C, Kakkar AK, Prins M, et al. Low-
et al. Long-term, low-intensity warfarin molecular-weight heparin versus a coumarin
therapy for the prevention of recurrent for the prevention of recurrent venous
venous thromboembolism. N Engl J Med. thromboembolism in patients with cancer.
2003;348(15):1425-34. N Engl J Med. 2003;349(2):146-53.
15. Kearon C, Ginsberg JS, Kovacs MJ, Anderson 24. Bergqvist D, Agnelli G, Cohen AT, Eldor
DR, Wells P, Julian JA, et al. Comparison A, Nilsson PE, Le Moigne-Amrani A, et
of low-intensity warfarin therapy with al. Duration of prophylaxis against venous
conventional-intensity warfarin therapy thromboembolism with enoxaparin
for long-term prevention of recurrent after surgery for cancer. N Engl J Med.
venous thromboembolism. N Engl J Med. 2002;346(13):975-80.
2003;349(7):631-9. 25. Prandoni P, Lensing AW, Prins MH, Frulla
16. Agnelli G, Prandoni P, Becattini C, Silingardi M, Marchiori A, Bernardi E, et al. Below-
M, Taliani MR, Miccio M, et al. Extended knee elastic compression stockings to
oral anticoagulant therapy after a first episode prevent the post-thrombotic syndrome: a
of pulmonary embolism. Ann Intern Med. randomized, controlled trial. Ann Intern Med.
2003;139(1):19-25. 2004;141(4):249-56.
17. Prandoni P, Lensing AW, Prins MH, 26. Eklof B, Kistner RL. Is there a role for
Bernardi E, Marchiori A, Bagatella P, et al. thrombectomy in iliofemoral venous
Residual venous thrombosis as a predictive thrombosis? Semin Vasc Surg. 1996;9(1):
factor of recurrent venous thrombo- 34-45.
embolism. Ann Intern Med. 2002;137(12): 27. Bller HR, Agnelli G, Hull RD, Hyers TM,
955-60. Prins MH, Raskob GE. Antithrombotic
18. Palareti G, Cosmi B, Legnani C, Tosetto A, therapy for venous thromboembolic
Brusi C, Iorio A, et al. D-dimer testing to disease: the Seventh ACCP Conference on
determine the duration of anticoagulation Antithrombotic and Thrombolytic Therapy.
therapy. N Engl J Med. 2006;355(17):1780-9. Chest. 2004;126(3 suppl):401S-28S.
28. Kearon C, Kahn SR, Agnelli G, Goldhaber of the efficacy and safety of once-daily
S, Raskob GE, Comerota AJ. Antithrombotic dose of enoxaparin versus unfractionated
therapy for venous thromboembolic disease: heparin in the treatment of proximal lower
American College of Chest Physicians limb deep-vein thrombosis. Thromb Res.
Evidence-Based Clinical Practice Guidelines 2004;114(3):149-53.
(8th ed.). Chest. 2008;133(6 suppl):454S-545S. 31. Gross PL, Weitz JI. New anticoagulants
29. Vedantham S, Vesely TM, Parti N, Darcy M, for treatment of venous thromboembolism.
Hovsepian DM, Picus D. Lower extremity Arterioscl Throm Vas. 2008;28(3):380-6.
venous thrombolysis with adjunctive 32. Baldwin ZK, Spitzer AL, Ng VL, Harken
mechanical thrombectomy. J Vasc Interv AH. Contemporary standards for the
Radiol. 2002;13(10):1001-8. diagnosis and treatment of heparin-
30. Ramacciotti E, Araujo GR, Lastoria S, Maffei induced thrombocytopenia (HIT). Surgery.
FH, Karaoglan de Moura L, Michaelis W, 2008;143(3):305-12.
et al. An open-label, comparative study
2
Deep Venous Thrombosis
and Pulmonary embolism
Pathophysiology
Peter K. Henke
19
Figure 2.1 Pathways of thrombosis. Two primary pathways that initiate thrombosis
are shown with the primary factors of platelets, tissue factor, and thrombin. These
converge on the coagulation pathway, and fibrin is formed, with thrombus produced.
PDI = protein disulfide isomerase; TF = tissue factor; gp = glycoprotein; vWF = von
Willebrand factor. Based on Furie and Furie, New Engl J Med. 2008;359:938(5).
for an effective hemostatic platelet plug to in the presence of calcium. Feedback ampli-
form, resulting in platelet activation. This fication occurs as VIIa, IXa, and Xa are all
interaction is mediated by von Willebrand capable of activating VII to VIIa, especially
factor (vWF), whose platelet receptor is gly- when bound to TF.7 Factor Xa is also capable
coprotein (Gp) Ib, and Gp VI and IX are of activating factor V to Va. Factors Xa, Va,
also involved. Direct TF de-encryption can and II (prothrombin) form on the platelet
occur by endothelial protein disulfide isom- phospholipid surface in the presence of Ca2+
erase, leading to platelet activation. Activa- to initiate the prothrombinase complex,
tion of platelets also leads to the release of the which catalyzes the formation of thrombin
prothrombotic contents of platelet granules, from prothrombin. Thrombin feedback
containing receptors for coagulation factors amplifies the system not only by activating
Va and VIIIa, as well as fibrinogen, vWF, factor V to Va, but also factors VIII (normally
and ADP. Fibrinogen forms bridges between circulating bound to vWF) to VIIIa and XI
platelets by binding to the GpIIb/IIIa recep- to XIa. After activation, factor VIIIa dissoci-
tor, resulting in further platelet aggregation. ates from vWF and assembles with factors
Platelet activation also leads to the elabora- IXa and X on the platelet surface in the pres-
tion of arachidonic acid metabolites such as ence of Ca2+ to form a complex called the
thromboxane A2, further promoting platelet Xase complex, which catalyzes the activa-
aggregation. tion of factor X to Xa. It is important to rec-
The extrinsic clotting pathway begins ognize that these events occur at the vessel
with TF complexing with factor VII, caus- wallplatelet interface.
ing activation (VIIa). The TF-VIIa complex Thrombin (factor II) is central to coagu-
then activates factors IX and X to IXa and Xa lation by its action of cleavage and release of
fibrinopeptide A (FPA) from the -chain of E-selectin has been found up-regulated at
fibrinogen and fibrinopeptide B (FPB) from later time points.11
the -chain of fibrinogen. This causes fibrin Microparticles (MPs) are involved in
monomer polymerization and cross-linking, the initiation and amplification of throm-
stabilizing the thrombus and the initial plate- bosis. MPs are small (< 1m) phospholipid
let plug. Thrombin also activates factor XIII vesicles shed from platelets, leukocytes, and
to XIIIa, which catalyzes this cross-linking endothelial cells in a calcium-dependent
of fibrin as well as that of other plasma pro- fashion.12-14 MPs lack DNA and RNA but are
teins, such as fibronectin and 2-antitrypsin. protein rich, and subpopulations of MPs rich
In addition, factor XIIIa activates platelets as in TF and phosphatidylserine have been
well as factors V and VIII, further amplify- identified.15 Fusion of MPs with activated
ing thrombin production. platelets is another mechanism that results
The physiologic importance of the in decryption of TF and the initiation of
intrinsic pathway is not completely clear thrombosis.16 Several circulating markers of
and is probably not as physiologically impor- inflammation once thought to be soluble are
tant in the venous system as the extrinsic actually carried by MPs.17
system. Intrinsic pathway activation occurs In veins, CAMs and MPs interface in
with activation of factor XI to XIa, which promoting thrombosis and inflammation.
subsequently converts factor IX to IXa, pro- Both venous stasis and hypoxia result in the
moting formation of the Xase complex and up-regulation of P-selectin, which localizes
ultimately thrombin. Another mechanism prothrombotic MPs to the area of stasis and
by which this occurs in vitro is through the promotes DVT formation.18-20 The P-selec-
contact activation system, in which factor tin receptor, P-selectin glycoprotein ligand
XII (Hageman factor) is activated to XIIa 1(PSGL-1), is expressed on leukocytes and
when complexed to prekallikrein and high- platelets, as well as on their derived MPs.
molecular-weight kininogen (HMWK) on a Indeed, MPs coexpressing TF and leukocyte
negatively charged surface; factor XIIa then markers have been shown to accumulate in
activates factor XI to XIa. growing thrombi in a P-selectin:PSGL-1
dependent fashion.10 Further, P-selectin:
PSGL-1 interactions stimulate the produc-
Inflammation and Thrombosis tion of thrombogenic MPs from leukocytes,
In the early 1970s, the relationship between along with platelets and endothelial cells.21
thrombosis and inflammation was first sug- The importance of P-selectin:PSGL-1
gested.8 Inflammation increases TF, mem- to thrombosis also depends on the nature of
brane phospholipids, platelet reactivity, and the stimulus and the role of TF, which is nor-
fibrinogen, while decreasing thrombomodu- mally abundant in the outer portion of the
lin (TM) and inhibiting fibrinolysis.9 Coin- vessel wall. With significant vascular injury
cidently, cell adhesion molecules (CAM) and the exposure of vein wall TF, this TF is
are up-regulated on inflamed endothelium likely more important in the thrombogenic
and allow leukocyte transmigration. The process than the TF that is brought to the
selectins (P- and E-selectin) are integrally point of thrombogenesis by activated MPs,
involved in venous thrombosis.10 P-selectin in contrast to arterial thrombosis.22 Further-
is up-regulated in the vein wall as early as more, monocyte-derived MPs deliver TF to
6 hours after thrombus induction, while areas of injury and inflammation by binding
to P-selectin mobilized to the surface of acti- aminoglycans, including both heparan and
vated platelets and endothelial cells, result- dermatan sulfate, but its deficiency is not
ing in the generation of fibrin. associated with increased VTE risk.23
These activated factors may also catalyze the thelium; hence, the endothelium is one of
release of bradykinin from HMWK, which the pivotal regulators of homeostasis (Figure
further augments tPA secretion. Finally, 2.2). Under normal conditions, endothelial
APC has been found to proteolytically inac- cells maintain a vasodilatory and local fibri-
tivate plasminogen activator inhibitor type nolytic state in which coagulation, platelet
1 (PAI-1), an inhibitor of plasmin activators, adhesion, and activation are suppressed.
released by endothelial cells in the presence A nonthrombogenic endothelial surface is
of thrombin.27 maintained through a number of mecha-
In plasma, plasminogen activator nisms, including: (1) endothelial production
inhibitor (PAI-1) is the primary inhibitor of of TM and subsequent activation of protein
plasminogen activators and is stored in the C; (2) endothelial expression of heparan sul-
alpha-granules of quiescent platelets.25 It fate and dermatan sulfate, which accelerate
is secreted in an active form from liver and AT and heparin cofactor II activity; (3) con-
endothelial cells, and stabilized by binding stitutive expression of TFPI; and (4) local
to vitronectin (and inhibits thrombin in this production of tPA and uPA. In addition, the
form).28 Plasminogen activator inhibitor-1 production of nitric oxide (NO) and pros-
levels are elevated by hyperlipidemia, and tacyclin by the endothelium inhibits the
PAI-1 elevation appears to synergize with fac- adhesion and activation of leukocytes and
tor V Leiden genetic abnormalities, possibly produces vasodilation.29
contributing to pathological thrombosis. During states of endothelial distur-
bances, a prothrombotic and proinflamma-
tory state of vasoconstriction is supported by
Endothelium and Hemostasis the endothelial surface. Release of platelet-
Most of the thrombosis-thrombolysis pro- activating factor and endothelin-1 promotes
cesses occur in juxtaposition to the endo- vasoconstriction, while production of vWF,
TF, PAI-1, and factor V augment thrombo- is the primary fibrinolytic mechanism in the
sis.30 Indeed, vWF is expressed to a greater PA, including the arterioles.38 Although ani-
extent on the endothelium of veins as com- mal models can mimic the acute to chronic
pared with arteries, and tPA is less commonly PE pathophysiology, DVT characteristics
expressed in venous endothelium. Loss of that predispose to PE as compared to no PE
endothelium likely also contributes to the have not yet been forthcoming, but would be
vein wall fibrosis, as well as the predisposi- of significant clinical interest.
tion to recurrent venous thrombosis. Experi-
mental DVT was associated with decreased
expression of homeostatic endothelial genes Differential Risk
such as NO and TM, as compared with
controls, and correlated with loss of vWF
of DVT vs. PE
positive cell luminal staining.31 Other inves- Approximately 50% of patients with a DVT
tigators have found that prolonged venous may suffer a PE, although many of the
stasis is associated with decreased plasmin- PEs are asymptomatic and of small magni-
ogen activators, probably related to loss of tude.33 Conversely, pulmonary symptoms
endothelium.32 often point the physician to a suspected PE.
Defining those at risk for PE after DVT is
diagnosed, given similar levels of risk, and
Experimental anticoagulation, is important, and data are
Pulmonary Embolism lacking. Adequate anticoagulation certainly
reduces the risk, but some patients present
Pulmonary embolism is the fatal complica- with primary PE without a DVT source de-
tion of DVT, primarily related to acute right fined and may have symptoms related to fail-
heart failure and less so to hypoxemia.33 ure of local DVT resolution.
While the basic processes of thrombolysis in From a large epidemiological study,
all vascular beds are similar, several experi- several factors including age and corticoste-
mental studies have suggested that pulmo- roid use are independently associated with
nary artery (PA) injury is different from in the PE, but not DVT.2 From a large patient
inferior vena cava (IVC). In an experimen- registry, predictors of fatal PE in patients
tal study in rats using an in vivogenerated with incident DVT included older age (>75
thromboembolism, the early PA response is years), cancer, and immobility related to
associated with elevated monocyte chemo- neurological disease.39 Recent studies also
tactic protein-1 (MCP-1), early monocyte suggest that very high D-dimer levels may be
influx, and later intimal hyperplasia.34 Little predictive of PE over isolated DVT, with a
elevation in PA selectins is observed, in con- threshold of >4000 ng/mL.40 Similarly, those
trast to the IVC response.35 Interestingly, the patients whose initial clinical event was a PE
mechanical obstruction in the PA induces rather than simply a DVT are at higher risk
significant injury, and the resulting intimal of recurrent PE.41 In hospitalized patients,
hyperplasia is independent of the thrombus the postoperative state may also confer an
effect.36 Moreover, LMWH, but neither tPA increased risk of PE, but studies specifically
nor a gIIb/IIIa antagonist, was associated defining the DVT prior to PE are few.42 It
with attenuated intimal hyperplasia.37 Con- is likely that the differences are related to
sistent with what is observed in the IVC, uPA the balance of thrombolysis, thrombusvein
wall adherence, and thrombus composition The most common inherited thrombophil-
related to undefined genetic factors. ias are divided among the deficiencies of the
natural anticoagulants, the genetic factor
abnormalities, and a miscellaneous category
Factors That (Table 2.1). The suggested screening tests
Increase VTE Risk to evaluate for the most common thrombo-
philias are as follows: antithrombin, protein
Most clinical VTEs have a proximate cause, C/S; factor V Leiden; prothrombin 20210A;
including environmental risks and genetic homocysteine; factor VIII; and antiphospho-
predispositions, which may account for lipid antibody panel.
most of the VTEs that manifest clinically.
The most common risk factors for VTE are
prior DVT, malignancy, immobility, intra- Biomarkers for VTE
venous catheters, increased age, major sur-
gery, trauma, infections such as pneumonia While atherosclerotic disease has numerous
and urinary tract, and certain chemothera- biomarkers for clinical symptom risk, such
pies.2,41,43 Some medications such as oral as C-reactive protein (CRP), cholesterol sub-
contraceptives and hormonal replacement types, Lipoprotein a, and more, markers for
therapies also increase the risk of VTE. A VT occurrence have been less forthcoming.
particularly aggressive acquired hyperco- Part of this has been a lack of understand-
agulable state is antiphospholipid antibody ing of VT pathophysiology and the nature
syndrome. of VTE occurrence as compared with arte-
Although beyond the scope of this rial events. Venous thrombosis biomarkers
chapter, it is important to emphasize that would be most useful to determine primary
under specific circumstances, evaluating risk, risk for recurrence, and perhaps defin-
for an inherited thrombophilic state is war- ing therapy duration of VTE.
ranted.44 Clinical scenarios include early age Venous thrombosis biomarkers are
of VTE (<40 years), recurrent unprovoked several and are variably studied clinically45
VTE, thrombosis at unusual sites, and a (Table 2.2). Some of these, such as factor
strong family history of VTE. The inter- VIII, may be etiologic for VTE, whereas
section of environmental stressor (eg, post- others such as D-dimer and thrombin-
surgery) with an inherited thrombophilia antithrombin complex (TAT) may reflect
is common; a form of the 2-hit hypothesis. the thrombotic process.46 D-dimer is cur-
Natural
AnticoagulantDeficiencies FactorGeneticAbnormalities Miscellaneous
Protein S Factor IX
Antiphospholipid syndrome
rently used to define the likelihood of VTE P-selectin levels correlate highly in patients
probability in the clinical setting.47 Studies with malignancy.41 Inhibition of P-selectin
are well established, confirming its use to can confer protection against VTE as sug-
effectively rule out, but not rule in, a VTE. gested by nonhuman primate studies.54
Defining VTE recurrence risk has been Specific anti-P-selectin agents are being
improved with D-dimer measurement and developed for human trials and may offer
now is included in the AACP guidelines.48 prophylaxis benefit with anticoagulation
For example, longer duration of VTE oral risks. Although E-selectin plays a role in
anticoagulation therapy is suggested by sig- DVT, it has not been examined as a bio-
nificantly elevated D-dimer levels after 3 marker in humans.
months of oral coagulant therapy.49 Throm- Both probrain natriuretic peptide
bin-antithrombin complex, although not (BNP) and troponin are useful biomarkers
widely available, is also useful for acute VTE reflecting the physiological stress of a PE.
probability46 and prediction of VTE recur- For example, a low BNP is associated with
rence.50 Other measures, such as assessment good clinical outcome, whereas BNP 1000
of fragment 1 + 2, may further improve these ng/mL is associated with a 12-fold increased
predictions.51 risk of cardiac complications in combination
Soluble P-selectin is an etiological fac- with an abnormal ECHO.55 Similarly, those
tor in VTE,52 and studies suggest its utility patients with significantly elevated troponin-
as additive for VTE prediction.53 Elevated I have a 5.9-fold increased risk of in-hospital
mortality. These biomarkers may be helpful The risk of VTE was increased 2.0-fold for
in selecting those who would benefit from obesity, 1.5-fold for hypertension, 1.4 for
aggressive thrombus removal and close clini- diabetes, 1.2 for smoking, and 1.2 for hyper-
cal support.56 cholesterolemia.62 Interestingly, mean HDL
On the horizon is potentially more levels were lower in patients with VTE as
specific VTE biomarker risk assessment by compared with controls, although this find-
proteonomic evaluation.57,58 Limiting this ing is not universal.63 Reports are forthcom-
will be ease of use and rapidity of test results. ing that HMG-CoA reductase inhibitors
However, refinements in technology may (statins) may be associated with decreased
make this very useful in the future for assess- DVT incidence. Registry cohort patient data
ment of VTE risk, particularly if specific suggest a 22% risk reduction of symptomatic
enough to allow risk prediction in asymp- VTE in patients on a statin, as compared
tomatic patients. to matched patients not on a statin.64 More
impressive are results of a randomized clini-
cal trial of patients with normal cholesterol
but elevated hsCRP who were assigned to
Commonalities rosuvastatin or placebo. A 45% highly sig-
nificant reduction of incident VTE was
of Arterial and observed.65 In addition to the endothelial
venous disease: venous thrombosis and venous vivo is dependent upon microparticle
trauma. J Vasc Surg. 2007;46 suppl S:25S-53S. P-selectin glycoprotein ligand 1 and platelet
4. Mackman N, Tilley RE, Key NS. Role of the P-selectin. J Exp Med. 2003;197(11):1585-98.
extrinsic pathway of blood coagulation in 16. Osterud B. The role of platelets in decrypting
hemostasis and thrombosis. Arterioscl Throm monocyte tissue factor. Semin Hematol.
Vas. 2007;27(8):1687-93. 2001;38(4 suppl 12):2-5.
5. Furie B, Furie BC. Mechanisms of thrombus 17. Ahn ER, Lander G, Jy W, Bidot CJ,
formation. N Engl J Med. 2008;359(9):938-49. Jimenez JJ, Horstman LL, Ahn YS.
6. Mackman N. Tissue-specific hemostasis in Differences of soluble CD40L in sera and
mice. Arterioscl Throm Vas. 2005;25(11): plasma: implications on CD40L assay as
2273-81. a marker of thrombotic risk. Thromb Res.
7. Dahlback B. Blood coagulation. Lancet. 2004;114(2):143-8.
2000;355(9215):1627-32. 18. Myers DD, Hawley AE, Farris DM, Wrobleski
8. Stewart GJ. Neutrophils and deep venous SK, Thanaporn P, Schaub RG, et al. P-selectin
thrombosis. Haemostasis. 1993;23 suppl 1:1 and leukocyte microparticles are associated
27-40. with venous thrombogenesis. J Vasc Surg.
9. Esmon CT. Inflammation and thrombosis. 2003;38(5):1075-89.
J Thromb Haemost. 2003;1(7):1343-8. 19. Myers DD, Wakefield TW. Inflammation-
10. Celi A, Pellegrini G, Lorenzet R, De Blasi A, dependent thrombosis. Front Biosci.
Ready N, Furie BC, et al. P-selectin induces 2005;10:2750-7.
the expression of tissue factor on monocytes. 20. Polgar J, Matuskova J, Wagner DD. The
P Natl Acad Sci USA. 1994;91(19):8767-71. P-selectin, tissue factor, coagulation triad.
11. Myers D Jr, Farris D, Hawley A, Wrobleski J Thromb Haemost. 2005;3(8):1590-6.
S, Chapman A, Stoolman L, et al. Selectins 21. Andre P, Hartwell D, Hrachovinova I,
influence thrombosis in a mouse model of Saffaripour S, Wagner DD. Pro-coagulant
experimental deep venous thrombosis. J Surg state resulting from high levels of soluble
Res. 2002;108(2):212-21. P-selectin in blood. P Natl Acad Sci USA.
12. Gilbert GE, Sims PJ, Wiedmer T, Furie 2000;97(25):13835-40.
B, Furie BC, Shattil SJ. Platelet-derived 22. Day SM, Reeve JL, Pedersen B, Farris DM,
microparticles express high affinity Myers DD, Im M, et al. Macrovascular
receptors for factor VIII. J Biol Chem. thrombosis is driven by tissue factor derived
1991;266(26):17261-68. primarily from the blood vessel wall. Blood.
13. Mesri M, Altieri DC. Endothelial cell 2005;105(1):192-8.
activation by leukocyte microparticles. 23. Corral J, Aznar J, Gonzalez-Conejero R, Villa
J Immunol. 1998;161(8):4382-7. P, Minano A, Vaya A, et al. Homozygous
14. Sabatier F, Roux V, Anfosso F, Camoin L, deficiency of heparin cofactor II: relevance of
Sampol J, Dignat-George F. Interaction of P17 glutamate residue in serpins, relationship
endothelial microparticles with monocytic with conformational diseases, and role in
cells in vitro induces tissue factor- thrombosis. Circulation. 2004;110(10):1303-7.
dependent procoagulant activity. Blood. 24. Hassouna HI. Laboratory evaluation of
2002;99(11):3962-70. hemostatic disorders. Hematol Oncol Clin N.
15. Falati S, Liu Q, Gross P, Merrill-Skoloff G, 1993;7(6):1161-249.
Chou J, Vandendries E, et al. Accumulation 25. Sidelmann JJ, Gram J, Jespersen J, Kluft
of tissue factor into developing thrombi in C. Fibrin clot formation and lysis: basic
mechanisms. Semin Thromb Hemost. al. Viral IL-10 gene transfer decreases
2000;26(6):605-18. inflammation and cell adhesion molecule
26. Singh I, Burnand KG, Collins M, Luttun expression in a rat model of venous
A, Collen D, Boelhouwer B, et al. Failure thrombosis. J Immunol. 2000;164(4):2131-41.
of thrombus to resolve in urokinase-type 36. Zagorski J, Debelak J, Gellar M, Watts JA,
plasminogen activator gene-knockout mice: Kline JA. Chemokines accumulate in the
rescue by normal bone marrow-derived cells. lungs of rats with severe pulmonary embolism
Circulation. 2003;107(6):869-75. induced by polystyrene microspheres.
27. Esmon CT. The regulation of natural J Immunol. 2003;171(10):5529-36.
anticoagulant pathways. Science. 37. Rectenwald JE, Deatrick KB, Sukheepod P,
1987;235(4794):1348-52. Lynch EM, Moore AJ, Moaveni DM, et al.
28. Stoop AA, Lupu F, Pannekoek H. Experimental pulmonary embolism: effects
Colocalization of thrombin, PAI-1, and of the thrombus and attenuation of pulm-
vitronectin in the atherosclerotic vessel wall: onary artery injury by low-molecular-
a potential regulatory mechanism of thrombin weight heparin. J Vasc Surg. 2006;43(4):
activity by PAI-1/vitronectin complexes. 800-8.
Arterioscl Throm Vas. 2000;20(4):1143-49. 38. Bdeir K, Murciano JC, Tomaszewski J,
29. Becker MD, ORourke LM, Blackman WS, Koniaris L, Martinez J, Cines DB, et al.
Planck SR, Rosenbaum JT. Reduced leukocyte Urokinase mediates fibrinolysis in the
migration, but normal rolling and arrest, in pulmonary microvasculature. Blood.
interleukin-8 receptor homologue knockout 2000;96(5):1820-6.
mice. Invest Ophth Vis Sci. 2000;41(7):1812-7. 39. Laporte S, Mismetti P, Decousus H, Uresandi
30. Aird WC. Phenotypic heterogeneity of the F, Otero R, Lobo JL, et al. Clinical predictors
endothelium: I. Structure, function, and for fatal pulmonary embolism in 15,520
mechanisms. Circ Res. 2007;100(2):158-73. patients with venous thromboembolism:
31. Moaveni DK, Lynch EM, Luke C, Sood V, findings from the Registro Informatizado
Upchurch GR, Wakefield TW, et al. Vein de la Enfermedad TromboEmbolica
wall re-endothelialization after deep vein venosa (RIETE) registry. Circulation.
thrombosis is improved with low-molecular- 2008;117(13):1711-6.
weight heparin. J Vasc Surg. 2008;47(3):616-24. 40. Tick LW, Nijkeuter M, Kramer MH, Hovens
32. Stenberg B, Bylock A, Risberg B. Effect MM, Buller HR, Leebeek FW, et al. High
of venous stasis on vessel wall fibrinolysis. D-dimer levels increase the likelihood
Thromb Haemost. 1984;51(2):240-2. of pulmonary embolism. J Intern Med.
33. Tapson VF. Acute pulmonary embolism. 2008;264(2):195-200.
N Engl J Med. 2008;358(10):1037-52. 41. Zhu T, Martinez I, Emmerich J. Venous
34. Eagleton MJ, Henke PK, Luke CE, Hawley thromboembolism: risk factors for recurrence.
AE, Bedi A, Knipp BS, et al. Southern Arterioscl Throm Vas. 2009;29(3):298-310.
Association for Vascular Surgery William J. 42. Henke PK, Ferguson E, Varma M, Deatrick
von Leibig Award. Inflammation and intimal KB, Wakefield GT, Woodrum DT. Proximate
hyperplasia associated with experimental versus nonproximate risk factor associated
pulmonary embolism. J Vasc Surg. primary deep venous thrombosis: clinical
2002;36(3):581-8. spectrum and outcomes. J Vasc Surg. 2007;
35. Henke PK, DeBrunye LA, Strieter RM, 45(5):998-1003; discussion 1003-4; quiz
Bromberg JS, Prince M, Kadell AM, et 1005-7.
43. Heit JA. The epidemiology of venous Longo C, Henke PK, Guire KE, et al.
thromboembolism in the community. D-dimer, P-selectin, and microparticles:
Arterioscl Throm Vas. 2008;28(3):370-2. novel markers to predict deep venous
44. Middeldorp S, Levi M. Thrombophilia: thrombosis. A pilot study. Thromb Haemost.
an update. Semin Thromb Hemost. 2005;94(6):1312-7.
2007;33(6):563-72. 54. Myers DD Jr, Wrobleski SK, Longo C, Bedard
45. Pabinger I, Ay C. Biomarkers and venous PW, Kaila N, Shaw GD, et al. Resolution of
thromboembolism. Arterioscl Throm Vas. venous thrombosis using a novel oral small-
2009;29(3):332-6. molecule inhibitor of P-selectin (PSI-697)
46. Bozic M, Blinc A, Stegnar M. D-dimer, other without anticoagulation. Thromb Haemost.
markers of haemostasis activation and soluble 2007;97(3):400-7.
adhesion molecules in patients with different 55. Binder L, Pieske B, Olschewski M, Geibel A,
clinical probabilities of deep vein thrombosis. Klostermann B, Reiner C, et al. N-terminal
Thromb Res. 2003;108:107-14. pro-brain natriuretic peptide or troponin
47. Wells PS, Anderson DR, Rodger M, Forgie testing followed by echocardiography for risk
M, Kearon C, Dreyer J, et al. Evaluation stratification of acute pulmonary embolism.
of D-dimer in the diagnosis of suspected Circulation. 2 005;112(11):1573-9.
deep-vein thrombosis. N Engl J Med. 56. Piazza G, Goldhaber SZ. Acute pulmonary
2003;349(13):1227-35. embolism: part II: treatment and prophylaxis.
48. Geerts W. Antithrombotic and thrombolytic Circulation. 2006;114(3):e42-7.
therapy. Chest. 2008;133 (8th ed: ACCP 57. Hong CC, Kume T, Peterson RT. Role
Guidelines):381s-451s. of crosstalk between phosphatidylinositol
49. Palareti G, Cosmi B, Legnani C, Tosetto A, 3-kinase and extracellular signal-regulated
Brusi C, Iorio A, et al. D-dimer testing to kinase/mitogen-activated protein kinase
determine the duration of anticoagulation pathways in artery-vein specification. Circ Res.
therapy. N Engl J Med. 2006;355(17):1780-9. 2008;103(6):573-9.
50. Hron G, Kollars M, Binder BR, Eichinger 58. Ganesh SK, Sharma Y, Dayhoff J, Fales HM,
S, Kyrle PA. Identification of patients at low Van Eyk J, Kickler TS, Billings EM, Nabel
risk for recurrent venous thromboembolism EG. Detection of venous thromboembolism
by measuring thrombin generation. JAMA. by proteomic serum biomarkers. PLoS One.
2006;296(4):397-402. 2007;2(6):e544.
51. Poli D, Antonucci E, Ciuti G, Abbate R, 59. Prandoni P, Bilora F, Marchiori A, Bernardi E,
Prisco D. Combination of D-dimer, F1+2 and Petrobelli F, Lensing AW, et al. An association
residual vein obstruction as predictors of VTE between atherosclerosis and venous
recurrence in patients with first VTE episode thrombosis. N Engl J Med. 2003;348(15):
after OAT withdrawal. J Thromb Haemost. 1435-41.
2008;6(4):708-10. 60. Prandoni P, Ghirarduzzi A, Prins MH, Pengo
52. Hrachovinova I, Cambien B, Hafezi- V, Davidson BL, Sorensen H, et al. Venous
Moghadam A, Kappelmayer J, Camphausen thromboembolism and the risk of subsequent
RT, Widom A, et al. Interaction of P-selectin symptomatic atherosclerosis. J Thromb
and PSGL-1 generates microparticles that Haemost. 2006;4(9):1891-6.
correct hemostasis in a mouse model of 61. Glynn RJ, Ridker PM, Goldhaber SZ,
hemophilia A. Nat Med. 2003;9(8):1020-5. Buring JE. Effect of low-dose aspirin on the
53. Rectenwald JE, Myers DD Jr, Hawley AE, occurrence of venous thromboembolism:
Mark g. Davies
33
Primary
Antithrombin III deficiency Heparan cofactor II deficiency
Protein C deficiency
Protein S deficiency
Factor V Leiden
Fibrinolysis deficiencies (dysplasminogenemia, decreased plasminogen activator
or increased plasminogen activator inhibitors, dysfibrinogenemia)
Factor VII deficiency
pathogenesis of pulmonary embolism and Thrombi found in veins when blood flow is
are known as Virchows triad.18,19 reduced are composed predominantly of
fibrin and entrapped blood cells with rela-
tively few platelets and are often termed red
Lower Limb thrombi. The friable ends of these thrombi
Venous stasis is the most important feature are the source of the material that eventu-
predisposing to venous thrombosis. The ally becomes pulmonary emboli. Formation
venous sinuses of the veins are especially of venous thrombi is typically asymptomatic
vulnerable to stasis and thrombosis. Propa- and may involve the superficial or deep ve-
gation of the thrombus may then follow up- nous systems. Deep venous thrombi can
stream, or the process may spread retrograde. propagate into the superficial system.
HighRisk Age >60 years and major surgery, with no additional risk factors
Age 4060 years and major surgery with additional risk factors, patients with
myocardial infarction, and medical patients with risk factors
HighestRisk Age >40 years and major surgery, plus prior thromboembolism, malignancy, or
hypercoagulable state
Major lower extremity orthopedic surgery
Hip fracture
Stroke
Multiple trauma
Spinal cord injury
initially involved limbs and rethrombosis of screening method of choice for the nonin-
partially occluded or recanalized segments vasive assessment of blood flow in leg veins
in 31% of extremities. After completion of and of valve cusp movement.33 It also can dif-
a course of anticoagulation, the theoreti- ferentiate between acute and chronic venous
cal rate of recurrence is 0.9% per month.24 thrombosis. All major deep veins of the lower
Following venous thrombosis, the venous limb can be assessed, but it cannot exclude
lumen is most often reestablished. Recanali- the presence of thrombi in small veins of
zation occurs in an exponential manner over the calf. With experience, ultrasonography
6 months.25 Up to 40% of occluded segments is accurate, repeatable, and inexpensive.34
recanalized within 7 days, while 100% re- Positive tests indicating above-knee DVT
canalization occurs within 90 days.26 It ap- do not require further follow-up unless the
pears that venous valvular incompetence condition of the leg significantly worsens.
occurs after deep venous thrombosis. This Below-knee DVT requires repeat scanning
is supported by natural history studies that and is recommended to confirm that there
have demonstrated a correlation between is no antegrade propagation on therapy. Du-
segment thrombosis and subsequent valvu- plex ultrasonography sensitivity and specific-
lar incompetence.27 The development of re- ity are 94% and 96%, respectively. Overall
flux is highest within the first 6 to 12 months accuracy is considered to be 95%. Cross-
after development of a DVT.28 However, up sectional imaging with computed tomogra-
to 30% of segments will develop reflux with- phy (CT) or magnetic resonance imaging
out evidence of an initial thrombus, and this (MRI) is a reliable method of diagnosing
phenomenon appears to be related to persis- venous thrombosis, especially in the pelvic
tent proximal obstruction.29 The interval be- veins.35 Venography is recommended to vali-
tween the development of these defects and date and diagnose a DVT when duplex ul-
symptoms and signs of chronic venous insuf- trasonography has failed to rule out a DVT
ficiency may range out to 20 years.30 and should be performed prior to iliofemo-
The clinical diagnosis of DVT can be ral thrombolysis or thrombectomy. A recent
unreliable, as the disease mimics the pat- review of the evidence strongly supports the
terns of many other disorders.18,31 More than use of clinical prediction rules, particularly
half of the patients who present with the the Wells model, for establishing the pretest
classical symptoms of DVT do not have the probability of DVT or pulmonary embolism
disease.18 Clinical suspicion in combination in a patient before ordering more definitive
with diagnostic imaging should be used to testing.36 Fifteen studies support the conclu-
confirm the diagnosis.32 The proportion of sion that when a D-dimer assay is negative
patients with clinically suspected DVT, in and a clinical prediction rule suggests a low
whom the diagnosis is confirmed by objec- probability of DVT or pulmonary embolism,
tive testing, increases with the number of the negative predictive value is high enough
risk factors identified. Approximately half to justify foregoing imaging studies in many
of the patients with deep venous thrombo- patients. The evidence in 5 systematic re-
sis who develop pulmonary embolism have views regarding the use of D-dimer, in isola-
no symptoms of deep venous disease. This tion, is strong and demonstrates sensitivities
causes a delay in the administration of ap- of the enzyme-linked immunosorbent assay
propriate prophylactic and therapeutic mea- (ELISA) and quantitative rapid ELISA,
sures. Duplex ultrasonography is a popular pooled across studies, of approximately 95%.
embolism as a cause of death; the changing 6th ed. New York: McGraw-Hill; 1989. p2207-
mortality in hospitalized patients. JAMA. 18.
1986;255:2039-42. 18. Anderson FA Jr, Wheeler HB, Goldberg
9. Salzman EW, Hirsh J. The epidemiology, RJ, Hosmer DW, Patwardhan NA, Jovanovic
pathogenosis and natural history of venous B, et al. A population-based perspective of the
thromboembolism. In: Colman RW, Hirsh hospital incidence and case fatality
J, Marder VJ, Salzman EW, eds. Hemostasis rates of deep venous thrombosisand
and Thrombosis. Philadelphia: JP Lippincott; pulmonary embolism. The Worcester
1994. p1275-96. DVT Study. Arch Intern Med. 1991;151:
10. Clagett GP, Anderson FA Jr, Levine MN, 933-8.
Salzman EW, Wheeler HB. Prevention 19. Anderson FAJ, Wheeler HB. Physician
of venous thromboembolism. Chest. practices in the management of venous
1992;102:391S-407S. thromboembolism: a community-wide survey.
11. Hill S, Berry R. Subclavian vein thrombosis: a J Vasc Surg. 1992;15:707-14.
continuing challenge. Surgery. 1990;108:1-9. 20. Hughes E. Venous obstruction in the upper
12. Smith V, Hallet J. Subclavian vein thrombosis extremity (Paget Schroetter Syndrome). Int
during prolonged catheterization for Abst Surg. 1949;88:89-127.
parenteral nutrition: early management 21. Adams J, DeWeese J. Effort thrombosis of
and long term follow-up. South Med J. the axillary and subclavian veins. J Trauma.
1983;76:603-6. 1971;11:923-30.
