Professional Documents
Culture Documents
David F Keren
Hodder Arnold
A MEMBER OF THE HODDER HEADLINE GROUP
First published in Great Britain in 2003 by
Hodder Arnold, a member of the Hodder Headline Group,
338 Euston Road, London NW1 3BH
http://www.hoddereducation.com
Whilst the advice and information in this book are believed to be true and
accurate at the date of going to press, neither the author[s] nor the publisher
can accept any legal responsibility or liability for any errors or omissions
that may be made. In particular (but without limiting the generality of the
preceding disclaimer) every effort has been made to check drug dosages;
however it is still possible that errors have been missed. Furthermore,
dosage schedules are constantly being revised and new side-effects
recognized. For these reasons the reader is strongly urged to consult the
drug companies printed instructions before administering any of the drugs
recommended in this book.
2 3 4 5 6 7 8 9 10
What do you think about this book? Or any other Hodder Arnold title?
Please send your comments to www.hoddereducation.com
To my wonderful family
This page intentionally left blank
Contents
Preface ix
Acknowledgements xi
1 Protein structure and electrophoresis 1
2 Techniques for protein electrophoresis 15
3 Immunoxation, immunosubtraction, and immunoselection techniques 33
4 Proteins identied by serum protein electrophoresis 63
5 Approach to pattern interpretation in serum 109
6 Conditions associated with monoclonal gammopathies 145
7 Examination of urine for proteinuria 217
8 Approach to pattern interpretation in cerebrospinal uid 259
9 Laboratory strategies for diagnosing monoclonal gammopathies 285
10 Case studies for interpretation 304
Index 395
This page intentionally left blank
Preface
This text presents the use of protein electrophoresis resulted in better detection of bisalbuminemia by
of serum, urine, and cerebrospinal uid in clinical laboratories using capillary zone electrophoresis.
diagnosis. It is a revision of two previous books on That technique also has enhanced our ability to
this subject with several substantive and many detect genetic variants such as 1-antitrypsin
trivial changes. The title has been changed from inhibitor deciencies (PiZZ and PiSZ) as well as
High-Resolution Electrophoresis and Immuno- benign variants.
xation: Techniques and Interpretations to the In addition to technical advances in the general
present one in recognition of the fact that the infor- detection and quantication of specic proteins, the
mation in this book may be useful to individuals techniques available to identify the monoclonal
who interpret a wide variety of electrophoretic gels proteins have also changed signicantly. Immuno-
(high-resolution or not). subtraction using capillary zone electrophoresis was
There have been several signicant changes in the not even mentioned in the second edition. Now,
eld since the second edition of High-Resolution immunosubtraction by automated technology is
Electrophoresis and Immunoxation: Techniques used in many laboratories that employ capillary
and Interpretation was published in 1994 by zone electrophoresis. It is highly efcient, but suf-
Butterworth-Heinemann. That text was largely fers from a lack of sensitivity and exibility when
written in 1993, making it 10 years old at the time compared with immunoxation. This is discussed
of publication of this book. At the time the second in the present text. For laboratories using some of
edition was written there was no high-resolution the newer semi-automated gel-based techniques,
technique available that provided automated or semi-automated immunoxation is now available.
semi-automated methods for laboratories with a Beyond the instruments, there are new reagents
large clinical volume of testing. There are now available such as the Penta (pentavalent) reagent
several products presented, with examples. Some of that one company supplies as a potential screen for
them are gel-based, others use capillary zone monoclonal proteins. Immunoselection was also
electrophoresis. Some automated gel-based systems just touched on in the second edition. Now that g
have achieved an excellent degree of resolution that heavy chain disease is more readily recognized, a
allows efcient performance of high-quality tech- broader discussion of immunoselection is provided.
niques by even moderately sized institutions. At the Advances have also been made in reagents avail-
time of the second edition, capillary zone electro- able for measurement of monoclonal free light
phoresis itself was quite new with no methods avail- chains by nephelometry in both serum and urine.
able that had been approved for use in the clinical Recent publications indicate that measurement of
laboratory by the Food and Drug Administration. I free light chains in serum may be useful not only to
discussed it only briey. Now two such instruments follow patients with known monoclonal free light
are already in place in many laboratories and there chains (Bence Jones proteins), but also to assist in
have been several publications about the advan- the difcult diagnoses of amyloid (AL) and even
tages, artifacts and limitations of these techniques. non-secretory myeloma. This is discussed in
Improved resolution especially of protein bands has Chapters 6 and 7.
x Preface
In the second edition, high-resolution agarose gels from our clinicians. Our current sign-outs are pre-
were preferred for detecting oligoclonal bands in sented in tabular form and readers are encouraged
the cerebrospinal uid from patients whose differ- to use them if they wish.
ential diagnosis included multiple sclerosis. Now, A couple of concepts were removed from this
the recommendation is the use of isoelectric focus- edition. The technique of two-dimensional electro-
ing or an immunoxation technique to identify the phoresis, while useful in research has not caught on
bands as immunoglobulins. One company cur- in the clinical laboratory and is not discussed in
rently offers a semi-automated immunoxation this volume. Further, the indirect immunouores-
method for detecting oligoclonal bands (O-bands) cence techniques reviewed in the previous texts has
in cerebrospinal uid on unconcentrated uid. been replaced by immunohistochemical and ow
There have also been improvements in the detec- cytometry studies.
tion of leakage of cerebrospinal uid into nasal and Finally, in this text I recommend the use of the
aural uids; these are discussed in Chapter 8. term monoclonal free light chain (MFLC) to
Beyond technical issues, there have been improve- replace the term Bence Jones proteins. It is both a
ments in our knowledge of proper utilization and practical and a technical improvement. Some indi-
interpretation of protein electrophoresis when viduals confuse the presence of an intact
searching for the presence of a monoclonal immunoglobulin monoclonal gammopathy in the
gammopathy. In 1998, the College of American urine with a Bence Jones protein. It is not. Bence
Pathologists Conference XXXII convened a panel Jones protein is a monoclonal free light chain and
of experts to provide recommendations for the has much greater signicance to the patients diag-
clinical and laboratory evaluation of patients sus- nosis and prognosis than an intact immunoglobulin
pected of having a monoclonal gammopathy. monoclonal gammopathy in the urine. The term
These guidelines were published in 1999 and MFLC says exactly what we found (i.e. a mono-
provide an important framework that emphasizes clonal free light chain). So it is technically correct
the partnership of clinicians with laboratory and removes any possible ambiguity. This has the
workers. advantage that individuals will no longer use a
Urine evaluation has improved with the newer hyphen that Dr Henry Bence Jones never used on
technologies and has been complicated by some any of the papers or books that he wrote.
current procedures. For example, individuals I hope you enjoy this book, or at least nd it
receiving a pancreas transplant may have the useful. Your comments and questions help to
exocrine pancreas drainage empty into the urinary provide a focus for making this text more relevant
bladder. As discussed in Chapter 7, this results in g- to the use of electrophoresis in clinical diagnosis.
migrating bands that may be mistaken for Please feel free to contact me via e-mail:
monoclonal gammopathies. They are exocrine kerend@wardelab.com.
pancreas secretions.