13. Fraschini G, Jadeja J, Lawson M, Holmes 22. Lee S, Neilberger R. Subclavian vein
FA, Carrasco HC, Wallace S. Local infusion stenosis: complication of subclavian vein
of urokinase for the lysis of thrombosis catheterization for hemodialysis. Child
associated with permanent central venous Nephrol Urol. 1991;11:212-4.
catheters in cancer patients. J Clin Oncol. 23. Schillinger F, Schillinger D, Montagnac R,
1987;5:672-8. Milcent T. Post-catheterization venous stenosis
14. Cosgriff TM, Bishop DT, Hershgold EJ, in hemodialysis: comparative angiographic
Skolnick MH, Martin BA, Baty BJ, et al. study of 50 subclavian and 50 jugular accesses.
Familial antithrombin III deficiency: its Nephrology. 1992;13:127-33.
natural history, genetics, diagnosis and 24. Meissner MH, Caps MT, Bergelin RO,
treatment. Medicine. 1983;62:209-20. Manzo RA, Strandness DE. Propagation,
15. Comp PC, Esmon CT. Recurrent venous rethrombosis and new thrombus formation
thromboembolism in patients with a partial after acute deep venous thrombosis. J Vasc
deficiency of protein S. N Engl J Med. Surg. 1995;22:558-67.
1984;311:1525-8. 25. VanRamshorst B, VanBemmelen PS,
16. Ridkar PM, Hennekens CH, Lindpaintner Honeveld H, Faber JAJ, Eikelbloom BC.
K, Stampfer MJ, Eisenberg PR, Miletich JP. Thrombus regression in deep venous
Mutation of the gene coding for coagulation thrombosis. Quantification of spontaneous
factor V and the risk of myocardial infarction, thrombolysis with duplex scanning.
stroke and venous thrombosis in apparently Circulation. 1992;86:414-9.
healthy men. N Engl J Med. 1995;332:912-7. 26. Killewich LA, Bedford GR, Beach KW,
17. Tollefsen DM. Antithrombotic deficiency. In: Strandness DE. Spontaneous lysis of deep
Scriver CR, Beaudet AL, Sly WS, Valle D, venous thrombi: rate and outcome. J Vasc
eds. The metabolic basis of inherited disease. Surg. 1989;9:89-97.
27. Meissner MH, Manzo RA, Bergelin RO, of venous thromboembolism: American
Markel A, Strandness DE Jr. Deep venous College of Chest Physicians Evidence-Based
insufficiency: the relationship between Clinical Practice Guidelines (8th ed.). Chest.
lysis and subsequent reflux. J Vasc Surg. 2008;133: 381S-453S.
1993;18:596-608. 38. Kearon C, Kahn SR, Agnelli G, Goldhaber
28. Markel A, Manzo R, Bergelin RO, Strandness S, Raskob GE, Comerota AJ. Antithrombotic
DE Jr. Valvular reflux after deep vein therapy for venous thromboembolic disease:
thrombosis: incidence and time of occurance. American College of Chest Physicians
J Vasc Surg. 1992;15:377-87. Evidence-Based Clinical Practice Guidelines
29. Caps MT, Meissner MH, Manzo R, Bergelin (8th ed.). Chest. 2008;133(6 suppl):
RO, Strandness DE. Venous valvular reflux 454S-545S.
in veins not involved at the time of acute deep 39. Schulman S, Granqvist S, Holmstrm M,
vein thrombosis. J Vasc Surg. 1995;22:524-31. Carlsson A, Lindmarker P, Nicol P, et al. The
30. Bulger CM, Jacobs C, Patel NH. duration of oral anticoagulant therapy after a
Epidemiology of acute deep vein thrombosis. second epsiode of venous thromboembolism.
Tech Vasc Interv Radiol. 2004;7(2):50-4. N Engl J Med. 1997;336:393-8.
31. Browse N. Deep vein thrombosis. Diagnosis. 40. Hyers TM, Hull RD, Weg JG. Antithrombotic
Brit J Med 1969;4:676-8. therapy for venous thromboembolic disease.
32. Hull R, Hirsh J, Sackett DL, Stoddard G. Cost Chest. 1995;108(suppl 4):335-51.
effectiveness of clinical diagnosis, venography 41. Elman EE, Kahn SR. The post-thrombotic
and non-invasive testing in patients with syndrome after upper extremity deep venous
symptomatic deep vein thrombosis. N Engl J thrombosis in adults: a systematic review.
Med. 1986;314:823-8. Thromb Res. 2006;117(8):609-14.
33. Appleman PT, DeJong TE, Lampmann 42. Molina JE, Hunter DW, Dietz CA.
LE. Deep venous thrombosis of the leg: US Protocols for Paget-Schroetter syndrome
findings. Radiology. 1987;163:743-6. and late treatment of chronic subclavian
34. Cronan JJ. Venous thromboembolic vein obstruction. Ann Thorac Surg.
disease. The role of ultrasound. Radiology. 2009;87(2):416-22.
1993;186:619-30. 43. Melby SJ, Vedantham S, Narra VR, Paletta
35. Erdman WA, Jayson HT, Redman HC, GAJ, Khoo-Summers L, Driskill M, et al.
Miller GL, Parkey RW, Peshock RW. Deep Comprehensive surgical management of
venous thrombosis of the extremities: role the competitive athlete with effort thrombosis
of MR imaging in the diagnosis. Radiology. of the subclavian vein (Paget-Schroetter
1990;174:425-31. syndrome). J Vasc Surg. 2008;47(4):
36. Segal JB, Eng J, Tamariz LJ, Bass EB. Review 809-21.
of the evidence on diagnosis of deep venous 44. Meissner MH, Wakefield TW, Ascher E,
thrombosis and pulmonary embolism. Ann Caprini JA, Comerota AJ, Eklof B, et al. Acute
Fam Med. 2007;5(1):63-73. venous disease: venous thrombosis and venous
37. Geerts WH, Bergqvist D, Pineo GF, Heit JA, trauma. J Vasc Surg. 2007;46(suppl S):
Samama CM, Lassen MR, et al. Prevention 25S-53S.
Diagnostic Approach
to Venous Thromboembolism
l. Bernardo Menajovsky,
Patricia Hightower lambden, and ruth l. Bush
43
angiography. However, the invasiveness and factors, including acute infectious disease,
the risks of these modalities have led to a previous history of VTE, or advanced age
steady increase in the use of noninvasive (>75 years).6
or minimally invasive VTE testing. These Endothelial injury has been attributed
tests should optimally be used after clinical to both overt as well as subtle insults. Direct
examinations and risk assessments reveal re- trauma such as soft tissue injury or fracture,
sults highly suggestive of VTE. This chapter as well as intravenous infusion of irritating
discusses the clinical symptoms and signs substances, are more obvious factors that
indicative of possible VTE and provides an cause endothelial injury. Chemical chang-
overview of the clinical models and diagnos- es, ischemia, anoxia, and inflammation are
tic strategies available to assess for thrombo- more subtle factors that may cause endothe-
embolic disease. lial tissue damage.
In theory, during episodes of immobi-
lization or prolonged limb dependency, ve-
nous flow becomes sluggish, which may give
Pathophysiology rise to platelet and fibrin deposit, thereby
and Risk Stratification initiating venous thrombosis development.
However, results from some studies sug-
VTE can occur as an idiopathic syndrome gest that the associated risk may not be as
or may be caused by an underlying condition significant.3,7,8 In one small study, Gatt and
that predisposes a patient to thrombosis. In a colleagues observed no difference in the risk
retrospective study of 366 validated cases of of VTE in immobilized vs. mobile patients.7
VTE, approximately 48% were idiopathic In a retrospective cohort analysis, Gatt and
and 52% were secondary to an underlying colleagues evaluated 18 mobile and 8 im-
condition.4 Rudolf Ludwig Karl Virchow was mobile patients with a mean age of 85 for a
the first to identify 3 primary clinical factors duration of 10 years. The immobile patients
associated with a substantial risk of throm- were bedridden for a prolonged period (>3
bosis. Together, these factors are known as months). No difference in baseline char-
Virchows triad, and include (1) vessel wall acteristics, which included the assessment
damage due to inflammation or trauma; (2) of risk factors, was observed between the 2
changes in blood flow or volume due to im- groups. The incidence of VTE was similar
mobility, ischemia, and other conditions; between the immobile and mobile patient
and (3) hypercoagulable factors present in groups (13.9 and 15.8, respectively; P = 0.77).
the blood, including inherited and acquired While these results are not consistent with
coagulation disorders.5 previous studies9,10an inconsistency that is
The consideration of risk factors in due, in part, to a relatively small study popu-
the assessment and diagnostic evaluation of lationthey do highlight the importance of
patients with potential DVT is important; considering the added threat conferred by
however, it is also important to consider the each risk factor both independently and in
relative risk for each factor independently. combination with other risk factors.
Some reports suggest that hospitalized Similarly, an analysis that was conduct-
medical patients subjected to an extended ed using data from the American College
period of immobility may be at risk for VTE, of Surgeons National Trauma Data Bank
especially if presenting with additional risk evaluated the frequency of VTE follow-
ing trauma. The results demonstrated that any force that creates an imbalance between
incidence of VTE in these patients is also platelets, clotting factors, and the fibrinolytic
relatively low.11 Data were collected from 131 system. There are many variables that con-
trauma centers. The following 6 risk factors tribute to hypercoagulability; however, the
were found to be associated with VTE: aged most common causes are hematologic disor-
40 years, lower extremity fracture with Ab- ders, malignancy, trauma, estrogen therapy,
breviated Injury Score (AIS) 3, head injury and surgical events.
with AIS 3, ventilator delays >3, venous in-
jury, and the performance of a major opera-
tive procedure. Of the 450,375 patients who
experienced trauma from 1994 to 2001, a Clinical Diagnosis
total of 1602 experienced DVT, resulting in
an incidence of 0.36%.11 of Venous
In contrast to the lower relative risk
conferred by immobilization and trauma,
Thromboembolism
a much stronger association exists between The vast majority of thrombi that originate
VTE and cancer.12 In one study, 26% of pa- within the lower extremities resolve sponta-
tients presenting with bilateral DVT were di- neously; however, propagation of fibrin net-
agnosed with cancer after the occurrence of works may lead to venous thrombosis. Such
DVT, and metastasis had occurred in 70% of thromboembolic events are classified based
these patients.12 Other reports cite as much upon the system involved (deep or superfi-
as a 2-fold increase in DVT risk among pa- cial) as well as by location. Classification of
tients with cancer and suggest that the risk thromboembolic events with regard to loca-
of recurrent thromboembolism is as much as tion refers to the thrombus in relation to the
3.5 times higher in patients with malignancy popliteal vein (distal or proximal).
vs. cancer-free patients.1,13,14 The deep venous system possesses the
Finally, hypercoagulabilityespecially ability to encompass a rather large amount
hypercoagulability due to a genetic predispo- of systemic flow. This heightened volume
sition, acquired syndromes, and certain med- capacity, as well as ease of effective collat-
ications, such as oral contraceptivescan eralization, allows many thromboembolic
increase the risk of thrombosis.15,16 Hereditary events to remain asymptomatic and undiag-
risk factors include factor V Leiden mutation; nosed. Unfortunately, the insidious nature
prothrombin G 20210A gene mutation; and of DVT allows for a great deal of vessel de-
deficiencies of protein C, protein S, and an- struction prior to diagnosis. In many cases,
tithrombin III.17 Hyperhomocysteinemia and the severity of symptomatology ineffectively
elevated levels of factors XIII and V, which represents the extent of the disease.
may be hereditary and/or acquired, are also It is important to note that 50% of pa-
risk factors.17 Additionally, ABO blood type tients who have VTE do not present with
is another VTE risk factor that was recently any symptoms.19 Clinical presentation of
noted in cancer patients and is believed to DVT most frequently involves unilateral
be associated with hypercoagulability result- lower extremity edema; however, iliofemoral
ing from influence on the levels of von Wil- obstruction may elicit bilateral edema. In-
lebrand factor (vWF) and factor VIII.18 creased intravascular blood volume, elevat-
Hypercogulability can be potentiated by ed hydrostatic pressure, and seepage of fluid
across the capillary membrane all produce DVT include immobility, dehydration, ma-
DVT-related edema. Increased tissue turgor lignancy, blood dyscrasias, estrogen therapy,
and elevated temperature of the affected postoperative status, trauma, pregnancy/
area are also secondary to edema and dila- postpartum state, congestive heart failure,
tion of superficial vessels. Mottling or cyano- and obesity.
sis of the extremity is often related to venous Wells and colleagues developed the first
stasis, increased oxygen consumption, and clinical model for the diagnosis of patients
reduction of hemoglobin. presenting with suspected DVT.20-26 This
Most individuals suffering from DVT model includes a thorough clinical exami-
report limb pain, described as aching or nation and identification of any risk factors
throbbing in nature, which is aggravated by that predispose patients to have increased
ambulation. Although considered somewhat risk of thrombosis. In accordance with this
unreliable, there are 2 techniques utilized to model, patients are first divided into 3 risk
assess limb tenderness. Homans sign entails categories (low, moderate, or high) and are
the dorsiflexion of the foot of the affected ex- further assessed through ultrasonography.23
tremity to elicit calf pain. With Lowenbergs Patients who are stratified to the high-risk
sign, calf pain is elicited by cuff inflation category, or who have abnormal ultrasonog-
over the affected limb. As mentioned previ- raphy results, are further assessed through
ously, each technique possesses a low pretest venography.23 Clinical practice guidelines
probability. for the diagnosis of DVT from the American
Long-term disability related to venous Thoracic Society concur with this strategy,
thrombosis involves the destruction of vessel recommending the use of venography as a
endothelium, which renders venous valves follow-up to inconclusive compression ul-
incompetent. Such valvular incompetence trasonography results, and the use of serial
allows for reversal of venous flow, stasis, and ultrasonography or impedance plethysmog-
recurrent thrombosis. Venous thrombosis raphy in patients with normal compression
most frequently involves the vessels within ultrasonography results.27
the calf; however, the proximal veins of the For patients presenting with PE, short-
lower extremities and the pelvic vessels are ness of breath, with or without leg pain, may
the primary etiology of pulmonary embolus, be the first symptom; however, a number of
the most common life-threatening compli- specific criteria allow for a more accurate
cation of deep venous thrombosis. Because diagnosis. Similar to the clinical model for
of the correlation of life-threatening pulmo- diagnosis of DVT, Wells and colleagues also
nary embolus with the presence of DVT, it defined a clinical algorithm for the diagno-
is of extreme importance to assess for pul- sis of PE (Figure 4.1),26 which when used
monary embolus in conjunction with DVT in conjunction with D-dimer testing, safely
evaluation. reduces the need for expensive imaging di-
Initial evaluation for DVT must include agnostics.26 This model for assessment of PE
screening for risk factors. There are many uses a point system for calculating the low,
common risk factors; however, any event moderate, or high pretest probability of PE.
that leads to the elements included within Points are assigned based on clinical symp-
Virchows triad (venous stasis, endothelial toms of DVT, including heart rate of >100
injury, and hypercoagulability) may lead beats per minute, immobilization for 3
to DVT. The most common risk factors for days or recent surgery in the past 4 weeks,
Figure 4.4 99m Tc macroaggregated albumin perfusion scan showing multiple large segmental defects.
segmental emboli.39 Although the clinical sensitivity of 85% for the detection of lobar
significance of these subsegmental emboli and segmental emboli, with less reliability
has not been established, a negative spiral in detecting peripheral subsegmental em-
CT cannot safely rule out thrombosis and boli (1 in 5).35 These results are consistent
must be confirmed with pulmonary angiog- with those of the PIOPED study, in which
raphy.39 In addition, a low sensitivity rate of interobserver agreement was 98% for lobar
70% has been reported in a previous study, PE, 90% for segmental PE, and only 66%
providing additional evidence that confirma- for subsegmental emboli.42 However, there
tory pulmonary angiography is required in has been some debate surrounding the true
patients with negative spiral CT findings.40 clinical importance of these peripheral em-
A 1-year follow-up study in patients with a boli.35 Some investigators posit that they are
normal spiral CT scan demonstrated a low caused by isolated calf vein thrombi and
2% rate of clinical PE, suggesting that ad- do not require treatment, whereas others
ditional data will be required to character- believe that they are a precursor to larger
ize the safety of a negative spiral CT without emboli.43,44
other confirmatory diagnostics.4
Pulmonary Angiography
Pulmonary angiography has been consid- Newer Diagnostic
ered a gold standard test for PE.27 It is an
invasive test and in some settings used to Techniques
detect PE in patients with indeterminate
V/Q scans. However, it is not necessary for Indirect CT Venography
diagnosis of PE in the acute setting when
the perfusion scan is normal.27 One study Indirect CT venography has been investi-
of pulmonary angiography demonstrated a gated as an adjunct to the pulmonary angi-
ography conducted during the differential
diagnosis of PE.45 This technique results
in a high rate of detection of DVTthe
underlying cause of many subsequent PE
eventsand requires no additional contrast
material.46,47
Magnetic Resonance
Direct Thrombus Imaging
MRDTI is a novel technique that has dem-
onstrated accuracy and reproducibility for
DVT diagnosis in limited studies.48 It detects
the presence of methemoglobin in clots, al-
Figure 4.5 Spiral computerized tomographic lowing visualization of thrombus without
pulmonary arteriography has demonstrated a high rate using intravenous contrast and making it
of sensitivity and specificity in detecting pulmonary
useful for detection of subacute thrombosis.49
embolism.
The major advantages of MRDTI over con-
ventional modalities include (1) early data fers limited specificity and cannot be solely
suggesting that it is highly accurate for the used to exclude a diagnosis of DVT. Such
detection of both DVT and PE, providing a results have been demonstrated in emer-
single imaging modality for the detection of gency department patients,51 patients who
VTE;49,50 (2) direct visualization of throm- are elderly, and those hospitalized for >3
bus, avoiding the pitfalls of conventional days.52 Another important issue is that some
techniques that have either identified throm- institutions do not have access to immedi-
bus as a filling defect or in terms of surro- ate D-dimer results. Nevertheless, D-dimer
gates; and (3) simultaneous imaging of the testing, especially the most specific enzyme-
legs and chest, allowing a comprehensive linked immunosorbent assay, still offers an
assessment of thrombus load, minimizing attractive bedside assay that is sensitive for
the importance of overlooked subsegmental DVT diagnosis if not specific under certain
PE, and potentially facilitating more titrated circumstances, and will still result in a re-
treatment. The safety of withholding treat- duction in the need for imaging diagnostics
ment in suspected DVT and PE on the basis when DVT can be diagnosed and treated
of negative MRDTI alone is being evaluated earlier.53 A positive D-dimer is not useful for
in ongoing outcome studies.50 Additionally, necessarily ruling in the diagnosis of VTE;
because it has proven useful in identifying rather, the value is found in the negative test,
complicated plaque in the carotid arteries in which, when it coincides with ultrasonogra-
the setting of transient and permanent cere- phy, can rule out the diagnosis.25
bral ischemia, MRDTI offers promise as a While measurement of D-dimers is
technique that is capable of detecting high- a recent addition to the diagnostic strategy
risk vessel wall disease prior to significant or of PE and has been shown to be a valu-
permanent end-organ damage.49 As costs for able tool with excellent sensitivity, there
this type of imaging decrease and institutions have been rare reports of patients with PE
gain wider access to the technology, it should but negative D-dimer tests. One investiga-
offer an attractive alternative to the invasive tion at an academic health center indicated
use of contrast venography with application that D-dimer measurement was of limited
in a wide range of vascular disease settings. utility in patients with suspected PE and
nondiagnostic lung scans or negative spiral
(helical) CT results.54 Another study of 150
Bedside D-dimer Assays patients admitted to hospital for PE who un-
derwent D-dimer measurement compared
A number of D-dimer assays have been results in patients with negative D-dimers
evaluated as diagnostic markers for VTE vs. raised D-dimers. The sensitivity of raised
and are considered to be inexpensive as well D-dimers for PE was high (96%), but the
as timely25; the predictive value of the assay finding of chest pain was statistically greater
depends on several attributes, one being the in the group with negative D-dimers (P =
prevalence of VTE in the population being 0.01). In these negative D-dimer cases, the
tested. Ruiz-Gimnez and colleagues found emboli were all distal (P = 0.0003), and the
that the VIDAS and ELISA D-dimers as a diagnostic value of ultrasound investigations
first diagnostic tool for the exclusion of DVT (echocardiography, ultrasonography of lower
are suitable approaches.22 To the contrary, limb veins) was less than in patients with
some data suggest that D-dimer testing of- higher D-dimers (P <0.0001). The authors
suggested that measurement of D-dimers by 2. Horlander KT, Mannino DM, Leeper KV.
the ELISA method may be nondiagnostic Pulmonary embolism mortality in the United
in distal PE, perhaps because of the less ex- States, 1979-1998: an analysis using multiple-
tensive thromboembolic process. They con- cause mortality data. Arch Intern Med.
cluded that in cases with negative D-dimers, 2003;163:1711-7.
a strong clinical suspicion of PE should sig- 3. Kroegel C, Reissig A. Principle mechanisms
nal a need for further investigation.55 underlying venous thromboembolism:
epidemiology, risk factors, pathophysiology
and pathogenesis. Respiration. 2003;70:
Conclusion 7-30.
4. Cushman M, Tsai AW, White RH, Heckbert
Clinical evidence indicates that patients SR, Rosamond WD, Enright P, et al. Deep
who are at moderate to high risk for develop- vein thrombosis and pulmonary embolism in
ing VTE include those with a history of can- two cohorts: the longitudinal investigation
cer, prior thrombosis, acquired syndromes, of thromboembolism etiology. Am J Med.
or genetic disorders that predispose them 2004;117:19-25.
to a hypercoagulative state, among others. 5. Nielsen H. Pathophysiology of venous
Among these patients, VTE should be high- thromboembolism. Semin Thromb Hemost.
ly suspected with the presentation of classic 1991;17 (suppl 3):250-3.
symptoms. Such cases require rapid assess- 6. Alikhan R, Cohen AT, Combe S, Samama
ment and accurate diagnosis to prevent the MM, Desjardins L, Eldor A, et al;
progression of DVT, long-term morbidity MEDENOX Study. Risk factors for venous
due to postthrombotic syndrome, and/or the thromboembolism in hospitalized patients
occurrence of potentially fatal PE. with acute medical illness: analysis of the
However, clinical presentation for 50% MEDENOX Study. Arch Intern Med.
of patients with VTE is often nonspecific, 2004;10;164:963-8.
and can be confused with a variety of other 7. Gatt M, Paltiel O, Bursztyn M. Is prolonged
conditions, including heart failure, cellulitis, immobilization a risk factor for symptomatic
hematoma, or edema due to an unrelated venous thromboembolism in elderly
condition. bedridden patients? Thromb Haemost.
New additions to the diagnostic battery 2004;91:538-43.
and increased awareness of risk stratification 8. Rosenow E. Concise review for primary-care
paradigms have the potential to allow more physicians. Mayo Clin Proc. 1995;70:45-9.
accurate identification of VTE in some pa- 9. Duggan C, Marriott K, Edwards R, Cuzick J.
tients, thereby greatly reducing incidence of Inherited and acquired risk factors for venous
mortality and morbidity. thromboembolic disease among women taking
tamoxifen to prevent breast cancer. J Clin
Oncol. 2003;21:3588-93.
references 10. Samama MM, Dahl OE, Qunlan DJ,
1. Prandoni P, Lensing AW, Cogo A, Cuppini Mismetti P, Rosencher N. Quantification of
S, Villalta S, Carta M, et al. The long- risk factors for venous thromboembolism: a
term clinical course of acute deep venous preliminary study for the development of a risk
thrombosis. Ann Intern Med. 1996;125: assessment tool. Haematologica. 2003;88:
1-7. 1410-21.
11. Knudson MM, Ikossi DG, Khaw L, Morabito of deep venous thrombosis. Fallibility of
D, Speetzen LS. Thromboembolism after clinical symptoms and signs. Arch Surg.
trauma: an analysis of 1602 episodes from 1976;111:34-6.
the American College of Surgeons National 21. Anderson DR, Wells PS, Stiell I, MacLeod
Trauma Data Bank. Ann Surg. 2004;240:490- B, Simms M, Gray L, et al. Thrombosis in
8. the emergency department: use of a clinical
12. Bura A, Cailleux N, Bienvenu B, Lger diagnosis model to safely avoid the need for
P, Bissery A, Boccalon H, Fiessinger JN, urgent radiological investigation. Arch Intern
Levesque H, Emmerich J. Incidence and Med. 1999;159:477-82.
prognosis of cancer associated with bilateral 22. Ruiz-Gimnez N, Friera A, Artieda P,
venous thrombosis: a prospective study of 103 Caballero P, Sanchez Molin P, Morales M, et
patients. J Thromb Haemost. 2004;2:441-4. al. for the Thromboembolic Disease Group.
13. Hansson PO, Sorbo J, Eriksson H. Recurrent Rapid D-dimer test combined a clinical model
venous thromboembolism after deep vein for deep vein thrombosis. Validation with
thrombosis: incidence and risk factors. Arch ultrasonography and clinical follow-up in 383
Intern Med. 2000;160:769-74. patients. Thromb Haemost. 2004;91:1237-46.
14. Schulman S, Lindmarker P, Johnsson H. A 23. Wells PS, Hirsh J, Anderson DR, Lensing AW,
comparison of six weeks with six months of Foster G, Kearon C, et al. Accuracy of clinical
oral anticoagulant therapy after a first episode assessment of deep-vein thrombosis. Lancet.
of venous thromboembolism. N Engl J Med. 1995;345:1326-30.
1995;332:1661-5. 24. Wells PS, Ginsberg JS, Anderson DR, Kearon
15. Allaart CF, Poort SR, Rosendaal FR, Reitsma C, Gent M, Turpie AG, et al. Use of a clinical
PH, Bertina RM, Brit E. Increased risk of model for safe management of patients with
venous thrombosis in carriers of hereditary suspected pulmonary embolism. Ann Intern
protein C deficiency defect. Lancet. Med. 1998;129:997-1005.
1993;341:134-8. 25. Wells PS, Anderson DR, Ginsberg J.
16. Hedenmalm K, Samuelsson E, Spigset Assessment of deep vein thrombosis or
O. Pulmonary embolism associated with pulmonary embolism by the combined use of
combined oral contraceptives: reporting clinical model and noninvasive diagnostic tests.
incidences and potential risk factors for Semin Thromb Hemost. 2000;26:643-56.
a fatal outcome. Acta Obstet Gyn Scand. 26. Wells PS, Anderson DR, Rodger M, Stiell
2004;83:576-85. I, Dreyer JF, Barnes D, et al. Excluding
17. Mazza JJ. Hypercoagulability and venous pulmonary embolism at the bedside without
thromboembolism: a review. WMJ. diagnostic imaging: management of patients
2004;103:41-9. with suspected pulmonary embolism
18. Streiff MB, Segal J, Grossman SA, Kickler presenting to the emergency department by
TS, Weir EG. ABO blood group is a potent using a simple clinical model and D-dimer.
risk factor for venous thromboembolism in Ann Intern Med. 2001;135:98-107.
patients with malignant gliomas. Cancer. 27. Tapson VF, Carroll BA, Davidson BL, Elliott
2004;100:1717-23. CG, Fedullo PF, Hales CA, et al. for the
19. Gathof BS, Picker SM, Rojo J. Epidemiology, American Thoracic Society. The diagnostic
etiology and diagnosis of venous thrombosis. approach to acute venous thromboembolism.
Eur J Med Res. 2004;9:95-103. Clinical practice guideline. Am J Resp Crit
20. Cranley JJ, Canos AJ, Sull WJ. The diagnosis Care Med. 1999;160:1043-66.
28. Stein PD, Terrin ML, Hales CA, Palevsky 36. The PIOPED Investigators. Value of the
HI, Saltzman HA, Thompson BT, et al. ventilation/perfusion scan in acute pulmonary
Clinical, laboratory, roentgenographic, and embolism. Results of the Prospective
electrocardiographic findings in patients with Investigation of Pulmonary Embolism
acute pulmonary embolism and no pre- Diagnosis (PIOPED). JAMA. 1990;263:
existing cardiac or pulmonary disease. Chest. 2753-9.
1991;100:598-603. 37. Kember PG, Euinton HA, Morcos SK.
29. Baker WF Jr. Diagnosis of deep venous Clinicians interpretation of the indeterminate
thrombosis and pulmonary embolism. Med ventilation-perfusion scan report. Br J Radiol.
Clin North Am. 1998;82(3):459-76. 1997;70:1109-11.
30. Elias A, Cadne A, Elias M, Puget J, 38. Mayo JR, Remy-Jardin M, Mller NL, Remy
Tricoire JL, Colin C, et al. Extended lower J, Worsley DF, Hossein-Foucher C, et al.
limb venous ultrasound for the diagnosis Pulmonary embolism: prospective comparison
of proximal and distal vein thrombosis of spiral CT with ventilation-perfusion
in asymptomatic patients after total hip scintigraphy. R adiology. 1997;205:447-52.
replacement. Eur J Vasc Endovasc Surg. 39. Remy-Jardin M, Remy J, Deschildre F, Artaud
2004;27:438-44. D, Beregi JP, Hossein-Foucher C, et al.
31. Stevens SM, Elliott CG, Chan KJ, Egger MJ, Diagnosis of pulmonary embolism with spiral
Ahmed KM. Withholding anticoagulation CT: comparison with pulmonary angiography
after a negative result on duplex and scintigraphy. Radiology. 1996;200:
ultrasonography for suspected symptomatic 699-706.
deep venous thrombosis. Ann Intern Med. 40. Perrier A, Howarth N, Didier D, Loubeyre P,
2004;140:985-91. Unger PF, de Moerloose P, et al. Performance
32. Costantini M, Bossone E, Renna R, of helical computed tomography in unselected
Sticchi G, Licci E, De Fabrizio G, et al. outpatients with suspected pulmonary
Electrocardiographic features in critical embolism. Ann Intern Med. 2001;135:88-97.
pulmonary embolism. Results from baseline 41. Tillie-Leblond I, Mastora I, Radenne F,
and continuous electrocardiographic Paillard S, Tonnel AB, Remy J, et al. Risk of
monitoring [in Italian]. Ital Heart J. pulmonary embolism after a negative spiral
2004;5:214-6. CT angiogram in patients with pulmonary
33. Rodger M, Makropoulos D, Turek M, disease: 1-year clinical follow-up study.
Quevillon J, Raymond F, Rasuli P, et al. Radiology. 2002;223:461-7.
Diagnostic value of the electrocardiogram 42. Stein PD, Athanasoulis C, Alavi A,
in suspected pulmonary embolism. Am J Greenspan RH, Hales CA, Saltzman HA, et
Cardiol. 2000;86:807-9. al. Complications and validity of pulmonary
34. Iles S, Le Heron CJ, Davies G, Turner JG, angiography in acute pulmonary embolism.
Beckert LE. ECG score predicts those with Circulation. 1992;85:462-9.
the greatest percentage of perfusion defects 43. Dalen JE. When can treatment be withheld in
due to acute pulmonary thromboembolic patients with suspected pulmonary embolism.
disease. Chest. 2004;125:1651-6. Arch Intern Med. 1993;153:1415-8.
35. Gupta A, Frazer CK, Ferguson JM, Kumar 44. Moser KM. Venous thromboembolism. Am
AB, Davis SJ, Fallon MJ, et al. Acute Rev Respir Dis. 1990;141:235-49.
pulmonary embolism: diagnosis with MR 45. Richman PB, Wood J, Kasper DM, Collins
angiography. Radiology. 1999;210:353-9. JM, Petri RW, Field AG, et al. Contribution of
5
Treatment of Acute
Deep Venous Thrombosis
What's new in the ACCP guidelines?
Anthony J. Comerota
57
the affected leg, and the significant benefit symptoms compared to patients treated with
of continued use of 30- to 40-mm Hg com- anticoagulation alone. Other, nonrandom-
pression stockings. ized series also reported favorable outcomes
of contemporary venous thrombectomy.
Long-term observational results from 10 re-
Treatment of Iliofemoral ports with the mean of 41 months of follow-up
Venous Thrombosis demonstrated a 76% patency, with 8 reports
The writing committee for the 2008 ACCP demonstrating functional venous valves in
guidelines recognized the excess morbid- the femoropopliteal segment in 63%.11
ity of this particular distribution of disease. Since venous thrombectomy is infre-
It is important to understand the anatomy quently performed, the committee sug-
of lower extremity venous drainage, which gested that catheter-directed thrombolysis
functionally resembles a funnel, with distal is usually preferable to operative venous
veins draining into larger but progressively thrombectomy (grade 2C).1
fewer veins as blood moves cephalad. The
common femoral and iliac veins represent
the spout of the funnel, which is the single Catheter-Directed Thrombolysis
common channel of lower extremity venous The recommendation for catheter-directed
drainage. If this channel is obstructed, it will thrombolysis (CDT) for acute iliofemoral
affect the entire leg, with adverse functional DVT in patients with a low risk of bleeding,
consequences on all distal veins. symptoms <14 days, good functional status,
The greatest change in guidelines is and a life expectancy 1 year is also pro-
the recommendation for consideration of posed to reduce acute symptoms and post-
a strategy of thrombus removal in patients thrombotic morbidity (grade 2B).
with iliofemoral DVT. This is a reversal of Rationale A small randomized trial of
the statements from guidelines published in catheter-directed lytic therapy vs. anticoagu-
2004.12 lation demonstrated significantly better pa-
tency and preservation of valve function in
patients treated with CDT vs. anticoagula-
Venous Thrombectomy tion.16 Large single-center series and multi-
The 2008 ACCP guidelines recommend center venous registries demonstrate an 80%
considering venous thrombectomy for acute to 90% success rate, with progressively lower
iliofemoral DVT in patients with symptoms bleeding complications over time.17-19 A case-
for <7 days, good functional status, and a life controlled cohort study, which followed the
expectancy 1 year (grade 2B). National Venous Registry, demonstrated sig-
Rationale This is based upon level 1 data nificantly improved quality of life (QOL) in
emanating from a large randomized study patients with iliofemoral DVT treated with
by Plate et al.13-15 Patients were followed up CDT compared to those treated with antico-
and reported at 6 months, 5 years, and 10 agulation.20 The improved QOL was directly
years following randomization to venous related to lytic success.
thrombectomy or anticoagulation. Patients The committee suggests correction of
randomized to thrombectomy demonstrated underlying venous lesions using balloon an-
improved patency, lower venous pressures, gioplasty and stents (grade 2C). While there
less leg swelling, and fewer postthrombotic are no objective data supporting this state-
iliofemoral venous thrombosis. A randomised 24. Brandjes DP, Heijboer H, Buller HR, de
clinical trial. Eur J Vasc Endovasc Surg. RM, Jagt H, ten Cate JW. Acenocoumarol
2002;24(3):209-14. and heparin compared with acenocoumarol
17. Bjarnason H, Kruse JR, Asinger DA, alone in the initial treatment of proximal-vein
Nazarian GK, Dietz CA Jr, Caldwell MD, thrombosis. N Engl J Med. 1992;327(21):
et al. Iliofemoral deep venous thrombosis: 1485-9.
safety and efficacy outcome during 5 years of 25. Hull RD, Raskob GE, Rosenbloom D, Panju
catheter-directed thrombolytic therapy. J Vasc AA, Brill-Edwards P, Ginsberg JS, et al.
Interv Radiol. 1997;8(3):405-18. Heparin for 5 days as compared with 10 days
18. Comerota AJ, Kagan SA. Catheter-directed in the initial treatment of proximal venous
thrombolysis for the treatment of acute thrombosis. N Engl J Med. 1990;322(18):
iliofemoral deep venous thrombosis. 1260-4.
Phlebology. 2000;15:149-55. 26. Levine MN, Hirsh J, Gent M, Turpie AG,
19. Mewissen MW, Seabrook GR, Meissner Weitz J, Ginsberg J, et al. Optimal duration
MH, Cynamon J, Labropoulos N, Haughton of oral anticoagulant therapy: a randomized
SH. Catheter-directed thrombolysis for lower trial comparing four weeks with three
extremity deep venous thrombosis: report of months of warfarin in patients with proximal
a national multicenter registry. Radiology. deep vein thrombosis. Thromb Haemost.
1999;211(1):39-49. 1995;74(2):606-11.
20. Comerota AJ, Throm RC, Mathias SD, 27. Schulman S, Rhedin AS, Lindmarker P,
Haughton S, Mewissen M. Catheter-directed Carlsson A, Larfars G, Nicol P, et al. A
thrombolysis for iliofemoral deep venous comparison of six weeks with six months of
thrombosis improves health-related oral anticoagulant therapy after a first episode
quality of life. J Vasc Surg. 2000;32(1): of venous thromboembolism. Duration of
130-7. Anticoagulation Trial Study Group. N Engl J
21. Martinez J, Comerota AJ, Kazanjian S, Med. 1995;332(25):1661-5.
DiSalle RS, Sepanski DM, Assi Z. The 28. Kearon C, Gent M, Hirsh J, Weitz J, Kovacs
quantitative benefit of isolated, segmental, MJ, Anderson DR, et al. A comparison of
pharmacomechanical thrombolysis three months of anticoagulation with extended
for iliofemoral DVT. J Vasc Surg. anticoagulation for a first episode of idiopathic
2008;48(6):1532-7. venous thromboembolism. N Engl J Med.
22. Lin PH, Zhou W, Dardik A, Mussa F, 1999;340(12):901-7.
Kougias P, Hedayati N, et al. Catheter-direct 29. Ridker PM, Goldhaber SZ, Danielson E,
thrombolysis versus pharmacomechanical Rosenberg Y, Eby CS, Deitcher SR, et al.
thrombectomy for treatment of symptomatic PREVENT Investigators. Long-term, low-
lower extremity deep venous thrombosis. Am J intensity warfarin therapy for the prevention
Surg. 2006;192(6):782-8. of recurrent venous thromboembolism.
23. Parikh S, Motarjeme A, McNamara T, N Engl J Med. 2003;348(15):1425-34.
Raabe R, Hagspiel K, Benenati JF, et al. 30. Kearon C, Ginsberg JS, Kovacs MJ,
Ultrasound-accelerated thrombolysis for the Anderson DR, Wells P, Julian JA, et al.
treatment of deep vein thrombosis: initial Extended Low-Intensity Anticoagulation
clinical experience. J Vasc Interv Radiol. for Thrombo-Embolism Investigators.
2008;19(4):521-8. Comparison of low-intensity warfarin
Anticoagulation Therapy
ruth l. Bush, Brandi Huf, and CleAnn Toner
67
Figure 6.1 The coagulation cascade and available anticoagulants. Adapted from Weitz JI, Bates SM. New
anticoagulants. J Thromb Haemost. 2005;3:1843-1853. Used with permission.
bioprosthetic heart valve replacement, myo- are more likely to affect the metabolism of
cardial infarction (MI), and hypercoagu- warfarin, including amiodarone, azole anti-
lable conditions.4 fungals, Flagyl, and Bactrim.