The method I suggested in the rst edition and David F. Keren, MD
expanded in the second for tailored reporting with Ann Arbor, Michigan, USA
detailed sign-outs has been improved by feedback January 12, 2003
Acknowledgements
I wish to thank my family and especially my wife Jeffrey S. Warren provided helpful suggestions in
for their great patience with me as I worked reviewing selected portions of this book. Their
through this task. input kept many outrageous statements out of the
I am deeply grateful to the many people who nal copy, and corrected several errors of omis-
helped in the development of the materials for this sion.
and previous editions of this book. They allowed Special thanks are due to Ron Gulbranson and
me to use their clinical material or illustrations to Debbie Hedstrom who prepared many of the elec-
give the reader a rst-hand view of the potential trophoretic gels, electropherograms, and special
uses and limitations of electrophoretic techniques studies that are in the present book. They have had
in the clinical laboratory. Figures, Tables or spe- endless patience with me while I cluttered their
cic cases were kindly provided by Dr R. S. work area with capillary zone electropherograms
Abraham, Dr Francesco Aguzzi, Dr Arranz-Pea, and gels containing interesting cases.
Dr Gary Assarian, Cynthia R. Blessum, Dr Xavier Finally, I also thank the many individuals who
Bossuyt, Dr Arthur Bradwell, Dr. Stephen O. have taken the time to write, telephone or speak to
Brennan, Chrissie Dyson, Dr Beverly Handy, me personally about some aspect of my work.
Margaret A. Jenkins, Carl R. Jolliff, Dr Jerry They have framed many questions about concepts
Katzmann, Dr Robert H. Kelly, Dr Joseph M. or omissions from previous work that I hope to
Lombardo, Dr A. C. Parekh, Dr Jeffrey Pearson, clarify in this text.
Dr Arthur J. Sloman, Dr Lu Song, Dr Tsieh Sun, Dr
E. J. Thompson, Dr Adrian O. Vladutiu and David F. Keren, MD
Dorothy Wilkins. Ann Arbor, Michigan, USA
Drs John M. Averyt, John L. Carey, III, and January 12, 2003
This page intentionally left blank
1
Protein structure and
electrophoresis
The term electrophoresis refers to the migration These are the D- and L-forms (Fig. 1.2). In proteins,
of charged particles in an electrical eld. The the L-form is almost always present.1
success of electrophoresis in separating serum,
L-Serine D-Serine
urine and cerebrospinal uid proteins into clini-
COOH COOH
cally useful fractions results from the heterogeneity | |
of the charges of these molecules. It is useful, there- H2N C H H C NH2
| |
fore, to review briey the structural features that CH2 CH2
result in the observed migration. | |
H2 COH H2 COH
not migrate in an electrical eld. Of course, the most hydrophobic amino acids exist as zwitterions
charge on each amino acid depends on both the R (Fig. 1.3). However, when large numbers of
group and the pH of the solution in which it is hydrophobic amino acids occur in other molecules,
dissolved. such as occasional products of malignant plasma
By altering the pH of an aqueous solution, the cell clones, they may exhibit properties that cause
charge on an amino acid can be changed. In acidic solubility problems. Hydrophobic amino acid
solution, the amino acid accepts a proton on its content is not the only factor that relates to solu-
carboxylate group, resulting in a net positive bility problems. For example, the complex cold-
charge on the molecule (Fig. 1.3). The positively precipitating property of cryoglobulins (see
charged cation thus formed will migrate toward Chapter 6) is not thought to be related to an excess
the negative pole (cathode) in an electrical eld. of hydrophobic amino acids.2
Conversely, in basic solution the ammonium group
gives up a proton, leaving the amino acid with a Aliphatic Amino Acids
net negative charge (Fig. 1.3). This negatively
Glycine Alanine Valine Leucine
charged anion migrates toward the positive pole | | | |
(anode) in an electrical eld. The pH of the solu- H CH3 CH CH2
|
tion and the nature of the R group (Fig. 1.4) have H3C CH3 CH
an important effect on the migration of individual
H3C CH3
amino acids. For example, both aspartic acid and
glutamic acid at pH 7.0 or greater are always non- Isoleucine Serine Threonine
protonated (negatively charged) and consequently | | |
HCCH3 CH2 H C OH
are referred to as aspartate and glutamate. At | | |
neutral pH, arginine, histidine and lysine are all CH2 H2COH CH3
|
protonated (positively charged). These differences CH3
in charge are important in determining the migra- (a)
tion of proteins during electrophoresis. Further, in
Aromatic Amino Acids
genetic variants such as a1-antitrypsin deciency,
the substitution of a charged amino acid for a Phenylalanine Tyrosine Tryptophan
neutral amino acid, or the converse, will alter the CH2 CH2 CH2
migration of the variant form that allows us to
detect the abnormality. NH
The solubility of proteins also is related to their
amino acid composition. For example, tyrosine, OH
threonine and serine have hydroxyl moieties in
their R group, and readily form hydrogen bonds
with water. Similarly, asparagine and glutamine
Sulfur-Containing Amnio Acids
have amide R groups that are hydrophilic. In con-
trast, several amino acids are hydrophobic. Large Cysteine Methionine
Basic Amino Acids the amino group of the next. During the process, a
Lysine Histidine Arginine water molecule is produced and the covalent
peptide bond is formed (Fig. 1.5). The resulting
CH2 CH2 CH2 molecule has the charge characteristics inherent in
NH the R groups, together with the carboxyl and
CH2 CH2
CH
amino terminal groups. When a group of amino
CH2 N+ CH2
H acids is linked through peptide bonds in a linear
CH2 NH array, it is referred to as a polypeptide. All
polypeptides have the same constituents on each
NH3+ C=NH2+ end, a C-terminal (with the free carboxyl group)
NH2 and the N-terminal (with the free amino group)
(Fig. 1.5). This linear sequence of amino acids is
termed the primary conformational structure of
Acidic Amnio Acids protein molecules.
Aspartate Glutamate When relatively few amino acids are included in a
polypeptide, they are called oligopeptides. There
CH2 CH2 are dramatic differences in the sizes of proteins
COO CH2
from tiny hormone oligopeptides like vasopressin
(nine amino acids) to massive molecules such as
COO immunoglobulin M (IgM), which contain 10
(c) polypeptide chains folded in a regular array con-
Polar Amino Acids taining over 6000 amino acids.
Aspargine Glutamine
SECONDARY STRUCTURE
CH2 CH2
The primary structure folds into a regular sec-
C CH2 ondary structure along one dimension in a simple
+
H3N O structure, such as an a-helix, random coils or b-
C
+ sheet. The secondary structure is held together
H3N O
mainly by hydrogen bonds that form between
peptide bonds.1
Immuno Acids
COOH
+
CH Proline R H
H OH H O
HN CH2
N C C N C C
H2C CH2 H O H OH
H R
(d)
Figure 1.4 (cd) R-Group structure of amino acids.
R H
H O
H
Peptide bonds and polypeptides N C C N C C + H2O
H OH
H O R
PRIMARY STRUCTURE
Figure 1.5 During peptide bond formation, a molecule of water
Amino acids may be linked by peptide bonds: the is given off. The resulting molecule has an N-terminal end with an
linkage of the carboxyl group of one amino acid to amino group, and a C-terminal end with a carboxyl group.
4 Protein structure and electrophoresis
When a protein is dissolved in a solution that is buffer. The buffer was placed in contact with elec-
acidic relative to that proteins pI, it will gain trodes, and the sensitive optical band method was
protons and migrate toward the cathode. In solu- used to monitor the progress of protein fractions
tions that are basic relative to that proteins pI, the during electrophoresis.
protein will donate protons and migrate toward When the current was turned off at the end of the
the anode. Although there are charge effects of run, a mixing of the boundaries occurred and only
sulfhydryl groups for proteins in the serum, urine, the fractions at the extreme cathodal and anodal
and cerebrospinal uid, the amount of charge due ends of the tube could be collected in a relatively
to free sulfhydryl groups is negligible with regard puried form. Therefore, originally, this moving
to electrophoretic migration. boundary electrophoresis was used to test the
success of protein purication that had been per-
formed by other means. Although crude by todays
standards, such early techniques were sufciently
ELECTROPHORETIC TECHNIQUES sensitive to allow Michaelis to determine isoelectric
IN CLINICAL LABORATORIES points of the marginally puried enzymes he had
for study.15 This moving boundary technique was
Moving boundary electrophoresis also instrumental in the original denition of the
major fractions of human serum proteins as
Analysis of proteins for clinical purposes began in albumin, al-, a2-, b-, and g-globulin.
the middle of the nineteenth century. Schmidt used
the term globulin in 1862 to describe proteins that
were insoluble in water.12 By the early twentieth Zone electrophoresis
century, serum proteins were broadly divided into
albumin and globulins, depending on the precipita- Zone electrophoresis involved the use of a solid
tion of globules after the addition of sodium support medium with staining by a protein dye to
sulfate to serum proteins. This process would visualize the major protein bands. One of the key
result in formation of a white residue, albumin practical problems with moving boundary electro-
(alba Latin for white), after the salt was removed phoresis was its inability to achieve a complete
by dialysis and the water was evaporated.13 The separation of electrophoretically adjacent major
albumin to globulin ratio was an early, crude index protein fractions. Also, the refractive index that
for evaluating liver pathology. was used to quantify the proteins in moving
Studies of the electrophoretic mobility of proteins boundary electrophoresis was limited in its dis-
were rst carried out by Arne Tiselius in the 1930s crimination of subtle differences. The development
employing a liquid medium.14 For these studies, of zone electrophoresis made it possible to over-
Tiselius devised a U-shaped electrophoretic cell come these difculties by providing a stable
and employed a Schlieren band optical system to support medium in which proteins could migrate,
detect the degree of refraction of light by proteins be stained and quantied. Zone electrophoresis,
as they moved through the tube. He found that then, offered the important feature of stabilizing
there was a difference in the refractive index of the migration of the proteins that moving bound-
light at boundary interfaces between major protein ary electrophoresis could not achieve.
fractions as they moved by the light detection The rst major supporting medium for electro-
device. Thus, this technique was called moving phoresis was lter paper. Studies using this support
boundary electrophoresis. For his technique, medium were begun as early as 1937. However, it
Tiselius dissolved a specic volume of a protein was not until the early 1950s that these techniques
mixture in buffer, and carefully layered the solu- were simplied and rigorously dened for practical
tion on the electrophoresis tube below the same use in clinical laboratories.16
6 Protein structure and electrophoresis
The use of lter paper as a support medium intro- toward the cathode (Fig. 1.7). This is because of
duced new variables into electrophoresis. As with movement of the cationic ions in the buffer
the moving boundary technique, the migration of toward the cathode. Depending on the amount of
individual proteins depended on the pI of the negative charge of the support medium, there will
molecule, the pH of the buffer, electrolyte concen- be an equal, but opposite (positive) charge of the
tration of the buffer, and amount of current buffer in the adjacent support medium. This will
applied. However, the texture of the lter paper pull the molecules that have a weaker negative
was found to be another important factor because charge, owing to their lower pI, toward the cath-
it offered substantially more resistance to the ode (Fig. 1.8). These molecules are not moving
movement of the proteins than was the case in the toward the cathode because they have a positive
free-moving boundary system. charge under the conditions of the assay; rather,
Early in the development of paper electropho- they are just weak swimmers (weak negative
resis, Kunkel and Tiselius observed that the texture charge) caught in a futile attempt to swim against
of paper products differed, and that this resulted in the ow of a strong river.
different migration of proteins depending on the There were some major problems with paper
brand and lot of paper used.17 Although the actual electrophoresis that limited its application. Paper
distance that human serum albumin would migrate electrophoresis was slow, requiring several hours
may be greater on one brand of paper than on (often overnight) in order to achieve adequate
another brand, the relationship between albumin separation of major protein fractions. Further-
and the subsequent major fractions was relatively more, it was opaque (frustrating early densito-
constant, and could be used to create a correction metric scanners), gave poor resolution, and had
factor specic for that preparation of lter paper. It signicant problems with non-specic protein
was also observed that the type of paper affected absorption.19 To quantify proteins from individual
the migration of smaller molecules less than larger
molecules.18
Origin
The use of paper as a support medium also intro-
duced the effect known as electroosmosis or end- Anode Cathode
bands, they would be cut out, eluted and subjected trophotometric measurement (Table 1.1).23
to a protein assay. Therefore, it is not surprising Unfortunately, the relatively poor resolution of
that a search was conducted for better support most commercially available cellulose acetate
media for protein electrophoresis. membranes limited the sensitivity of the technique.
Cellulose acetate and agarose became popular Subtle abnormalities such as heterozygotes for a1-
stabilizing media for the clinical laboratories in the antitrypsin deciency and small monoclonal gam-
1960s and 1970s. With these media, electrophore- mopathies (especially those in the a2- or b-regions)
sis could be performed in less than an hour, and the were often undetectable. The insensitivity of some
clarity of the media facilitated densitometric scan- earlier low-resolution methods was demonstrated
ning to estimate the protein concentration of the by a College of American Pathologists Survey
major fractions. report.24 It disclosed that many systems had a lower
For several years, cellulose acetate electrophore- detection rate of a subtle monoclonal gammopathy
sis was the most popular method for performing that was picked up by most participants using elec-
routine serum protein electrophoresis.20 Cellulose trophoretic systems with better resolution (Table
acetate electrophoresis has advantages over paper 1.2).