Vitamin K antagonists (VKAs) exert Patients should be advised to moni-
their anticoagulation effect by inhibiting the tor their diet when initiating warfarin due
enzyme vitamin K oxide reductase, block- to the adverse affect it can have on the pa-
ing the conversion of vitamin K epoxide to tients therapy. Stressing a diet consistent in
vitamin K. This inhibits the carboxylation vitamin K foods can have a considerable im-
and activation of the vitamin Kdependent pact on stabilizing patients early on in their
coagulation factors II, VII, IX, and X, and therapy. See Figure 6.2 for a list of vitamin K
proteins C and S.5 Warfarin is a racemic foods divided into low, moderate, and high
mixture of 2 active isomers, the R and S vitamin K content. In our facility, we watch
enantiomers. Of the 2 enantiomers, the S closely for changes in consumption of dark
enantiomer is 2.7 to 3.8 times more potent leafy green vegetables, green tea, nutritional
than the R enantiomer. The S enantiomer supplements, and multivitamin use. On the
is metabolized by the CYP2C9 enzyme of other side of the spectrum, 2 foods we ad-
the cytochrome P450 system, whereas the vise patients to stay away from are cranber-
R enantiomer is metabolized by CYP 1A2 ries and grapefruit, along with their juices.
and 3A4.5 Therefore, medications that are Grapefruit can inhibit warfarin metabolism,
metabolized through the CYP2C9 enzyme and the mechanism behind the interaction
Figure 6.2 Vitamin K foods. Adapted from USDA National Nutrient Database, http://www.nal.usda.gov/fnic
/foodcomp/Data/SR17/wtrank/sr17a430.pdf.
with cranberries is largely unknown; how- In patients at high risk for developing
ever, it is thought to also be involved in me- another event and in need of immediate
tabolism. anticoagulation, heparin/LMWH is used
Contraindications to using this medica- alongside warfarin for at least 4 to 5 days
tion are typically based on risk vs. benefit. If and pending 2 therapeutic INR levels. The
the beneficial use of warfarin is outweighed therapeutic INR range for most indications
by a higher risk of bleeding, the patient can- continues to be 2.0 to 3.0 (target 2.5) and for
not be placed on warfarin. Disease states high-risk mechanical valves 2.5 to 3.5 (target
and processes included in this consideration 3.0).5,10,11
are certain blood disorders, active bleeding, Monitoring is usually done within the
any disorder causing the patient to be more first few days of therapy initiation and then
prone to bleed, and malignant hypertension. is spaced out depending on the patients re-
If a patient cannot be properly monitored sponse and stabilization of warfarin. If a pa-
due to personal/psychological issues or non- tient develops a sub/supratherapeutic INR
compliance, it is more often than not safer due to diet, noncompliance, interacting
for the patient to be taken off warfarin. As medication, hold for surgery, or for any other
with all medications, if the patient develops reason, it is prudent to monitor more closely
an allergic reaction to warfarin, other op- and restart the cycle of monitoring. A patient
tions regarding anticoagulation will need should never go more than 4 weeks without
to be evaluated. Warfarin is teratogenic and having an INR drawn.5 Dose adjustment pro-
cannot be used during pregnancy. When a tocols vary depending on the institution.12,13
patient already on warfarin is scheduled for
any type of surgery or procedure where there
is a high risk of bleeding, the patient will Parenteral
need to hold warfarin for a sufficient period
of time prior to the procedure with or with- Anticoagulants
out an alternative anticoagulant.4
Warfarin dosing is highly individualized Heparins
and must take into account comorbid disease First discovered in 1916, unfractionated
states. The recommended initiation dose for heparin (UFH) exerts its anticoagulant
warfarin is 5 mg, which was shown to achieve effect by binding to antithrombin.14 The
therapeutic INR as fast as the 10-mg loading resulting heparin-antithrombin complex in-
dose.6 This was contested in an outpatient hibits the activity of factor Xa and thrombin.
study that demonstrated faster achievement Only one-third of the molecules of UFH
of target INR with a 10-mg loading dose.7 have the required sequence for binding to
However, the study also demonstrated in- antithrombin.15 UFH also contains mol-
creased rates of bleeding associated with the ecules that bind to other plasma proteins
higher loading dose. A lower starting dose that are present in variable amounts in each
of 2 to 3 mg daily may be used in patients individual. Thus, UFH has an unpredict-
with disease states or illnesses causing an in- able effect in different people.16 UFH is me-
creased response to warfarin, including the tabolized by the liver and has side effects of
elderly population.5,8 Table 6.1 is an example bleeding, heparin-induced thrombocytope-
of a warfarin dose adjustment protocol used nia (HIT), and osteoporosis.16 The dosing
in our facility for initiation of warfarin.9 of UFH is based on 1 of 2 schedules and
tive and safe as enoxaparin for the treatment Table 6.3 FondaparinuxDosing
of DVT.34,35 The drug is also used for prophy-
Dose
laxis of venous thromboembolism and is con- Weight(kg) (mg)
traindicated in patients with CrCl <30 mL/
min, hemorrhage, bacterial endocarditis, Prophylaxis 2.5 mg 2.5
thrombocytopenia, and allergy to the drug Treatment <50 kg 5.0
itself, and for prophylaxis in patients weigh-
50100 kg 7.5
ing less than 50 kg.36 Fondaparinux is dosed
based on body weight for treatment and at a >100 kg 10.0
fixed dose for prophylaxis (Table 6.3).
As with LMWHs, fondaparinux is cus-
tomarily not monitored.22
initial dose should be 44 kg, and the maxi-
mum infusion dose should be 16.5 mg/h.37
Direct Thrombin Inhibitors The dose of lepirudin does need to be re-
Direct thrombin inhibitors also have the ad- nally adjusted starting at a CrCl of 60 mL/
vantages of predictable dose response and min.37 Monitoring of lepirudin is done with
reduced incidence of thrombocytopenia. the aPTT. A baseline value should be ob-
However, they do not have an antidote for tained, and lepirudin should not be started
cases of severe bleeding. Argatroban and if the baseline aPTT is more than 2.5 times
lepirudin are 2 intravenous direct thrombin the normal aPTT control.37 Target range for
inhibitors that are FDA approved for the pre- aPTT during lepirudin treatment should be
vention and/or treatment of VTE in patients 1.5 to 2.5 times the normal aPTT control,
with heparin-induced thrombocytopenia.33 with the value taken 4 hours after the initial
Ximelagatran is an oral direct thrombin in- dose of lepirudin.37
hibitor that has been withdrawn from the Argatroban is dosed initially at 2 mcg/
market because of an increased risk of he- kg/min, as a continuous infusion. The dose
patic toxicity. may be adjusted up to 10 mcg/kg/min to an
These medications bind directly to aPTT of 1.5 to 3 times the baseline.38 The
thrombin to inhibit the activation of fibrin. aPTT levels need to be drawn 1 to 3 hours
Lepirudin and argatroban are indicated for after the dose is given. No dosage adjustment
patients with HIT.37,38 Bivalirudin is indi- is necessary in renal impairment. However,
cated for patients undergoing percutaneous adjustments are needed in hepatic impair-
transluminal coronary angioplasty (PTCA) ment to 0.5 mcg/kg/min as an initial dose,
who have unstable angina.39 Contraindica- and then adjusted according to the aPTT.38
tions to the direct thrombin inhibitors in- Initial dose of bivalirudin is an IV bolus
clude hypersensitivity to the products37-39 and of 0.75 mg/kg, followed with an infusion of
major bleeding.38,39 Antidotes are not avail- 1.75 mg/kg/h during the PTCA.39 An activat-
able for the direct thrombin inhibitors.10 ed clotting time (ACT) should be drawn 5
Lepirudin is dosed 0.4 mg/kg body minutes after the bolus, and an extra 0.3 mg/
weight (up to 110 kg) IV as a bolus dose, kg should be given to achieve the target ACT
then 0.15 mg/kg body weight/h IV continu- of 300 to 350 seconds if needed. Dose adjust-
ous infusion for 2 to 10 days.37 If a patients ments are necessary in renal impairment. In
body weight exceeds 110 kg, the maximum patients with CrCl of 30 to 59 mL/min, the
infusion rate should be 1.75 mg/kg/h; CrCl researched to attempt to alleviate some of
less than 30 mL/min, reduce rate to 1 mg/ the current drawbacks.
kg/h.39 When looking at developing agents
for anticoagulation, another look to the co-
agulation cascade provides a graphical un-
New Anticoagulants derstanding of their mechanism of action
Current medications on the market for (Figure 6.3).
anticoagulation have many drawbacks. As mentioned at the beginning of the
Vitamin K antagonists require intensive chapter, there are 2 pathways of the coagula-
monitoring and have multiple drug interac- tion cascade: the contact activation pathway
tions. Heparin usually requires admission and the tissue factor pathway.3 These 2 path-
to the hospital, it is administered paren- ways meet to form the final common path-
terally, and there is the possibility for the way. The first step in this pathway to initiate
development of heparin-induced thrombo- coagulation is the binding of tissue factor to
cytopenia. Low-molecular-weight heparins factor VII, which activates factor VIIa/tissue
are also parenterally administered, have factor complex. New agents that are being
the potential for developing HIT, and cost studied target this step to inhibit the initia-
considerably more than the other alterna- tion of coagulation, including Tifacogin and
tives. Several new medications are being NAPc2, which are both parenteral medica-
Figure 6.3 New anticoagulants. Adapted from Weitz JI, Bates SM. New anticoagulants. J Thromb Haemost.
2005;3:1843-1853. Used with permission.
11. Kearon C, Kahn SR, Agnelli G, Goldhaber S, 21. Prandoni P, Carnovali M, Marchiori A;
Raskob GE, Comerota AJ; American College Galilei Investigators. Subcutaneous adjusted-
of Chest Physicians. Antithrombotic therapy dose unfractionated heparin vs fixed-dose
for venous thromboembolic disease: the low-molecular-weight heparin in the initial
eighth ACCP conference on antithrombotic treatment of venous thromboembolism. Arch
and thrombolytic therapy. Chest. Intern Med. 2004;164(10):1077-83.
2008;133:454S545S. 22. Hirsh J, Bauer KA, Donati MB, Gould M,
12. Ebell M. A systematic approach to managing Samama MM, Weitz JI; American College of
warfarin doses. Family Practice Management. Chest Physicians. Parenteral anticoagulants:
2005:77-83. http://www.aafp.org/fpm American College of Chest Physicians
/20050500/77asys.html. Evidence-Based Clinical Practice Guidelines
13. Franke CA, Dickerson LM, Carek PJ. Improv- (8th ed.). Chest. 2008;133(6 suppl):
ing anticoagulation therapy using point-of- 141S-59S.
care testing and a standardized protocol. Ann 23. Basu D, Gallus A, Hirsh J, Cade J. A
Fam Med. 2008;6 suppl 1:S28-32. prospective study of the value of monitoring
14. McLean J. The thromboplastic action of heparin treatment with the activated partial
cephalin. Am J Physiol. 1916;41:250-7. thromboplastin time. N Engl J Med.
15. Choay J, Lormeau JC, Petitou M, Sina P, 1972;287(7):324-7.
Fareed J. Structural studies on a biologically 24. Brill-Edwards P, Ginsberg JS, Johnston M,
active hexasaccharide obtained from heparin. Hirsh J. Establishing a therapeutic range
Ann NY Acad Sci. 1981;370:644-9. for heparin therapy. Ann Intern Med.
16. McRae S, Ginsberg J. Initial treatment of 1993;119(2):104-9.
venous thromboembolism. Circulation. 2004; 25. Bounameaux H, de Moerloose P. Is laboratory
(9 suppl 1):I3-9. monitoring of low-molecular-weight heparin
17. Heparin Sodium Injections [package insert]. therapy necessary? No. J Thromb Haemost.
Schaumburg, IL: APP Pharmaceuticals, 2004;2:551-4.
LLC, 2008. http://www.apppharma.com 26. The Columbus Investigators. Low-molecular-
/our-products/alphabetical/product-55.html. weight heparin in the treatment of patients
18. Raschke RA, Reilly BM, Guidry JR, Fontana with venous thromboembolism. N Engl J
JR, Srinivas S. The weight-based heparin Med. 1997;337:657-62.
dosing nomogram compared with a standard 27. Mismetti P, Quenet S, Levine M, Merli G,
care nomogram. A randomized controlled Decousus H, Derobert E, et al. Enoxaparin
trial. Ann Intern Med. 1993;119(9):874-81. in the treatment of deep vein thrombosis
19. Cruickshank MK, Levine MN, Hirsh J, with or without pulmonary embolism: an
Roberts R, Siguenza M. A standard heparin individual patient data meta-analysis. Chest.
nomogram for the management of heparin 2005;128:2203-10.
therapy. Arch Intern Med. 1991;151(2):333-7. 28. Quinlan DJ, Mcquillan A, Eikelboom JW.
20. Kearon C, Ginsberg JS, Julian JA, Douketis Low-molecular-weight heparin compared
J, Solymoss S, Ockelford P, et al; Fixed-Dose with intravenous unfractionated heparin for
Heparin (FIDO) Investigators. Comparison treatment of pulmonary embolism: a meta-
of fixed-dose weight-adjusted unfractionated analysis of randomized, controlled trials. Ann
heparin and low-molecular-weight heparin for Intern Med. 2004;140:175-83.
acute treatment of venous thromboembolism. 29. OBrien JA, Caro JJ. Direct medical cost of
JAMA. 2006;296(8):935-42. managing deep vein thrombosis according
Pharmacological Thrombolysis
Indications, Techniques, and outcomes
Mark g. Davies
79
chain proteins and then secrete them. These which retains clottability but forms clots
single-chain proteins then assemble into with reduced tensile strength that stimulate
functional complexes and act as serine prote- plasminogen activation by tPA more than
ases. There are 2 forms of plasminogen acti- fibrin clots.4 It accumulates after treatment
vators (PA): the urokinase type (uPA), which with tPA but not with tPA given with 2 anti-
activates plasminogen in the fluid phase, plasmin. It does not accumulate with use of
and tissue PA (tPA), which is most active more fibrin-specific agents.5
when bound to fibrin.2,3 Normal endothelial
cells express tPA. However, if stimulated by
a variety of cytokines and other factors, they Pharmacology
preferentially synthesize uPA and down-
regulate tPA synthesis. In addition to these Thrombolysis is the pharmacological dis-
2 fibrinolytic enzymes, endothelial cells solution of fibrin thrombus by exogenously
also secrete 2 PA inhibitors, PAI-1 and PAI- delivered agents. The characteristics of these
2. Both are serine protease inhibitors and agents are shown in Table 7.1. These agents
form equimolar complexes with active uPA form plasmin, which leads to the degradation
or tPA molecules. PAI-1 requires the pres- of fibrin, fibrinogen, factor V, and factor VII.
ence of vitronectin in the extracellular ma- Plasminogen is found free in the circulation
trix to maintain its active conformation and and within the thrombus, which allows for
is therefore inactive outside the matrix. The activation of plasmin within a thrombus by
thrombin-thrombomodulin pathway also regional perfusion. Streptokinase binds plas-
is involved in the regulation of fibrinolysis. minogen to form an active complex, which
Thrombin activatable fibrinolysis inhibitor then activates another plasminogen to form
(TAFI), a carboxypeptidase, is a plasma pro- plasmin. Use of streptokinase is associated
tein that is a substrate, like protein C, for the with an increased incidence of complications
thrombin thrombomodulin complex. TAFIa such as allergic reactions or hemorrhages and
suppresses glu-plasminogen but not lys-plas- has fallen out of favor. The success of uroki-
minogen activation by tPA in the presence nase was due to its nonantigenic properties
of a fibrin analogue that had been exposed and its direct activation of plasminogen with-
to plasmin (clotted fragment X). Through out the formation of an intermediate complex.
the activation of protein C and TAFI, the As a nonspecific activator, however, it can, in
thrombin/thrombomodulin complexes can high doses, induce systemic fibrinolysis. rtPA
down-regulate coagulation and fibrinolysis. or alteplase is a recombinant version of the
The general schema of fibrinolysis is shown naturally occurring tPA protein; its activity is
in Figure 7.1. Bleeding, the most serious enhanced by the presence of fibrin. The final
complication of thrombolytic therapy with product is recombinant plasminogen activa-
tissue-type plasminogen activator (tPA), is tor rPA or reteplase. rPA is a recombinantly
thought to result from lysis of fibrin in hemo- derived mutant lacking the 3 nonprotease do-
static plugs and from the systemic lytic state mains of rtPA. A comparison of the efficacy,
caused by unopposed plasmin. One mecha- safety, and costs associated with catheter-
nism by which systemic plasmin can impair directed thrombolysis (CDT) with urokinase
hemostasis is by partially degrading fibrino- (UK) and the recombinant agents alteplase
gen to fragment X, a high-molecular-weight (tissue plasminogen activator [tPA]) and re-
clottable fibrinogen degradation product, teplase (recombinant plasminogen activator
Figure 7.1 Control of intravascular coagulation.The endothelium coordinates the procoagulant and anticoagulant
pathways to maintain vascular fluidity. It synthesizes several factors: protein S, urokinase-like plasminogen activator
(uPA), tissue plasminogen activator (tPA), and plasminogen activator inhibitors (PAI). uPa and PAI are localized on
the membrane by uPAR. Several pathways involved in fibrinolysis and anticoagulation reside on the membrane of the
endothelial cells (plasminogen activator/inhibitors and the protein C pathway). Clot is formed by the activation of
tissue factor, which leads to thrombin activity that cleaves fibrinogen to fibrin and is quickly associated with activated
platelets to form a secondary hemostatic plug. At this time the plasmin-activating system is inhibited at various levels
by the cells. Once fibrinolysis is activated, plasmin is formed from plasminogen by the actions of tPA and uPA whose
activity is in turn modulated by PAI-1 and 2 antiplasmin. Activated plasmin breaks fibrin into fibrin degradation
products, one of which is termed fragment X and can lead to clot instability in the presence of long-term infusions of
exogenous plasminogen activators.
generation thrombolytic agents such as mon- which can occur months to years after the
teplase, tenecteplase, reteplase, lanoteplase, initial event.9 Conventional therapy with
pamiteplase, and staphylokinase result in a heparin has been effective in diminishing
greater angiographic patency rate in patients and preventing the propagation of throm-
with acute myocardial infarction, although, bus;10 however, it does not diminish the de-
thus far, mortality rates have been similar velopment of postthrombotic syndrome.11
for those few drugs that have been studied In fact, it has been estimated that 30% to
in large-scale trials. Bleeding risk, however, 40% of existing thrombi will demonstrate
may be greater. propagation even at therapeutic levels of
heparin.12 A pooled analysis of 13 studies by
Comerota13 showed that 4% of patients who
Indications received heparin therapy had significant or
complete lysis of thrombus compared with
Systemic anticoagulation has been the main- 45% who were randomized to systemic strep-
stay of therapy to prevent the occurrence of tokinase therapy. The ideal therapy should
pulmonary embolism. However, the long- dissolve existing thrombus, restore flow, and
term effects of venous thromboembolism preserve venous valve function, which will
have largely gone unnoticed, specifically, the diminish the incidence of venous hyperten-
development of postthrombotic syndrome, sion. Catheter-directed venous thrombolysis
Figure 7.2 Algorithm for thrombolysis of venous thrombotic disease.In the patient presenting with a diagnosed
deep venous thrombosis in the lower extremity, the acuity of the lesion, the extent of the thrombosis, and the impact
of the thrombosis on distal circulation must be considered. In the absence of contraindications, thrombolysis can be
instituted. The use of an adjunct inferior vena cava (IVC) filter is institutional-dependent. Lysis of the clot may reveal a
lesion that will require intervention, either endoluminally or surgically. With resolution or no documented improvement,
thrombolysis may be stopped and a standard heparin/oral anticoagulation protocol instituted.
has the potential to meet these goalsin of postthrombotic syndrome and its associ-
particular, the retention of venous valvular ated complications.
function, which can prevent the formation Percutaneous clot removal using throm-
introduced. Baseline venography is then per- Low-dose tPA continuous infusion at 0.5 to
formed. After passage of a 0.035-in guidewire 1.0 mg/h is administered through the infu-
through the occluded venous segment, a di- sion sheath with intravenous heparin at a
agnostic catheter is advanced past the occlud- rate of 500 U/h. Others recommend giving
ed segment. This catheter is then exchanged a lacing bolus of tPA 3 to 5 mg followed by
for a multiple side hole infusion catheter. It a continuous tPA infusion at 0.05 mg/kg/h.
is important that the total volume of throm- Adding intermittent pneumatic compres-
bus is traversed with an infusible segment of sion to CDT for DVT treatment of the leg
catheter so that the thrombus can come into resulted in better early and late outcomes
direct contact with the thrombolytic agent. compared with CDT alone and was not asso-
ciated with an increased risk of symptomatic strates that a larger proportion of patients
pulmonary emboli.18 with acute deep venous thrombosis (86%)
vs. chronic (68%) were able to achieve grade
II or III lysis. This is important because the
Monitoring degree of lysis was found to be predictive of
Currently, there is no need to monitor the early and continued patency. For instance,
changes in coagulation induced by low-dose 75% of limbs with complete lysis remained
thrombolytics after the physician has estab- patent at 1 year compared to 32% of limbs
lished a baseline coagulation profile. Many in which there was <50% lysis. A presenta-
centers will follow fibrinogen and seek to tion of acute deep venous thrombosis (<10
maintain a target of >200 mg/dL. The pa- days) was predictive of a better lysis grade
tient does require monitoring of the infu- when compared with chronic deep venous
sion site to ensure that there is no exit site thrombosis, although a significant percent-
bleeding or disruption of the thrombolytic age of patients with chronic iliofemoral
delivery system. Progress in thrombolysis deep venous thrombosis were found to have
may be assessed by interval duplex ultraso- significant lysis (grade IIIII). The cathe-
nography or more commonly by repeat ve- ter-directed group demonstrated a much
nography. Repeat imaging is performed at higher degree of significant lysis (83% vs.
8- to 12-hour intervals, and infusions may be 20%) when compared with the 19% of pa-
continued as long as 72 hours, although the tients who had a pedal infusion. This point
mean infusion time is generally 56 hours.19 underscores the necessity of having the
The incidence of major clinical hemorrhage thrombus in direct contact with the throm-
after fibrinolysis for DVT is between 6% and bolytic agent. Adjunctive procedures with
30%, a 3-fold increase compared to standard stent placement and/or venoplasty were per-
heparin therapy. Infusion of thrombolysis is formed in 33% of affected limbs, which indi-
associated with increased risk of local he- cates that an underlying lesion often needs
matoma formation at the site of catheter in- to be treated in conjunction with throm-
sertion and a lower risk of distant bleeding bolysis. Major bleeding complications were
(Figure 7.3). Hypersensitivity reactions are noted in 4% of patients enrolled. Subgroup
low with both tPA or rPA. analysis reveals that the best results occurred
in patients with acute iliofemoral deep ve-
nous thrombosis with no prior history of
Outcomes deep venous thrombosis. Complete lysis oc-
curred in 65% of patients with 96% patency
In the National Multicenter Venous Reg- at l year. Conversely, the worst results were
istry,17 catheter-directed thrombolysis was seen in patients with chronic femoral-pop-
performed on acute (66%), chronic (45%), liteal deep venous thrombosis. The groups
and a mixture of acute and chronic throm- were not large enough to establish statistical
bus (19%). The degree of lysis achieved was significance, but the investigators believed
categorized as grade I (<50% lysis), grade II these results could serve as a guide to patient
(50%90% lysis), and grade III (complete selection for catheter-directed thrombolysis.
lysis). Continued patency was followed at 3, Patients most likely to achieve complete lysis
6, and 12 months with duplex ultrasonog- and increased patency were those with an
raphy. Analysis of the registry data demon- acute iliofemoral deep venous thrombosis
and no prior history of deep venous throm- eligibility criteria has improved the safety
bosis. Follow-up data are limited because a and acceptability of this treatment. The
significant proportion of patients were lost optimum drug, dose, and route of adminis-
to follow-up at 12 months.17 Following treat- tration have yet to be determined. A more
ment, patients receiving catheter-directed recent review of articles on catheter-directed
thrombolysis reported better overall physi- thrombolysis (CDT) from PubMed and the
cal functioning, less stigma, less health dis- Cochrane library was recently performed.23
tress, and fewer postthrombotic symptoms CDT reduced clot burden and DVT recur-
compared to those patients treated with rence and may prevent the formation of
anticoagulation alone.20 Within the throm- postthrombotic syndrome. Indications for
bolysis group, successful lysis correlated with its use include younger individuals with a
health-related quality of life.20,21 long life expectancy and few comorbidities,
The results of the current Cochrane limb-threatening thromboses, and proximal
database included 12 studies.22 Complete iliofemoral DVTs. These results suggest that
clot lysis occurred significantly more often the outcomes of CDT in DVT management
in the treatment group in early follow-up are encouraging in selected patient cohorts,
(relative risk [RR] 0.24; 95% confidence but further evidence is required to establish
interval [CI] 0.070.82) and in late follow- longer-term benefits and cost-effectiveness.
up (RR 0.37; 95% CI 0.25 0.54). A similar There is a marked lack of randomized
effect was also seen for any degree of im- controlled trials comparing CDT-related
provement in venous patency. Significantly mortality and long-term outcomes com-
less postthrombotic syndrome occurred in pared to anticoagulation alone. The current
those receiving thrombolysis (RR 0.66; 95% ATTRACT trial sponsored by the NIH will
CI 0.470.94). Leg ulceration was reduced, hopefully answer many of the questions re-
although the data were limited by small lated to catheter-directed thrombolysis.
numbers (RR 0.53; 95% CI 0.122.43). Ve-
nous function was improved at late follow- references
up, but not significantly (RR 0.43; 95% CI 1. Gertler JP, Abbott WM. Prothrombotic and
0.063.17). Out of 668 patients, those re- fibrinolytic functions of normal and perturbed
ceiving thrombolysis had significantly more endothelium. J Surg Res. 1992;52:89-95.
bleeding complications (RR 1.73; 95% CI 2. Henkin K, Marcotte P, Yang H. The
1.042.88). Two strokes occurred in the plasminogen-plasmin system. Prog Cardiovasc
treatment group (RR 1.70; 95% CI 0.21 Dis. 1991;34:135-62.
13.70). The incidence of bleeding appears to 3. Robbins K. The Plasminogen-Plasmin
have reduced over time with the introduc- Enzyme System. New York: Lippincott; 1995.
tion of stricter selection criteria. There was 4. Schaeerf AV, Leslie BA, Rischke JA, Stafford
no significant effect on mortality detected AR, Fredenburgh JC, Weitz JI. Incorporation
in either early or late follow-up. Data on of fragment X into fibrin clots renders
occurrence of pulmonary embolism (PE) them more susceptible to lysis by plasmin.
and recurrent DVT were inconclusive. The Biochemistry. 2006;45(13):4257-65.
conclusions were that thrombolysis appears 5. Weitz JI. Limited fibrin specificity of tissue-
to offer advantages in terms of reducing type plasminogen activator and its potential
postthrombotic syndrome and maintaining link to bleeding. J Vasc Interv Radiol. 1995;6
venous patency after DVT.22 Use of strict (Pt 2 suppl):19S-23S.
6. Grunwald MR, Hofmann LV. Comparison 16. Kearon C, Kahn SR, Agnelli G, Goldhaber S,
of urokinase, alteplase, and reteplase for Raskob GE, Comerota AJ; American College
catheter-directed thrombolysis of deep of Chest Physicians. Antithrombotic therapy
venous thrombosis. J Vasc Interv Radiol. for venous thromboembolic disease: American
2004;15(4):347-52. College of Chest Physicians Evidence-Based
7. Longstaff C, Williams S, Thelwell C. Fibrin Clinical Practice Guidelines (8th ed.). Chest.
binding and the regulation of plasminogen 2008;133(6 suppl):454S-545S.
activators during thrombolytic therapy. 17. Mewissen M, Seabrook G, Meissner M,
Cardiovasc Hematol Agents Med Chem. Cynamon J, Labropoulos N, Haughton S.
2008;6(3):212-23. Catheter-directed thrombolysis for lower
8. Verstraete M. Third-generation thrombolytic extremity deep venous thrombosis: report
drugs. Am J Med. 2000;109(1):52-8. of a nation multicenter registry. Radiology.
9. Strandness DE Jr, Langlois Y, Cramer M, 1999;211:39-49.
Randlett A, Thiele BL. Long-term sequelae 18. Ogawa T, Hoshino S, Midorikawa H, Sato
of acute venous thrombosis: a clinical review. K. Intermittent pneumatic compression of
JAMA. 1983(250):1289-92. the foot and calf improves the outcome of
10. Johnson BF, Manzo RA, Bergelin RO, catheter-directed thrombolysis using low-dose
Strandness DE Jr. Relationship between urokinase in patients with acute proximal
changes in the deep venous system and the venous thrombosis of the leg. J Vasc Surg.
development of the post-thrombotic syndrome 2005;42(5):940-4.
after an acute episode of lower limb deep 19. Mewissen M, Seabrook GR, Meissner MH,
venous thrombosis: a one-to-six year follow-up. Cynamon J, Labropoulos N, Haughton SH.
J Vasc Surg. 1995;21:307-12. Catheter-directed thrombolysis for lower
11. Meissner MH, Manzo RA, Bergelin RO, extremity deep venous thrombosis: report
Markel A, Strandness DE Jr. Deep venous of a nation multicenter registry. Radiology.
insufficiency: the relationship between 1999;211:39-49.
lysis and subsequent reflux. J Vasc Surg. 20. Comerota AJ. Quality-of-life improvement
1993;18:596-608. using thrombolytic therapy for iliofemoral
12. Strandness DE. Thrombus propagation deep venous thrombosis. Rev Cardiovasc Med.
and level of anticoagulation. J Vasc Surg. 2002;3 (suppl 2):S61-7.
1990;12:497-8. 21. Comerota AJ, Throm RC, Mathias SD,
13. Comerota A, Aldridge S. Thrombolytic Haughton S, Mewissen M. Catheter-directed
therapy for deep vein thrombosis: a clinical thrombolysis for iliofemoral deep venous
review. Can J Surg. 1993;36:359-64. thrombosis improves health-related quality of
14. Meissner MH, Wakefield TW, Ascher E, life. J Vasc Surg. 2000;32(1):130-7.
Caprini JA, Comerota AJ, Eklof B, et al. Acute 22. Watson LI, Armon MP. Thrombolysis for acute
venous disease: venous thrombosis and venous deep vein thrombosis. Cochrane Database
trauma. J Vasc Surg. 2007;46 (suppl S):25S- Syst Rev. 2004; Oct 18(4):CD002783.
53S. 23. Gogalniceanu P, Johnston CJ, Khalid U, Holt
15. Molina JE, Hunter DW, Dietz CA. PJ, Hincliffe R, Loftus IM, et al. Indications
Protocols for Paget-Schroetter syndrome for thrombolysis in deep venous thrombosis.
and late treatment of chronic subclavian Eur J Vasc Endovasc Surg. 2009;38(2):
vein obstruction. Ann Thorac Surg. 192-8.
2009;87(2):416-22.
8
Pharmacomechanical Thrombolysis
for Acute Deep Venous Thrombosis
Indications, Techniques, and Clinical Data
89
tive lesions and need for treatment; and being alternatives. Initial venography is per-
the poor results of lytic therapy for chronic formed to document the extent of thrombus
thrombosis.4 and a wire is advanced across the thrombus.
The limitations of both systemic and If caval thrombus is suspected, some also
catheter-directed thrombolysis have led to recommend a computed tomography (CT)
a variety of pharmacomechanical strategies venogram to document thrombus extent
directed toward increasing lytic efficiency prior to bringing the patient to the angiog-
while reducing procedural times. Theo- raphy suite. Further aspects of the procedure
retically, this would translate into reduced are more device-specific.
bleeding complications, decreased hospital The AngioJet rheolytic thrombectomy
and intensive care unit stays, and reduced catheter (MEDRAD Inc., Warrendale, PA)
cost. is an over-the-wire system using a high-
velocity (350/450 km/h) saline jet directed
backward from the tip of the catheter. A
Pharmacomechanical low-pressure zone is created via a Venturi
Thrombolytic Strategies effect, causing fragmentation of the throm-
bus, which is aspirated through the efflu-
The isolated use of mechanical devices is ent lumen (Figure 8.1). In the power-pulse
rarely successful, and their use in combi- mode, the effluent port is closed with a stop-
nation with thrombolytic drugs is usually cock and a dilute lytic agent, usually tPA at
required.5 However, use of a mechanical a dose of 0.4 mg/cc is used in place of the
device to speed thrombolysis is theoretically saline jet. After infusion, the lytic agent is
attractive since it has the potential to reduce allowed to dwell for 20 to 25 minutes prior
thrombolytic infusion times and thereby to aspiration on withdrawal of the catheter
decrease bleeding complications, hospital with the effluent port open. A variety of
resource utilization, and cost while improv- catheter sizes are available, allowing treat-
ing efficacy. Although there is little solid ment of different vessels.6
data that this has been achieved, use of such The Trellis-8 infusion catheter (Covi-
devices would optimally allow the entire ve- dien) has proximal and distal occlusion bal-
nous system to be cleared of thrombus in a loons with a drug-infusion port between the
single session in the angiography suite. balloons, and a wire that oscillates within the
Amongst several mechanical throm- catheter (Figure 8.2). This allows the throm-
bolytic devices, 3 have been widely utilized bolytic drug to be isolated between balloons
in the venous systemthe AngioJet power in direct contact with the thrombus. The
pulse, the Trellis-8 infusion catheter, and the oscillating wire macerates the thrombus, in-
EKOS EndoWave. The mechanical throm- creasing exposure to the thrombolytic agent
bolytic devices share many technical details. and theoretically reducing the dose required
If inflow to the popliteal vein is patent, ul- for lysis. After treatment, the thrombolytic
trasound-guided puncture of the popliteal agent and clot fragments are aspirated. Sev-
vein using a micropuncture set is preferred eral thrombolytic agents had been used, in-
by many interventionalists, with more distal cluding tenecteplase, with doses between 5
access via the posterior tibial vein, small sa- and 10 mg and tPA, with a recommend dose
phenous vein, or great saphenous vein (with of 5 to 10 mg. The current available cath-
deep system access through a perforator) eters are 80 and 120 cm in length, and the
Figure 8.1 AngioJet rheolytic thrombectomy catheter. Image Courtesy of MEDRAD Interventional.
9
endovascular Intervention for lower
extremity Deep Venous Thrombosis
Techniques, Devices, and outcomes
95
clot lysis.
considering the lower thrombolytic dosages With the aspiration port clamped, in-
administered and decreased length of ICU fusate is released into the thrombosed ve-
stay compared to CDT,19 and the decreased nous segment during a slow pull-back of
long-term morbidity from postthrombotic the catheter, effectively lacing the clot with
syndrome compared to traditional antico- thrombolytic drug. After 10 minutes, the as-
agulation. piration function of the catheter is turned
on. The catheter is then advanced through
AngioJet Rheolytic the thrombosed segment a second time, re-
moving macerated thrombus through the
Thrombectomy System aspiration ports as the catheter is advanced.
Power-Pulse Spray Technique This process may be repeated if there is re-
The AngioJet catheter system is comprised maining thrombus burden at the end of the
of a single-use catheter, a single-use pump first pass. Alternatively, as is often our prefer-
set, and a drive unit. The catheter, which is ence, the patient may then undergo catheter-
available in working lengths of 60, 100, and directed thrombolysis in the intensive care
120 cm, contains a central lumen for in- unit (ICU) overnight and return to the oper-
fusate and a larger lumen encompassing the ative room the following day for reevaluation
central channel, the guidewire, and aspirate with venography and possible repeat throm-
from the thrombus. The drive unit generates bectomy or venous stenting, if indicated.
10,000 psi of pulsatile infusion flow, which Success in thrombus removal, restora-
is released from the catheter in retrograde- tion of venous patency, and preservation of
directed high-velocity saline jets. These jets valvular function have been demonstrated
create a localized low-pressure zone (Ber- with the use of the AngioJet power-pulse
noullis principle) at the catheter tip, mac- spray technique. While Kasirajan reported
erating thrombus and redirecting flow and only 24% of patients had >90% clot resolu-
debris into outflow channels directed behind tion, 35% had 50% to 90% resolution, and
the catheter tip for aspiration and removal 41% had less than 50% resolution,21 im-
(see Figure 8.1 on page 91). proved results have been demonstrated with
Access for the AngioJet system requires the addition of lytics to the infusate, as dis-
a 6F introducer sheath. The AngioJet cath- cussed previously. Bush et al. reported com-
eter is then advanced over a 0.035 guidewire plete thrombus resolution in 65% of patients,
through the thrombus load. While this sys- with at least partial resolution seen in all of
tem was originally intended for use without the remaining patients.22 Lin et al. demon-
adjunctive thrombolytics, it has been dem- strated that PMT with the AngioJet system
onstrated that the addition of lytics to the was at least as effective as CDT in treating
infusion solution results in decreased treat- lower extremity DVT. They showed com-
ment time and improved results. We recom- plete clot lysis in 75% of patients treated with
mend that thrombolytics be routinely used AngioJet vs. 70% in patients treated with
except when contraindicated, as is our prac- CDT (P = NS) with similar patency at 1-year
tice. While thrombolytic choice and dose follow-up of 64% and 68%, respectively. In
vary according to surgeon preference, we addition, they demonstrated reduced ICU
have experienced good results using 10 mg stay, reduced total in-hospital length of stay,
of tenecteplase in 50 mL of sodium chloride and reduced costs in the PMT cohort.19 In
infusing solution.20 our series, we demonstrated a 90% venous
patency restoration and maintenance of ve- tial venogram, with the goal of minimizing
nous valvular function in 88% at a mean treatment length of nonthrombosed vein.
follow-up of 6 months.20 The drive unit is attached to the sinusoidal
This therapy is associated with a low dispersion wire, which creates catheter os-
incidence of hemorrhagic complications. cillatation at 500 to 3500 rpm, causing dis-
Isolated case reports of pancreatitis resulting persion of lytics within the thrombus load
from massive hemolysis with use of the An- and mechanical clot disruption. Aspiration
gioJet system have been reported but appear of thrombus debris and lytic remaining in
to be rare occurences.23 the isolated segment completes treatment of
the isolated venous segment (Figure 9.2).
Trellis-8 Infusion System Access for the Trellis-8 infusion sys-
tem requires and 8F introducer sheath. A
Pharmacomechanical Thrombectomy 0.035 Glidewire (Terumo, Sommerset, NJ)
The Trellis-8 Periphperal Infusion System is used to cross the thrombosed venous seg-
incorporates the use of both chemical throm- ment and the Trellis-8 catheter is advanced
bolysis and mechanical thrombectomy. The over the Glidewire. With proximal and dis-
Trellis device consists of a single-use cath- tal balloons inflated, 5 to 10 mg of lytics are
eter, a dispersion wire, and an integral drive infused within the thrombus. After 10 min-
unit. The catheter contains proximal and utes, the dispersion wire is inserted into the
distal occlusion balloons that allow infu- catheter. Catheter vibration between the oc-
sion of thrombolytics to an isolated segment clusion balloons aids in clot maceration and
of thrombosed vein. Catheters are available increases the thrombus surface area exposed
in lengths of 80 or 120 cm, with varied dis- to the lytics. The dispersion wire may fur-
tances between occlusion balloons allowing ther be advanced and retracted once a min-
treatment of 10-, 15-, or 30-cm venous seg- ute during the treatment interval to further
ments. Selection of which catheter to use ensure mixing of the lytics with the throm-
will depend on the location and length of bus. After 5 to 15 minutes, the distal balloon
the thrombosed segment determined on ini- is deflated and the catheter aspirated via a
Adjunctive Procedures
Conclusion
Recalcitrant thrombus after initial treatment
with PMT may require further therapy. Percutaneous mechanical thrombectomy
While small residual thrombus may respond and ultrasound-accelerated thrombolysis are
to venoplasty and stenting, larger amounts at least as effective as catheter-directed throm-
of residual thrombus may require use of a bolysis with reduced ICU and hospital stays
second PMT device or overnight catheter- and decreased overall costs. Further, use of a
directed thrombolysis.15,20,21,26 Use of an ad- second PMT device, adjunctive CDT, and/or
junctive device or CDT should not be re- stenting of underlying venous stenoses may
garded as a failure of the first device, but be needed to achieve optimal clinical results.
Figure 9.3 A. Following an IVC filter placement the left iliac vein occlusion is treated with ultrasound-facilitated
thrombolysis in the prone position. B. PTA and stenting of May-Thurner syndrome results in an excellent venographic
result with long-term patency.
SH. Catheter-directed thrombolysis for lower 19. Lin PH, Zhou W, Dardick A, Mussa F,
extremity deep venous thrombosis: report of Kougias P, Hedayati N, et al. Catheter-directed
a national multicenter registry. Radiology. thrombolysis versus pharmacomechanical
1999;211:39-49. thrombectomy for treatment of symptomatic
12. Lieberman S, Safadi R, Aner H, Verstandig lower extremity deep venous thrombosis. Am J
A, Sasson T, Bloom AI. Local thrombolysis Surg. 2006;192:782-8.
for the treatment of patients with proximal 20. Arko F, Davis CM, Murphy EH, Smith ST,
deep vein thrombosis of the leg. Harefuah. Timaran CH, Modrall JG, et al. Aggressive
2002;141:424-9. percutaneous mechanical thrombosis: Early
13. AbuRahma AF, Pekins SE, Wulu JT, Ng clinical results. Arch Surg. 2007;142:513-8.
HK. Iliofemoral deep vein thrombosis: 21. Kasirajan K, Grey B, Ouriel K. Percutaneous
conventional therapy versus lysis and AngioJet thrombectomy in the management
percutaneous transluminal angioplasty and of extensive deep venous thrombosis. J Vasc
stenting. Ann Surg. 2001;233:752-60. Interv Radiol. 2001;12(2):179-85.
14. Ouriel K, Grey B, Clair DG, Olin J. 22. Bush RL, Lin PH, Bates JT, Mureebe L,
Complications associated with the use Zhou W, Lumsden AB. Pharmacomechanical
of urokinase and recombinant tissue thrombectomy for treatment of symptomatic
plasminogen activator for catheter directed lower extremity deep venous thrombosis:
peripheral aterial and venous thrombolysis. safety and feasibility study. J Vasc Surg.
J Vasc Interv Radiol. 2000;11:295-8. 2004;40:965-70.
15. McLafferty RB. Endovascular management of 23. Piercy KT, Ayerdi J, Geary RL, Hansen
deep venous thrombosis. Perspect Vasc Surg KJ, Edwards MS. Acute pancreatitis: a
Endovasc Ther. 2008;20:87-91. complication associated with rheolytic
16. Braaten JV, Goss RA, Francis CW. Ultrasound mechanical thrombectomy of deep venous
reversibly disaggregates fibrin fibers. Thromb thrombosis. J Vasc Surg. 2006;44(5):1110-3.
Haemost. 1997;78:1063-8. 24. Hilleman DE, Pharm D, Razavi MK. Clinical
17. Francis CW, Blinc A, Lee S, Cox C. and economic evaluation of the Trellis-8
Ultrasound accelerates transport of infusion catheter for deep vein thrombosis.
recombinant tissue plasminogen activator J Vasc Interv Radiol. 2008;19:377-83.
into clots. Ultrasound Med Biol. 1995;21: 25. OSullivan GJ, Lohan DG, Gough N,
419-24. Cronin CG, Kee ST. Pharmacomechanical
18. Parikh S, Motarjeme A, McNamara T, Raabe thrombectomy of acute deep vein thrombosis
R, Hagspiel K, Benenati JF, et al. Ultrasound- with the Trellis-8 isolated thrombolysis
accelerated thrombolysis for the treatment catheter. J Vasc Interv Radiol. 2007;715-24.
of deep vein thrombosis: initial clinical 26. McLafferty RB. Endovascular management of
experience. J Vasc Interv Radiol. 2008;19: deep venous thrombosis. Perspect Vasc Surg
521-8. Endovasc Ther. 2008;20:87-91.
10
Surgical Thrombectomy
Indications, Techniques, and outcomes
Bo eklf
103
another 60% partly recanalize, or at least de- matosclerosis, and ulceration), but these too
velop adequate enough collaterals that they can be managed, in the compliant patient,
do not test positive for obstruction on non- with elastic stockings and elevation. Howev-
invasive study. Thus, in reality, only about er, those with both obstruction and valvular
20% remain significantly occluded enough incompetence have severe postthrombotic
or have such poor collaterals to produce sequelae, so severe, in fact, that conservative
symptomatic venous outflow obstruction de- management is difficult even in a compliant
tectable on noninvasive physiologic testing, patient.
if time (3 to 6 months) is allowed for com- The importance of the proximal ob-
plete resolution. Nevertheless, a significant struction for the development of distal valvu-
degree of obstruction persists for these 3 to 6 lar incompetence has been carefully studied
months in the majority of patients with ilio- by D. Eugene Strandness Jr. and his group.
femoral venous thrombosis. What is not well In a series of articles, they have shown the
appreciated is that persisting proximal ob- following: from 20% to 50% of initially un-
struction, even if it is ultimately relieved by involved distal veins become incompetent
partial recanalization or collateral develop- by 2 years; the combination of reflux and
ment, can lead to progressive breakdown of obstruction, as opposed to either alone, cor-
distal valves, resulting in reflux. This is im- related with the severity of symptoms and
portant because, if the distal venous valves was present in 55% of symptomatic patients;
(and particularly those in the popliteal vein) 25% of all venous segments developed reflux
remain competent, the postthrombotic se- in time, of which 32% were documented
quelae are relatively modest and controllable to not have been previously involved with
by conservative measures.9 The severity of thrombosis; and finally, in a study of the pos-
postthrombotic sequelae correlates with the terior tibial veins located below a popliteal
level of ambulatory venous pressure (AVP), segment involved with thrombosis, 55% of
as shown by Nicolaides et al.10 It is impres- the distal veins became incompetent if the
sive how steadily the frequency of ulceration segment remained obstructed, compared
climbs with increasing levels of AVP. It sug- to 7.5% of those below a popliteal vein that
gests that anything that significantly reduces recanalized.12 These changes in the distal
AVP will reduce the severity of the PTS, and veins occur early enough that they cannot
vice versa. Nicolaides and Sumner11 have be blamed on proximal valvular reflux, al-
also shown that the highest levels of AVP are though this also comes into play with time.
found in patients with both obstruction and Thus, it would appear that the early relief
reflux. of obstructing thrombus by thrombolysis
These 2 observations underscore a or thrombectomy should prevent more ex-
major point to be made in regard to the tensive postthrombotic sequelae if only by
disturbed pathophysiology associated with protecting the distal veins against progres-
iliofemoral venous thrombosis and the sive valvular incompetence. This is a point
pathogenesis of postthrombotic sequelae. that has not been well recognized not only
Obstruction alone is rarely sufficient enough in terms of basic pathogenesis but in judging
to cause venous claudication and mostly the results of thrombolysis and thrombecto-
causes increased swelling with activity. Val- my, where critics have largely ignored the res-
vular incompetence alone causes most of toration of proximal patency while focusing
the stasis sequelae (pigmentation, lipoder- on the presence or absence of valve reflux.
However, even these patients, if faced with presented 6 patients, 5 of whom had an ex-
the threat of venous gangrene, may deserve cellent result with rapid disappearance of leg
prompt clot removal. swelling, very little late morbidity, and mini-
In terms of the choice of method of clot mum leg edema. There was no PE prior or
removal, CDT is an appropriate choice for subsequent to surgery. Mahorner claimed
removing obstructing thrombus and thereby that this method restores vein function with
preserving valve function, although the lat- preservation of the vein lumen and vein
ter has been presumed rather than proven. valves. In a follow-up paper in 1957,17 he re-
If CDT cannot be achieved, the clot re- ported 16 patients where TE was performed
moval or dissolution is unsuccessful or does in 14 legs and 2 arms with excellent results
not progress satisfactorily, or the concomi- in 12, good in 2, and poor in 2 patients. The
tant anticoagulation is contraindicated (eg, enthusiasm for TE created by Mahorner re-
IFVT in young women in the peripartum ceived strong support by the report by Haller
period, or in certain postoperative or trauma and Abrams in 1963.18 They presented 45
patients), then surgical TE or PMT is an ap- patients with IFVT who underwent TE. In
propriate choice. 34 patients with short history (<10 days), ex-
cellent bidirectional flow was established in
31 patients (91%). At follow-up after an aver-
age of 18 months, 26 out of these 31 patients
Historical Background of (84%) had normal legs, and where ascend-
ing venography was permitted in 13 patients,
Venous Thrombectomy normal patency of the deep venous system
(3) incompetence of the valves in the femo- pressure (PEEP) added during manipulation
ral and popliteal veins cannot reliably be as- of the thrombus to prevent perioperative PE.
sessed from an ascending venographic study The involved leg, contralateral groin, and
in the supine position. abdomen are prepared. A cell saver is used to
In a 1969 paper, Iliofemoral venous minimize blood loss. A longitudinal incision
thrombosis: reappraisal of thrombectomy,20 is made in the groin to expose the great sa-
William Edwards argued with Lansings phenous vein (GSV), which is followed to its
results and concluded that venous TE of- confluence with the common femoral vein
fers an effective and safe method of restor- (CFV), which is dissected up to the ingui-
ing flow in the deep venous system; when nal ligament. The superficial femoral artery
the thrombus is less than 10 days in dura- is cleared off 3 to 4 cm below the femoral
tion and is of the iliofemoral segment, TE bifurcation for construction of the arteriove-
is recommended; venograms at operation to nous fistula (AVF). In primary IFVT, where
determine the patency of the deep venous the thrombus originates in the iliac vein
system will aid in complete removal of the with subsequent distal progression of the
thrombus and give a basis for later compari- thrombus, a longitudinal venotomy is made
son and evaluation of long-term patency. In in the CFV, and a venous Fogarty TE cath-
the discussion, Lansing repeated his findings eter is passed upward through the thrombus
from the 5-year follow-up, still questioning into the IVC. The balloon is inflated and
the value of TE, but Haller, who was never withdrawn, these maneuvers being repeated
consulted about the follow-up report, stated until no more thrombotic material can be
that, at a recent visit to Louisville, he had extracted. With the balloon left inflated in
studied 17 patients in whom total removal the common iliac vein, a suction catheter is
of the thrombus had been possible and none introduced to the level of the internal iliac
had significant residual edema. Despite this vein to evacuate thrombi from this vein.
rebuttal, the impact of Lansings study, com- Backflow is not a reliable sign of throm-
bined with Karp and Wylies subsequent bus clearance since a proximal valve in the
1-page report21 in the Surgical Forum of 10 external iliac vein may be present in 25% of
patients in whom 8 had reocclusion of the cases, preventing retrograde flow in a cleared
femoral vein before discharge (even though vein. On the other hand, backflow can be
all had phlegmasia cerulea dolens and ex- excellent from the internal iliac vein and its
tensive thrombosis) was profound: only a few tributaries despite a remaining occlusion of
series on TE were subsequently published the common iliac vein. Therefore, an intra-
from the United States, in spite of the fact operative completion venogram is mandato-
that they all showed very good clinical re- ry. An alternative is the use of an angioscope,
sults in >75% of patients. which enables removal of residual thrombus
material under direct vision, or intravascular
ultrasound (IVUS). In early cases, the distal
Surgical Thrombectomy thrombus is usually readily extruded through
the venotomy by manual massage of the leg
The first TE for IFVT was performed by distally, starting at the foot. The Fogarty
Lwen, in Germany in 1937.22 Surgery today venous catheter, with a soft flexible tip, can
is performed under general intubation an- sometimes be advanced in retrograde fash-
esthesia with 10-cm positive end-expiratory ion without significant trauma. The aim is to
remove all fresh thrombi from the leg. The If phlegmasia cerulea dolens is the in-
venotomy is closed with continuous suture dication, because of the threat of impend-
and an AVF created using the saphenous ing venous gangrene, we start the operation
vein, anastomosing it end-to-side to the su- with fasciotomy of the calf compartments
perficial femoral artery (for illustrations, see to release the pressure and improve the cir-
Elkf et al12). An intraoperative venogram is culation immediately. If there is extension
performed through a catheter inserted in a of the thrombus into the IVC, the cava is
branch of the AVF. After a satisfactory com- approached transperitoneally through a
pletion venogram the wound is closed in lay- subcostal incision. The IVC is exposed by
ers and a closed suction drain is placed in the deflecting the ascending colon and duo-
wound to evacuate blood and lymphatic fluid denum medially. Depending on the veno-
that may accumulate after the operation. graphic findings relative to the top of the
In IFVT secondary to ascending thrombus, the IVC is controlled, usually just
thrombosis from the calf, the thrombus in below the renal veins. The IVC is opened
the femoral vein may be old and adherent to and the thrombus is removed by massage,
the venous wall. In such cases, the chance of especially of the iliac venous system; then if
preserving valve function has already been the femoral segment is involved, the opera-
lost and the opportunity to restore patency tion is continued in the groin as described
significantly diminished. A femoral segment previously. As an alternative, a retrievable
without functioning valves will lead to distal caval filter can be introduced before the TE
valve dysfunction in time, much as will fail- to protect against fatal PE. Heparin is con-
ure to achieve proximal patency. However, tinued at least 5 days postoperatively, and
a patent iliac venous outflow plus a compe- warfarin is started the first postop day and
tent profunda collateral system will most of continued routinely for 6 months. The pa-
the time achieve normal venous function. tient is ambulated the day after the operation
Therefore, if iliac patency is established but wearing a compression stocking and is usu-
the thrombus in the femoral vein is too old ally discharged after a week, to return after 6
to remove, it is preferable to ligate the femo- weeks for closure of the fistula.
ral vein. If normal flow in the femoral vein The objectives of a temporary AVF are
cannot be reestablished, we recommend ex- to increase blood flow in the thrombecto-
tending the incision distally and exploring mized iliac segment to prevent immediate
the orifices of the deep femoral branches. rethrombosis, to allow time for healing of
These are isolated, and venous flow is re- the endothelium, and to promote develop-
stored with a small Fogarty catheter. The ment of collaterals in case of incomplete
femoral vein is then ligated distal to the clearance or immediate rethrombosis of
profunda branches. In a 13-year follow-up the iliac segment. Usually the AVF is per-
after femoral vein ligation in this setting, formed between the saphenous vein and
Masuda et al23 found excellent clinical and the superficial femoral artery. More distally
physiological results without PTS. Finally, if placed AVFs have not been functional, in
there is evidence of iliac vein compression our experience.
on the completion venogram, which can A new percutaneous technique for fis-
occur in about 50% of left-sided IFVT, we tula closure was developed by Endrys in Ku-
recommend intraoperative endovenous iliac wait.24 Through a puncture of the femoral
angioplasty and stenting. artery on the opposite, surgically untouched
Results of Surgical
New Developments
Thrombectomy
One reason that induced surgeons to aban-
to Improve TE
don thrombectomy in the 1960s was the high There has been a significant improvement
mortality associated with early thombecto- of the results of surgical TE with the un-
my. In our series of more than 200 patients, derstanding to immediately restore iliac
mortality was less than 1%. There was no vein outflow in patients with iliac vein ob-
case of fatal PE in the perioperative period. struction as the major cause of their DVT.
In an RCT from Sweden,25 perfusion lung Control of iliac vein outflow immediately
scans were positive on admission in 45% of after TE can be achieved by intraoperative
all patients, with additional defects seen after venogram, angioscopy, or IVUS. In remain-
1 and 4 weeks in the conservatively treated ing obstructions, intraoperative endovenous
group, in 11% and 12% respectively, and in angioplasty and stenting are recommended.
the thrombectomized group in 20% and 0%, Schwarzbach et al. report excellent results
respectively. Tables with results from the lit- in 18 of 20 patients who maintained patency
erature are available in Rutherfords sixth after 21 months of followup.29 To improve
edition of Vascular Surgery.12 In the RCT patency and preservation of the valves in
from Sweden where TE was combined with the deep veins of the leg Blttler et al. has
a temporary AVF, 13% of patients had early combined TE of the iliac vein with throm-
rethrombosis of the iliac vein.26 In the Swed- bolysis applied under ischemic conditions
ish RCT, iliac vein patency at 6 months was to the leg veins.30 None of the 33 patients
76% in the surgical group compared with experienced clinically apparent recurrence
35% in the conservative group, as demon- within the first year. Clinical signs of the
strated by venography.26 This significant dif- postthrombotic syndrome were absent in all
ference was upheld after 5 and 10 years, with but 1 patient.
77% and 83% patency in the surgical group, With improved outflow through the
respectively, vs. 30% and 41% in the conser- iliac system by angioplasty and stenting
and increased inflow from the improved 200633the following recommendations are
distal patency by regional thrombolysis, the suggested:
temporary AVF may not be necessary. An
excellent review of contemporary venous Catheter-directed thrombolysis should
thrombectomy is published by Comerota be considered for proximal DVT,
and Gale 2006.31 especially iliofemoral thrombosis in
active patients at low risk for bleeding,
where the risk of the postthrombotic
Evidence-Based syndrome is higher than for more
Recommendations distal DVT (grade B). Systemic
thrombolysis should be avoided
The ACCP recommendations from 2004 because it is less effective, and because
(Bller, Agnelli, Hull, Hyers, Prins, Ras- the longer duration of therapeutic
kob)32 for treatment of acute venous throm- infusion required increases the risk of
boembolism concerning early thrombus hemorrhagic complications.
removal are the following: Surgical venous thrombectomy
should be considered for patients with
In patients with DVT or PE the symptomatic iliofemoral DVT who are
routine use of systemic thrombolytic not candidates for catheter-directed
treatment is not recommended (grade thrombolysis (grade C).
1A). Data concerning the short- and
In selected DVT patients, such as long-term effects of catheter-based
those with massive iliofemoral DVT mechanical intervention on the vessel
at risk of limb gangrene secondary to wall, venous valve, and pulmonary
venous occlusion, IV thrombolysis is vasculature are lacking and are
suggested (grade 2C). required before its role can be clearly
In patients with DVT, we recommend defined. This technique needs further
against the routine use of catheter- short- and long-term evaluation and
directed thrombolysis (grade 1C). eventually randomized controlled
In DVT patients confining catheter- trials before any recommendations can
directed thrombolysis to selected be made.
patients, such as those requiring limb
salvage is suggested (grade 2C). The ACCP recommendations from 2008
In patients with DVT, we recommend (Kearon, Kahn, Agnelli, Goldhaber, Raskob,
against the routine use of venous Comerota)34 are more in favor of early throm-
thrombectomy (grade 1C). bus removal:
In selected patients, such as patients
with massive iliofemoral DVT at risk In selected patients with extensive
of limb gangrene secondary to venous acute proximal DVT (eg, iliofemoral
occlusion, venous thrombectomy is DVT, symptoms for <14 days, good
suggested (grade 2C). functional status, life expectancy
>1 year) who have a low risk of
In the international consensus statement bleeding, we suggest that CDT may
guidelines according to scientific evidence be used to reduce acute symptoms
Figure 10.1 Treatment algorithm for acute iliofemoral thrombosis. CDT = catheter-
directed thrombolysis; PMT = percutaneous mechanical thrombectomy.
10. Nicolaides AN, Hussein MK, Szendro G, 21. Karp RB, Wylie EJ. Recurrent thrombosis
Christopoulos D, Vasdekis S, Clarke H. The after iliofemoral venous thrombectomy. Surg
relation of venous ulceration with ambulatory Forum. 1966;17:147-9.
venous pressure measurements. J Vasc Surg. 22. Lwen A. Uber thrombectomie bei
1993;17:414-9. Venenthrombose und Arteriespasmus.
11. Nicolaides AN, Sumner DS, eds. Investigation Zentralbl Chir. 1937;64:961-8.
of Patients with Deep Vein Thrombosis and 23. Masuda EM, Kistner RL, Ferris EB. Long-
Chronic Venous Insufficiency. Los Angeles: term effects of superficial femoral vein
Med-Orion Publishing Company; 1991. ligation: thirteen-year follow-up. J Vasc Surg.
12. Eklf B, Rutherford RB. Surgical 1992;16:741-9.
thrombectomy for acute deep venous 24. Endrys J, Eklf B, Negln P, Zka I, Peregrin
thrombosis. In: Rutherford RB, ed. Vascular J. Percutaneous balloon occlusion of surgical
Surgery. 6th ed. Elsevier, Saunders; 2005. arteriovenous fistulae following venous
p2188-98. thrombectomy. Cardiovasc Inter Rad.
13. See-Tho K, Harris EJ Jr. Thrombosis with 1989;12:226-9.
outflow obstruction delays thrombolysis and 25. Plate G, Ohlin P, Eklf B. Pulmonary
results in chronic wall thickening of rat veins. embolism in acute iliofemoral venous
J Vasc Surg. 1998;28:115-22. thrombosis. Br J Surg. 1985;72:912.
14. Kistner RL. Valve repair and segment 26. Plate G, Einarsson E, Ohlin P, Jensen R,
transposition in primary valvular insufficiency. Qvarfordt P, Eklf B. Thrombectomy with
In: Bergan JJ, Yao JST, eds. Venous Disorders. temporary arteriovenous fistula: the treatment
Philadelphia: WB Saunders; 1991. p261-72. of choice in acute iliofemoral venous
15. Homans J. Exploration and division of the thrombosis. J Vasc Surg. 1984;1:867-76.
femoral and iliac veins in the treatment 27. Plate G, Akesson H, Einarsson E, Ohlin
of thrombophlebitits of the leg. JAMA. P, Eklf B. Long-term results of venous
1941;224:179-86. thrombectomy combined with a temporary
16. Mahorner H. New management for arterio-venous fistula. Eur J Vasc Surg.
thrombosis of deep veins of extremities. Am 1990;4:483-9.
Surg. 1954;20:487-98. 28. Plate G, Eklf B, Norgren L, Ohlin P,
17. Mahorner H, Castleberry JW, Coleman WO. Dahlstrm JA. Venous thrombectomy for
Attempts to restore function in major veins iliofemoral vein thrombosis: 10-year results
which are the site of massive thrombosis. Ann of a prospective randomized study. Eur J
Surg. 1957;146:510-22. Endovasc Surg. 1997;14:367-74.
18. Haller JAJ, Abrams BL. Use of thrombectomy 29. Schwarzbach MH, Schumacher H, Bckler
in the treatment of acute iliofemoral venous D, Frstenberger S, Thomas F, Seelos R,
thrombosis in forty-five patients. Ann Surg. et al. Surgical thrombectomy followed by
1963;158:561-9. intraoperative endovascular reconstruction for
19. Lansing AM, Davis WM. Five-year follow-up symptomatic iliofemoral venous thrombosis.
study of iliofemoral venous thrombectomy. Eur J Vasc Endovasc Surg. 2005;29:58-66.
Ann Surg. 1968;168:620-8. 30. Blttler W, Heller G, Largiadr J, Savolainen
20. Edwards WH, Sawyers JL, Foster JH. H, Gloor B, Schmidli J. Combined regional
Iliofemoral venous thrombosis: reappraisal thrombolysis and surgical thrombectomy for
of thrombectomy. Ann Surg. 1970;171: treatment of iliofemoral vein thrombosis. J
961-70. Vasc Surg. 2004;40:620-5.
31. Comerota AJ, Gale SS. Technique of 33. Nicolaides AN, Fareed J, Kakkar AK, Breddin
contemporary iliofemoral and infrainguinal HK, Goldhaber SZ, Hull R, et al. Prevention
venous thrombectomy. J Vasc Surg. and treatment of venous thromboembolism:
2006;43:185-91. international consensus statement. (guidelines
32. Bller HR, Agnelli G, Hull RD, Hyers TM, according to scientific evidence). Int Angiol.
Prins MH, Raskob GE. Antithrombotic 2006;25:101-61.
therapy for venous thromboembolic 34. Kearon C, Kahn SR, Agnelli G, Goldhaber
disease: the seventh ACCP conference on S, Raskob GE, Comerota AJ. Antithrombotic
antithrombotic and thrombolytic therapy. therapy for venous thromboembolic disease.
Chest. 2004;126:401S-28S. Chest. 2008;133:454S-545S.
11
IVC Filters
Indications, Techniques, and outcomes
117
of placement, increased patient and practi- treatment of DVT and PE, and in this chap-
tioner awareness due to educational efforts, ter, we discuss the indications, techniques,
and medicolegal concerns about thrombo- and outcomes for placement of IVC filters.
embolism. However, relatively strict adher-
ence to conservative indications for filter
placement is advised. Evolving indications IVC Filter Indications
such as PE prophylaxis in trauma or bariatric
surgery patients are of great interest since ve- Only one trial, by Decousus and colleagues,
nous thromboembolic disease is a significant published in the New England Journal of
complication in the management of these Medicine in 1998, has been performed that
patients. However, prospective randomized established the role of inferior vena cava fil-
data in these subgroups are, in general, lack- ter use.1-3 This trial demonstrated that IVC
ing, and while interest exists in studying filters are effective in preventing pulmonary
these indications, they are not without con- embolism in patients with proximal deep
troversy within the pulmonary, hematologic, venous thrombosis. Four hundred patients
general medical, surgical, and endovascular with proximal DVT were randomly assigned
communities. Retrievable filters have further to placement of IVC filter or no placement of
lowered our threshold to utilize IVC filters, IVC filter, with 200 patients in each group.
especially in these subgroups where the indi- They found significant differences in inci-
cations are less clear. Retrievability does not dence of all PE between the 2 groups at day
equate with actual retrieval. It is thought that 12. In the filter group, 2 patients, or 1.1%,
since they can be retrieved and are tempo- had either a symptomatic or asymptomatic
rary, most complications due to the filter PE, whereas 9 patients in the no-filter group,
can be reduced or eliminated with the use of or 4.8%, had a symptomatic or asymptom-
retrievable filters. atic PE. Four types of permanent IVC filters
The retrievable filters have unique lia- were placed, and the authors concluded that
bilities and these devices do not yet have in high-risk patients with proximal DVT,
long-term follow up data. With the recogni- there was a significant early beneficial effect
tion of the fact that patients fail to follow up of vena cava filter placement for the preven-
and undergo retrieval, and the technical in- tion of PE. However, this was in part offset
ability to retrieve the filter either due to place- by an increased rate for recurrent DVT in
ment exceeding the indicated dwell time or the filter group and an absence of any dif-
to simply technical inability to remove the ference in mortality between the groups. To
filter, a relative minority of filters are ever ac- this day, this remains the only prospective
tually retrieved. Thus, although retrievabil- randomized controlled trial regarding the
ity lends a new and exciting dimension to use of IVC filters to prevent PE.
IVC filter placement and creates the oppor- Despite this, the use of IVC filters has
tunity to do so in a temporary fashion, these increased dramatically, especially in the
devices will require further study to define United States. This increase is due partially
their exact long-term role in the treatment of to greater awareness of venous thromboem-
patients with venous thromboembolism and/ bolic disease (VTE) in general. However,
or prevention of complications from venous it is clear that further work to evaluate the
thromboembolism. Previous chapters have rational role of IVC filter placement in the
discussed the pathophysiology and medical treatment of patients with VTE or of pa-
tients in need of VTE prophylaxis is neces- discharge from the hospital within an average
sary. The clearest indications for IVC filter of 10 days. These authors concluded that the
placement are recurrent acute or chronic super-obese male patient is at a very high risk
VTE on adequate therapeutic anticoagula- of developing PE and that their relative risk
tion, contraindication to or complication significantly exceeds that of other Roux-en-Y
of anticoagulation in the presence of VTE, gastric bypass patients. It was also noteworthy
and an inability to achieve or maintain that they found that the PE risk lasted several
therapeutic anticoagulation in the presence weeks after discharge, therefore extending
of documented VTE. Relative indications PE prophylaxis for several weeks after surgery
include the following: chronic PE treated may be warranted. In the group that had a
with thromboendarterectomy; massive PE BMI >55 kg/m2, patients were assigned to ei-
treated with thrombolysis and/or thrombec- ther receive or not receive an IVC filter. In
tomy, filter placement for thrombolysis for the group with the BMI >55 kg/m2 that had
iliocaval DVT (controversial); large, free- no IVC filter placed, the PE rate was 28%
floating proximal or caval DVT; VTE with and the mortality was 11%. In the BMI >55
limited cardiopulmonary reserve; recurrent kg/m2 group who received filters, there was
PE with a filter in place; difficulty establish- a 0% incidence of PE and a 0% PE mortal-
ing therapeutic anticoagulation; and poor ity rate. Their conclusion was that there was
compliance with anticoagulation or high a significant reduction in the perioperative
risk for anticoagulation such as unsteady gait PE rate when the patient with a BMI >55 kg/
or frequent falls. Prophylactic indications for m2 had an IVC filter placed. IVC filters were
IVC filters are usually cited when primary used in addition to subcutaneous heparin and
prophylaxis is not feasible but PE prevention sequential compression devices (SCD). In
is deemed necessary. Examples of this are a the morbidly obese, technical issues for IVC
trauma patient with a high risk for VTE, a filter placement can arise related to imaging
patient with a high risk for VTE undergoing table weight restrictions, inadequate imag-
a surgical procedure, or a medical condition ing due to body mass, and x-ray penetration
in a patient with a high risk for VTE.4 issues. Alternative imaging techniques such
Surgery for morbid obesity is an area as intravascular ultrasonography may be use-
in which prophylactic IVC filters are often ful in this setting. Reports exist of the use of
entertained and placed. Recent data indi- carbon dioxide cavography in these patients.
cate that the fatal PE rate with morbid obe- However, published evidence documenting
sity surgery within 60 days may be as high this practice so far is limited.
as 0.85%. Risk factors that contribute to this IVC filter placement in the hands of
are underlying venous insufficiency, sleep an experienced practitioner is an extremely
apnea, and the patients baseline body mass safe and low-risk procedure; however, there
index (BMI).5 Abou-Nukta and colleagues are contraindications to placement that in-
compared age and BMI of morbidly obese clude the following: an IVC that is either
patients in a randomly selected Roux-en-Y occluded or completely filled with thrombus
gastric bypass patient control group. Patients with no safe region to deploy a filter below
were broken down into BMI >55 kg/m2 and the right atrium; active septicemia or bacte-
BMI <55 kg/m2. In super-obese men with remia, which may provide an opportunity to
BMI >55 kg/m2 the incidence of PE was 4%. colonize the filter with active infection; and
Nine of the 11 patients developed PE after a thrombus presenting between the proposed
access site and site of deployment of the fil- record for safety and durability than other,
ter. Sepsis or bacteremia is often considered better known familiar permanent filter de-
a relative contraindication only, and it is con- signs, can and often do at this point remain
troversial as to whether infection should in- as permanent filters. Consequently, addi-
terfere with IVC filter placement. tional studies to document their long-term
In patients who have a retrievable filter safety, durability, and functionality still need
in place, and if during the course of treat- to be done.
ment it is concluded that the filter will no Suprarenal IVC filter placements may
longer be required or is no longer effective, need to be considered in some cases. Situa-
there are several considerations for removal tions in which this might be considered are
of retrievable filters. If the patients baseline renal vein thrombus, pregnancy, IVC clots
low risk for PE has returned after resolu- either above the renal veins or above an IVC
tion of other conditions or if the patient filter, duplicated IVC, PE that has been doc-
can return to routine anticoagulation, con- umented to occur from a gonadal vein, or an
sideration should be given to filter removal. excessively short infrarenal IVC.6
Retrievable filters allow the option for re- Occasionally, superior vena cava (SVC)
moval if it has become ineffective from mi- filters are entertained. In our experience,
gration, excessive angulation, or fracture. If these are rarely, if ever, necessary, and even
a continued need for the filter exists, these proponents of this treatment acknowledge
filters could be replaced in a more optimal the controversy surrounding it.7 However,
position or location. Active bacteremia or some literature does exist that documents
septicemia is cited by some as a reason not their use in certain cases and the interven-
to place a filter; however, if a patient is ac- tionalist or surgeon should be aware of these
tively bacteremic at the time filter placement reports, if only to address the concerns of
becomes necessary, a retrievable filter may consulting physicians who request their
be a good choice in such a patient. Another placement. Upper extremity DVT is in the
contraindication to filter retrieval is a filter vast minority, comprising less than 10% of
that is largely or fully laden with clot that all DVTs. However, upper extremity and
cannot be remedied by endovascular means. major central thoracic vein thromboses are
This filter cannot be removed since it pres- becoming more common with the increased
ents an undue risk of PE during the retrieval use of vascular access devices. Although its
procedure. For this and various other rea- incidence is uncommon, upper limb DVT
sons, many retrievable filters that are initially is more likely to lead to PE, but these emboli
placed with the intent of future removal will are smaller and less likely to be symptomatic
become permanent. An additional contra- or fatal. Considerable controversy contin-
indication to filter removal is persistence of ues to surround these questions with regard
clot in the pelvic or lower extremity veins to management of upper extremity DVT.
unless the patient becomes a candidate for Placement of a vena cava filter in the reverse
anticoagulation therapy. Until the patient orientation in the SVC is conceptually a po-
with the retrievable filter has their risk for PE tential method of managing these thrombo-
returned to normal baseline, or if the patient ses in patients who have a contraindication
becomes a candidate for anticoagulation, fil- to anticoagulation or have higher fall risk,
ters likely will not be removed. Retrievable which is not uncommon in these popula-
filters, although they have a shorter track tions. Filter placement in this location and
in reverse orientation is technically chal- from the left iliac vein excludes a duplicated
lenging. There is a very short landing zone cava. The width of the vena cava should be
for placement of these filters in the SVC documented. After documenting caval anat-
below the confluence of the major thoracic omy and excluding caval thrombosis, du-
veins and above the right atrium. Any com- plication, megacava, renal vein anomalies,
plications, such as inadvertent deployment and/or the presence of a left-sided vena cava,
even slightly beyond the intended landing filter placement can proceed. While the in-
area for the filter, could result in emboliza- dividual types of IVC anomalies and condi-
tion to the right atrium, and any perforations tions are rare, the incidence of an anatomic
that occur would likely be more serious. A variation of the IVC, taken all together, ap-
femoral insertion kit is used if the filter is to proaches 15%. The filter is typically placed
be placed from the jugular position, and a with its apex at or just below the level of the
jugular insertion kit is used if the filter is to lowest renal vein. Some manufacturer in-
be placed from the femoral approach. structions for use (IFU) indicate placement
of the upper portion of the cone of the filter
within the outwash of the renal veins to aid
IVC Filter Techniques the resolution of thrombi captured by the fil-
ter. Placement above the renal veins should
IVC filter implantation is a relatively straight- be reserved for the indications stated earlier.
forward procedure that ought to be well Deployment protocols for specific filters are
studied and mastered by the vascular inter- per the IFU for the individual filters. Table
ventionalist or endovascular surgeon. Patient 11.1 contains filter device details, and fluoro-
preparation, intraprocedural imaging, and scopic images of several types appear in Fig-
meticulous technique are required. Most ure 11.1. The practitioner should be familiar
implants are performed either via the com- with the manufacturers IFU of the various
mon femoral or the jugular vein approaches. filters available at their institution that they
Additional approaches can be via the antecu- routinely use. Successful deployment is con-
bital or the subclavian veins, as well as other sidered routine once the above conditions
veins in unique situations. Extremely careful have been assessed and should occur 95%
technique should be utilized when obtain- to 99% of the time in appropriately selected
ing percutaneous access since many patients patients.
are either on anticoagulation, have had one In the event that any of the abovemen-
venous thrombosis, or have had a bleeding tioned anomalies of the vena cava are pres-
complication. The location and side of any ent, significant alterations of filter placement
DVT should be determined before access is are required. In duplicated cava, a filter in
obtained. Ultrasound-guided access to the each cava should be considered. Alterna-
vein is recommended. An initial cavogram tively, a filter in the suprarenal IVC is an
is obtained using a pigtail catheter, another option. If megacava is encountered, options
multiside hole catheter, or the kits delivery include the use of a Birds Nest Filter (Cook)
sheath, which is optimally placed in the or placing 2 filters with 1 filter in each com-
confluence of the iliac veins either from mon iliac vein. With a left-sided vena cava,
the femoral or jugular vein approach; or al- the filter can be placed in the solitary left-
ternatively, with a catheter tip placed in the sided vena cava. Careful documentation of
left iliac vein since opacification of the cava the presence of the aorta to the right of the
SheathOuter
Retrievable Diameter(Fr) MaximumIVC Jugular
Filter Manufacturer (Route) Material [Jugular] Diameter(mm) Placement
Opt Ease Cordis Endovascular Yes (4 weeks) Nitinol hypotube 8.5 30 Yes
Venous Thromboembolic Disease
(Femoral)
5/13/11 10:27 AM
IVC Filters 123
Figure 11.1 Fluoroscopic imaging of 6 common IVC filter designs in use today. 1. Greenfield Titanium (Boston
Scientific). 2. Trap Ease (Cordis). 3. VenaTech LP (VenaTech). 4. Gnther Tulip (Cook). 5. Opt Ease (Cordis). 6. G2 X
(CR Bard). The first 3 are examples of permanent filter designs. The latter 3 are retrievable and have hooks at the top
or bottom (for either transjugular or transfemoral access for retrieval, respectively) via dedicated retrieval systems or
interventional sheath/loop snare combinations.
left-sided vena cava is needed before deploy- injection of contrast at 20 mL/s for a total of
ment of the filter. For renal vein anomalies, 40 mL. Another less established technique is
such as a circumaortic left renal vein or the to document the major vena caval branches
presence of multiple renal veins on one side, with selective catheterization without con-
the filter should be placed below the level of trast administration.
the lowest renal vein. Many practitioners have chosen to uti-
A marker pigtail catheter or sizing pig- lize retrievable filters on a fairly regular basis
tail specifically designed to measure the size given their apparent reliability as long-term
of the vena cava can be used if it was not filters from early experience.8,9 Retrieval pro-
clearly documented prior to the procedure. cedures are reviewed in Figures 11.2 and 11.3.