electrophoresis: only minimal adsorption of serum Agar is a polysaccharide product that is produced
proteins occurs upon application, and a sharper commercially by boiling red algae, ltering out the
separation of the major serum protein bands is larger impurities, and removing the water-soluble
obtained much more quickly than by paper elec- impurities by freeze-thawing.2527 After precipita-
trophoresis. However, the resolution on tradi- tion in ethanol, the mixture consists mainly of 14
tional cellulose acetate systems is inferior to that linked 3,6-anhydro-a-L-galactose and 13 linked
obtained with most agarose gel electrophoresis b-n-galactose.28 The nal agar gel is a chemically
systems (see below).21 Cellulose acetate has been complex structure. For practical purposes, it con-
popular in the clinical laboratory because of its tains varying quantities of agarose and
simplicity, reproducibility, reliable quantication agaropectin. Agaropectin has a relatively high
of protein fractions by densitometry, and rela- sulfate, pyruvate, and glucuronate content, which
tively low cost.22 imparts a strong negative charge to the gel and
The successful replacement of paper electro- results in considerable endosmotic ow; pure
phoresis by cellulose acetate electrophoresis was agarose has few anionic groups.29
facilitated by the demonstration that accurate Most commercial preparations of agar gels used
estimates could be made of each major protein for electrophoresis today contain relatively pure
fraction by densitometric scanning. In 1964, Briere preparations of agarose. This minimizes non-
and Mull demonstrated that densitometric scan- specic adsorption of some proteins (such as b-
ning of serum proteins separated by cellulose lipoprotein and thyroglobulin) to the agar and the
acetate electrophoresis gave the same measure of amount of endosmotic ow.30,31 Although puried
the major protein fractions as did elution and spec- agarose preparations substantially reduce the
Table 1.1 Comparison of densitometric scans on cellulose acetate with eluted protein concentrationa
Densitometry 4.5 0.36 0.27 0.08 0.62 0.10 0.64 0.12 0.95 0.27
Elution 4.8 0.34 0.22 0.06 0.52 0.09 0.59 0.13 1.01 0.25
a
Data from Briere and Mull.23
8 Protein structure and electrophoresis
a
Data expresses the number of laboratories using the indicated technique. From CAP survey 1991.24
endosmotic effect, some manufacturers use sub- For CZE, a small volume of sample (15 l) is
stantial quantities of agaropectin to promote aspirated into a thin fused silica capillary 2550 m
endosmosis. Some endosmotic ow is desirable in diameter (Fig. 1.9). The strong negative charge
for electrophoresis of serum proteins because it on the interior of the capillary, together with the
pulls the g-globulins cathodally. With these narrow lining, provides a large net negative surface
systems, most serum monoclonal gammopathies area. Under conditions of electrophoresis, this sets
migrate cathodally as do oligoclonal bands seen up a strong endosmotic ow of cations toward the
in cerebrospinal uid from patients with multiple cathode. In this system, the pull of this endosmotic
sclerosis. By moving these important bands away ow is stronger than the pull of the anode for the
(cathodal) from the origin, these systems mini-
mize the effect that minor distortions, often
present at the point of application, have on inter- Capillary Zone Electrophoresis
pretation of g-region abnormalities. Distortions at
the point of sample application can be especially (+)
Capillary
problematic when one is dealing with a cryoglo-
bulin that often precipitates at the origin (see Alb Detector
(+)
Chapter 6). a1
a2
b1
(+) b2
Capillary zone electrophoresis g
+ Anode Cathode
Sample
In the last decade, capillary zone electrophoresis
Positive buffer ions (+) flow to cathode
(CZE) has been developed for use in clinical labo-
ratories.3242 This technique is a liquid-based system Figure 1.9 Capillary zone electrophoresis, schematic. The sample
is aspirated into the anodal end of the 2550 mm diameter fused
that bears some similarities to the early Tiselius
silica capillary. The negative charge () on the interior of the
system in that no permanent gel is produced in the capillary sets up a strong endosmotic ow of cations (+) toward
process, although available systems may create the cathode. An ultraviolet detector evaluates absorbance at
virtual gel images (see below).42 200215 nm by the fractions indicated.
Early clinical applications of electrophoresis 9
anionic proteins being evaluated. The proteins then Capillary Zone Electrophoresis
migrate toward the cathode, but are variously
Albumin
impeded in their migration, based upon the nega-
tive charge of the proteins. Thus, the electrical eld
separates the proteins by charge. An ultraviolet
detector that evaluates the absorbance at
200215 nm (in various systems) determines the
a1-region
a2-region
b-region
Transthyretin
g-region
protein concentration (Fig. 1.10). Since peptide
bonds absorb at 214 nm, these systems provide
quantitative measurements of the various proteins
that are not inuenced by the presence of carbohy-
drate groups. However, this method of measure-
ment suffers from the disadvantage that other Figure 1.10 Capillary zone electropherogram of serum
substances that absorb at this wavelength will pro- performed on a Beckman Paragon CZE 2000.
a
Decreased beta in massive liver necrosis.
10 Protein structure and electrophoresis
A small decrease in serum albumin was quickly were occasionally too insensitive to detect the
recognized as a relatively nonspecic occurrence abnormal serum protein in some of these patients.48
found in a variety of conditions that cause meta- Using these early techniques, some studies reported
bolic stress and as a feature of the acute-phase as many as 22 per cent of patients with multiple
reaction pattern (discussed later). However, the myeloma had no signicant abnormality in the
level of serum albumin in patients with liver serum.49 Some of their patients had light chain
disease gave clinically useful information because it disease with monoclonal free light chains (MFLC,
was signicantly correlated with the amount of formerly termed Bence Jones proteins) in their
tissue damage.46 Further, it was noted that patients serum and, more frequently, in their urine. Since
with severe liver disease had a broad elevation of g- immunoglobulin light chains are relatively small
globulin,47 although the immunological signi- molecules (25 kDa as monomers to 50 kDa as
cance of that observation would not be understood dimers), they would pass into the urine and only a
for several years. minimal, or no monoclonal restriction was seen
Detection of monoclonal gammopathies in serum with the older zone serum protein electrophoresis
and urine is one of the most important uses of clin- techniques. Many of their patients might have been
ical protein electrophoresis. Many different types expected to have decreased g-globulins; this
of electrophoretic patterns can result from prod- decrease is subtle, probably below the ability of the
ucts of the neoplastic B-lymphocyte and plasma early methods to discern. The MFLC were detected
cell proliferation that occur in chronic lymphocytic best by electrophoresis of urine (see Chapter 7).50
leukemia and multiple myeloma, respectively. A major step forward in understanding the basic
Most frequently, patients with multiple myeloma immunology of the human immune system
have markedly elevated g-globulin regions with a involved the early application of serum protein
restriction in the migration (Fig. 1.11), although electrophoresis by Colonel Ogden Bruton at the
abnormalities may be found anywhere from the al- Walter Reed Army Medical Center.51 One of his
to the g-globulin region. patients was a boy with a history of recurrent pyo-
The older, paper zone electrophoresis techniques genic infections. When protein electrophoresis was
performed on this childs serum, it was discovered
that the g-globulin region was absent (in reality it
was very low, but undetectable with the zone elec-
trophoretic techniques available in 1952). By using
g-globulin replacement therapy, Bruton was able to
successfully treat this individual. Because we
understand the inheritance pattern of this condi-
tion, today it is referred to as X-linked agamma-
globulinemia.52 Early workers also recognized the
existence of other forms of immunodeciency dis-
eases associated with low g-globulin levels on elec-
trophoresis of serum. When seen in infants, they
often represented transient hypogammaglobuline-
mia of infancy; in young adults, the common vari-
Figure 1.11 Three electrophoretic patterns are shown with the able immunodeciency syndrome was the most
anode at the left and the cathode at the right. The location of the frequent cause of isolated hypogammaglobuline-
anode can be assumed to be on the same side as the albumin band.