Typically, a cavogram is performed with an The Bard recovery filter has been updated,
Figure 11.2 Recovery sequence of G2 X (CR Bard) IVC filter using a recovery cone (Bard): 1. Initial cavagram
and positioning of recovery cone at apex of filter. The filter is seen to be free of embolic debris or thrombus. 2. After
positioning of the cone over the filter, the sheath is advanced to close the cone. 3. Retraction of the filter into the cone.
4. After withdrawal of the
filter and sheath removal,
a final cavogram
demonstrates complete
removal of the filter.
Figure 11.3 Recovery sequence of Gnther Tulip (Cook) IVC filter using a loop snare and sheath: 1. Initial
cavagram is performed. The filter is free of embolic debris or thrombus. 2. Positioning of snare at apex of filter. 3.
After positioning of the snare over the filter and capture of the hook, the sheath is advanced to close the filter and
begin recovery. 4. Retraction of the filter into the sheath by the snare. 5. After withdrawal of the filter, final cavagram
is performed to demonstrate absence of spasm, thrombus, or extravasation. 6. Recovered filter with mix of chronic
fibrinous material at the apex and acute thrombotic material that is typically seen at retrieval and probably forms while
filter is in sheath with static blood.
which is now the Bard G2 X, and it can be re- These retrieval procedures are performed
covered with either a recovery cone or with a with the patient under anticoagulation in
snare from a jugular approach. Likewise, the many cases.10 The patient is ideally accessed
Gnther Tulip and the Cook Celect filters under sedation via the right internal jugu-
can be recovered with a snare from a jugu- lar vein with ultrasound guidance starting
lar approach. The Cordis Opt Ease is recov- with a micropuncture technique and then
ered from a femoral approach also using a upsizing to a short, interventional sheath.
snare. Anecdotal evidence indicates that the Using standard techniques, the inferior vena
Cordis Opt Ease has a shorter recovery win- cava is accessed, and a catheter or sheath is
dow (Figure 11.4). For this reason, we would placed at or below the level of the filter. A
typically prefer any of the other retrievable cavogram is obtained, and documentation
filters, as there is accumulating evidence of patency of the filter or determination of
that they can often be safely retrieved well clot burden is performed. In cases of small
beyond the manufacturers IFU. This may clots within the filter cone (25% or less cone
extend the opportunity for filter retrieval in volume), retrieval may be performed safely.
a number of patients whose risk for PE does However, when there is large clot burden
not return to baseline early, who are initially (>25%), filters are usually not removed, and
lost to follow-up, or whose ability to tolerate patients are typically anticoagulated and
anticoagulation does not return for a longer brought back at a later time for interval reas-
period of time. However, this needs to be sessment for possible later retrieval. If a pat-
done on an individualized basis, with great ent filter is documented, a recovery cone in
care and attention to signs that retrievability the case of the Bard G2, a loop snare in the
may not be possible at the time of removal. case of the Bard G2 X, or a loop snare or
Figure 11.4 Opt Ease (Cordis Endovascular) IVC filter retrieval attempted with resulting caval stenosis and inability
to remove filter. The patient presented late for retrieval but strongly wished to undergo attempted retrieval despite
counseling about decreased potential for successful filter removal and increased risk of stenosis or injury. A. Caval
stenosis and retained IVC filter. The patient was maintained on warfarin, in accordance with his history of recent PE
and DVT. B. Three-week CT venogram confirming resulting stenosis (rather than spasm) and continued filter patency.
C. Four-month CT venogram again documenting continued filter patency and some favorable remodeling of the IVC. It
was planned to continue warfarin in this patient (who suffered DVT and PE perioperatively after oncologic surgery) for
at least a full 12 months and follow the IVC morphology on subsequent CT scans during oncologic followup.
the respective recovery systems for the Cook filter placement under fluoroscopy alone.
Celect or Tulip filters, is brought into place. Follow-up IVUS imaging can also provide
A snare is brought around the hook and the an additional confirmation of correct filter
sheath is advanced over the recovery cone or deployment and the relative position to renal
the snare to collapse the filter within the de- veins following deployment. This combined
vice. The filter is then withdrawn carefully IVUS and fluoroscopic approach can be a
under fluoroscopy to document that it is not single-access site technique. Ashley and col-
dislodged from the sheath. The catheter or leagues have argued that IVUS more accu-
sheath is then reinserted into the abdominal rately localizes the renal veins and measures
vena cava and a cavogram is obtained to rule the diameter of the vena cava than contrast
out perforation, stenosis, or thrombus forma- venography.15 Transabdominal ultrasonog-
tion in the cava following the procedure. raphy can fail to identify proper anatomical
Careful attention to both meticulous access landmarks in up to 10% of patients, but in
technique and hemostasis of the internal centers that have made a committed effort to
jugular vein, particularly if the patient is an- this approach, transabdominal ultrasonogra-
ticoagulated, is of course critical. phy can be utilized to document anatomy
Rosenthal and others have identified and adequate filter positioning.16 If technical
techniques to place inferior vena cava filters support and expertise exist to obtain quality
using intravascular ultrasonography (IVUS) transabdominal venous images, and body
at the bedside of patients in the ICU, pro- habitus, bowel gas, or ileus does not preclude
viding a way to insert an IVC filter in pa- adequate visualization, this is a valuable al-
tients who cannot be easily transported. The ternative technique.
IVUS bedside ICU technique has been ad-
opted at other institutions as well.11-14 Typi-
cally, this requires a 2-vein accessvia both IVC Filter Outcomes
the right and left common femoral veins
and specialized standing protocols to bring Complications of IVC filter placement along
IVUS and filter deployment equipment to with their incidence are important data to
the ICU.11 More recently, beyond the ini- incorporate into the decision to place, the
tial IVUS studies done by Rosenthal, other placement procedure, and follow-up of these
workers have documented a single-vein ac- devices. Migration and access site throm-
cess technique using markings and dimen- bosis are 2 frequent and potentially highly
sions obtained by using IVUS and using significant adverse events. Conservative esti-
this length data to guide deployment. An- mates place access site thrombosis at a rate
other use for IVUS is in a procedure suite of 0% to 6% of placements. Other reports
with fluoroscopy in cases where avoidance cite access site thrombosis occurring at 2%
of contrast administration, such as in cases to 35%. Access site thrombosis appears to be
of contrast allergy or renal insufficiency, is greater with the femoral route. Ultrasound
required. Vascular landmarks are assessed guidance, use of careful micropuncture
using IVUShepatic veins, left renal artery, needle technique, and the smallest-diameter
renal veins, and iliac veins. Then, these land- delivery system appropriate to a given pa-
marks, fluoroscopy of the IVUS catheter in tient may limit such thromboses. Access site
the planned filter landing zone can identify thrombosis is a particularly worrisome out-
a bony landmark, which then serves to guide come when placement was for a purely pro-
phylactic indication. Filter migration is also into the cardiac chambers, although rare, is
a worrisome problem with an uncertain in- a dreaded complication of filter placement.
cidence. Estimates of migration occurrence Injury to adjacent retroperitoneal structures,
vary from 0%18% to 3%69%. Migration which fortunately is extremely rare, is also
may be limited to some extent by careful compelling evidence of a perforation relat-
IVC sizing and meticulous filter deployment ed to filter placement. Occasionally, a filter
procedure in accordance with each devices strut has been seen to migrate through the
instructions for use. Reports aiming to quan- caval wall and into adjacent vessels, verte-
tify both access site thrombosis and filter brae, or viscera. In the absence of bleeding
migration are heavily influenced by follow- or gastrointestinal complications, this prob-
up protocol and imaging modality as well ably can best be managed by serial moni-
as whether or not asymptomatic patients are toring. Gastrointestinal complications of
screened for these adverse outcomes. filter placement may include bleeding or
Numerous other complications occur perforation of the adjacent viscera, particu-
related to these devices: death may occur, larly the duodenum. These rare cases may
but well under 1% of cases; recurrent clinical present either as abnormalities on CT scan-
PE may occur in 2% to 5%; filter emboliza- ning for related or other reasons or upper gas-
tion may occur in 2% to 5%; IVC occlusion trointestinal bleeding when visceral ulcers
may occur in 2% to 30%; depending on the are created by local strut trauma. Central
series examined, IVC penetration can occur venous guidewire manipulation or central
in 0% to 41%; and filter fracture can occur in line insertions have also contributed to the
2% to 10% (Table 11.2). IVC penetration is a dislodgement of previously placed IVC fil-
controversial issue. A computed tomography ters. Guidewires may become entangled
(CT) scan may poorly identify a collapsed in the vena cava filter during other venous
vena cava around the limbs of an IVC fil- interventional procedures with a J or other
ter and present the image of caval penetra- wire advanced well into the vena cava and/or
tion, when in fact, this is an underfilled cava during central line placement. It is possible
collapsed around an IVC filter. It may be that many of these may be unrecognized
more reasonable to be concerned with caval by the operator and manifest only as mild
penetration when clear evidence of this ex- to moderate difficulty removing the wire at
ists and/or a complication of such penetra- some point during the procedure. If such a
tion occurs, such as a visceral perforation or wire becomes impossible to remove or filter
bleeding complication. Filter misplacement entanglement is suspected, intervention to
or migration may occur despite meticulous remove it using catheter techniques may be
technique and careful deployment. Rapid necessary. Such conditions are likely under-
and precise deployment of the filter reduces recognized and probably underreported. PE
the incidence of strut asymmetry, incom- or recurrent PE despite the presence of a
plete opening (Figure 11.5), and tilting. Mis- filter can occur. Anticoagulation should be
placement of the filter at insertion may occur maintained whenever possible to minimize
through a number of mechanisms, and these recurrent PE. Renal complications include
filters are at an increased risk for migration. renal vein thrombosis related to suprarenal
Migration of either fractured components or filter placement. Rarely, obstructive uropa-
the entire device may occur. Central migra- thy due to a pelvic obstruction of the ureter
tion of a filter into the right atrium or further related to vena cava wall strut perforation
AccessSite
Major Postfilter Thrombosis
Procedural RecurrentPE (Symptomatic) CavalThrombosis
Hoppe et al., 23
2007 5% 2% 5%
Figure 11.5 Fluoroscopic (A) and CT (B) imaging of partial incomplete opening seen with a Greenfield IVC filter
(Boston Scientific). Asking the patient to cough or individual strut manipulation with an angled tipped catheter are
options to try to correct this during the initial deployment procedure, but strut tips are usually firmly placed in the IVC
wall immediately after release. This deployment configuration will afford some protection from large PE. However, if PE
occurs in this situation or more optimal filter protection from PE is needed, a second IVC filter below (if cava and filter
are patent) or a suprarenal IVC filter (if the filter itself is shown to be a possible source or caval thrombus is present)
may be considered.
Filter 30day
Filter Migration Filter 30dayMortality Mortality Recurrent
Fracture (Major) Infection (duetofilter) (overall) DVT
0.0%59.3%a <<1%
0.0%10.9%c
6%8%d (Rare)
0%32%
<1% f
<1% 20%
2% 6% h
a
All migration, includes single-device series. bFilter misplacement. cNot limited to 30 days. dSingle, early device data
cited here. eAll migration. fWith clinical consequences. gIatrogenic PE. hCombined migration and malposition.
may occur. This may manifest as either he- experience has been gained through multiple
maturia or hydronephrosis. Careful imaging models of the Greenfield filter. The Green-
and active questioning of position during field filter is widely regarded as a filter with
deployment will avoid other such highly an excellent long-term safety and efficacy re-
unusual mishaps such as placement in the cord. Greenfield and colleagues have main-
mesenteric vein via the portal system and/ tained large registries of filters that they have
or aortic placement, both of which have placed, and their 20-year experience with
been documented to occur in extremely rare these filters documented an incidence rate of
cases.24 Frank phlegmasia complicating pro- recurrent PE at 4% and a caval patency rate
phylactic filter placement is a real, though of 96%, and caval movement of little or no
uncommon, entity.25 Concern regarding this clinical significance was seen in 8% of their
is one of the factors driving increased consid- cases. They reported minimal procedural
eration for retrievable filters. Fortunately, in morbidity and mortality, and they concluded
patients who can undergo either mechanical in their analysis that insertion of the stainless
or pharmacomechanical thrombolysis with steel Greenfield vena cava filter provides pro-
the filter in place, phlegmasia from a filter tection from PE while maintaining patency
occlusion can be treated with a minimally of the IVC over long-term follow-up.
invasive approach to attempt to resolve the Reports about experience with retriev-
thrombus and then anticoagulation can be able filters that have been removed after ex-
resumed (Figure 11.6). tended dwell times past the manufacturers
The IVC filter with the longest clinical IFU is beginning to become available.27
experience is the Greenfield filter.26 Extensive When retrieval is attempted and the patient
Figure 11.6 A patient presenting with bilateral sudden onset limb edema and discomfort occurring late after a
filter placed for VTE immediately after spine surgery. There was no frank phlegmasia or compromised lower extremity
viability. A CT venogram was the initial study obtained (A)and was to some extent equivocal as is often the case due
to venous contrast opacification patterns on CT. Correlation with catheter venography did show that the filling defect
below the outflow of the right renal vein was in fact filter-associated thrombus (marked by arrow in A ). Initial catheter
venogram (B) confirms massive illiocaval thrombosis. Extensive pharmacomechanical debulking of clot was performed
with iliofemoral and lower IVC improvement (C) in an effort to avoid long-term complications related to lower extremity
venous insufficiency. Due to contrast volume and sedation time, it was elected to bring the patient back for a staged
second procedure to complete thrombus removal. He was maintained on anticoagulation and had symptomatic
improvement but required coronary angiography and intervention. Due to this and transient renal impairment, it was
elected not to reintervene on his cava and IVC filter. He was discharged on warfarin anticoagulation with continued
symptomatic improvement.
was not lost to follow-up or had secondarily patients are either lost to follow-up, the deci-
requested to keep the filter in place, techni- sion to leave the filter in place is made, or the
cal success approaches 85% or greater. Fac- patients themselves decline to undergo the
tors correlating with failure of filter retrieval removal procedure.
were due to device angulation, tilt, or cant- Filter placement and its use in trauma
ing within the IVC. Most clinical series of patients have been assessed by Quirke and
retrievable filters are reported to be free or colleagues by questionnaire.28 They que-
relatively free of symptomatic PE, bleeding, ried respondents that were largely at level I
thrombosis, or caval stenosis. These proce- trauma centers that have over 1000 trauma
dure times should be short when retrieval is admissions per year. These centers reported
straightforward. It is reasonable to conclude a low complication rate, with most of these
that when filter retrieval is attempted, it is centers reporting one or fewer complications
performed successfully in a large proportion per year despite their relatively high volume.
of cases, despite growing evidence that many Interesting trends were observed in these
caval filters for the prevention of pulmonary placement in the military multitrauma
embolism. Cochrane Database Syst Rev. patients: a single center experience. Vasc
2007;17(4):CD006212. EndovascSurg. 2009;43(5):497-501.
4. Kaufman JA. Development of a research 13. Wellons ED, Rosenthal D, Shuler FW, Levitt
agenda for IVC Filters. Endovascular Today. AB, Matsuura J, Henderson VJ. Real-time
2007;6(10):67-70. intravascular ultrasound-guided placement of
5. Abou-Nukta F, Alkhoury F, Arroyo K, Bakhos a removable inferior vena cava filter. J Trauma.
C, Gutweiler J, Reinhold R, et al. Clinical 2004;57(1):20-5.
pulmonary embolus after gastric bypass 14. Ebaugh JL, Chiou AC, Morasch MD,
surgery. Surg Obes Relat Dis. 2006;2(1): Matsumura JS, Pearce WH. Bedside vena cava
24-8. filter placement guided with intravascular
6. Georgiades CS, Hong K. Inferior vena cava ultrasound. J Vasc Surg. 2001;34(1):21-6.
filters. In: Cameron JL, ed. Current Surgical 15. Ashley DW, Gamblin TC, Burch ST, Solis
Therapy. 9th ed. Philadelphia: Mosby; 2008. MM. Accurate deployment of vena cava filters:
p908-13. comparison of intravascular ultrasound and
7. Usoh F, Hingorani A, Ascher E, Shiferson contrast venography. J Trauma. 2001;50(6):
A, Tran V, Patel N, et al. Superior vena cava 975-81.
perforation following the placement of a 16. Amankwah KS, Seymour K, Costanza M,
superior vena cava filter in males less than 60 Berger J, Gahtan V. Transabdominal duplex
years of age. Vascular. 2009;17(1):44-50. ultrasonography for bedside inferior vena
8. Berczi V, Bottomley JR, Thomas SM, Taneja cava filter placement: examples, technique,
S, Gaines PA, Cleveland TJ. Long-term and review. Vasc Endovascular Surg.
retrievability of IVC filters: should we abandon 2009;43(4):379-84.
permanent devices? Cardiovasc Intervent 17. Becker DM, Philbrick JT, Selby JB.
Radiol. 2007;30(5):820-7. Inferior vena cava filters: indications,
9. Comerota AJ. Retrievable IVC filters: a safety, and effectiveness. Arch Intern Med.
decision matrix for appropriate utilization. 1992;152(10):1985-94.
Perspect Vasc Surg Endovasc Ther. 18. Mohan CR, Hoballah JJ, Sharp WJ, Kresowik
2006;18(1):11-7. TF, Lu CT, Corson JD. Comparative efficacy
10. Hoppe H, Kaufman JA, Barton RE, Petersen and complications of vena caval filters. J Vasc
BD, Lakin PC, Deloughery TG, et al. Surg. 1995;21(2):235-45.
Safety of inferior vena cava filter retrieval in 19. Athanasoulis CA, Kaufman JA, Halpern
anticoagulated patients. Chest. 2007;132(1): EF, Waltman AC, Geller SC, Fan CM.
31-6. Inferior vena caval filters: review of a 26-year
11. Rosenthal D, Wellons ED, Levitt AB, single-center clinical experience. Radiology.
Shuler FW, OConner RE, Henderson VJ. 2000;216(1):54-66.
Role of prophylactic temporary inferior 20. Sharafuddin MJ, Corson JD. Percutaneous
vena cava filters placed at the ICU bedside devices for vena cava filtration. In: Ernst CB,
under intravascular ultrasound guidance in Stanley JC, eds. Current Therapy in Vascular
patients with multiple trauma. J Vasc Surg. Surgery. 4th ed. St. Louis: Mosby; 2001. p884-
2004;40(5):958-64. 91.
12. Aidinian G, Fox CJ, White PW, Cox MW, 21. Patel NC. Inferior vena cava interruption.
Adams ED, Gillespie DL. Intravascular In: Waldman DL, Saad WEA, Patel NC,
ultrasound-guided inferior vena cava filter eds. Interventional Radiology Secrets.
Philadelphia: Hanley & Belfus; 2004. inferior vena cava filters. J Vasc Interv Radiol.
p206-15. 2009;20(9):1193-9.
22. Hann CL, Streiff MB. The role of vena 28. Quirke TE, Ritota PC, Swan KG. Inferior
caval filters in the management of venous vena caval filter use in U.S. trauma centers:
thromboembolism. Blood Reviews. a practitioner survey. J Trauma. 1997;43(2):
2005;19(4):179-202. 333-7.
23. Hoppe H, Kaufman JA. Inferior vena cava 29. Giannoudis PV, Pountos I, Pape HC, Patel
filters. In: Kandarpa K, ed. Peripheral Vascular JV. Safety and efficacy of vena cava filters in
Interventions. Philadelphia: Lippincott trauma patients. Injury. 2007;38(1):7-18.
Williams & Wilkins; 2007. p401-15. 30. Girard P, Stern J-B, Parent F. Medical
24. Ascher E, Hingorani A, Yorkovich WR. literature and vena cava filters: so far so weak.
Complications of vena cava filters. In: Towne Chest. 2002;122:963-7.
JB, Hollier L, eds. Complications in Vascular 31. Kaufman JA, Kinney TB, Streiff MB, Sing RF,
Surgery. 2nd ed. New York: Marcel Dekker; Proctor MC, Becker D, Cipolle M, Comerota
2004. p. 569-79. AJ, Millward SF, Rogers FB, Sacks D, Venbrux
25. Harris EJ Jr, Kinney EV, Harris EJ Sr, Olcott AC. Guidelines for the use of retrievable
C 4th, Zarins CK. Phlegmasia complicating and convertible vena cava filters: report
prophylactic percutaneous inferior vena caval from the Society of Interventional Radiology
interruption: a word of caution. J Vasc Surg. multidisciplinary consensus conference. J Vasc
1995;22(5):606-11. Interv Radiol. 2006;17(3):449-59.
26. Greenfield LJ, Proctor MC. Twenty-year 32. Protack CD, Bakken AM, Patel N, Saad
clinical experience with the Greenfield filter. WE, Waldman DL, Davies MG. Long-term
Cardiovasc Surg. 1995;3(2):199-205. outcomes of catheter directed thrombolysis
27. Cantwell CP, Pennypacker J, Singh H, Scorza for lower extremity deep venous thrombosis
LB, Waybill PN, Lynch FC. Comparison without prophylactic inferior vena cava filter
of the recovery and G2 filter as retrievable placement. J Vasc Surg. 2007;45(5):992-7.
12
Percutaneous
Pulmonary embolectomy
Indications, Techniques, and outcomes
135
dication for thrombolytic therapy is not sup- teplase, Genetech, South San Francisco,
ported by a large patient survey that revealed CA) for the treatment of acute massive PE.
that 50% of patients with high-probability The findings of this study suggested that the
lung scans or pulmonary angiographic evi- intrapulmonary infusion of rtPA does not
dence for PE are acceptable candidates for offer significant benefit compared to intra-
treatment with IV thrombolysis.13 venous administration. However, this study
did not use the standard catheter-directed
technique currently used by most interven-
Catheter-Directed Thrombolysis tionalists, which includes embedding the
Catheter-directed thrombolytic therapy with infusion catheter directly into the thrombus,
intrapulmonary administration of a throm- while attempting to fragment the clot. Since
bolytic agent has been described by several rtPA must cleave clot-bound plasminogen
investigators with encouraging results.14,15 to create the active enzyme plasmin, infu-
Catheter-directed techniques aim to acceler- sion of rtPA in the main pulmonary artery
ate clot lysis and achieve rapid reperfusion confers minimum benefit to simple IV infu-
of the pulmonary arteries and lung paren- sion as most of the drug does not penetrate
chyma. In one study, 13 patients were treated the clot, but rather, flows in the path of
with urokinase (Abbott Labs, Abbott Park, least resistance (the patent pulmonary arte-
IL) for angiographically proven PE within rial segments). In addition, experience with
14 days of major surgery.14 The catheter peripheral arterial bypass grafts and throm-
was positioned in the pulmonary artery clot bosed dialysis fistulas has demonstrated that
and 2200 IU/kg of urokinase were injected catheter-directed thrombolysis is associated
directly into the clot. Continuous infusion with better rates of lysis, more rapid lysis, the
of urokinase at 2200 IU/kg/h until the clot need for lower doses of the lytic agent, and
lysed or up to a maximum of 24 hours was fewer complications.17 Ultrasound-enhanced
then performed. Follow-up pulmonary angi- catheter-directed thrombolysis using the
ography at 24 hours revealed that 98% of the EndoWave system (EKOS, Bothell, WA) has
clots had completely disappeared from the been used by some investigators to acceler-
pulmonary vasculature. No deaths or bleed- ate the thrombolysis. In one study the me-
ing complications occurred. In another se- dian infusion time was 24 hours with 76%
ries, 16 patients with massive PE were given complete resolution of clot burden.18
a bolus of 50,000 IU of urokinase directly
into the clot.15 An infusion of 1,000,000 IU
of urokinase was then given into the right Surgical Embolectomy
atrium over a 12-hour period. Cardiac out- For patients with massive central PE with
put, total pulmonary vascular resistance, marked compromised hemodynamics who
and mean pulmonary artery pressures all are too unstable to undergo or have a contra-
improved following the thrombolytic thera- indication to thrombolytic therapy, the only
py. One patient did suffer a severe bleeding remaining therapeutic options are percuta-
complication. neous techniques with clot fragmentation or
In 1988, Verstraete et al16 published the surgical embolectomy. Traditionally, surgi-
results of a multicenter comparative study cal embolectomy has been associated with
of intravenous vs. intrapulmonary infusion a high perioperative mortality rate. With
of rtPA (100 mg over a 7-hour period) (Al- improvements in anesthesia and cardiopul-
monary bypass technology, 30-day survival ventricular function. In the absence of one
rates after surgical pulmonary artery embo- of these noninvasive imaging studies (ie, a
lectomy have been reported to be as high as ventilation/perfusion nuclear medicine lung
75% at centers specializing in the treatment scan), a pulmonary angiogram is performed
of thromboembolic disease. In one study, 20 using a minimum amount of contrast to
patients who presented with acute PE and document the presence and distribution of
cardiogenic shock had a 5-year survival of PE and to measure pulmonary arterial and
100% following surgical embolectomy. In right heart pressures. However, pulmonary
this study, patients undergoing surgical em- arterial pressures (PAP) should be measured
bolectomy had a more favorable New York prior to the injection of any contrast and
Heart Association classification level and a prior to any intervention. Most operators
lower incidence of chronic pulmonary ar- agree that the injection of 10 mL of contrast
terial hypertension at follow-up when com- is sufficient to obtain a baseline pulmonary
pared to a similar cohort of patients receiving angiogram, especially since the diagnosis of
medical therapy.19 In a more recent study, 47 a massive PE has often already been made
patients underwent surgical thrombectomy with another imaging modality. The use of
at a single center with 6% perioperative mor- a large amount of contrast (ie, >20 mL with
tality and 83% 3-year survival.20 one injection) may be dangerous due to the
Yet, very few centers in the United risk of worsening right ventricular failure
States have widespread experience with this and cardiogenic shock from acute volume
type of surgery. Therefore, various percuta- overload.7
neous transvenous devices designed to re- If the patient is unable to lay supine,
move or fragment centrally obstructing PE the pulmonary angiogram and catheter-di-
have been developed. rected intervention can be performed from
a basilic or brachial vein approach. Other-
wise, the study is usually performed from
Percutaneous Embolectomy the femoral vein or less commonly, the in-
ternal jugular vein approach. The diagnostic
Technique pulmonary arterial catheter is removed over
We advocate that all patients undergoing a guidewire, maintaining access in the pul-
percutaneous embolectomy be intubated monary artery. An appropriate-sized 70- or
and on mechanical ventilation, preferably 90-cm introducer sheath (usually 6F11F
under heavy sedation or general anesthesia. depending on the device used) is inserted
Meticulous attention is also required for on- into access vein and positioned with its tip
going vasopressor, fluid, electrolyte, and ven- in the main pulmonary artery. The sheath
tilator management. Patient management can be used to perform suction embolecto-
includes continuous assessment of volume my, but more importantly, it is used to mini-
status, urine output, end-organ perfusion, mize trauma to the heart and stabilize the
and ECG monitoring during the procedure. various mechanical thrombectomy devices
Most often, a computed tomographic (CT) during their use in the pulmonary arterial
or magnetic resonance (MR) pulmonary segments. The sheath is connected to a hep-
arterial study or an echocardiogram dem- arinized saline flush (4000 IU heparin/1000
onstrates the clot distribution and burden cc normal saline) and infused at 30 cc/hr.
and the presence of compromised right The sheath side port can also be connected
Figure 12.1. 71-year-old male with acute shortness of breath. A. Oblique coronal MPR CT. B. Left pulmonary
angiogram demonstrates left PA embolus. Mean PAP was 42 mm Hg. C. Angiogram post AngioJet and catheter-
directed thrombolysis with tPA over 12 hours. Mean PAP decreased to 30 mm Hg.
InterventionsforAcutePulmonaryEmbolism
devices
Hydrolizer
Local lytics 12 47 97 46 30 92 8
Systemic plus local 8 43 36 100 12
AngioJet
No lytics 25 42 30 75 0
Local lytics 21 87 13
Combinationoftechniques
FragmentationplusApirex 18 74 88 37 31 89 11
Fragmentation,aspiration, 12 83 17
hydrodynamicdevicespluslytics
Adapted from Kucher N. Catheter embolectomy for acute pulmonary embolism. Chest. 2007;132:657-63. With permission.
5/13/11 10:27 AM
Percutaneous Pulmonary Embolectomy 141
sults from these series reveals an overall ported on their experience using this device
clinical success rate with catheter-directed in 46 patients over a 22-year period.21 Em-
therapy for acute PE of >80% (Table 12.1), boli were extracted in 35 (76%) of 46 pa-
with clinical success being defined as im- tients. There was an average reduction in
mediate hemodynamic improvement. The mean pulmonary artery pressure of 8 mm
reported mortality rates range from 0% to Hg and a significant increase in mean car-
25%, again reflecting the wide variations in diac output after embolectomy. The 30-day
the patients being treated. mortality rate was 30%. When subgroups of
The ideal thrombectomy device patients were analyzed, embolectomy was
should: (1) be easy to use and position with- most successful for major and submassive
in the pulmonary artery clots, (2) be highly PE, and least likely to be helpful in patients
maneuverable to allow rapid right heart with chronic, recurrent PE. Experience
passage and advancement into major pul- with the Greenfield suction embolectomy
monary arteries, (3) be able to promote com- catheter has been relatively small. The de-
plete removal of clots or fragmentation of vice is somewhat bulky, requires familiarity
clots into very small particles, and (4) have with the control handle, and is designed for
a low profile and be safe to use in the pul- insertion via a surgical venotomy, since it
monary circulation.7,8 A review of the most requires insertion through a 22F sheath.
commonly used devices is detailed below. The device was not widely used, so it is no
None of the currently available devices de- longer manufactured.
scribed below are FDA-approved for applica-
tion in the pulmonary arterial system, and
their use for acute PE represents an off-label
use of these devices.
Figure 12.4. A. Left pulmonary angiogram demonstrates left main pulmonary embolus.B. Amplatz
Thrombectomy Device (ATD) was used for 7 minutes. C. Post-angiogram demonstrates decrease in clot
burden.
Initial experience with the ATD device with local thrombolytic therapy and
showed clinical improvement in a limited achieved clinical success in one patient with
group of patients, with reduction of the re- no complications. However, we encountered
spiratory symptoms and improvement of an intraprocedural death in one patient with
hypotension.22,23 Transient hemoptysis and massive main pulmonary artery clot who was
arrhythmias have been described as compli- hypotensive and moderately hypoxemic prior
cations associated with the use of the ATD. to initiation of the catheter-directed treat-
Hemolysis also occurs commonly with the ment. Upon activation of the over-the-wire
use of the ATD, but there are no reported Trerotola device in the main pulmonary, the
cases of associated renal failure. The use of patient developed electrical-mechanical dis-
the ATD has been limited recently due to its sociation on our procedure table and could
bulkiness and lack of steerability. not be resuscitated. The experience with this
device remains very limited and its safety in
native pulmonary arteries is unproven.
Arrow-Trerotola Percutaneous
Thrombolytic Device
Thrombus Fragmentation Catheters
The Arrow-Trerotola device (Arrow, Read-
ing, PA) is a low-speed (3000 rpm) rotational and Balloons
basket designed for thrombectomy in dialy- The theoretical advantage of the fragmenta-
sis grafts. It scrapes the walls of the vessel tion technique is that the central pulmonary
and fragments the thrombus. The device artery volume is roughly 50% of the volume
was modified for the treatment of PE in an of the branch pulmonary arterial segments.
animal model and was shown to be effective Therefore, by achieving immediate redistri-
in fragmenting the clots and relatively safe bution of the occlusive thrombus from the
for treating large acute central pulmonary central main pulmonary artery to the more
emboli. A modified 8F, 120-cm-long device peripheral pulmonary artery branches, the
was redesigned with a nitinol wire basket afterload on the right ventricle is immedi-
that measures 9 to 15-mm in diameter when ately reduced. In addition, following clot
expanded. When this device was applied in fragmentation, a greater surface area of the
a porcine pulmonary artery, histologic speci- thrombus is exposed to allow greater activa-
mens showed that there was moderate acute tion of clot-bound plasminogen to plasmin
intimal injury, but no evidence of pulmonary by the infused lytic agents, if thrombolysis is
artery disruption. In a case report on the use used in combination with the fragmentation
of the Trerotola device (7F 80-cm-long over- technique.
the-wire device) for mechanical thrombec-
tomy of massive pulmonary embolism,24 the Balloon Catheters
device was found to be difficult to direct into Balloon angioplasty (616 mm in diameter)
some of the vessels being treated. In this for fragmentation of large central pulmonary
case report, there was no improvement in emboli has been used in association with
pulmonary pressures and large portions of local thrombolytic infusion with encourag-
clot remained untreated, although clinical ing results. Recovery rates of 87.5%, as mea-
improvement was observed. We have per- sured by pulmonary artery pressures, blood
formed 2 procedures at our institution com- O2 values, and clinical outcomes, have been
bining the use of the over-the-wire Trerotola reported with use of this technique.25
presented in cardiogenic shock and nonfatal pressure jets. The effect of adenosine released
hemoptysis occurred in one patient.35 In an- from lysed cells on the atrioventricular node
other study, 51 patients were treated with the has also been suggested as a possible mecha-
AngioJet device, 21% of whom also received nism. Most typically, the bradyarrhythmias
local thrombolytic therapy. The in-hospital will stop within 10 seconds of deactivating
mortality rate was 15%.36 the device. In addition, activation of the de-
The use of this device close to the heart vice for short bursts (ie, less than 10 seconds)
is commonly associated with bradyarrhyth- with 15- to 20-second pauses between device
mias, including transient asystole, which activation minimizes the occurrence of the
can cause significant symptoms and hinder bradyarrhythmias. A few procedural-related
its use in some patients. The incidence of deaths with use of the AngioJet for acute
the bradyarrhythmias appears to increase PE have been described in the literature,
with the proximity of the device to the heart which leads many operators to recommend
and the duration of device activation. The against using this device in the pulmonary
incidence of bradyarrhythmias with the use circulation. It is worth mentioning, how-
of this device is 20% to 79% in patients un- ever, that we have used the AngioJet device
dergoing coronary artery thrombectomy.37 In in combination with local thrombolytics
one series, 2 of 17 patients (12%) who un- for pulmonary embolectomy in 16 patients
derwent pulmonary thrombectomy with the and have had no procedural-related mortali-
AngioJet had bradyarrhythmias.34 In another ties. Bradyarrhythmias were common, but
series 8% of patients required tranvenous transient (Figure 12.6). We emphasize that
pacing for significant bradycardia.36 The careful monitoring of the cardiac rhythm
cause of AngioJet-induced bradyarrhythmia and hemodynamic parameters is paramount
remains unknown. Some authors suggest during the use of this device. As mentioned
that the arrhythmias could be related to earlier, mechanical ventilation is encour-
activation of stretch receptors by the high- aged as the activation of this device in the
Figure 12.6. 72-year-old with acute chest pain. A. Coronal MPR CT demonstrates a saddle embolus extending into
right main PA. Echocardiogram showed moderate right atrium dilatation and right ventricular dysfunction.B.Right
pulmonary angiogram demonstrates right PA embolus. Mean PAP on initial study was 34 mm Hg. C.Post AngioJet and
12 hours of thrombolysis angiogram demonstrates significant decrease in clot burden. Mean PAP was 15 mm Hg.
the treatment of acute massive pulmonary massive pulmonary embolism. J Vasc Interv
embolism: a retrospective multicenter case Radiol. 2002;13:163-9.
series. J Vasc Interv Radiol. 2008;19( 3):372-6. 28. Schmitz-Rode T, Janssens U, Schild
19. Meyer G, Tamisier D, Sors H, Stern M, HH, Basche S, Hanrath P, Gnther RW.
Vouh P, Makowski S, et al. Pulmonary Fragmentation of massive pulmonary
embolectomy: a 20-year experience at one embolism using a pigtail rotation catheter.
center. Ann Thorac Surg. 1991;51:232-6. Chest. 1998;114:1427-36.
20. Leacche M, Unic D, Goldhaber SZ, Rawn JD, 29. Yoshida M, Inoue I, Kawagoe T, Ishihara
Aranki SF, Couper GS, et al. Modern surgical M, Shimatani Y, Kurisu S, et al. Novel
treatment of massive pulmonary embolism: percutaneous catheter thrombectomy in acute
results in 47 consecutive patients after rapid massive pulmonary embolism: rotational
diagnosis and aggressive surgical approach. J bidirectional thrombectomy (ROBOT).
Thorac Cardiovasc Surg. 2005;129(5):1018-23. Catheter Cardiovasc Interv. 2006;68:112-7.
21. Greenfield LJ, Proctor MC, Williams DM, 30. Lang EV, Barnhart WH, Walton DL, Raab
Wakefield TW. Long-term experience with SS. Percutaneous pulmonary thrombectomy. J
transvenous catheter pulmonary embolectomy. Vasc Interv Radiol. 1997;8:427-32.
J Vasc Surg. 1993;18:450-8. 31. Tajima H, Murata S, Kumazaki T, Nakazawa
22. Uflacker R, Strange C, Vujic B. Massive K, Kawamata H, Fukunaga T, et al. Manual
pulmonary embolism: preliminary results of aspiration thrombectomy with a standard
treatment with the Amplatz thrombectomy PTCA guiding catheter for treatment of acute
device. J Vasc Interv Radiol. 1996;7:519-28. massive pulmonary thromboembolism. Radiat
23. Mller-Hlsbeck S, Brossmann J, Jahnke Med. 2004;22:168-72.
T, Grimm J, Reuter M, Bewig B, et al. 32. Timsit JF, Reynaud P, Meyer G, Sors H.
Mechanical thrombectomy of major and Pulmonary embolectomy by catheter device
massive pulmonary embolism with use of the in massive pulmonary embolism. Chest.
Amplatz thrombectomy device. Invest Radiol. 1991;100:655-8.
2001;36:317-22. 33. Siablis D, Karnabatidis D, Katsanos K,
24. Rocek M, Peregrin J, Velimsky T. Mechanical Kagadis GC, Zabakis P, Hahalis G. AngioJet
thrombectomy of massive pulmonary rheolytic thrombectomy versus local
embolism using an Arrow Trerotola intrapulmonary thrombolysis in massive
percutaneous thrombolytic device. Eur Radiol. pulmonary embolism: a retrospective
1998;8:1683-5. data analysis. J Endovasc Ther. 2005;12:
25. Fava M, Loyola S, Flores P, Huete I. 206-14.
Mechanical fragmentation and pharmacologic 34. Zeni PT, Blank BG, Peeler DW. Use of
thrombolysis in massive pulmonary embolism. rheolytic thrombectomy in treatment of acute
J Vasc Interv Radiol. 1997;8:261-6. massive pulmonary embolism. J Vasc Interv
26. Schmitz-Rode T, Janssens U, Hanrath P, Radiol. 2003;14:1511-5.
Gnther RW. Fragmentation of massive 35. Chauhan MS, Kawamura A. Percutaneous
pulmonary embolism using a pigtail rotation rheolytic thrombectomy for large pulmonary
catheter: possible complication. Eur Radiol. embolism: a promising treatment option.
2001;11:2047-9. Catheter Cardiovasc Interv. 2007;70:
27. De Gregorio MA, Gimeno MJ, Mainar 123-30.
A, Herrera M, Tobio R, Alfonso R, et al. 36. Chechi T, Vecchio S, Spaziani G, Giuliani
Mechanical and enzymatic thrombolysis for G, Giannotti F, Arcangeli C, et al. Rheolytic
thrombectomy in patients with massive and 39. Skaf E, Beemath A, Siddiqui T, Janjua
submassive acute pulmonary embolism. M, Patel NR, Stein PD. Catheter-tip
Catheter Cardiovasc Interv. 2009;73:506-13. embolectomy in the management of acute
37. Dwarka D, Schwartz SA, Smyth SH, massive pulmonary embolism. Am J Cardiol.
OBrien MJ. Bradyarrhythmias during use 2007;99:415-20.
of the AngioJet system. J Vasc Interv Radiol. 40. Kuo WT, van den Bosch MA, Hofmann LV,
2006;17:1693-5. Louie JD, Kothary N, Sze DY. Catheter-
38. Eid-Lidt G, Gaspar J, Sandoval J. Combined directed embolectomy, fragmentation, and
clot fragmentation and aspiration in patients thrombolysis for the treatment of massive
with acute pulmonary embolism. Chest. pulmonary embolism after failure of systemic
2008;134:45-60. thrombolysis. Chest. 2008;134;250-4.
Thoracic Outlet
Syndrome
153
13
Primary Axillosubclavian
Venous Thrombosis
observational Care
Kaj H. Johansen
155
Indeed, the major debate has been outcome following axillosubclavian vein
not whether to carry out such surgical re- decompression and reconstruction with an
construction for primary ASVT or not, but immediate operation vs. a period of observa-
whether to wait for a period for postthrombo- tion. Noting recent data,19-25 a skeptical ob-
sis inflammation to clear rather than operat- server might wonder whether, if observation
ing immediately.30 continued long enough, a substantial num-
It has been proposed that an aggressive ber of such patients would be asymptomatic
operative reconstructive approach to prima- and might not need an operation at all. One
ry ASVT may be particularly appropriate for study comparing outcomes of operations for
certain individuals who might be expected venous and neurogenic thoracic outlet syn-
to be sensitive to even minimal amounts of drome by means of a validated functional
arm swellingyounger individuals or elite tool 32 noted that primary ASVT patients
athletes,31 for example. This assertion may were minimally symptomatic preoperatively
be correct. However, it is impossible to prove and were not improved by operation.33
that a subset of primary ASVT patients exist Optimally, given the fact that this is
for whom aggressive operative reconstruc- a relatively stereotypic patient population
tion should be the norm because of the ab- that presents soon after the onset of upper
sence of a control group of such individuals extremity symptoms, a prospective random-
who have not been reconstructed in this ized trial of immediate operation vs. observa-
fashion. tion should be launched. An excellent venue
Indeed, the dilemma in what to advise for this could be the nascent Consortium
patients who have suffered primary ASVT, on Outcomes Research and Education for
after they have undergone thrombolysis and Thoracic Outlet Syndrome (CORE-TOS),
3 months of anticoagulation, is that no level chaired by Robert Thompson, MD, of St.
I, or even level II, evidence is available to Louis, and Julie Freischlag, MD, of Balti-
permit a truly informed decision by the pa- more. A multicenter enrollment of primary
tient. Instead, if a primary ASVT patient ASVT patients, all of them undergoing ini-
presents to a major academic medical center tial anticoagulation and thrombolysis and
in Baltimore, St. Louis, or Denver, the pa- then randomized either to operation or ob-
tient is likely to undergo thoracic outlet de- servation, should soon yield credible results
compression and, if necessary, major upper regarding whether there is an advantage
extremity venous reconstructive surgery. to operative intervention for such patients.
The same patient appearing at a similar in- Such a randomized control trial is under
stitution in Palo Alto, Seattle, or Vancouver discussion.
will likely simply be urged to exercise vigor-
ously and will be offered an operation only references
if, during long-term follow-up, persistent or 1. Linenburger ML. Catheter-related
recurrent postthrombotic symptoms persist thrombosis: risks, diagnosis and management.
or recuran extremely uncommon event. J Natl Compr Cancer Netw. 2006;4:889-901.
At the very least, the necessity for im- 2. Mandolfo S, Piazza W, Galli F. Central
mediately decompressing and reconstruct- venous catheter and the hemodialysis
ing primary ASVT patients axillosubclavian patient: a difficult symbiosis. J Vasc Access.
veins should be reexamined. There are no 2002;3:64-73.
compelling data that demonstrate improved 3. Rozmus G, Danbert JP, Huang DT, Rosero
S, Hall B, Francis C. Venous thrombosis and 14. Adelman MA, Stone DH, Riles TS,
stenosis after implantation of pacemakers and Lamparello PJ, Giangola G, Rosen RJ. A
defibrillators. J Interv Card Electrophysiol. multidisciplinary approach to the treatment
2005;13:9-19. of Paget-Schroetter syndrome. Ann Vasc Surg.
4. Monreal M, Lafoz E, Ruiz J, Valls R, 1997;11:149-54.
Alastrue A. Upper-extremity deep venous 15. Molina JE. Need for emergency treatment in
thrombosis and pulmonary embolism. Chest. subclavian vein effort thrombosis. J Am Coll
1991;99:280-3. Surg. 1995;181:414-20.
5. Kammen BF, Soulen MC. Phlegmasia 16. Melliere D, Becquemin JP, Etienne G,
cerulean dolens of the upper extremity. J Vasc Lecheviller B. Severe injuries resulting from
Interv Radiol. 1995;6:283-6. operations for thoracic outlet syndrome: can
6. Marshall BJ, Warren JR. Unidentified curved they be avoided? J Cardiovasc Surg (Torino).
bacilli in the stomach of patients with gastritis 1991;32:599-603.
and peptic ulceration. Lancet. 1984;1: 17. Chang DC, Lidor AO, Matsen SL, Freischlag
1311-5. JA. Reported in-hospital complications
7. Tilney NL, Griffiths HJG, Edwards EA. following rib resections for neurogenic
Natural history of major venous thrombosis of thoracic outlet syndrome. Ann Vasc Surg.
the upper extremity. Arch Surg. 1970;101: 2007;21:564-70.
792-6. 18. Hingorani A, Ascher E, Lorenson E, DePippo
8. Gloviczki P, Kazmier FS, Hollier LH. P, Salles-Cunha S, Scheinman M, et al.
Axillary-subclavian venous occlusion: the Upper extremity deep venous thrombosis and
morbidity of a non-lethal disease. J Vasc Surg. its impact on morbidity and mortality rates
1986;4:333-7. in a hospital-based population. J Vasc Surg.
9. Kunkel JM, Machleder HI. Treatment of 1997;26:853-60.
Paget-Schroetter syndrome. Arch Surg. 19. Heron E, Lozinguez O, Emmerich J, Laurian
1989;124:1153-8. C, Fiessinger JN. Long-term sequelae of
10. Green RM, Waldman D, Ouriel K, Riggs P, spontaneous axillary-subclavian venous
Deweese JA. Claviculectomy for subclavian thrombosis. Ann Intern Med. 1999;131:510-3.
venous repair: long-term functional results. 20. Lechner D, Wiener C, Weltermann
J Vasc Surg. 2000;32:315-21. A, Eischer L, Eichinger S, Kyrle PA.
11. Urschel HC, Razzuk MA. Improved Comparison between idiopathic deep vein
management of the Paget-Schroetter thrombosis of the upper and lower extremity
syndrome secondary to thoracic outlet regarding risk factors and recurrence. J
compression. Ann Thorac Surg. 1991;52: Thromb Haemost. 2008;6:
1217-21. 1269-74.
12. Machleder HI. Evaluation of a new treatment 21. Martinelli I, Battaglioli T, Bucciarelli
strategy for Paget-Schroetter syndrome: P, Passamonti SM, Mannucci PM. Risk
spontaneous thrombosis of the axillary- factors and recurrence rate of primary deep
subclavian vein. J Vasc Surg. 1993;17:305-17. vein thrombosis of the upper extremities.
13. Thompson RW, Schneider PA, Nelken NA, Circulation. 2004;110:566-70.
Skioldebrand CG, Stoney RJ. Circumferential 22. Elman EE, Kahn SR. The post-thrombotic
venolysis and paraclavicular thoracic outlet syndrome after upper extremity deep venous
decompression for effort thrombosis of the thrombosis in adults: a systematic review.
subclavian vein. J Vasc Surg. 1992;16:723-32. Thromb Res. 2006;117:609-14.
23. Lee WA, Hill BB, Harris EJ Jr, Semba CP, a Palmaz stent in the subclavian vein. AJR
Olcott C IV. Surgical intervention is not Am J Roentgenol. 1993;160:1123-4.
required for all patients with subclavian vein 30. Capparelli DJ, Freischlag J. A unified
thrombosis. J Vasc Surg. 2000;32:57-67. approach to axillosubclavian venous
24. Lokanathan R, Salvian AJ, Chen JC, Morris thrombosis in a single hospital admission.
C, Taylor DC, Hsiang YN. Outcome after Semin Vasc Surg. 2005;18:153-7.
thrombolysis and selective thoracic outlet 31. Melby SJ, Vedantham S, Narra VR, Paletta
decompression for primary axillary vein GA Jr, Khoo-Summers L, Driskill M, et al.
thrombosis. J Vasc Surg. 2001;33:783-8. Comprehensive surgical management of
25. Johansen K. Does axillosubclavian (effort) the competitive athlete with effort thrombosis
thrombosis oblige first-rib resection? Arch. of the subclavian vein (Paget-Schroetter
Surg., in press. syndrome). J Vasc Surg. 2008;47:809-20.
26. Chan FKL, Leung WK. Peptic ulcer disease. 32. Beaton DE, Wright JG, Katz JG. The Upper
Lancet. 2002;360:933-41. Extremity Collaborative Group. Development
27. Taylor LM, McAllister WR, Dennis DL, of the QuickDASH: comparison of three
Porter JM. Thrombolytic therapy followed by item-reduction approaches. J Bone Joint Surg.
first rib resection for spontaneous (effort) 2005;87A:1038-46.
subclavian vein thrombosis. Am J Surg. 33. Cordobes-Gual J, Lozano-Vilardell P,
1985;149:644-7. Torreguitart-Mirada N, Lara-Hernandez
28. Glanz S, Gordon DH, Lipkowitz GS, R, Riera-Vazquez R, Julia-Montoya J.
Butt KM, Hong J, Sclafani SJ. Axillary Prospective study of the functional recovery
and subclavian vein stenosis: percutaneous after surgery for thoracic outlet syndrome.
angioplasty. Radiology. 1988;168:371-3. Eur J Vasc Endovasc Surg. 2008;35:
29. Bjarnason H, Hunter DW, Crain MR, Ferral 79-83.
H, Miltz-Miller SE, Wegryn SA. Collapse of
14
Axillary-Subclavian
Venous effort Thrombosis
Surgical Care
163
Figure 14.1 Pathogenesis of axillary-subclavian venous thrombosis. A. Normal anatomy of the thoracic outlet
illustrating relationship of the internal jugular vein (IJV) and subclavian vein (SCV) to the clavicle and first rib.
B. Vigorous activities requiring overhead positions of the arm are associated with the development of axillary-
subclavian venous effort thrombosis. C. Subclavian vein compression between the clavicle and first rib results in focal
vein wall injury. D. Chronic repetitive compression injury of the subclavian vein leads to formation of circumferential
perivenous scar tissue, which can severely constrict the lumen. E. Thrombus formation within the lumen of the constricted
subclavian vein causes complete obstruction of the subclavian vein, with extension of thrombus to the axillary vein
causing obstruction of collateral veins. F. Symptomatic presentation of axillary-subclavian venous effort thrombosis.
Adapted and redrawn from Melby et al., Comprehensive surgical management of the competitive athlete with effort
thrombosis of the subclavian vein (Paget-Schroetter syndrome). J Vasc Surg 2008; 47:809-820, Supplement A Figure III.
vein in the neck, and is highly technician- Direct venography is also required to un-
dependent. Expanded collateral veins may dertake catheter-based venous thromboly-
also be mistaken for the subclavian vein, and sis, the preferred initial step in treatment
indirect hemodynamic measures of venous of almost all patients presenting with effort
flow may not accurately reflect the status of thrombosis. Taking all of these factors into
the proximal subclavian vein. Duplex imag- consideration, we believe the most practi-
ing studies have a false-negative rate as high cal, efficient, and cost-effective approach to
as 30% for effort thrombosis and are thereby evaluating the patient with suspected effort
not sufficiently accurate to exclude the diag- thrombosis is to go directly to catheter-based
nosis of venous TOS if negative. venography, rather than utilize duplex stud-
ies or other noninvasive imaging tests.
Radiologic Imaging
In current practice, either contrast-enhanced Blood Coagulation Tests
computed tomography (CT) or magnetic Although considered to be a mechani-
resonance (MR) angiography are being used cal anatomical problem unrelated to an
with greater frequency as the initial nonin- increased propensity toward thrombosis, it
vasive diagnostic studies for axillary-subcla- has been reported that up to 70% of patients
vian venous effort thrombosis. Both of these with venous TOS may have associated ab-
studies are highly accurate in detecting axil- normalities in coagulation tests.3 Since such
lary-subclavian vein occlusion and/or focal abnormalities may influence subsequent
stenosis at the level of the first rib (Figure patient management, a panel of coagula-
14.2A-B). They can also demonstrate the tion studies should be obtained either dur-
presence or absence of enlarged collateral ing the initial diagnostic evaluation or in
veins, may provide information on the age of follow-up, including tests for protein C and
any thrombus present, and can be performed protein S activities, antithrombin III levels,
with the arms in elevated positions as well plasma homocysteine, the presence of anti-
as at rest to elucidate evidence of positional cardiolipin antibodies and lupus anticoagu-
subclavian vein obstruction. Additional in- lant, and mutations in the genes encoding
formation can also be obtained by compari- prothrombin (G20210A), factor V (Leiden,
sons with the contralateral upper extremity. G1692A), plasminogen activator inhibitor-1
Since CT or MR venography provide more (PAI-1, 4G/5G), and methyltetrahydrofolate
anatomic information than venous duplex reductase (MTHFR, C677T). These tests
imaging, these studies can be used to accu- are usually negative and thereby add little to
rately exclude the diagnosis of venous TOS the initial diagnosis or management of effort
when negative. thrombosis.
The most direct and definitive means to
confirm the diagnosis of axillary-subclavian
venous effort thrombosis and venous TOS is Goals of Treatment and
through catheter-directed contrast venogra-
phy. Direct venography provides complete Initial Management
anatomic information regarding the site and
extent of thrombosis, and it allows definitive There are 4 principal goals of treatment for
evaluation of the collateral venous pathways. axillary-subclavian venous effort thrombo-
Figure 14.2 Imaging studies in venous TOS.AandB: Positional magnetic resonance venography. With the arms at
rest (A), there is a patent right subclavian vein and an occluded left subclavian vein. With the arms elevated overhead
(B) , there is moderate stenosis of the right subclavian vein and persistent occlusion of the left subclavian vein. Cand
D : Thrombolytic therapy for left-sided axillary-subclavian venous effort thrombosis. Initial venogram (C) demonstrates
a segmental occlusion of the subclavian vein with dense collaterals. Following thrombolytic therapy (D) , most of the
axillary and subclavian vein has been cleared of thrombus, with a residual focal high-grade stenosis at the level of the
first rib. EandF: Thrombolytic therapy for right-sided axillary-subclavian venous effort thrombosis. Initial venogram
(E) demonstrates a long occlusion of the entire axillary and subclavian veins, with relatively limited development of
collaterals. Following thrombolytic therapy (F) , most of the axillary and subclavian vein has been cleared of thrombus,
with a moderate residual stenosis at the level of the first rib.
sis: (1) provide prompt relief of acute symp- requiring monitoring in an acute-care set-
toms of upper extremity venous congestion ting (eg, intermediate-care or intensive-care
and prevention of pulmonary embolism; (2) unit) for several days. In recent years, it has
reduce the likelihood of recurrent venous become more typical to perform thromboly-
thrombosis following initial management; sis with catheter-based pharmacomechani-
(3) avoid the development of upper extrem- cal thrombectomy, in which a mechanical
ity postthrombotic syndrome; and (4) return device on the tip of the catheter is used to
to normal upper extremity activity without rapidly break up the clot, along with local-
medications. The first of these goals is best ized infusion of a much smaller amount of
met by prompt anticoagulation and throm- thrombolytic agent.5 The great advantage of
bolytic therapy. this approach is that it can usually be com-
pleted in a single stage, often within several
hours, thereby avoiding a long stay in a mon-
Anticoagulation itored hospital setting.
In the absence of any contraindications,
once the diagnosis of axillary-subclavian ve-
nous effort thrombosis is suspected, almost Balloon Angioplasty
all patients should be anticoagulated with The goal of thrombolysis is to clear any fresh
intravenous or subcutaneous heparin. This or recent clot from the axillary-subclavian
can be done while additional diagnostic and collateral veins. This usually results in
studies are being performed and/or prior to a marked improvement in the venographic
patient transfer from one hospital to another, appearance of the vein and a prompt reduc-
and is important to help prevent the exten- tion in symptoms of venous obstruction.
sion of thrombus within the axillary and sub- Following thrombolysis a focal occlusion or
clavian veins. Treatment with an antiplatelet high-grade stenosis of the proximal subcla-
agent, such as aspirin or clopidogrel (Plavix), vian vein is usually identified at the level of
is often included. the first rib; this is not composed of throm-
bus, but represents the underlying scar tis-
sue caused by subclavian vein compression,
Thrombolytic Therapy injury, and tissue repair. In some cases, bal-
Current approaches to venous TOS empha- loon angioplasty may be used at the same
size early diagnosis by contrast venography time in an attempt to reduce the degree of
and prompt use of catheter-based thrombo- stenosis in the subclavian vein. However,
lytic therapy to reduce the amount of throm- because the vein is usually obstructed by
bus within the axillary and subclavian veins scar tissue in the wall of the vein as well as
(Figure 14.2C-F).4 Venous thrombolysis has external compression between the clavicle
traditionally been performed by continuous and first rib, balloon angioplasty is often
infusion of the thrombolytic drug directly unsuccessful, and even when improvement
into a multihole catheter that has been is obtained it is usually short-lived. There is
placed within the axillary-subclavian vein at also abundant evidence demonstrating that
the time of the initial venogram. Infusion is vascular stents should not be placed in the
continued with repeat venograms performed subclavian vein, at least prior to surgical de-
at follow-up intervals for a period of 24 to 48 compression, due to an inevitably high rate
hours, until a maximum effect is achieved, of failure.6 We therefore rarely recommend
the use of balloon angioplasty for subclavian anatomy. It is notable that there is still a sig-
vein stenosis following thrombolysis. Follow- nificant risk of recurrent thrombosis follow-
ing thrombolysis, the patient should remain ing thrombolysis and anticoagulation alone,
on systemic anticoagulation. with published estimates ranging from 50%
to 70%. As summarized by Aziz et al., Med-
ical therapy, consisting either of anticoagu-
Nonsurgical lation or thrombolytic therapy, results in an
Management unsatisfactory clinical outcome because it
does not correct the underlying mechanical
Conservative treatment of axillary-subclavi- abnormality.12
an venous effort thrombosis has traditionally
consisted of chronic anticoagulation, inter-
mittent arm elevation, long-term restrictions Indications and
in arm activity, and the use of compression
sleeves, with the hope that increased collat- Protocols for Surgical
eral development will eventually compen-
sate for axillary-subclavian vein occlusion. Treatment
Many studies in the medical literature have Surgical treatment provides definitive man-
shown that this approach rarely results in agement for axillary-subclavian venous ef-
symptom-free use of the arm and is associat- fort thrombosis and venous TOS, and should
ed with a significant incidence of chronic ve- be considered in almost all patients with this
nous congestion, particularly with active use condition. Operative treatment is centered
of the arm, requiring considerable limita- on 2 goals: (1) decompression of the sub-
tions in young, active patients (Table 14.1).7- clavian vein and collateral venous pathways
11
Unlike lower extremity DVT, the proper through the thoracic outlet by removal of the
duration of anticoagulation treatment for first rib and associated scalene and subcla-
subclavian venous effort thrombosis is not vius muscles, and (2) restoration and main-
known. Because this condition is caused by tenance of normal blood flow through the
compression of the vein rather than a dis- subclavian vein, by removing constricting
order of blood clotting, many recommend scar tissue from around the vein, by adjunc-
lifelong anticoagulation unless there has tive balloon angioplasty, or by direct venous
been a defined alteration in the underlying reconstruction when necessary.
Donayre et al. 9
1986 41 Disabling symptoms >50%
Gloviczki et al. 10
1986 95 Late sequelae in 30%
this approach we offer operative decompres- The anterior scalene muscle is circum-
sion to virtually all patients with symptom- ferentially dissected at the level of its inser-
atic venous TOS or recent effort thrombosis, tion upon the first rib and sharply divided
regardless of the interval between initial under direct vision (Figure 14.2B). The re-
diagnosis and referral, previous treatment, maining muscle is lifted superiorly and de-
or adverse findings on contrast venography. tached from the underlying tissues, carrying
Operative procedures based on paraclavicu- the dissection superiorly to the level of its
lar exposure thereby provide the most versa- origin on the C6 transverse process, and the
tile, comprehensive, and safe approach to the muscle is divided and removed. Each of the
treatment of venous TOS.27,28 5 nerve roots comprising the brachial plexus
(C5, C6, C7, C8, and T1) are identified, mo-
bilized, and protected from injury, and any
Paraclavicular Thoracic aberrant fibrous bands, ligaments, or fascial
attachments that may contribute to neuro-
Outlet Decompression vascular compression are resected.
With the brachial plexus nerve roots
Thoracic outlet decompression for venous gently retracted in an anteromedial direc-
TOS begins with supraclavicular exposure tion, the attachment of the middle scalene
(Figure 14.3A). The patient is positioned su- muscle is carefully divided from the top of
pine with the head of the bed elevated 30 the first rib (Figure 14.2C). The intercostal
degrees, and the neck, chest, and affected muscles along the posterolateral aspect of
upper extremity are prepped into the field the first rib are divided, and the tip of a right-
to permit movement of the arm during the angle clamp is passed underneath the poste-
operation and access to the forearm and rior neck of the first rib. The posterior aspect
wrist. A transverse supraclavicular incision of the first rib is divided with a modified
is made, subplatysmal flaps are developed to Stille-Giertz rib cutter, and a Kerrison bone
expose the scalene fat pad, and the omohy- rongeur is used to smooth the posterior end
oid muscle is divided. The scalene fat pad of the bone to a level medial to the course
is detached and progressively elevated in a of the T1 nerve root (Figure 14.3D-E). The
medial to lateral direction, exposing the sur- proximal end of the first rib is elevated and
face of the anterior scalene muscle and the a fingertip is passed underneath the rib to
phrenic nerve. Small blood vessels and lym- bluntly separate additional extrapleural fas-
phatics are secured between ligatures and, if cia and intercostal muscle attachments. The
necessary, the thoracic duct is ligated and di- anterior portion of the first rib is not yet di-
vided. Lateral mobilization of the scalene fat vided at this stage.
pad continues until there is ample exposure To accomplish complete resection of
of the anterior scalene muscle and phrenic the anteromedial portion of the first rib, a
nerve, the brachial plexus nerve roots, and second transverse skin incision is made one
the middle scalene muscle. The long tho- fingerbreadth below the medial clavicle
racic nerve is observed emerging from the (Figure 14.3F). A plane of separation is cre-
middle scalene muscle and crossing the pos- ated between the upper and middle portions
terolateral aspect of the first rib. The scalene of the pectoralis major muscle, and the carti-
fat pad is held in position with several retrac- laginous portion of the first rib is identified.
tion sutures. This is facilitated by applying downward
INFRACLAVICULAR
INCISION
Figure 14.3 Paraclavicular thoracic outlet decompression.A. Paraclavicular decompression begins with a
supraclavicular incision. B. The anterior scalene muscle is divided from the first rib and removed. C. The middle
scalene muscle is divided from the first rib and removed. DandE. The posterior aspect of the first rib is divided (D)
with protection of the C8 and T1 brachial plexus nerve roots under direct vision, and the remaining edge of the rib is
trimmed to a level proximal to the nerve roots (E). The anterior first rib is not yet divided at this stage of the procedure,
as it is in operations for neurogenic or arterial TOS. F. An infraclavicular incision is made overlying the anteromedial
aspect of the first rib and the rib is exposed, facilitated by pressure on the posterior end of the rib to separate the
costoclavicular space. G. The anteromedial aspect of the first rib is divided at the edge of the sternum and the entire
first rib is removed, allowing complete dissection of the axillary-subclavian vein throughout the thoracic outlet, to its
junction with the internal jugular and innominate veins. Adapted and redrawn from Thompson RW, Venous Thoracic
Outlet Syndrome: Paraclavicular Approach. Operative Techniques in General Surgery 2008; 10:113-121 (Figures 1-7).
pressure to the divided posterior segment of constricted segment of the vein, and if the
the first rib with a finger placed within the vein is soft and easily compressible to pal-
supraclavicular incision, placing the attach- pation, and shows evidence of rapid filling
ments between the medial first rib and clav- and emptying with respiratory variation, it
icle under tension and allowing the medial is likely that no further venous reconstruc-
portion of the first rib to be dissected from tion is necessary (in our experience, this is
its soft tissue attachments through the infra- the case in approximately 50% of patients
clavicular incision. with venous TOS, even in those with long-
The subclavius muscle tendon, the cos- segment stenosis prior to operation). When
toclavicular ligament, and the muscles of external venolysis does not alleviate subclavi-
the first intercostal space are divided under an vein obstruction, or when intraoperative
direct vision and the cartilaginous portion venography demonstrates a residual stenosis
of the first rib is divided adjacent to the ster- despite the apparent success of external ve-
num, with the first rib removed from the nolysis, additional venous reconstruction is
operative field as a single specimen (Figure performed. Following systemic anticoagula-
14.3G). The axillary-subclavian vein is iden- tion (Dextran and heparin), clamp control
tified underneath the clavicle and carefully is obtained of the distal subclavian and in-
separated from the subclavius muscle, with ternal jugular veins and a pediatric Satinsky
ligation and division of any collateral vein clamp is passed around the upper portion of
branches that enter the subclavian vein, and the innominate vein. A longitudinal venoto-
the subclavius muscle is resected. Further my is created along the superior aspect of the
exposure of the subclavian vein is undertak- subclavian vein and the lumen is thoroughly
en through the supraclavicular exposure and inspected. If the luminal surface is smooth
continued medially toward the junction of and free of thrombus and/or irregularity, a
the subclavian and internal jugular veins to vein patch angioplasty is performed using
form the innominate vein. The internal jug- greater saphenous vein or a segment of cryo-
ular vein is fully exposed several centimeters preserved femoral vein (Figure 14.4B). It is
superior to its junction with the subclavian considered important to construct the patch
vein, and the innominate vein is exposed for to span the entire length of the affected sub-
several centimeters into the upper mediasti- clavian vein, with an extension into the lat-
num. The course of the phrenic nerve into eral aspect of the internal jugular vein or the
the upper mediastinum is also noted, and anteromedial aspect of the innominate vein.
the nerve is protected where it passes under- When dense fibrosis remains within
neath the subclavian vein. the wall of the subclavian vein despite ex-
Any pathological changes in the central ternal venolysis, the affected segment of the
portion of the subclavian vein are assessed subclavian vein is replaced by interposition
visually and by digital palpation. As the bypass (Figure 14.4C). The intervening seg-
subclavian vein is typically found to harbor ment of the native subclavian vein is excised
a focal area with fibrous wall thickening re- (Figure 14.4D), and an interposition graft
sulting from chronic repetitive injury, any is constructed using a widely beveled end-
residual scar tissue surrounding the vein is to-end anastomosis to the unaffected distal
completely excised (circumferential exter- axillary-subclavian vein. The proximal anas-
nal venolysis) (Figure 14.4A). This often tomosis is constructed in a wide end-to-side
results in reexpansion of the previously anastomosis, using an extension of the graft
Figure 14.4 Subclavian vein reconstruction in venous TOS.A. Circumferential external venolysis to remove
constricting fibrous scar tissue from around the subclavian vein, which may allow the vein to re-expand to a normal
diameter. B. When subclavian vein reconstruction is necessary and the lumen of the vein is found to be smooth
and free of chronic thrombus, vein patch angioplasty is sufficient for reconstruction. C. When subclavian vein
reconstruction is necessary but the luminal surface is unsuitable for patch angioplasty, replacement of the subclavian
vein is accomplished with an interposition bypass using a saphenous vein panel graft. D. Microscopic cross-sectional
appearance of the chronically occluded subclavian vein in effort thrombosis. The lumen is occluded with chronic
organized thrombus and inflammatory cell infiltrates, and the vein wall is markedly thickened with fibrosis, loss of the
normal elastic architecture, and previous intramural hemorrhage (Verhoeff van Giesen stain). EandF. Venography in
right-sided axillary-subclavian venous effort thrombosis. The initial venogram (E) demonstrates long occlusion of the
axillary and subclavian veins with widespread collaterals. Venogram performed 12 weeks after surgical treatment (F),
which had involved creation of an axillary-innominate subclavian vein interposition bypass reconstruction. Adapted and
redrawn from Melby et al., Comprehensive surgical management of the competitive athlete with effort thrombosis of the
subclavian vein (Paget-Schroetter syndrome). J Vasc Surg 2008; 47:809-820, Supplement A Figure III; and Thompson
RW, Venous Thoracic Outlet Syndrome: Paraclavicular Approach. Operative Techniques in General Surgery 2008;
10:113-121 (Figures 1-7).
onto the lateral aspect of the internal jugu- Pneumothorax or pleural effusion
lar vein or the anteromedial aspect of the Postoperative lymph leak
innominate vein. Because the caliber of the Residual subclavian vein obstruction
saphenous vein is usually too small to match or early postoperative rethrombosis
the subclavian vein, use of the saphenous Postoperative bleeding/wound
vein requires creation of a panel graft with hematoma/excessive anticoagulation
twice the diameter of the native saphenous Late postoperative axillary or
vein. Alternatively, subclavian vein interposi- subclavian vein obstruction or
tion bypass can be performed using a suit-
rethrombosis
able segment of cryopreserved femoral vein.
Finally, intraoperative venography is
used to confirm satisfactory subclavian vein The expected hospital stay is 5 days, with the
reconstruction, typically performed through closed-suction drain removed approximately
the cephalic vein in the distal forearm. Our 7 days after operation. Inpatient physical
operative approach also includes frequent therapy is started the day after operation
construction of a temporary radiocephalic to maintain range of motion, with postop-
arteriovenous (AV) fistula between the end erative rehabilitation then overseen by a
of the distal cephalic vein and the side of the physical therapist with expertise in the man-
radial artery, used as an adjunct to increase agement of TOS, and no restrictions placed
upper extremity venous blood flow during on upper extremity activity beyond 12 weeks
the first several months after operation. This after operation. Therapeutic anticoagulation
is subsequently ligated under local anesthe- (heparin/warfarin plus clopidogrel) is initiat-
sia at 12 weeks after surgical treatment, at ed several days after operation and then dis-
which time a follow-up contrast venogram is continued at 12 weeks. Recovery is typically
also performed (Figure 14.4E-F). complete within 3 months of operation, and
Postoperative care includes ample use a full return to previous levels of function
of pain medications, muscle relaxants, and can usually be expected.
anti-inflammatory agents. The potential
complications of surgery are similar to those
considered in other operations for thoracic references
outlet syndromes, as well as those related to 1. Hughes ESR. Venous obstruction in the upper
venous reconstruction, as follows: extremity (Paget-Schroetters syndrome). Int
Abstr. Surg. 1949;88:89-127.
Subclavian artery injury/intraoperative 2. Sanders RJ. Thoracic Outlet Syndrome:
hemorrhage A Common Sequelae of Neck Injuries.
Subclavian vein injury/intraoperative Philadelphia: J. B. Lippincott Company; 1991.
hemorrhage 3. Cassada DC, Lipscomb AL, Stevens SL,
Brachial plexus nerve injury or Freeman MB, Grandas OH, Goldman MH.
postoperative paresis The importance of thrombophilia in the
Phrenic nerve injury or postoperative treatment of Paget-Schroetter syndrome. Ann
paresis Vasc Surg. 2006;20:596-601.
Long thoracic nerve injury or 4. Rutherford RB. Primary subclavian-axillary
postoperative paresis vein thrombosis: the relative roles of
thrombolysis, percutaneous angioplasty, stents,
and surgery. Semin Vasc Surg. 1998;11: spontaneous thrombosis of the axillary-
91-5. subclavian vein. J Vasc Surg. 1993;17:305-15.
5. Schneider DB, Curry TK, Eichler CM, 16. Lee WA, Hill BB, Harris EJ Jr, Semba CP,
Messina LM, Gordon RL, Kerlan RK. Olcott CI. Surgical intervention is not
Percutaneous mechanical thrombectomy for required for all patients with subclavian vein
the management of venous thoracic outlet thrombosis. J Vasc Surg. 2000;32:57-67.
syndrome. J Endovasc Ther. 2003;10:336-40. 17. Lee JT, Karwowski JK, Harris EJ, Haukoos
6. Urschel HC Jr, Patel AN. Paget-Schroetter JS, Olcott C IV. Long-term thrombotic
syndrome therapy: failure of intravenous recurrence after nonoperative management
stents. Ann Thorac Surg. 2003;75:1693-6. of Paget-Schroetter syndrome. J Vasc Surg.
7. Adams JT, McEvoy RK, DeWeese JA. 2006;43:1236-43.
Primary deep venous thrombosis of the upper 18. Urschel HC Jr, Razzuk MA. Improved
extremity. Arch Surg. 1965;91:29-42. management of the Paget-Schroetter
8. Tilney NL, Griffiths HJG, Edwards E. Natural syndrome secondary to thoracic outlet
history of major venous thrombosis of the compression. Ann Thorac Surg. 1991;52:1217-
upper extremity. Arch Surg. 1970;101:792-5. 21.
9. Donayre CE, White GH, Mehringer SM, 19. Angle N, Gelabert HA, Farooq MM, Ahn SS,
Wilson SE. Pathogenesis determines late Caswell DR, Freischlag JA, et al. Safety and
morbidity of axillosubclavian vein thrombosis. efficacy of early surgical decompression of the
Am J Surg. 1986;152:179-84. thoracic outlet for Paget-Schroetter syndrome.
10. Gloviczki P, Kazmier FJ, Hollier LH. Axillary- Ann Vasc Surg. 2001;15:37-42.
subclavian venous occlusion: the morbidity of 20. Caparrelli DJ, Freischlag J. A unified approach
a nonlethal disease. J Vasc Surg. 1986;4:333-7. to axillosubclavian venous thrombosis in a
11. Lindblad B, Tengborn L, Bergqvist D. Deep single hospital admission. Semin Vasc Surg.
vein thrombosis of the axillary-subclavian 2005;18:153-7.
veins: epidemiologic data, effects of different 21. Thompson RW, Schneider PA, Nelken NA,
types of treatment and late sequelae. Eur J Skioldebrand CG, Stoney RJ. Circumferential
Vasc Surg. 1988;2:161-5. venolysis and paraclavicular thoracic outlet
12. Aziz S, Straehley CJ, Whelan TJ Jr. Effort- decompression for effort thrombosis of the
related axillosubclavian vein thrombosis. A subclavian vein. J Vasc Surg. 1992;16:723-32.
new theory of pathogenesis and a plea for 22. Molina JE. Surgery for effort thrombosis of the
direct surgical intervention. Am J Surg. 1986; subclavian vein. J Thorac Cardiovasc Surg.
152:57-61. 1992;103:341-6.
13. Kunkel JM, Machleder HI. Spontaneous 23. Molina JE. Need for emergency treatment in
subclavain vein thrombosis: a successful subclavian vein effort thrombosis. J Am Coll
combined approach of local thrombolytic Surg. 1995;181:414-20.
therapy followed by first rib resection. Surgery. 24. Thompson RW, Petrinec D, Toursarkissian
1989;106:114. B. Surgical treatment of thoracic outlet
14. Kunkel JM, Machleder HI. Treatment compression syndromes. II. Supraclavicular
of Paget-Schroetter syndrome: a staged, exploration and vascular reconstruction. Ann
multidisciplinary approach. Arch Surg. Vasc Surg. 1997;11:442-51.