mia.53
Note the prominent area of restriction in the slow g-region of the
top pattern (arrow). This pattern is typical of patients with
Broad increases in the g-globulin region were
multiple myeloma. (Paragon SPE2 gel system stained with Paragon found to be associated with the immune response
Blue.) to infectious agents and occasionally were reported
Early clinical applications of electrophoresis 11
in autoimmune diseases. For instance, Coburn and many central nervous system disorders, it was
Moore54 found that patients with clinically active logical to use electrophoresis to examine cerebro-
systemic lupus erythematosus had elevated levels of spinal uid for any new clues that might aid
g-globulin. These ndings preceded by 5 years the clinicians. Analysis of concentrated cerebrospinal
demonstration by Hargraves et al.55 of the auto- uid from patients with multiple sclerosis and
immune phenomenon called the LE (lupus erythe- neurosyphilis showed markedly elevated g-globulin
matosus) cell, and represent one of the earliest content.56 Similar elevations in the g-globulin were
laboratory observations suggesting the complicity found in central nervous system infections.
of g-globulin in the pathogenesis of this disease. Despite these and many other observations, the
The availability of zone electrophoresis encour- clinical applications of zone electrophoresis were
aged many investigators to examine a wide variety limited to obvious extreme elevations or reductions
of uids and extracts. Not surprisingly, consider- of major protein components. Many diseases were
ing its ready availability and long history of use in associated with more subtle alterations of proteins
the diagnostic laboratory, urine was one of the rst that were beyond the limitations of the early zone
uids studied. Urine from patients with nephrotic electrophoretic methods. Better resolution of
syndrome contained considerable albumin and protein bands, simpler methods to quantify the
some globulins in amounts that gave an inverse fractions, and greater sensitivity were required for
correlation with the serum concentrations of these protein electrophoresis to aid in the diagnosis of
proteins.44 Monoclonal free light chains had long these conditions.
since been described by Henry Bence Jones, but With the earlier methods, described above, it was
with the advent of zone electrophoresis it became arguable whether agarose gels could provide better
apparent that they had greater heterogeneity than resolution of the major protein bands than cellu-
previously thought. Although MFLC were always lose acetate. The heterogeneity of agar prepara-
globulins, they migrated anywhere from the a- tions and the ready availability of pure,
through the g-region (Fig. 1.12). This nding raised commercially prepared cellulose acetate caused
important questions about the structure of these much wider usage of the latter. In addition, cellu-
molecules, which were previously assumed to be lose acetate allowed more rapid electrophoresis
homogeneous by virtue of their peculiar thermo- and could be dried, stained, and cleared more
precipitating characteristics (see Chapter 7).50 easily than agar.29
Because of the difculty involved in diagnosing Around 1970, however, reports appeared that
demonstrated advantages of the careful, high-reso-
lution agarose electrophoresis system described by
Wieme.31 One could achieve diagnostically useful
results on agarose with techniques and equipment
well within the capabilities of the clinical labora-
tory.57 The availability of more highly puried
A
agarose preparations, which minimized endos-
motic ow and offered good optical clarity and
unimpeded migration because of their porosity,
was credited with some of the improvement in res-
olution.58 By the mid-1970s, using the modica-
Figure 1.12 Two electrophoretic patterns of concentrated urine tions described in the next section, a few agarose
from patients with multiple myeloma. Both samples have relatively
and cellulose acetate methodologies had evolved
little albumin (A), but large restrictions in the slow g- (top sample)
and fast g- (bottom sample) regions. These are both monoclonal
that facilitated the consistent demonstration of up
free light chains (MFLC) which have signicantly different to 12 distinct protein fractions. It is certainly
migrations. (Paragon SPE2 gel stained with Paragon Blue.) debatable whether one needs to evaluate so many
12 Protein structure and electrophoresis
proteins for clinical purposes. Most of the proteins 11. Harrison HH, Miller KL, Abu-Alfa A, Podlasek
demonstrated by electrophoresis are readily avail- SJ. Immunoglobulin clonality analysis.
able for specic assay by nephelometry or other Resolution of ambiguities in immunoxation
immunoassay techniques. The better quality of the electrophoresis results by high-resolution, two-
more recent electrophoretic techniques has, dimensional electrophoretic analysis of
however, improved the ability to detect subtle paraprotein bands eluted from agarose gels. Am J
monoclonal proteins, as documented by the results Clin Pathol 1993;100:550560.
of an independent study of electrophoretic results 12. Joliff CR. Analysis of the plasma proteins. J Clin
by a College of American Pathologists survey Immunoassay 1992;15:151161.
sample (Table 1.2). Systems with the better overall 13. Harrison HH, Levitt MH. Serum protein
resolution had higher detection rates of the small electrophoresis: basic principles, interpretations,
M-protein than systems with poorer resolution. and practical considerations. Check Sample
(ASCP) 1987;7:116.
14. Tiselius A. A new apparatus for electrophoretic
REFERENCES analysis of colloidal mixtures. Trans Faraday Soc
1932;33:524531.
1. Jones AJS. Analysis of polypeptides and proteins. 15. Tiselius A. Introduction. In: Bier M, ed.
Adv Drug Del Rev 1993;10:2990. Electrophoresis, theory, methods and
2. Levo Y. Nature of cryoglobulinaemia. Lancet applications. New York: Academic Press, 1959.
1980;1:285287. 16. Durrum EL. A microelectrophoretic and
3. Johnson AM, Rohlfs EM, Silverman LM. microionophoretic technique. J Am Chem Soc
Chapter 20. Proteins. In: Burtis CA, Ashwood, 1950;72:29432948.
ER, eds. Tietz textbook of clinical chemistry. 17. Kunkel HG, Tiselius A. Electrophoresis of
Philadelphia: WB Saunders, 1999. proteins on lter paper. J Gen Physiol
4. Carrell RW, Lomas DA. Alpha1-antitrypsin 1951;35:39118.
deciency a model for conformational diseases. 18. McDonald HJ. Polyvinylpyrrolidine: the
N Engl J Med 2002;346:4553. electromigration characteristics of the blood
5. Zaret DL, Morrison N, Gulbranson R, Keren plasma expander. Circ Res 1953;1:396404.
DF. Immunoxation to quantify beta 2- 19. Johansson BG. Agarose gel electrophoresis.
transferrin in cerebrospinal uid to detect Scand J Clin Lab Invest Suppl 1972;124:719.
leakage of cerebrospinal uid from skull injury. 20. Kohn J. A cellulose acetate supporting medium
Clin Chem 1992;38:19081912. for zone electrophoresis. Clin Chim Acta
6. Hotchkiss A, Reno CJ, Leonard CK, et al. The 1957;2:297304.
inuence of carbohydrate structure on the 21. Cawley LP, Minard B, Penn GM. Electrophoresis
clearance of recombinant tissue-type plasminogen and immunochemical reactions in gels.
activator. Thromb Haemost 1988;60:255261. Techniques and interpretations. Chicago: ASCP
7. Furie B, Furie BC. Molecular basis of vitamin K- Press, 1978.
dependent gamma-carboxylation. Blood 22. Kohn J. Small-scale membrane lter
1990;75:17531762. electrophoresis and immunoelectrophoresis. Clin
8. Huttner WB. Tyrosine sulfation and the secretory Chim Acta 1958;3:450454.
pathway. Annu Rev Physiol 1988;50:363376. 23. Briere RO, Mull JD. Electrophoresis of serum
9. Eipper BA, Mains RE. Peptide alpha-amidation. protein with cellulose acetate. A method for
Annu Rev Physiol 1988;50:333344. quantitation. Am J Clin Pathol 1964;34:
10. James G, Olson EN. Fatty acylated proteins as 547551.
components of intracellular signaling pathways. 24. College of American Pathologists. Survey Report
Biochemistry 1990;29:26232634. EC-07. 1991.