1989;124:1153-7. 25. Azakie A, McElhinney DB, Thompson RW,
15. Machleder HI. Evaluation of a new treatment Raven RB, Messina LM, Stoney RJ. Surgical
strategy for Paget-Schroetter syndrome: management of subclavian vein effort
Central Venous
Disease
179
180
15
Iliofemoral
Deep Venous Thrombosis
Mark g. Davies
181
goal of venous thrombectomy is to remove was 82%, and the assisted patency rate was
thrombus from the iliofemoral segments of 91%, which remained unchanged over a
a lower extremity in patients who are un- mean follow-up of 22 months. There was
suitable for catheter-directed thrombolysis mild reflux with few clinical symptoms of
(CDT) and in danger of limb loss.1 The postthrombotic syndrome.3 In the only pub-
chapter by Bo Eklf clearly illustrates the lished randomized trial, Plate et al. showed
role of surgical venous thrombectomy in the that, although the operated group (venous
modern management of iliofemoral DVT. thrombectomy) had better outcomes than
Surgical extraction of venous thrombi should the group treated with anticoagulation, nei-
be used in the management of patients with ther of the 2 groups did very well in the long
extensive iliofemoral disease in which limb run.4
loss is imminent, such as in phlegmasia
cerulea dolens. In these patients, there are
benefits to avoiding limb loss, preventing Endovascular
pulmonary embolism, and reducing the se-
verity of postthrombotic syndromes. Surgi-
Management
cal thrombectomy should not be performed The management of iliofemoral DVT is un-
in DVT that does not involve the iliofemoral dergoing an evolution with increased recog-
segment, in nonambulatory patients or in nition that early intervention may benefit the
high-risk surgical patients, and is generally well-selected patient. The indications for en-
reserved for young patients. Early mortality dovascular intervention are phlegmasia ceru-
is generally 1% if proper selection criteria lea dolens, acute/subacute inferior vena cava
are applied. Early iliac vein rethrombosis is (IVC) thrombosis, acute iliofemoral DVT,
13%. Cumulative patency is 75% at 4 years. acute femoropopliteal DVT, and subacute/
In the presence of nonadherent clots, the pa- chronic iliofemoral DVT. This evolution is
tency is 92%, while in patients with adherent prompted by accruing evidence of benefit
clots, the patency is 45%.2 In patients with and spurred on by the 2008 ACCP guide-
successful venous thrombectomy, 37% are lines as discussed by Comerota in chapter 5.
symptom-free compared to 18% in patients As described in previous chapters by Henke,
treated conservatively. Those patients with Wakefield, and Bush in this book, the etiol-
no symptoms had significantly lower am- ogy of DVT is multifactorial and must be
bulatory venous pressures, improved venous individualized for each patient. Iliofemoral
emptying, and a better calf pump function. DVT may be due to extrinsic compression,
These physiological changes portend to a wall damage secondary to catheters, and
lower incidence of chronic venous insuffi- transient and permanent hypercoaguable
ciency in the longer term. In a study that ex- states. Iliac vein compression is the most
amined the efficacy of stent placement after probable cause of iliofemoral DVT. One-
infrainguinal loco-regional thrombolysis half to two-thirds of patients with left-sided
and iliac thrombectomy of acute DVT in pa- iliofemoral DVT have intraluminal webs or
tients with May-Thurner syndrome, techni- spurs from chronic extrinsic compression of
cal success defined as complete vein patency the left iliac vein at the crossing point of the
and normal valve function was documented right common iliac artery. Approximately
in all patients. The calculated cumulative 2% to 5% of those with chronic deep ve-
primary patency rate for venous iliac stents nous insufficiency of the left leg may have
iliac vein compression syndrome. Iliac vein Patients suitable for catheter-directed
compression syndrome occurs when com- thrombolysis include young, active individu-
pression of the common iliac vein is severe als who have an acute (<10 days) deep venous
enough to inhibit the rate of venous outflow. thrombosis and those with isolated infrain-
In its more severe manifestation, iliac vein guinal disease. Additionally, any patient who
compression syndrome is known to cause has signs/symptoms of phlegmasia cerulea
acute iliofemoral DVT. Iliac vein compres- dolens, regardless of their clinical condition
sion syndrome is caused by the combination or age, should be considered for thrombolyt-
of compression and the vibratory pressure of ic therapy. The algorithm for care is shown
the right iliac artery on the iliac vein, which in Figure 15.1. The incidence of major clini-
is pinched between the artery and the pel- cal hemorrhage after fibrinolysis for DVT is
vic bone. In oncology patients requiring between 6% and 30%, a 3-fold increase com-
staging CT scans of the thorax, abdomen, pared to standard heparin therapy. Infusion
and pelvis, there is a prevalence of 6.8% of thrombolysis is associated with increased
unsuspected iliofemoral, 1.2% unsuspected risk of local hematoma formation at the site
common iliac, and 0.3% unsuspected infe- of catheter insertion and a lower risk of dis-
rior vena cava DVT.5 In a series of patients tant bleeding. The National Multicenter
treated medically for iliofemoral DVT, 50% Venous Registry recommends that any acute
of the limbs had a pathological (deep reflux deep venous thrombosis without a prior his-
or obstructive change) finding in the poplite- tory of thrombosis will yield the greatest de-
al segment after a 20-month follow-up. The gree of lysis, which is predictive of long-term
rate of recanalization was high. There was benefit and a significant decrease in the like-
no difference between calf and more proxi- lihood of the development of chronic venous
mal DVTs. Pain (62%), edema (46%), and insufficiency. Endovascular management
pigmentation (35%) were common, and only of iliofemoral DVT due to May-Thurner
27% of the legs with DVT were asymptomat- syndrome in patients with protein C and/
ic. One study suggests that the development or S deficiency is clinically effective in the
of the postthrombotic syndrome begins quite short term.11 The incidence of clinical signs
early. The frequency of the subjective symp- and symptoms of venous insufficiency and
toms is high.6 The introduction of catheter- duplex-scan findings of valvular reflux was
directed thrombolysis, rheolytic catheters, significantly lower in the patients in which
and the combination pharmacomechanical lytic therapy succeeded and patency was
thrombectomy (PMT) were initiated with kept, compared with patients experiencing
the goals of (1) eliminating iliofemoral ve- acute therapeutic failure or rethrombosis (P
nous thrombus, (2) providing unobstructed <.01).12
venous drainage from the affected limb, and Catheter-directed thrombolytic therapy
(3) preventing recurrent thrombosis; these for the treatment of acute extensive iliofem-
approaches appear to be associated with bet- oral DVT due to May-Thurner syndrome is
ter clinical outcome compared with either an effective method for restoring venous pa-
systemic fibrinolysis and standard anticoag- tency, preventing valvular insufficiency, and
ulation (Table 15.1).7-10 The specifics of the providing relief of the acute symptoms.13 In
catheters, techniques, and results are illus- a study by Dake et al., initial technical suc-
trated in chapters by Davies, Meissner, and cess was achieved in 34 of 39 patients (87%).
Arko in this book. The overall patency rate at 1 year was 79%.
Catheter-DirectedTherapy
Semba 1994 Observational 21 UK 85 85
Bjarnason 1997 Prospective Cohort 77 UK 79 79
Verhaeghe 1997 Observational 25 TPA 76 76
Raju 1998 Observational 24 UK 88 -
Mewissen 1999 Prospective 287 UK 83 -
Registry
Patel 2000 Observational 10 UK 100 100
OSullivan 2000 Observational 39 UK 87 -
Kasirajan 2001 Observational 17 UK/TPA/RPA 82 82
Chang 2001 Observational 10 TPA 100 100
Ouriel 2001 Prospective 11 RPA 91 -
Registry
Shortell 2001 Observational 31 UK/tPA 80 -
AbuRahma 2001 Prospective 51 UK/tPA 89 -
Controlled
Castaneda 2002 Prospective Cohort 25 rPA 92 -
Vedantham 2002 Observational 20 UK/tPA/rPA 89 82
Razavi 2002 Prospective Cohort 31 TNK 89 -
Elsharawy 2002 Randomized Trial 35 SK 100 -
Sugimoto 2003 Observational 54 UK/tPA - 85
Grunwald 2004 Observational 74 UK/tPA/rPA 98 -
Vedantham 2004 Observational 18 RPA 100 96
Lin 2006 Observational 46 tPA 64 -
Kim 2006 Observational 23 UK 81 -
Martinez 2008 Observational 21 tPA 60 -
PharmacomechanicalTherapy
Kasirajan 2001 Observational 17 tPA 80 82
Vedantham 2002 Observational 20 UK/rPA/RPA 62 -
Bush 2004 Observational 22 UK/rPA/RPA 65 100
Jackson 2006 Observational 28 TNK 100 80
Lin 2006 Observational 52 tPA 68 -
Kim 2006 Observational 14 UK 84 -
Arko 2007 Observational 14 TNK 80 90
Martinez 2008 Observational 22 tPA 80 -
Rao 2009 Observational 43 tPA 95 93
Shi 2009 Observational 16 UK 89 -
Li 2010 Observational 36 tPA 80 83
EKOS
No large studies
Abbreviations: EKOS = EkoSonic Endovascular System; rPA = reteplase; SK = streptokinase; TNK = tenecteplase;
tPA = tissue plasminogen activator; UK = urokinase. Adapted from Vedantham S et al. J Vasc Interv Radiol. 2009
Jul;20(7 suppl):S227-39.
Symptomatically, 85% of patients were com- was achieved in 30% for medical therapy
pletely or partially improved compared with vs. 78% endovascular therapy.15 Kwak et al.
findings before treatment. There were no demonstrated a technical success rate of
deaths, pulmonary embolus, cerebral hem- 96% (26 of 27 stents) and a clinical success
orrhage, or major bleeding complications.14 rate of 95% (21 of 22 patients). The causes
AbuRahma et al. reported a technical suc- of common iliac vein obstruction were May-
cess in 16 of 18 patients (89%) receiving Thurner syndrome (n = 16), pelvic mass (n
CDT and primary iliofemoral venous pa- = 2), and unknown (n = 4). Overall, the
tency rates at 1, 3, and 5 years of 24%, 18%, 1-year and 2-year primary patency rates were
and 18% and 83%, 69%, and 69% for medi- both 95%, and the 1-year and 2-year second-
cal therapy and endovascular therapy, re- ary patency rates were both 100%.16 From a
spectively. Long-term symptom resolution national registry of patients (n = 473) with
IliofemoralDVT
Stent
Figure 15.1 Algorithm for iliofemoral DVT prevention. Patients diagnosed with an iliofemoral DVT should be
stratified into symptomatic and asymptomatic and then assessed for contraindications to thrombolysis. Concomitant
with this, a risk-benefit analysis should also be performed. If in the clinicians opinion declotting needs to be
done and there are no contraindications for lysis, percutaneous lysis is performed with use of simple lyse and
wait, mechanical thrombectomy, pharmacomechancial (PMT), or ultrasound-assisted lysis (EKOS). If the patient
is symptomatic with a significant clot burden and contraindications to thrombolysis, open thrombectomy should
be performed. Venography and correction of underlying lesions with stent placement is recommended. Standard
therapy thereafter is anticoagulation, ambulation, and compression stockings.
symptomatic lower limb DVT, results of that all of those who initially had complete
312 urokinase infusions in 303 limbs of 287 lysis remained patent at 20 months, whereas
patients were analyzed. After thrombolysis, those with only partial lysis had a lower pa-
grade III (complete) lysis was achieved in 96 tency rate.21,22
(31%) infusions; grade II (50%99% lysis) Use of rheolytic mechanical throm-
in 162 (52%); and grade I (< 50% lysis) in bectomy catheters has been shown to be ef-
54 (17%). For acute thrombosis, grade III fective in the treatment of ilifemoral deep
lysis occurred in 34% of cases of acute and venous thrombosis. Additional infusion of
in 19% of cases of chronic DVT (P < .01). thrombolytic agents via the device creates
Major bleeding complications occurred in a novel treatment strategy of pharmacom-
54 (11%) patients, most often at the punc- echanical thrombectomy, which further en-
ture site. At 1 year, the primary patency hances thrombectomy efficacy (Figure 15.2).
rate was 60%. Lysis grade was predictive of In a preliminary experience, pharmacome-
1-year patency rate (grade III, 79%; grade II, chanical catheter-directed iliofemoral DVT
58%; grade I, 32%; P < .001).17 There was thrombectomy with early stent placement
no difference found in physical functioning was safe and effective.23 The Amplatz device
and well-being between the groups before is reported as having removal of thrombus at
the development of deep venous thrombo- 75% to 83% in lower extremity acute DVT
sis. Following treatment, patients receiving within 6 months and a patency of 77%.24,25
catheter-directed thrombolysis reported bet- The Arrow-Trerotola device, when used clini-
ter overall physical functioning, less stigma, cally in addition to thrombolytic therapy and
less health distress, and fewer postthrom- angioplasty with stents, has been reported to
botic symptoms compared to those patients have technical and clinical success in 100%
treated with anticoagulation alone.18 A 10- of patients with a 16-month clinical success
year study by Klbel et al. demonstrated a of 92%. Concerns about valve and intimal
technical success of 100% and a clinical suc- damage, although justified, have not been
cess of 96%. While cumulative patency was reported.26 In a study without adjunctive
89%, clinically 68% of limbs were asymp- preprocedural thrombolytic therapy, Kasira-
tomatic, 18% of limbs were moderately im- jan et al27 reported that half of the patients
proved, and the remainder were unchanged treated with the AngioJet device had 50%
or moderately worse.19 In another study to of their thrombus removed. Patency was re-
assess venous reflux and the obstruction stored in 77% of those patients with 50%
pattern after catheter-directed and systemic thrombus removal. Improved results have
thrombolysis of deep iliofemoral venous been observed in a similar study in which
thrombosis, valvular competence was pre- the AngioJet was used without preprocedural
served in 44% of patients treated with cath- thrombolytics. Sixty-five percent of patients
eter-directed thrombolysis compared with had complete thrombus removal while par-
13% of those treated with systemic throm- tial thrombus removal was observed in the
bolysis (P = 0.049). Reflux in any deep vein remaining 35%.28 The Trellis catheter can
was present in 44% of patients treated by provide segmental and controlled pharma-
catheter-directed lysis compared with 81% comechanical thrombectomy. It is associ-
of patients receiving systemic thrombolysis ated with a greater technical success rate, a
(P = 0.03).20 Bjarnason et al. have reported lower rate of bleeding, and a lower cost than
a 2-year patency of 78% and Ly et al. found that reported for CDT.29,30 The emission of
Figure 15.2 32-year-old patient presenting with symptomatic iliofemoral DVT. She is 6 weeks postpartum.
She underwent transpopliteal venogram in the prone position. The venograms show iliofemoral DVT ( A-
D ). The final result after PMT and 24 hours of EKOS-supplemented tPA infusion, which reveals a stenosis
consistent with May-Thurner syndrome ( EandF ). Final image shows a patent vein and an intact, fully
deployed stent indicated by the arrows ( G ).
ultrasound waves from an infusion catheter DVT treated with thrombolysis and best
delivering lytic agent is an interesting new medical therapy, a significantly greater pro-
adjunct to catheter-directed thrombolysis. A portion of iliofemoral patients (73%) than
recent study evaluated the success of lysis and femoral patients (31%) remained asymptom-
clinical outcomes in patients treated with ul- atic at the end of their follow-up (P <0.025);
trasound (US)-accelerated thrombolysis for 82% of iliofemoral limbs showed partial or
DVT using an EKOS device. Complete lysis complete lysis 4 weeks after diagnosis of clot.
(90%) was seen in 70% cases and overall No significant difference in reflux develop-
lysis (complete plus partial) was seen in 91%. ment was observed between the 2 groups.
No lysis occurred in 5 cases (9%), 4 of which Although the extent of reflux development
were chronic. It appears that the addition of was similar in both groups, iliofemoral pa-
ultrasound reduces total infusion time and tients still showed fewer clinical symptoms
provides a greater incidence of complete after follow-up.38
lysis.31 In another retrospective study, patients In a small randomized trial of local
undergoing CDT showed complete or partial thrombolysis and anticoagulation vs. antico-
thrombus removal in 32 (70%) and 14 (30%) agulation alone in patients with iliofemoral
cases, respectively. When compared to CDT DVT, the patency rate at 6 months was bet-
in this study, PMT provides similar treat- ter in cases treated with thrombolysis (13/18
ment success with reduced ICU and total [72%] vs. 2/17 [12%], P < 0.001). Venous
hospital length of stay and hospital costs.32 reflux was higher in patients treated with
Subclinical thrombus embolization during anticoagulant (7 patients [41%] vs. 2 [11%],
CDT/PMT is a common phenomenon in P = 0.04).39 In an open multicenter, random-
patients with iliofemoral DVT.33 A study by ized, controlled trial, 103 patients (64 men,
Protack et al. demonstrated that when using mean age 52 years) were allocated to CDT
a mixed modality (CDT/PMT) approach, (n = 50) or standard treatment alone (n =
83%, 83%, and 75% of patients were free of 53). After CDT, grade III (complete) lysis
recurrent DVT at 1, 2, and 3 years, respec- was achieved in 24 and grade II (50%90%)
tively. Furthermore, it appears that CDT/ lysis in 20 patients. After 6 months, iliofemo-
PMT without universal prophylactic IVC fil- ral patency was found in 32 patients (64.0%)
ter placement is safe and effective in treating in the CDT group vs. 19 (35.8%) controls,
acute iliofemoral DVT. Selective rather than corresponding to an absolute risk reduction
routine IVC filter placement is a safe and ap- (RR) of 28.2% (95% CI: 9.7%46.7%; P =
propriate approach.34 0.004). Venous obstruction was found in 10
Several reports suggest that long-term patients (20.0%) in the CDT group vs. 26
results of CDT are not satisfactory because (49.1%) controls; absolute RR 29.1% (95%
of the high recurrence rate of DVT, with CI: 20.0%38.0%; P = 0.004). Femoral ve-
stent length more than 6 cm being a poor nous insufficiency did not differ between
prognostic factor35 and only a marginal re- the two groups.40
duction in the incidence and severity of
postthrombotic limb syndromes.36 Further-
more, extent of thrombolysis is a statistically Conclusion
significant factor affecting the freedom of
rethrombosis and chronic change.37 In a ret- CDT reduced clot burden and DVT recur-
rospective study of iliofemoral and femoral rence, and it may prevent the formation of
16
Management of
Central Venous Stenosis
Michael J. reardon and Mark g. Davies
193
venous stenosis has been reserved for signifi- malignancy.10 When replacing the superior
cantly symptomatic patients who have an ob- vena cava combined with resection of medi-
struction or stenosis that is not amenable to astinal malignancies, reconstruction of a left
an endovascular approach, who have failed brachiocephalic vein alone results in a sig-
endovascular intervention, or who have ve- nificant rate of occlusion and development
nous resection as part of a planned en bloc of superior vena cava syndrome. Single right
tumor resection.4 brachiocephalic vein reconstruction or bilat-
eral brachiocephalic vein reconstruction in
this setting, and separate reconstruction of
Intrathoracic the veins, is preferable to use of a Y graft.11
Replacement grafts can include PTFE,1 spi-
Malignant Disease ral vein graft,12 or pericardium, although no
data exist to support the superiority of one
In the presence of known malignant disease over the other. Our preference for superior
of the thorax, percutaneous management of vena cava reconstruction has been a self-
complete superior vena cava (SVC) occlu- constructed pericardial tube graft (Figure
sion with thrombolysis and/or clot aspiration 16.1). Vessel involvement by soft tissue sarco-
followed by stent insertion is safe and effec- ma can classified as type I, artery and vein;
tive, giving sustained symptomatic relief.5 type II, artery only; type III, vein only; and
Primary patency in malignant and benign type IV, neither artery nor vein. In patients
cases at 1 year was 64% and 76%, respec- with retroperitoneal soft tissue sarcoma, the
tively. Overall symptom-free survival ranged most common vascular involvement pattern
from 1 to 34 months.6 Resection for malig- was vein only (type III, 64%). The inferior
nant tumor involvement was long considered vena cava (6 ePTFE tube grafts, 3 ePTFE
an absolute contraindication to resection. In- patches, 2 venoplasties), iliac vein (1 ePTFE
creasingly, resection and graft replacement bypass, 1 Dacron bypass, 1 venous patch),
have been used in selected cases in which en and superior mesenteric vein (1 anastomo-
bloc resection could be achieved.7,8 Infiltra- sis, 1 Dacron bypass) were restored in 80%
tion of the SVC due to advanced non small of the patients (n = 16). Morbidity was 36%
cell lung cancer (NSCLC) or thymoma can and mortality was 4%. At a median follow-up
be treated by prosthetic replacement or tan- of 19.3 months, the venous patency rate was
gential resection. It should not be considered 93.8% (primary and secondary).13
as palliative treatment because of the peri-
operative risks. SVC tangential resection
involves fewer surgical problems. However, Intrathoracic
since this procedure is used mostly for N2
NSCLC subjects, patients have a low mean Benign Disease
survival in spite of adjuvant therapy.9 Surgi-
cal reconstruction of the superior vena cava The need for intervention for benign etiolo-
with an ePTFE (expanded polytetrafluo- gies is increasing as the use of indwelling
roethylene) prosthesis provided immediate catheters for dialysis and cardiac therapy
and long-term relief of symptoms of superior such as pacemakers and implantable defi-
vena cava obstruction with a low surgical brillators increases. Endovascular treatment
morbidity, even in patients with unresectable of benign iliocaval occlusive disease is a safe
A B C
Figure 16.1 Central venous bypass. A.Superior vena cava reconstruction with a self-constructed pericardial
tube graft (arrow).B. Innominate vein tandom stenosis on venography resistant to angioplasty or stent placement.
C. Venogram of a superior vena cava reconstruction with a self-constructed pericardial tube graft.
caval vein obstruction or the May-Thurner of 8.3%. The variable prognosis of the vari-
syndrome. Technical success of 90% or ous IVC tumor lesions depends on tumor
greater can be achieved with this approach, entity, stage, resection status, and individual
and a 3-year assisted primary patency ex- risk factors.18 Palliative relief can be achieved
ceeds 80%. Venous reconstructions for il- with covered stents to obtain short-term de-
iofemoral or IVC obstruction offers 3-year compression, while other therapies (radia-
patency rates of 62%.14 The Palma procedure tion or chemotherapy) are considered or the
with autologous saphenous vein has the best natural course of the disease occurs (Figure
long-term patency, whereas long-term suc- 16.2).
cess with ePTFE is more moderate.17 Tumor
lesions of the inferior vena cava can originate
from the vein or can develop by malignant Budd-Chiari Syndrome
tumor infiltration from the surrounding tis-
sue. The resection rate is 83%, with surgical Primary Budd-Chiari syndrome (BCS) is
reconstruction of the IVC achievable in all characterized by a blocked hepatic venous
cases. The perioperative morbidity for such outflow tract at various levels from small
a strategy is 33%, with a hospital mortality hepatic veins to inferior vena cava, resulting
Figure 16.2.Stent grafting of the IVC. A and B demonstrate occlusion of the IVC and the common iliac veins.
Crepresents the patency of the IVC and common iliac veins after lytic therapy.
is determined by the balance between the found in the asymptomatic general popula-
acuteness of the disease, extent of the de- tion. However, the clinical syndrome, vari-
velopment of collateral circulation, involve- ously known as May-Thurner syndrome,
ment of one or both kidneys, and the origin Cockett syndrome, or iliac vein compression
of the underlying disease. Renal vein oc- syndrome, is thought to be a relatively rare
clusion is generally a complication of some contributor of chronic venous disease. Iliac
other condition but may also be a primary vein compression syndrome (May-Thurner
disease. Renal vein thrombosis is relatively syndrome) is the most probable cause of il-
rare. CT angiography is considered the in- iofemoral deep venous thrombosis (DVT).
vestigation of choice; alternatives include One-half to two-thirds of patients with left-
MR angiography or renal venography in sided iliofemoral DVT have intraluminal
highly selected patients. webs or spurs from chronic extrinsic com-
As the condition is relatively uncom- pression of the left iliac vein at the crossing
mon, consensus on the best form of therapy point of the right common iliac artery. Ap-
for this condition has been slow to evolve. proximately 2% to 5% of those with chronic
The trend in management has shifted to deep venous insufficiency of the left leg may
nonsurgical therapies, particularly systemic have May-Thurner syndrome. May-Thurner
anticoagulation, except in a highly select- syndrome occurs when compression of the
ed group of patients. The principal mode common iliac vein is severe enough to in-
of treatment includes correction of fluid hibit the rate of venous outflow. In its more
and electrolyte imbalance, dialysis, antihy- severe manifestation, May-Thurner syn-
pertensive drugs, anticoagulation, and in drome is known to cause acute iliofemoral
certain cases, thrombolysis. Percutaneous DVT.
catheter-directed thrombectomy with or May-Thurner syndrome is caused by
without thrombolysis for acute RVT is as- the combination of compression and the vi-
sociated with a rapid improvement in renal bratory pressure of the right iliac artery on
function and low incidence of morbidity. It the iliac vein that is pinched between the
is feasible for native and allograft renal veins artery and the pelvic bone. With the advent
and should be considered in patients with of catheter-directed thrombolytic therapy for
acute RVT, particularly in the setting of de- patients presenting with iliofemoral DVT,
teriorating renal function.24 the underlying cause has been unveiled, and
May-Thurner syndrome is gaining recogni-
tion. Patients presenting with symptoms of
May-Thurner chronic venous insufficiency often fail con-
servative treatment, and because of their
Syndrome crippling symptoms, they may have a high
rate of work absence or are on permanent
The obstruction of the left common iliac disability. If May-Thurner syndrome can be
vein by the pressure of the anteriorly posi- identified as the cause and corrected, pa-
tioned right common iliac artery, with in- tients quality of life would improve. With
timal changes, was first described by May the advent of endovascular stenting, the un-
and Thurner in 1956.25 Nonthrombotic derlying cause can be easily corrected, and
iliac vein lesions, such as webs and spurs de- long-term patency is acceptable26,27 (Figure
scribed by May and Thurner, are commonly 16.3).
Figure 16.3. May-Thurner syndrome. A. Stenosis of the common iliac vein with collaterals. B. An angioplasty
balloon inflated in the area. C. Final result: a patent common iliac vein without collaterals.
Figure 16.5. Venous thoracic outlet syndrome. Example of a patent subclavian vein with extrinsic
compression and collateral formation representing TOS.
venous stenosis or ongoing symptoms.29,30 dialysis access sites.39 Thus, patients with a
Angioplasty with or without stenting is dis- reasonable life expectancy or who are unable
couraged without concomitant anatomic to return for subsequent procedures should
decompression but does have an adjunctive be considered for alternative therapy.40 The
role in patients undergoing first rib resection right atrial bypass grafting has been used to
and can avoid the need for jugular turn- restore central venous patency in the careful-
down or patch angioplasty. ly selected patient, in whom all other access
sites are exhausted and in whom percutane-
ous dilation and/or stenting has failed.41
Dialysis Access-Related
Central Venous Stenosis Conclusion
Central venous obstruction is a common Central venous disease remains a small but
problem in patients with chronic renal fail- significant area of venous disease. The data
ure who undergo maintenance hemodialysis. in the literature are composed of small series
Incidences of central vein stenosis reported and case reports. In most cases, treatment
within hemodialysis patients have ranged has not been validated by level I evidence.
from 11% to 40%.31-35 Significant stenosis or There remains significant potential for clini-
occlusion of the subclavian vein is known to cal research and reporting. The practicing
occur in 20% to 50% of patients who have vascular specialist should be cognizant of
had central venous catheters inserted into each of the entities and be prepared to offer
the subclavian vein or the internal jugular the spectrum of medical, radiological, and
vein.36 Stents provide a temporary benefit surgical therapy.
in most patients with central or peripheral
upper extremity stenosis and obstruction.
Incidence of central vein stenosis reported
within hemodialysis patients have ranged references
from 11% to 40%.31-33,37,38 Regular follow-up 1. Huu TC. Central venous stenosis: review of
and reinterventions are required to maintain the literature from 1980 to 2000. Nephrologie.
patency and achieve long-term clinical suc- 2001;22(8):479-85.
cess. Stents used for central venous lesions 2. Schifferdecker B, Shaw JA, Piemonte TC,
have higher clinical success rates than stents Eisenhauer AC. Nonmalignant superior
used for peripheral venous lesions. However, vena cava syndrome: pathophysiology
endovascular therapy for central venous ste- and management. Catheter Cardio Inte.
nosis in the dialysis access patient, whether 2005;65:416-23.
via angioplasty or stenting, is safe, with low 3. Marlier S, Bonal J, Cellarier G, Bouchiat C,
rates of technical failure. Multiple addi- Talard P, Dussarat GV. Superior vena cava
tional interventions are the rule with both syndromes of benign etiology. Presse Med.
treatment modalities. Although neither of- 1996;25(26):1203-7.
fers truly durable outcomes, stent placement 4. Picquet J, Blin V, Dussaussoy C, Jousset Y,
does not improve upon the patency rates Papon X, Enon B. Surgical reconstruction of
provided by balloon angioplasty and does the superior vena cava system: indications and
not add to the longevity of ipsilateral hemo- results. Surgery. 2009;135:93-9.
5. Crowe MT, Davies CH, Gaines PA. 14. Hartung O, Otero A, Boufi M, Decaridi
Percutaneous management of superior G, Barthelemy P, Juhan C, et al. Mid-
vena cava occlusions. Cardiovasc Inter Rad. term results of endovascular treatment for
1995;18(6):367-72. symptomatic chronic nonmalignant iliocaval
6. Barshes NR, Annambhotla S, El Sayed venous occlusive disease. J Vasc Surg.
HF, Huynh TT, Kougias P, Dardik A, et al. 2005;42(6):1138-44.
Percutaneous stenting of superior vena cava 15. Rizvi AZ, Kalra M, Bjarnason H, Bower TC,
syndrome: treatment outcome in patients Schleck C, Gloviczki P. Benign superior vena
with benign and malignant etiology. Vascular. cava syndrome: stenting is now the first line of
2007;15(5):314-21. treatment. J Vasc Surg. 2008;47(2):372-80.
7. Suzuki K, Asamura H, Watanbe S, Tsuchiya 16. Teo N, Sabharwal T, Rowland E, Curry P,
R. Combined resection of superior vena cava Adam A. Treatment of superior vena cava
for lung carcinoma: prognostic significance of obstruction secondary to pacemaker wires with
patterns of superior vena cava invasion. Ann balloon venoplasty and insertion of metallic
Thoracic Surg. 2004;78:1184-9. stents. Eur Heart J. 2002;23(18): 1465-70.
8. Thomas P, Magnan PE, Moulin G, Giudicelli 17. Jost CJ, Gloviczki P, Cherry KJJ, McKusick
R, Fuentes P. Extended operation for the lung MA, Harmsen WS, Jenkins GD, et al. Surgical
cancer invading the superior vena cava. Eur J reconstruction of iliofemoral veins and the
Cardiothorac Surg. 1994;8:177-82. inferior vena cava for nonmalignant occlusive
9. Politi L, Crisci C, Montinaro F, Andreani disease. J Vasc Surg. 2001;33(2):320-8.
M, Podzemny V, Borzellino G. Prosthetic 18. Eder F, Halloul Z, Meyer F, Huth C, Lippert
replacement and tangential resection of H. Surgery of inferior vena cava associated
the superior vena cava in chest tumors. malignant tumor lesions. Vasa. 2008;37(1):68-
J Cardiovasc Surg (Torino). 2007;48(3):363-8. 80.
10. Magnan PE, Thomas P, Giudicelli R, Fuentes 19. Valla DC. Primary Budd-Chiari syndrome.
P, Branchereau A. Surgical reconstruction J Hepatol. 2009;50(1):195-203.
of the superior vena cava. Cardiovasc Surg. 20. Senzolo M, Cholongitas EC, Patch D,
1994;2(5):598-604. Burroughs AK. Update on the classification,
11. Shintani Y, Ohta M, Minami M, Shiono assessment of prognosis and therapy of
H, Hirabayashi H, Inoue M, et al. Long- Budd-Chiari syndrome. Nat Clin Pract Gastr.
term graft patency after replacement of 2005;2(4):182-90.
the brachiocephalic veins combined with 21. Lee BB, Villavicencio L, Kim YW, Do YS,
resection of mediastinal tumors. J Thorac Koh KC, Lim HK, et al. Primary Budd-
Cardiovasc Surg. 2005;129(4):809-12. Chiari syndrome: outcome of endovascular
12. Erbella J, Hess PJ, Huber TS. Superior vena management for suprahepatic venous
cava bypass with superficial femoral vein for obstruction. J Vasc Surg. 2006;43(1):101-8.
benign superior vena cava syndrome. Ann 22. Mahmoud AE, Helmy A, Billingham L, Elias
Vasc Surg. 2006;20:834-8. E. Poor prognosis and limited therapeutic
13. Schwarzbach MH, Hormann Y, Hinz U, options in patients with Budd-Chiari syndrome
Leowardi C, Bckler D, Mechtersheimer and portal venous system thrombosis. Eur
G, et al. Clinical results of surgery for J Gastr Hepat. 1997;9(5):485-9.
retroperitoneal sarcoma with major 23. Asghar M, Ahmed K, Shah SS, Siddique MK,
blood vessel involvement. J Vasc Surg. Dasgupta P, Khan MS. Renal vein thrombosis.
2006;44(1):46-55. Eur J Vasc Endovasc Surg. 2007;34(2):217-23.
24. Kim HS, Fine DM, Atta MG. Catheter- RH, Saeed M, Dennis VW. Hemodialysis-
directed thrombectomy and thrombolysis for associated subclavian vein stenosis. Kidney Int.
acute renal vein thrombosis. J Vasc Interv 1988;33:1156-9.
Radiol. 2006;17(5):815-22. 34. Surratt RS, Picus D, Hicks ME, Darcy MD,
25. May R, Thurner J. Em Gefsssporn in Kleinhoffer M, Jendrisak M. The importance
der v.iliaca corn. sin. als wahrscheinliche of preoperative evaluation of the subclavian
Ursacheder uberwiegend liknsseitigen vein in dialysis access planning. Am J
Beckenvenenthrombose. Z Kreisl-Forsch. Roentgenol. 1991;156:623-5.
1956;45:912. 35. Lumsden AB, MacDonald MJ, Isiklar H,
26. Mussa FF, Peden EK, Zhou W, Lin PH, Martin LG, Kikeri D, Harker LA, et al.
Lumsden AB, Bush RL. Iliac vein stenting for Central venous stenosis in the hemodialysis
chronic venous insufficiency. Tex Heart Inst J. patient: incidence and efficacy of endovascular
2007;34(1):60-6. treatment. Cardiovasc Surg. 1997;5:504-9.
27. Lin PH, Zhou W, Dardik A, Mussa FF, 36. Chandler NM, Mistry BM, Garvin PJ.
Kougias P, Hedayati N, et al. Catheter-direct Surgical bypass for subclavian vein occlusion
thrombolysis versus pharmacomechanical in hemodialysis patients. J Am Coll Surg.
thrombectomy for treatment of symptomatic 2002;194(4):416-21.
lower extremity deep venous thrombosis. Am J 37. Lumsden AB, MacDonald MJ, Isiklar H,
Surg. 2006;192(6):782-8. Martin LG, Kikeri D, Harker LA, et al.
28. Landry GJ, Liem TK. Endovascular Central venous stenosis in the hemodialysis
management of Paget-Schroetter syndrome. patient: incidence and efficacy of endovascular
Vascular. 2007;15(5):290-6. treatment. Cardiovasc Surg. 1997;5:504-9.
29. Molina JE, Hunter DW, Dietz CA. 38. Surratt RS, Picus D, Hicks ME, Darcy MD,
Protocols for Paget-Schroetter syndrome Kleinhoffer M, Jendrisak M. The importance
and late treatment of chronic subclavian of preoperative evaluation of the subclavian
vein obstruction. Ann Thorac Surg. vein in dialysis access planning. Am J
2009;67(2):416-22. Roentgenol. 1991;156:623-5.
30. Molina JE, Hunter DW, Dietz CA. 39. Bakken AM, Protack CD, Saad WEA, Lee
Paget-Schroetter syndrome treated with D, Waldman DL, Davies MG. Longterm
thrombolytics and immediate surgery. J Vasc outcomes of primary angioplasty versus
Surg. 2007;45(2):328-34. primary stenting in the management of central
31. Schumacher KA, Wallner B, Weidenmaier venous stenoses in the hemodialysis patient.
W, Friedrich JM. Venous occlusions J Vasc Surg. 2007;45:776-83.
distant to the shunt as malfunction factors 40. Oderich GS, Treiman GS, Schneider P,
during hemodialysis. Fortschr Rntgenstr. Bhirangi K. Stent placement for treatment of
1989;150:198-201. central and peripheral venous obstruction: a
32. Haage P, Vorwerk D, Piroth W, Schuermann long-term multi-institutional experience.
K, Guenther RW. Treatment of hemodialysis- J Vasc Surg. 2000;32(4):760-9.
related central venous stenosis or occlusion: 41. El-Sabrout RA, Duncan JM. Right atrial
results of primary wallstent placement bypass grafting for central venous obstruction
and follow-up in 50 patients. Radiology. associated with dialysis access: another
1999;212:175-80. treatment option. J Vasc Surg. 1999;29(3):
33. Schwab SJ, Quarles LD, Middleton JP, Cohan 472-8.
17
205
However, the most common cause of Mutation of the prothrombin gene has
secondary MVT remains thrombophilia or also been associated with increased risk of
hypercoagulable conditions. In fact, more thrombotic events. This mutation leads to
than half of all cases of MVT have an higher plasma prothrombin levels and is 4
identifiable hypercoagulable condition.6,12 times as common in patients with a history
Hypercoagulable conditions can be either of venous thrombosis compared to their con-
inherited when the subject is born with a ge- trols.15 In one small study, this mutation was
netic defect or acquired. The acquired types found in 40% of cases with idiopathic portal
of hypercoagulable conditions refer to those vein thrombosis.16
systemic diseases that have been known and
proven to be associated with thrombosis.
However, the true frequency of these dis- Acquired Hypercoagulable Conditions
orders in patients with MVT is difficult to Multiple risk factors are associated with ac-
estimate since most studies in the literature quired hypercoagulable states. The most
included patients with all forms of deep ve- common of these conditions are malignan-
nous thrombosis. One study suggested that cy, hormonal therapy, and antiphospholipid
an acquired or inherited hypercoagulable antibodies.
state was more likely in patients with isolated Malignancy has been a known risk
MVT compared with those with concomi- factor for hypercoagulable state and venous
tant involvement of the portal or splenic thrombosis. This is particularly more com-
veins.13 A list of the hypercoagulable condi- mon in older patients and in those malig-
tions associated with MVT can be found in nancies associated with advanced disease at
Table 17.1. the time of diagnosis (eg, pancreas).17 Cer-
tain cancers are particularly more associated
with MVT. Myeloproliferative disorders,
Inherited Hypercoagulable Conditions especially polycythemia vera and essential
The available data suggest that the most thrombocythemia, are strongly associated
common inherited hypercoagulable condi- with MVT.1 In a small series of patients with
tion associated with MVT is the activated MVT, 10% had associated polycythemia
protein C resistance, secondary to factor vera.18 Pancreatic cancer and metastatic ab-
V Leiden mutation. In one epidemiologic dominal cancers are also strongly associated
study of cases with documented MVT, fac- with MVT. In one series, abdominal cancer
tor V Leiden mutation was present in 45% of was present in 24% of all cases of document-
those cases.6 Activated protein C resistance ed MVT, with pancreatic cancer present in
can also occur due to other forms of muta- half of those cases.6
tions in another 10% of cases.14 Hormonal therapy is also highly associ-
Another cause of inherited hyperco- ated with venous thrombosis in general and
agulable conditions is the deficiency of the MVT in particular. Historically, oral con-
natural anticoagulant proteins that maintain traceptive use accounted for 9% to 18% of
the balance between the coagulant and anti- the episodes of MVT in young women.12,19
coagulant activity of the blood. Deficiencies In a more recent study, oral contraceptives
of protein C, protein S, and antithrombin or estrogen hormonal replacement therapy
III (AT III), collectively, are present in about was present in 12% of women diagnosed
10% of cases with MVT.6,14 with MVT.6 Antiphospholipid antibodies
are directed against plasma proteins bound ing mesenteries, supplying the wall of the
to anionic phospholipids and are associated gut. The venous drainage follows a similar
with venous thrombosis. The antiphospho- pattern, with the venae rectae forming the
lipid antibodies, including the cardiolipin venous arcades, which drain the gut into the
antibodies and the lupus anticoagulants, are inferior mesenteric and superior mesenteric
present in another 5% to 10% of cases.6,8 veins. The inferior mesenteric vein joins the
splenic vein before the confluence with the
superior mesenteric vein behind the neck of
Pathophysiology the pancreas to form the portal vein that sup-
plies the liver parenchyma (Figure 17.1).
The celiac, superior mesenteric, and inferior There are multiple extrinsic and intrin-
mesenteric arteries are the visceral branches sic mechanisms that control the blood flow
of the aorta supplying the gut. The branches to the bowel wall. This is mainly controlled
from these arteries terminate as the arteriae via changes in the resistance of the mesenter-
rectae in the distal parts of the correspond- ic arterioles, which accounts for the marked
Figure 17.1 Normal mesenteric venous circulation. Reprinted with permission from Agur AM, Dailey, AF, Boileau
Grant JC. Grants Atlas of Anatomy. 11th ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2004.
variation in the splanchnic blood flow, be- keeps flowing into the bowel wall in patients
tween 10% and 35% of the cardiac output, with venous compromise. This leads to in-
depending on variable conditions such as creased intravascular hydrostatic pressure,
fasting and postprandial states and systemic which dilates the blood vessels and widens
hypotension. The bowel can withstand an the fenestrations between the vascular en-
acute 75% reduction of the mesenteric blood dothelial cells. This results in extravasation
flow for up to 12 hours without significant of plasma and red blood cells into the bowel
damage occurring.20 When ischemic injury wall or lumen. Tension in the submucosal
of the intestine develops, progressive vaso- extravascular compartment or prolonged
constriction develops in the affected vas- stasis-induced thrombosis of the microvas-
cular bed, thereby reducing collateral flow. culature may result in interruption of the ar-
Vasoconstriction can persist even after blood terial blood flow.23 Grossly, the bowel wall is
flow has been restored, leading to continued swollen and looks cyanotic, with intramural
bowel ischemia and providing the rationale hemorrhage. These changes also affect the
for therapeutic infusion of the vasodilator nearby mesentery. The transition from the
papaverine for treatment of acute mesenteric affected bowel segment to the normal bowel
ischemia.20,21 is usually gradual, unlike that seen with arte-
The location of the initial thrombosis rial occlusion, which makes it more difficult
within the mesenteric venous circulation to accurately evaluate viable from nonviable
varies with the etiology. Thrombosis due bowel during surgery. The arterial pulsations
to intra-abdominal causes such as cirrhosis, are present up to the bowel wall, but arte-
neoplasm, or operative injury starts at the rial vasoconstriction frequently intervenes,
large vessels (eg, superior mesenteric vein which accentuates the compromised intra-
or portal vein) and progresses peripherally mural blood flow in the congested bowel.
to involve the smaller venous arcades and Later on, transmural infarction of the bowel
arcuate channels. However, thrombosis due sets in, and at that point it may be impos-
to an underlying hypercoagulable state be- sible to differentiate venous from arterial
gins in the small vessels and progresses to occlusion as a cause of the bowel infarction.
involve the larger vessels.1,13 Intestinal infarc- Transmural bowel infarction happens in
tion rarely occurs unless the branches of the 16% of patients suffering from acute MVT.24
peripheral arcades and the microcirculation Serosanguineous peritoneal fluid accompa-
are thrombosed, even when the junction of nies early hemorrhagic infarction. When
the portal vein and the superior mesenteric a large segment of the bowel is affected, it
vein is occluded. In a recent large series of leads to extensive third space fluid loss with
patients diagnosed with MVT, bowel infarc- systemic hypovolemia.13,20,21
tion occurred in 33% of them that required Two factors contribute to the bowel in-
bowel resection.22 Inferior mesenteric vein jury when affected by acute MVT, the first
thrombosis leading to infarction has been being transmural ischemia secondary to ces-
reported in fewer than 6% of cases with mes- sation of the blood flow. The second factor
enteric venous thrombosis.21 is reperfusion injury if restoration of flow is
When acute MVT happens, the ve- achieved. Most of the damage is due to reper-
nous drainage of the affected segment of fusion injury after a brief period of ischemia,
the bowel is compromised. The intramural while the detrimental effects of hypoxia pre-
blood volume increases as the arterial blood dominate with longer periods of ischemia.25
MVT can lead to presinusoidal portal is present in almost all patients with acute
hypertension in the long run and is classi- and subacute MVT, no other symptoms are
fied as chronic mesenteric venous throm- pathognomonic of MVT. The pain is usu-
bosis. This usually happens when patients ally central abdominal and colicky in nature.
have no symptoms at the time of the ini- The duration of pain varies, but 75% of cases
tial thrombosis but secondary pathologic had the symptoms for more than 48 hours,
changes develop over the time. This usually a very important distinction from mesenter-
manifests as gastrointestinal bleeding from ic arterial causes where the presentation is
esophageal and/or intestinal varices. Most of much earlier.7 Mathews and White27 found
these patients have associated thrombosis of that approximately 50% of their patients had
the major veins such as the portal or splenic pain from 5 to 30 days before seeking medi-
vein. It was reported that portal hypertension cal attention, and 27% reported abdominal
develops in 25% of patients who suffered pain for more than 1 month. It was also
from acute MVT.24 reported that patients who presented with
symptoms more than 5 days had a better
prognosis than patients presenting earlier,
Clinical Presentation possibly due to a more benign disease form
that allowed them to delay seeking medical
MVT can present with acute onset within attention for so long.21
hours, subacute within days to even weeks, Anorexia, nausea, vomiting, and di-
or chronic that is usually asymptomatic until arrhea are also common in acute MVT,
late complications occur. Patients with the occurring in more than 50% of cases. Clini-
acute presentation are at the highest risk for cal gastrointestinal bleeding in the form of
developing bowel infarction and peritoni- bloody diarrhea or melena is present in 15%
tis. On the other hand, the subacute form of cases. Hematemesis is also present in an
is more indolent, where abdominal pain is additional 13% of cases. Actually, the pres-
prominent but neither infarction nor vari- ence of hematemesis, as well as bleeding per
ceal bleeding is likely. In less common situa- rectus, should alert the physician to the pos-
tions, the subacute form will evolve over the sibility of mesenteric ischemic catastrophe.21
period of days or weeks into intestinal infarc- However, occult blood in stool is found in
tion and peritonitis, blurring the distinction nearly 50% of cases.12
between the acute and the subacute presen- Initial physical findings in acute MVT
tation.1 In the chronic form of the disease, vary greatly, reflecting both different stages
patients are essentially asymptomatic, and and degrees of bowel injury. Most of the pa-
their presentation can be upper or lower gas- tients will have some degree of abdominal
trointestinal bleeding secondary to esopha- tenderness and decreased bowel sounds with
geal or intestinal varices. Those patients can abdominal distention. However, a minor-
also develop hypersplenism with pancytope- ity of patients will develop peritoneal signs
nia or secondary thrombocytopenia.21,26 in the form of rebound tenderness and ab-
The hallmark of mesenteric ischemia, dominal rigidity, which signifies the pres-
whether it is due to arterial or to venous ence of intestinal infarction with peritonitis.
thrombosis, is abdominal pain that is out In a large series of patients with MVT, this
of proportion to the physical findings. With occurred in 33% of cases.22 Hemodynamic
the exception of the abdominal pain, which instability can also occur if a large segment
of the bowel is affected due to fluid seques- suggest the diagnosis of intestinal ischemia.21
tration within the bowel lumen or the ab- A useful clinical guideline is that any patient
dominal cavity. It also appears later in the with acute abdominal pain and metabolic
disease as a manifestation of septic shock acidosis is suspected of having acute mesen-
secondary to the bowel infarction and septic teric ischemia until proven otherwise.
peritonitis. It was consistently shown that the
presence of hemodynamic instability with a
systolic blood pressure of less than 90 mm Workup
Hg denotes a poor prognosis.28 About half
of the patients have a personal or family his- In the absence of any reliable and specific
tory of deep venous thrombosis, pulmonary symptoms, signs, and laboratory studies,
embolism, or a heritable hypercoagulable the preoperative diagnosis of acute MVT is
state, and this is an important point to in- difficult. Also, the severe variability in the
quire about as it can guide the diagnosis and duration and severity of the manifestations
therapy of the case.7 significantly adds to the difficulty. The aim
Laboratory studies in all forms of intes- of the workup is to reach the correct diag-
tinal ischemia have low sensitivity and speci- nosis of MVT before bowel infarction devel-
ficity. While abnormal laboratory values may ops, in which case the prognosis becomes
be helpful in bolstering suspicion for acute significantly worse. In the past, the correct
mesenteric ischemia, normal laboratory val- diagnosis of MVT was reached in more than
ues do not exclude it and do not justify delay- 90% of cases at laparotomy for patients with
ing management when clinical suspicion is peritoneal signs secondary to bowel infarc-
present. There is no specific laboratory test tion, signifying that the diagnosis was severe-
that is pathognomonic of MVT. The pres- ly delayed. In recent years, there has been
ence of increased serum lactate levels and enormous improvement of diagnostic radio-
metabolic acidosis may serve to identify pa- graphic modalities that helped in achieving
tients with established bowel infarction, but this goal. Nowadays, most of the patients are
this is a late finding.1 Elevated serum amy- diagnosed prior to surgery, with significant
lase levels have been observed in half the pa- improvement in morbidity and mortality.
tients with intestinal ischemia. An elevated
phosphate level has also been observed in Projectional Radiography
80% of cases. Creatine kinase (CK)-BB iso-
enzyme elevations have been found to cor- Plain films of the abdomen are nonspecific
relate with bowel infarction while total CK and can be normal in 25% of patients. Posi-
levels were not found to be useful. Lactate tive findings include intestinal dilatation,
dehydrogenase (LDH) elevation was present small bowel pseudo-obstruction pattern,
in 75% of cases with bowel infarction but did or paralytic ileus. More specific but far less
not differentiate between ischemia and in- common findings include thumbprinting,
farction.29-31 In fact, the simple leukocytosis in which multiple round, smooth, soft tissue
with a white cell count above 12,000/cu mm densities project into the intestinal lumen as
with an increase in the proportion of poly- a result of mucosal and submucosal edema
morphonuclear cells is the most clinically and hemorrhage. Specific but usually late
helpful lab test and is present in two-thirds signs include the presence of air in the wall
of cases. Currently, laboratory tests can only of the bowel (peumatosis intestinalis) (Fig-
enlarged superior mesenteric vein and per- agents.23 MR, however, is best used in the
sistent enhancement of the bowel wall due nonacute setting. The critically ill patient
to congestion, as well as thickened mesen- with suspected acute mesenteric ischemia
tery (Figure 17.3). In late cases, it will show usually has life support apparatus that is in-
pneumatosis, portal venous gas, and possibly compatible with the MR scanner. Those pa-
intraperitoneal free air and fluid.21 tients are best scanned with MDCT, which
also offers better spatial resolution.35
fusion of vasodilator agents (papaverine), the pain persist for several days before pre-
thrombolytic agents (streptokinase, uroki- sentation, should have MDCT as their initial
nase, recombinant tissue plasminogen acti- evaluation after starting aggressive resuscita-
vator), and angioplasty.36 The angiographic tion. In cases where MVT was diagnosed,
findings of MVT have been determined the patient should immediately be started
experimentally and clinically and include21: on heparin. If the patient has obvious peri-
toneal signs, the patient should immediately
go the operating room for exploratory lapa-
Demonstration of a thrombus in the rotomy. A mesenteric angiogram is needed
superior mesenteric vein (SMV) on the in case of a negative CT scan for MVT or in
venous phase of the study with partial cases where endovascular interventions are
or complete occlusion. needed. At any time, if the patient starts de-
Failure to visualize the SMV or portal veloping peritoneal signs, the patient has to
vein on delayed views. undergo exploratory laparotomy.
Slow or absent filling of the mesenteric
veins.
Mesenteric arterial spasm with back Management
flushing into the aorta on intra-arterial
contrast injection. The principles of management of MVT are
Failure of the arterial arcades to empty. 3-fold:
Prolonged blush of the wall of the
involved bowel segment. Prevention of further development of
The angiography can also show slow venous thrombosis,
reconstitution of the mesenteric resection of necrotic bowel, and
venous blood flow proximal to the prevention of recurrence of venous
thrombus, which is important in the thrombosis.
therapeutic decisions.
Historically, the treatment of acute and sub-
acute MVT included anticoagulation with
The workup of patients with suspected exploratory laparotomy with resection of the
MVT is summarized in the guidelines of necrotic bowel. Laparotomy was an essential
management of intestinal ischemia from the part of the management both for diagnostic
American Gastroenterological Association as and therapeutic reasons, as preoperative di-
illustrated in the algorithm in Figure 17.4.37 agnosis of MVT was not possible. Recently,
Patients presenting with abdominal pain that with the advent of MDCT, the preopera-
is severe enough to call to the attention of a tive diagnosis of MVT was possible in most
physician, whose pain persists for more than cases. Adding to that the rapid and extensive
2 or 3 hours, and whose clinical picture does development of endovascular techniques
not suggest other abdominal problems (eg, as minimally invasive therapy options, cur-
cholecystitis or diverticulitis) should be eval- rently the use of exploratory laparotomy is
uated and treated for acute mesenteric isch- limited to patients who have persistent peri-
emia. In those patients, suspicion of MVT is toneal signs for possible bowel resection. In
signaled by a history of DVT or familial or a one study, that was only needed in 33% of
personal hypercoagulability state, or having their patients with acute MVT.22
laparotomy
resuscitate the patient and no yes
correct any predisposing or Plain film History of DVT or familial
Peritoneal findings Abdominal
precipitating factors hypercoagulable state no angiogram
other yes
cause
Dynamic CT scan
no persistent Persistent
peritoneal findings peritoneal findings
Heparin with or
without thrombolytic laparotomy
agents
Short ischemic
extensive ischemic Involvement
segment
Second look
+/ resection
Thrombolytic Agents
Figure 17.4 Diagnosis and treatment of acute mesenteric venous thrombosis. Solid lines indicate accepted
management plan; dashed lines indicate alternate management plan. DVT, deep venous thrombosis. (From American
Gastroenterological Association Medical Position Statement: guidelines on intestinal ischemia. Gastroenterology
118(5): 951-3.) With permission from Elsevier.
Currently, the management of acute MVT. The theoretical aim of heparin in such
and subacute MVT includes: a situation is to prevent progression of the
thrombosis and limit the amount of bowel
Anticoagulation resection, if needed. Contrary to heparin use
Exploratory laparotomy in acute mesenteric arterial thrombosis and
Endovascular options for recanalizing embolism, where the use is delayed for 48
the thrombosed veins hours postoperatively to avoid postoperative
Adjunctive supportive measures bleeding, in acute and subacute MVT it is
started immediately with the diagnosis of the
condition, continues intraoperatively, and in
Anticoagulation the immediate postoperative period. Antico-
Heparin is an essential component of ther- agulation should be started even in the pres-
apy in all patients with acute and subacute ence of gastrointestinal bleeding, as the risk
turing the liver parenchyma both through thrombectomy device (ev3, Plymouth, MN),
the transjugular approach and especially and the Arrow Trerotola device (Arrow In-
through the direct percutaneous approach. ternational Inc, Reading, PA). Aspiration
Very few case reports of thrombolysis via a thrombectomy using a large angulated cath-
surgically placed mesenteric venous cath- eter (at least 8 Fr) has also been described.42
eter into one of the small venous tributar- The aim of these devices is to debulk the
ies during exploratory laparotomy are found mesenteric venous clot, thereby reducing
in the literature.45 Most of the earlier stud- the dose and duration needed for throm-
ies of mesenteric venous thrombolysis used bolysis, especially in patients with high risk
urokinase. Nowadays, because urokinase of bleeding complications. Also, balloon an-
is no longer available on the market, tPA is gioplasty has also been described for cases
typically used. The usual dose is an initial in which there have been venous stenotic le-
bolus of 2 mg followed by infusion of 1 to 2 sions or large clot burden not responding to
mg/h for 24 hours. At that time, a catheter thrombolysis or mechanical thrombectomy
check and repeat angiogram to evaluate the because of a large chronic component in the
response are performed. If there is still re- clot. The angioplasty is used to relieve the
sidual thrombus left, continued infusion for stenosis with or without use of stents or cov-
another 24 hours is performed, after which ered stents. They are also useful to fragment
time the catheter is removed. In the percuta- and displace the clot to restore flow within
neous trans-hepatic approach, the track has the portal venous system to relieve the con-
to be sealed with haemostatic glue and coils gestion in the affected bowel.42
to avoid intraperitoneal bleeding from the
liver track, which can be life-threatening. Intra-arterial Vasodilator Infusion
The patient should be closely moni- Mesenteric arterial vasospasm is commonly
tored for manifestations of bleeding and present in cases of acute and subacute MVT.
fibrinogen levels to avoid overconsumption It is specifically useful in cases following ex-
and excessive thrombolysis. Heparin infu- ploration where there is doubtful bowel left
sion continues throughout the thrombolytic behind for a second-look laparotomy. In this
infusion, however, at a smaller dose (500 situation, the use of the vasodilators can im-
units/h), to resume full dose heparinization prove the blood flow to the affected bowel,
once thrombolytic infusion has stopped. which in addition to anticoagulation and
other endovascular interventions, can help
Percutaneous Mechanical Thrombectomy limit the amount of bowel to be resected.1
This is another endovascular adjunct proce- Multiple vasodilators can be used for this
dure for clearing acute MVT. It can be used indication, including tolazoline, nitroglyc-
only when there is trans-hepatic access into erin, glucagon, and isoproterenol. However,
the portal venous system. It is also indicated the most-used vasodilator for this purpose
when there is a large clot burden in the main is papaverine. Papaverine is a nonaddictive
mesenteric veins. Several commercially opium derivative extracted from the poppy
available percutaneous mechanical throm- plant. It acts by inhibiting the phosphdiester-
bectomy devices have been reportedly used ase enzyme that metabolizes cyclic adenos-
for that indication including the AngioJet ine monophosphate (cAMP). The net effect
rheolytic thrombectomy device (MEDRAD, is to increase the tissue levels of cAMP,
Inc., Warrendale, PA), the Helix Clot Buster which is a direct vascular smooth muscle re-
laxant. Ninety percent of papaverine under- require surgery and bowel resection are sick-
goes first-pass metabolism by the liver, which er and have longer hospital stays and a more
limits its systemic toxicity and side effects.38 complicated course than those who do not
The usual dose is a bolus of 30 to 60 mg, fol- require surgery.1 However, there has been no
lowed by infusion at a rate of 30 to 60 mg/h. correlation between the length of resected
The duration of therapy is usually 24 to 48 bowel and the mortality rate, although it has
hours and is guided by clinical and angio- a detrimental effect on long-term morbidity
graphic response. Patients receiving papaver- in the form of short bowel syndrome and its
ine infusion should be closely monitored in sequelae.21
an ICU setting for hypotension. The sudden In recent years, the mortality rate has
onset of hypotension in such patients indi- been slightly lower than those from earlier
cates that the catheter has been dislodged series, now ranging between 13% and 30%.22
from the SMA and that the drug is being in- This may be due, in part, to early diagnosis
fused systemically and requires another trip and aggressive management. However, this
to the angiography suite for catheter check might also be a selection bias, because we
and replacement.38 are catching more patients with milder and
more benign disease due to the diagnosis
of the condition using CT scan instead of
Supportive Measures the previous reports in which the diagnosis
All patients with acute and subacute MVT was mostly made at laparotomy, indicating
should be managed in an ICU setting. Ade- a more severe and advanced disease.22 Only
quate resuscitation is mandatory. They must a few studies have described the long-term
have complete bowel rest with NPO and outcome in patients who developed MVT.
nasogastric tube suction. Broad-spectrum One of the largest series included 60 patients
antibiotic coverage is not indicated in the who were followed for a median of 3.5 years.
absence of bowel perforation or peritonitis.1 The overall survival at 1 and 5 years was 82%
and 78%, respectively. These values were
86% and 82%, respectively, after excluding
Outcome patients with an underlying malignancy.46
Long-term survival depends primarily on
The true incidence of MVT is unknown, the cause of the thrombosis. If cancer is the
mainly because it is not diagnosed in many underlying cause, survival is short and deter-
patients because of the vague, mild, and mined by the nature of the cancer.1 The use
nonspecific manifestations. Because of that, of endovascular techniques in the manage-
the true outcomes of the condition are es- ment of acute MVT may contribute to the
sentially unknown, as well. Historically, the improving mortality rates; however, the use
mortality rate associated with acute MVT of those techniques remains limited, and
ranged between 20% and 50%.1 Although their true value remains to be evaluated
high, it was still better than other cases of when more experience is accumulated.
acute arterial mesenteric ischemia. Survival Mesenteric venous thrombosis is a dis-
depends on multiple factors, including age, ease with a high rate of recurrence, and re-
the presence or absence of coexisting comor- currences are most common within the first
bid conditions, and the timing of the diag- 30 days after presentation.47 Coumadin ther-
nosis and surgical intervention. Patients who apy significantly reduces the risk of recur-
rence from 20%25% to 13%15%. There intestine due to occlusion of the mesenteric
is no evidence that long-term Coumadin vessels. Ann Surg. 1895(21):9-23.
therapy is beneficial except in cases where 4. Cokkinis A. Mesenteric Vascular Occlusions.
there is a persistent hypercoagulable state London: Bailliere, Tindall and Cox; 1926. p
when lifelong therapy is indicated.19,38 1-93.
The natural history of chronic MVT is 5. Warren S, Eberhard TP. Mesenteric venous
not known, but most of the patients appear thrombosis. Surg Gynecol Obstet. 1935;141:
to be asymptomatic. The percentage of pa- 740-2.
tients with chronic MVT who develop late 6. Acosta S, Alhadad A, Svensson P, Ekberg O.
gastrointestinal bleeding or hypersplenism Epidemiology, risk and prognostic factors
has not been determined but is probably in mesenteric venous thrombosis. Br J Surg.
small. As MVT is recognized more frequent- 2008;95(10):1245-51.
ly on CT scans done for other disorders, our 7. Rhee RY, Gloviczki P. Mesenteric
understanding of the natural history of this venous thrombosis. Surg Clin North Am.
disorder should improve.21 1997;77(2):327-38.
MVT remains an underdiagnosed and 8. Martinelli I, Mannucci PM, De Stefano V,
understudied condition that we need to Taioli E, Rossi V, Crosti F, et al. Different
learn more about. Although there has been risks of thrombosis in four coagulation defects
significant improvement in the diagnosis associated with inherited thrombophilia:
and management of the disease in recent a study of 150 families. Blood.
years, this has not been translated into sig- 1998;92(7):2353-8.
nificant reduction in mortality and morbid- 9. Bagot CN Arya R. Virchow and his triad:
ity, which tells us that there is still a long way a question of attribution. Br J Haematol.
to go to achieve that goal. More is needed 2008;143(2):180-90.
to study the natural history of the disease. 10. Stamou KM, Toutouzas KG, Kekis PB, Nakos
Also, we need to evaluate the efficacy and S, Gafou A, Manouras A, et al. Prospective
refine the new modalities of therapy, espe- study of the incidence and risk factors of
cially the minimally invasive endovascular postsplenectomy thrombosis of the portal,
techniques, as well as developing new thera- mesenteric, and splenic veins. Arch Surg.
pies that can prevent or limit bowel injury, 2006;141(7):663-9.
thus reducing the need for open surgery with 11. Hatoum OA, Spinelli KS, Abu-Hajir M, Attila
bowel resection. T, Franco J, Otterson MF, et al. Mesenteric
venous thrombosis in inflammatory bowel
disease. J Clin Gastroenterol. 2005;39(1):27-31.
references 12. Harward TR, Green D, Bergan JJ, Rizzo RJ,
1. Kumar S, Sarr MG, Kamath PS. Mesenteric Yao JS. Mesenteric venous thrombosis. J Vasc
venous thrombosis. N Engl J Med. 2001; Surg. 1989;9(2):328-33.
345(23):1683-8. 13. Kumar S, Kamath PS. Acute superior
2. Schoots IG, Koffeman GI, Legemate DA, mesenteric venous thrombosis: one disease
Levi M, van Gulik TM. Systematic review or two? Am J Gastroenterol. 2003;98(6):
of survival after acute mesenteric ischaemia 1299-304.
according to disease aetiology. Br J Surg. 14. Provan D, OShaughnessy DF. Recent
2004;91(1):17-27. advances in haematology. BMJ. 1999;318
3. Elliot J. The operative relief of gangrene of (7189):991-4.
15. Margaglione M, Brancaccio V, Giuliani N, 25. Zimmerman BJ, Granger DN. Reperfusion
DAndrea G, Cappucci G, Iannaccone L, et injury. Surg Clin North Am. 1992;72(1):65-83.
al. Increased risk for venous thrombosis in 26. Warshaw AL, Jin GL, Ottinger LW.
carriers of the prothrombin GA20210 gene Recognition and clinical implications of
variant. Ann Intern Med. 1998;129(2):89-93. mesenteric and portal vein obstruction
16. Chamouard P, Pencreach E, Maloisel F, in chronic pancreatitis. Arch Surg
Grunebaum L, Ardizzone JF, Meyer A, et 1987;122(4):410-5.
al. Frequent factor II G20210A mutation 27. Mathews JE, White RR. Primary mesenteric
in idiopathic portal vein thrombosis. venous occlusive disease. Am J Surg 1971;
Gastroenterology. 1999;116(1):144-8. 122(5):579-83.
17. Elting LS, Escalante CP, Cooksley C, 28. Boley SJ, Kaleya RN, Brandt LJ. Mesenteric
Avritscher EB, Kurtin D, Hamblin L, et al. venous thrombosis. Surg Clin North Am.
Outcomes and cost of deep venous thrombosis 1992;72(1):183-201.
among patients with cancer. Arch Intern Med. 29. Lange H, Jackel R. Usefulness of plasma
2004;164(15):1653-61. lactate concentration in the diagnosis of acute
18. Alvi AR, Khan S, Niazi SK, Ghulam M, abdominal disease. Eur J Surg. 1994;160(6-
Bibi S. Acute mesenteric venous thrombosis: 7):381-4.
improved outcome with early diagnosis and 30. Jamieson WG, Marchuk S, Rowsom J,
prompt anticoagulation therapy. Int J Surg. Durand D. The early diagnosis of massive
2009;7(3):210-3. acute intestinal ischaemia. Br J Surg. 1982;69
19. Abdu RA, Zakhour BJ, Dallis DJ. Mesenteric suppl:S52-3.
venous thrombosis1911 to 1984. Surgery. 31. Fried MW, Murthy UK, Hassig SR, Woo J,
1987;101(4):383-8. Oates RP. Creatine kinase isoenzymes in the
20. Rosenblum JD, Boyle CM, Schwartz LB. diagnosis of intestinal infarction. Dig Dis Sci.
The mesenteric circulation. Anatomy 1991;36(11):1589-93.
and physiology. Surg Clin North Am. 32. Baker SR, Cho KC. The Abdominal Plain
1997;77(2):289-306. Film with Correlative Imaging. 2nd ed.
21. Boley SJK. Mesenteric ischemic disorders. Stamford (CT): Appleton & Lange; 1999. p
In: Zinner MJ, ed. Maingots Abdominal 315-29.
Operations, 10th ed. Upper Saddle River, NJ: 33. Dietrich CF, Jedrzejczyk M, Ignee A.
Appleton & Lange; 1997. Sonographic assessment of splanchnic arteries
22. Morasch MD, Ebaugh JL, Chiou AC, and the bowel wall. Eur J Radiol. 2007;64(2):
Matsumura JS, Pearce WH, Yao JS. 202-12.
Mesenteric venous thrombosis: a changing 34. Horton KM, Fishman EK. Computed
clinical entity. J Vasc Surg. 2001;34(4):680-4. tomography evaluation of intestinal ischemia.
23. Gore RM, Thakrar KH, Mehta UK, Berlin Semin Roentgenol. 2001;36(2):118-25.
J, Yaghmai V, Newmark GM. Imaging 35. Michaely HJ, Dietrich O, Nael K, Weckbach
in intestinal ischemic disorders. Clin S, Reiser MF, Schoenberg SO. MRA of
Gastroenterol Hepatol. 2008;6(8): 849-58. abdominal vessels: technical advances. Eur
24. Brunaud L, Antunes L, Collinet-Adler S, Radiol. 2006;16(8):1637-50.
Marchal F, Ayav A, Bresler L, et al. Acute 36. Kim ST, Nemcek AA, Vogelzang RL.
mesenteric venous thrombosis: case for Angiography and interventional radiology of
nonoperative management. J Vasc Surg. the hollow viscera. In: Gore RM, Levine MS,
2001;34(4):673-9. eds. Textbook of Gastrointestinal Radiology.
Practice
Management
223
18
225
pathway separately. For example, if a physi- nature of your complex work components.
cian accesses the right common femoral For example, when performing a right leg
vein with advancement to the inferior vena venogram and a left leg venogram from two
cava and then superior vena cava to the right separate venous access sites, both are report-
brachiocephalic vein vascular family with ed (36005 + 36005). However, a payor will
final selective manipulation to the right sub- rarely recognize both access sites with in-
clavian vein and right internal jugular vein, jection codes without a modifier to show 2
each of the 2 separate pathways would be access sites were performed. This is primar-
reported. The second-order right subclavian ily because the codes are relatively generic.
vein (36012) and separate pathway end posi- Code 36005 is a venous access for extremity
tion to the right internal jgular (36012) are contrast injection venogram. The code is not
reported. This represents 2 pathways and 2 specific to right or left leg access and injec-
end positions of a diagnostic catheter place- tion. Thus, a modifier 59 (distinct procedure
ment or interventional device. Again, report or separate access site with contrast injection
both work components. in this case) must be appended. The modifier
It is important to note that only inten- will only need appended to one of the codes
tional catheter positions are reported. that are duplicated (ie, 36005 + 36005-59).
Duplicate Codes
Phase 2: Diagnostic
When reporting the same procedure code Venogram Imaging
twice on a claim, you will invariably need a
modifier appended to one of your codes to There are 3 key components to diagnostic
ensure that a payor recognizes the duplicate venography imaging. First, contrast must be
injected. Second, images must be captured gram, and thus, can be used as an additional
and stored. Third and finally, the physician imaging component if the 3 components for
must document an interpretation of the vas- diagnostic venography imaging are met (ie,
cular vessels. contrast injection, image capture, physician
Guiding injections to ensure that an in- interpretation).
terventional device is in position prior to the
interventional technique is not considered
diagnostic imaging. Road-mapping imaging
Venous Imaging Performed at the
techniques are also not considered diagnos-
tic imaging. Postinterventional follow-up im- Same Time/Same Session as the
aging is considered part of the interventional Venous Intervention
technique and again not diagnostic imaging
because these imaging techniques tend not Guidelines have been established in the
to have revealed the initial diagnostic inter- past few years by payors to limit diagnos-
pretation of the endovascular disease. tic imaging reimbursements at the same
A physician must create an initial dis- time of the interventional procedure. Thus,
ease assessment interpretation of a contrast when a physician performs a diagnostic ve-
injection whether patent, totally occluded, nography at the same time as the interven-
and anything in between to qualify for re- tional procedure the physician must append
porting a code for diagnostic imaging. And a modifier to reveal that venous imaging
remember, a stored image will validate your was not done previously. The modifier es-
work components. tablished to reveal that the venogram was a
Venography reporting is not impacted diagnostic study that leads to the interven-
by the amount of contrast injected or the tion is 59 (distinct procedure). For example,
number of images stored. It is solely based if a right lower extremity unilateral veno-
on the disease assessment of the vessel. gram (75820) was performed and a venous
angioplasty followed at the same session,
the physician must append the modifier 59
Selective Venography to the venography (75820-59) to show that
There are some venography codes that re- the interpretation lead to the interventional
flect the narrative of selective venography in procedure.
the narrative of the procedure. For example, If the procedure is staged where the
selective bilateral renal venography reflects venography was performed at a separate
this narrative. When the narrative of the ve- session than the intervention, the repeated
nography code reflects the selective termi- venogram should not be reported a second
nology, a catheter position within the vessel time. However, if there is a change in the
is necessary to report the imaging procedure. clinical findings, it would be appropriate to
As in the selective bilateral renal venography report the venography with the modifier 59
(75833) procedure, 2 catheter positions are to establish it as a distinct imaging compo-
required within the renal veins (ie, right nent at the time of intervention.
renal vein and left renal vein) to accompany
the imaging code. It is noted that the IVC
gram (75825) is not included in the code
narrative for selective bilateral renal veno-
nous angioplasty. Each vessel zone treated and right subclavian as separate work zones
by intravascular stent is reported. It is impor- for venous stent deployment.
tant to note that a physician can only report
the initial vessel treated by stent deployment Modifier Adjustments
(37205) once during an interventional pro- As discussed earlier, when a code is dupli-
cedure. Each vessel after the initial vessel cated but reported correctly, most payors will
treated by stent deployment is considered an require a modifier appended (ie, modifier 59
additional vessel treated (37206). distinct procedure) to establish the separate
The venous vessels zones of treatment work zones. Our recommendation is offered
are established from venous bifurcation to because we find it most consistent with cod-
bifurcation. It is not reported for each size ing guidelines and payor acceptance across
(ie, diameter or length) of stent placed in the the country. It would be appropriate to con-
venous vessel. Nor does it reflect any differ- sult each of your payor guidelines for modi-
ence in the deployment methods (ie, balloon fier submission requirements.
expandable or self-expanding). It is also not It is important to note that each of the
reported for the number of stents placed in other phases of procedure capture are not in-
the same vessel zone. It is reported once for cluded in this interventional technique and
each vessel treated by stent deployment. One should be reported separately. Add the cath-
initial (37205) vessel treated and the remain- eter placement to the interventional zone as
ing vessels treated with stent deployment well as any diagnostic imaging performed to
past separate bifurcations are the additional assess disease prior to intervention in the ve-
vessels (37206). The additional vessel treated nous vessels.
(37206) can be reported multiple times but
only once within a vessel.
There is a supervision and interpreta- Venous Angioplasty Followed by
tion imaging component that accompanies Stent Deployment
the surgical work (75960). These 2 com-
panion codes will travel together for each When combining multiple interventional
vessel treated with stent deployment (37205 techniques like venous angioplasty followed
+ 75960 initial vessel, 37206 + 75960 addi- by stent deployment, documentation is nec-
tional vessel, 37206 + 75960 additional ves- essary to establish each interventional pro-
sel, etc). cedure as its own separate work component.
For example, if a physician places a stent For instance, if the balloon angioplasty yield-
in the right brachiocephalic vein (37205 + ed a suboptimal recoiled residual stenosis, a
75960) and a right subclavian vein (37206 + predilation and postdilation percentage as-
75960), each is reported separately because sessment of the disease would reveal the rea-
of the bifurcation from the brachiocephalic son for further intervention with the stent.
to the subclavian with the internal jugular. Also, if a balloon angioplasty resulted in a
In fact, if a venous stent was also performed vessel dissection, then the multiple inter-
in the superior vena cava (37206 + 75960), ventional techniques are established by the
it would be established as its own zone of result of the initial procedure. In each of the
treatment and added as a separate work com- balloon results (ie, residual stenosis, dissec-
ponent. Thus, a physician would report the tion) both interventions are reported. It is not
superior vena cava, right brachiocephalic, appropriate to modify and angioplasty from
reporting can be divided into the 3-phase physician must have a high-quality support
approach outlined in this document (cath- system (ie, certified medical coders) that
eter placement, diagnostic imaging, and ve- understands the guidelines for each payor.
nous intervention). The above is an attempt Without a high-quality support system, a
to guide a physician though the maze of complex venous procedure involving mul-
complex procedure reporting requirements, tiple work components may reflect payments
as well as offer suggestions for documenta- that do not reveal all the procedures per-
tion quality. In no way is it possible to list formed by the physician.
all scenarios for venous procedures. Each
235
idraparinux, 14, 75
G2 X IVC filter (CR Bard), 123, 123f, 124f, 125 iliac vein compression syndrome, 18283, 198. See
gastrointestinal bleeding, 210 also May-Thurner syndrome
Glidewire guidewire (Terumo), 99 iliocaval occlusive disease, 19495
Greenfield embolectomy device (Boston iliofemoral deep venous thrombosis, 18189
Scientific), 141, 141f about, 181
Greenfield IVC filter (Boston Scientific), 127, ACCP guidelines (2008), 5960, 182
128f, 129 catheter-directed thrombolysis, 183, 18586
Greenfield Titanium IVC filter (Boston Scientific), diagnosis, 106
123f early thrombus removal, rationale, 1034
Gnther Tulip IVC filter (Cook), 123f, 124f, 125, endovascular management, 18288
126 iliac vein compression, 18283, 198
indications for intervention, 1067
interventions and clinical outcomes, 183, 184t
Hamptons hump, 49, 49f medical therapy, 18182
hematemesis, 210 natural history vs. distal DVT, 104
hemodialysis, access-related venous stenosis in, pathophysiology, 1045
201 pharmacomechanical thrombectomy, 186,
hemostasis, 22f, 2324 187f, 188