References 13
25. Serwer P. Agarose gels: properties and use for 39. Katzmann JA, Clark R, Sanders E, Landers JP,
electrophoresis. Electrophoresis 1983;4:375382. Kyle RA. Prospective study of serum protein
26. Serwer P. Improvements in procedures for capillary zone electrophoresis and immunotyping
electrophoresis in dilute agarose gels. Anal of monoclonal proteins by immunosubtraction.
Biochem 1981;112:351356. Am J Clin Pathol 1998;110:503509.
27. Serwer P, Hayes SJ. Exclusion of spheres by 40. Keren DF. Capillary zone electrophoresis in the
agarose gels during agarose gel electrophoresis: evaluation of serum protein abnormalities. Am J
dependence on the spheres radius and the gels Clin Pathol 1998;110:248252.
concentration. Anal Biochem 1986;158:7278. 41. Bienvenu J, Graziani MS, Arpin F, et al.
28. Rees DA. Structure, conformation, and Multicenter evaluation of the Paragon CZE 2000
mechanism in the formation of polysaccharide capillary zone electrophoresis system for serum
gels and networks. Adv Carbohydr Chem protein electrophoresis and monoclonal
Biochem 1969;24:267332. component typing. Clin Chem 1998;44:
29. Nerenberg ST. Electrophoretic screening 599605.
procedures. Philadelphia: Lea and Febiger, 1973. 42. Jolliff CR, Blessum CR. Comparison of serum
30. Guo Y, Xinhui L, Yong F. The effects of protein electrophoresis by agarose gel and
electroendosmosis in agarose electrophoresis. capillary zone electrophoresis in a clinical setting.
Electrophoresis 1998;19:13111313. Electrophoresis 1997;18:17811784.
31. Wieme R. Agar gel electrophoresis. Amsterdam: 43. Bossuyt X, Mewis A, Blanckaert N. Interference
Elsevier, 1965. of radio-opaque agents in clinical capillary zone
32. Jenkins MA, OLeary TD, Guerin MD. electrophoresis. Clin Chem 1999;45:129131.
Identication and quantitation of human urine 44. Longsworth LG, MacInnes DA. Electrophoretic
proteins by capillary electrophoresis. J study of nephrotic sera and urine. J Exp Med
Chromatogr B Biomed Appl 1994;662:108112. 1940;71:7786.
33. Jenkins MA, Kulinskaya E, Martin HD, Guerin 45. Lenke SE, Berger HM. Abrupt improvement of
MD. Evaluation of serum protein separation by serum electrophoretic pattern in nephrosis after
capillary electrophoresis: prospective analysis of ACTH-induced diuresis. Proc Soc Exp Biol Med
1000 specimens. J Chromatogr B Biomed Appl 1951;78:366369.
1995;672:241251. 46. Wajchenberg BL, Hoxter G, Segal J, et al.
34. Jenkins MA, Guerin MD. Quantication of Electrophoretic patterns of the plasma proteins in
serum proteins using capillary electrophoresis. diffuse liver necrosis. Gastroenterology
Ann Clin Biochem 1995;32:493497. 1956;30:882893.
35. Jenkins MA, Guerin MD. Optimization of serum 47. Franklin M, Bean WB, Paul WD, Routh JI, de la
protein separation by capillary electrophoresis. Hueraga J, Popper H. Electrophoretic studies in
Clin Chem 1996;42:1886. liver disease. I. Comparison of serum and plasma
36. Jenkins MA, Guerin MD. Capillary electrophoretic patterns in liver disease, with
electrophoresis as a clinical tool. J Chromatogr B special reference to brinogen and gamma
Biomed Appl 1996;682:2334. globulin patterns. J Clin Invest
37. Jenkins MA. Clinical applications of capillary 1951;30:718728.
electrophoresis. Status at the new millennium. 48. Moore DH. Clinical and physiological
Mol Biotechnol 2000;15:201209. applications of electrophoresis. In: Bier M, ed.
38. Jenkins MA. Three methods of capillary Electrophoresis theory, methods and
electrophoresis compared with high-resolution applications. New York: Academic Press, 1959.
agarose gel electrophoresis for serum protein 49. Reiner M, Stern KG. Electrophoretic studies on
electrophoresis. J Chromatogr B Biomed Sci the protein distribution in the serum of multiple
Appl 1998;720:4958. myeloma patients. Acta Haematol 1953;9:1929.
14 Protein structure and electrophoresis
50. Moore DH, Kabat EA, Gutman AB. Bence Jones 55. Hargraves MM, Richmond H, Morton R.
proteinemia in multiple myeloma. J Clin Invest Presentation of two bone marrow elements. The
1943;22:6775. tart cell and the LE cell. Mayo Clin Proc
51. Bruton OC. Agammaglobulinemia. Pediatrics 1948;23:2528.
1952;9:722727. 56. Kabat EA, Moore DH, Landow H. An
52. Noordzij JG, de Bruin-Versteeg S, Comans-Bitter electrophoretic study of the protein components
WM, et al. Composition of precursor B-cell in cerebrospinal uid and their relationship to the
compartment in bone marrow from patients with serum proteins. J Clin Invest 1942;21:571577.
X-linked agammaglobulinemia compared with 57. Rosenfeld L. Serum protein electrophoresis. A
healthy children. Pediatr Res 2002;51:159168. comparison of the use of thin-layer agarose gel
53. Gitlin D. Low resistance to infection: relationship and cellulose acetate. Am J Clin Pathol.
to abnormalities in gamma globulins. Bull NY 1974;62:702706.
Acad Med 1955;31:359365. 58. Elevitch FR, Aronson SB, Feichtmeir TV,
54. Coburn AF, Moore DH. The plasma proteins in Enterline ML. Thin gel electrophoresis in
disseminated lupus erythematosus. Bull Johns agarose. Tech Bull Regist Med Technol
Hopkins Hosp 1943;73:196214. 1966;36:282287.
2
Techniques for protein
electrophoresis
This excessive heat production plays havoc with are attracted to the cathode during electrophoresis
good resolution of electrophoretic bands. One of and tend to retard the progress of albumin toward
the major effects of heat is to increase the thermal the anode. This accumulation of positive charges in
agitation and hence the diffusion of the protein the buffer around the negatively charged albumin
molecules. Diffusion broadens the width of a band, is known as the diffuse double layer. This is why it
thereby decreasing the resolution. Heat production is important to control evaporation with resultant
can also decrease the viscosity of the medium. concentration of ions in the buffer during elec-
Although this does permit a more rapid elec- trophoresis.1
trophoretic migration (m) of the proteins through Another factor limiting the effective separation of
the gel, it is more than counterbalanced by an even protein bands is adsorption of the molecules to the
greater increase in diffusion, with a resulting agar gel itself. Because of the negative charges pos-
decrease in resolution. Before closed systems were sessed by the relatively puried agarose solutions
common, the heat generated further complicated used today, pH < 5.0 is impractical. Below this pH,
resolution by causing enough evaporation to serum proteins would have a positive charge and
change ionic strength. would precipitate in the gel.
The ionic strength of the buffer is also an impor- Depending upon buffer strength, voltage, heat
tant factor in the resolution of individual protein dissipation, purity and thickness of the gel, avail-
bands. As the concentration of the salt ions in a able electrophoretic systems display from ve to as
buffer increases, the velocity of electrophoretic many as 12 protein bands, which encompass more
migration decreases for each protein being assayed. than 95 per cent of the total mass of serum pro-
There is no effect, however, on the relative migra- teins.3 A comparison of the resolutions generally
tion of serum proteins as a result of ionic strength. available is shown in Fig. 2.2. Better resolution is
The effect of ionic concentration on the migration an important part of detecting monoclonal gam-
of proteins in the electric eld is largely the result mopathies.4,5
of interaction of the buffer ions with the surface To improve the detection of monoclonal gam-
charges on the protein. mopathies, the Protein Commission of the Societa
Consider a buffer in which we increase the con- Italiana di Biochimica Clinica published guidelines
centration of NaCl. At the typical pH 8.6 of agar for criteria for performance of sensitive electro-
gel electrophoresis, human serum albumin has a
negative surface charge. The positive sodium ions
are attracted to the negative charges on albumin
Levels of
and diminish its effective net negative charge in the Transferrin C3
resolution
solution (Fig. 2.1). Further, positively charged
ions, now in immediate proximity to the albumin, High-resolution
Na+ Na+
Anode + Na+ Cathode
Na
+
Albumin
Na+
Na+
Na+ Na+
Na+
Low-resolution
Figure 2.1 At the pH of the typical serum protein
electrophoresis gel (8.6), albumin has a strong negative charge. The Figure 2.2 This composite gure illustrates the resolution of
positive ions in the buffer, in this case sodium, are attracted to the different samples run on different electrophoretic analyses.
negative charges on albumin and diminish its migration toward the Examining the b-region bands (transferrin and C3) helps to
anode. evaluate the resolution available by the different techniques.
Electrophoresis on agarose 17
phoresis procedures. Aguzzi et al.5 summarized kits usually have a uniform thin (about 1 mm)
these recommendations as follows: layer of agarose on an inert plastic support.
The specimen must be applied to the agarose
1. It should be possible to see the faint
surface in a very narrow band. In manual systems,
transthyretin (prealbumin) band in the serum
excess moisture is removed from the surface of the
of all healthy persons.
gel by blotting with lter paper. The blotting is
2. It should be possible to detect, if present, the
needed to help the proteins diffuse into the gel and
heterozygosity of a1-antitrypsin.
to prevent excessive lateral movement at the point
3. It should be possible to recognize the two
of application. An inadequately blotted gel will
main components of the a2-zone (haptoglobin
have distortions in all the bands (Fig. 2.3). A
and a2-macroglobulin).
plastic template with uniform narrow slits for
4. The two main components of the b-zone
sample application is rmly layered onto the
(transferrin and C3) should be clearly resolved
blotted gel. The template should be applied evenly
as two distinct bands.
to the surface of the agarose so that no air pockets
5. In the g-zone, it should be possible to
are present; these may distort the application of the
recognize the presence of small monoclonal
sample. In most systems, 35 l of sample are
components (< 1 g/l, i.e. < 100 mg/dl) and
placed over each slit and allowed to diffuse into the
possibly the oligoclonal pattern.
gel for 57 min. Consistency and attention to
6. Zonal terminology and densitometric
detail in sample application are extremely impor-
reporting should not be used, and results
tant in manual techniques because the nal
should be expressed in terms of qualitative
bandwidth and conguration are determined by
and semiquantitative variations of specic
the initial application. For example, in Fig. 2.4, a
proteins.
small drop of serum fell on the top pattern prior to
Similarly, recently published conclusions from a electrophoresis, affecting the pattern.
conference on guidelines for clinical and labora-
tory evaluation patients with monoclonal
gammopathies recommends the use of systems
that provide resolution sufcient to separate b1
(transferrin) and b2 (C3) proteins to improve
detection of monoclonal gammopathies.4
Measurements of proteins present in concentra-
tions smaller than 10 mg/dl require immunoassay
techniques.6
Figure 2.3 Serum protein electrophoretic strip of two sera
stained from an inadequately blotted gel. Note the distortion
(irregularities) in all bands of these samples. The occurrence of the
ELECTROPHORESIS ON AGAROSE distortion in all bands indicates that this is an artifact due to an
application problem (in this case insufcient blotting). (Paragon
The basic principles of electrophoresis apply to SPE2 system stained with Paragon Violet.)
both the manual and the automated systems (see
below). The method of Wieme, as modied by
Johansson, is commonly used with agarose.7 A 1 Occasional gel preparations have distortions
per cent concentration of agarose is used in arising from their initial preparation, or from
0.075 M, pH 8.6 barbital buffer containing 2 mM problems with storage. A distortion in the gel may
calcium lactate. The calcium ions are especially give a pattern like that in Fig. 2.5. The top sample
useful for improving the resolution in the b-region. shows a normal migration. However, the albumin
Commercially available agarose electrophoresis bands of the next three samples show a distinctive
18 Techniques for protein electrophoresis
bowing. The origin artifact (O) and C3 in the b-2 urine with the electrophoretic pattern to be sure
region do not have this distortion. This indicates the sample applied. This is not a problem in
the problem was not in the application, rather in serum both because the lack of visible bands is an
the gel itself toward the anode. obvious problem for any serum, but even in con-
On the semi-automated, gel-based systems, centrated normal urine, albumin may not be seen
application devices are used along with auto- (see Chapter 7).
mated washing to streamline the technical After sample application is complete, some mech-
process. The Helena Rep Unit (Helena anism for cooling the gel is used. With some
Laboratories, Beaumont, MI, USA) has been systems, the gels are cooled by convection. In other
available for several years and the Sebia Hydrasis systems, gels are oriented such that their plastic
(Sebia, Issy-les-Moulineaux, France) has become backing is in direct contact with a cooling block
available more recently. The Sebia Hydragel b1b2 (typically kept at 4C prior to electrophoresis). The
15/30 method provides serum and urine gel cooling block must be properly prepared and
results with crisp separation of the b1b2 region as stored; if it is not at the proper temperature, the
mentioned above. However, when processing heat generated from the voltage applied to these
urine samples the technologist needs to be careful samples will produce the effects described above
of samples that contain solid matter (cells, crys- and poor resolution of bands will result. A Peltier
tals, etc.). This may interfere with the wicking of cooling device is used to control the heat in some of
the sample onto the gel. If the sample does not the automated systems.
wick properly, no sample is applied to the gel. The amount of buffer in the reservoir is another
The electrophoretic result on a urine containing a important factor. If there is too much buffer in the
large amount of protein will then appear as reservoir, the migrating g-region, upon reaching
though no protein were present. Because of this, I the buffer, will form an artifactual slow g-band
recommend comparing the total protein on the (Fig. 2.6). Usually, this is obvious because all of the
Electrophoresis on cellulosic media 19
g-regions have bands. However, when one sample our laboratory, we just play recordings of my old
has a relatively large amount of polyclonal, slow lectures!).
migrating g-globulin (as may occur in patients with For examination of serum and urine proteins, we
chronic active hepatitis), a slow g restriction may prefer staining the gel with Amido black. Some
be seen that could be mistaken for a monoclonal commercial suppliers have their own versions of
gammopathy. similar dye. Whereas both Coomassie Brilliant
Most agarose systems run with an electrical eld Blue and Amido Black give similar patterns, the
of about 20 V/cm (a setting of 200 V for each 10- Amido Black has less background between major
cm length of agarose) and a current of about bands, which makes interpretation of serum and
100120 mA. Under these conditions, the typical urine electrophoretic patterns more straightfor-
run lasts 3050 min. When electrophoresis has ward.7,9 Part of the difference is that Coomassie
been completed, the proteins are xed with an Brilliant Blue is more sensitive than Amido Black
acid xative. Some systems still use picric acid for and stains small protein molecules at these sites.
this step but most new methods do not. (Note For cerebrospinal uid, where sensitivity can be a
that picric acid can become explosive when stored problem even after concentrating a sample 80-fold,
for long periods of time.8 Good laboratory tech- Coomassie Brilliant Blue is preferred. Coomassie
nique, including checking the bottle for the expiry Brilliant Blue-stained gels give better results for
of the reagent and for crystallization around black and white photography.
bottle caps, is especially important with this
reagent.) After xing the proteins, the gel is dried
with a gentle stream of hot air for 510 min (in ELECTROPHORESIS ON
CELLULOSIC MEDIA
Cellulose acetate has been available as a support-
ing media for protein electrophoresis since the late
1950s.10 The clarity of the background made this a
considerable improvement over lter paper.
Cellulose acetate has the advantage of uniform
porosity. As with agarose, a wide variety of
systems are available with cellulosic media. The
membrane is obtained by dissolving, in a volatile
organic solvent, the product of mixing carbonic
anhydride with cellulose.11 The resulting mem-
branes provide consistent ve-band resolution.
However, by using gelication to prevent the mem-
branes from drying, gelled cellulose acetate is
created, with improved resolution of protein
bands.11 Preparations of gelled cellulose acetate
Figure 2.6 High-resolution electrophoretic strip with four (Cellogel; Chemetron, Milan, Italy) can separate
samples that all appear to have a small, slow g-restriction (arrow in serum proteins into the same fractions seen with
restriction of top lane only). If such a band were present in only the high-resolution agarose methods.1215 Unlike
one sample, a monoclonal gammopathy should be suspected. agarose electrophoresis, cooling is not required to
However, when it occurs in more than one, it represents an
provide optimal resolution.11 Furthermore,
artifact. In this case, the artifact results from excess buffer in the
reservoir. (Paragon SPE2 system stained with Paragon Blue. Note immunoxation analysis may also be performed
lighter staining than seen in Fig. 2.3 which used the same system on these gels. Under certain circumstances, the
with Paragon Violet.) Cellogel strips may be reused.16
20 Techniques for protein electrophoresis
electrophoresis on serum, a sample volume of as well as the display and output of results.20 The
10 l (plus dead space) is needed. It is placed into apparatus uses pressure injection to sample un-
a bar-coded tube in one of 10 sectors. Each sector diluted serum from the sector. By use of a lower
can accommodate seven samples. Data from each ionic strength buffer to carry the sample than the
sector can be viewed as a virtual gel sector analy- ionic strength of the running buffer, a stacking
sis (Fig. 2.9). After the samples are aspirated, they effect (like airplanes queuing up for landing) is
pass into seven 20-cm long uncoated fused silica produced.26 The individual samples travel into one
capillaries (inner diameter 25 m). Software con- of the six capillaries. At the alkaline pH of the
trols both the physical operation of the instrument Beckman buffer system (pH 9.9 by this labora-
torys pH meter), endosmotic ow is created by
high voltage and the narrow bore of the nega-
tively charged silica capillary. This propels the
proteins toward the cathode where a deuterium
lamp emits ultraviolet light to all seven channels.26
Separation results from the individual isoelectric
points, tertiary structure and charge of the pro-
teins under the conditions of the electrophoresis.
This light passes through a 214 nm interference
lter and ultraviolet silicon detectors sense the
absorbance. The software automatically provides
delimits for the ve major protein fractions, how-
ever, the operator can readjust these as needed.
Also, the operator can dene limits of M-proteins
for measurement (Fig. 2.10). A copy of this mea-
surement is stored with the patients le in order
to measure the M-protein in the same manner on
subsequent samples.
In the serum protein electrophoresis mode, the
instrument has a throughput of 42 samples per
S hour. Automated immunosubtraction may be per-
formed on the Paragon CZE 2000 (see Chapter 3).
For the immunosubtraction mode, ve samples are
processed in 1 h. Results are displayed as the elec-
tropherogram and measurements of the percentage
of major protein fractions.
Figure 2.9 Virtual gel image photographed directly from the The software allows the operator to concentrate
video monitor of the Paragon CZE 2000. A is just the left of the
on specic areas of the electropherogram by use of
albumin band, 1 is in the a1-region, 2 is at the start of the a2-
region, T is just left of the transferrin band and C is to the right of the zoom feature. With the zoom feature selected,
the C3 band. In the second lane (counting from the top), note that an area of the curve may be indicated for closer
the C3 band appears dark and broad. This reects the presence of study. This may be useful in examining small dis-
an IgA k monoclonal gammopathy migrating in this location. Note tortions in the b- or g-region for the presence of
also the fth lane from a patient that was receiving heparin. The
M-proteins (Fig. 2.11).
broad diffuse protein slur (s) anodal to the albumin band reects
a1-lipoprotein that often migrates here in serum from heparinized
Recently, the Sebia Capillarys system received
patients (see Chapter 4). Also, there is a very subtle diffuse band Food and Drug Administration (FDA) approval for
overlying the C3 region (arrow). This was a subtle IgA k clinical use.27 As in the Beckman instrument, the
monoclonal gammopathy. inner diameter is 25 m, however, the Capillarys
22 Techniques for protein electrophoresis
(a)
(a)
(b)
2 Figure 2.11 (a) The arrow indicates a slight irregularity in the fast
g-region of this electropherogram (Paragon CZE 2000). (b) The
designated area was expanded to allow a better look at the fast g-
region. The rounded area indicated by the line turned out to be
polyclonal. (Paragon CZE 2000.)
(b)
a
Data modied from Bossuyt et al.28 Range in for all three instruments expressed as 95% condence intervals. CZE, capillary zone
electrophoresis.
(Paragon CZE 2000), the lower and upper ranges method. This difference likely relates to the differ-
for a1-globulin were twice as high as the ranges for ences in the gels and stains used by the agarose
the gel-based systems. The a2- and b-globulin commercial products, since the CZE instrument
fractions for CZE were slightly lower than those used was the same in both cases. Despite minor dif-
from cellulose and agarose gels. Similar results ferences in the fractions from CZE to gel-based
were reported when Katzmann et al.23 compared techniques, Petrini et al.30 found good agreement
CZE (Paragon CZE 2000; Beckman Instruments) between interpretation of results when they com-
with agarose gel electrophoresis (Helena REP pared CZE to a high-resolution cellulose acetate
system). In their study, there was a 46 per cent electrophoresis on one thousand sera.
increase in the a1-globulin fraction in the CZE
samples compared with the agarose. They also
reported a 36 per cent decrease in a2-globulin frac- Pediatric reference ranges
tion and a 10 per cent decrease in b-globulin
fraction.23 Reference ranges for the ve major protein frac-
The increase in a1-globulin likely reects the tions in a pediatric population was established by
increased ability of CZE to detect both a1-lipo- Bossuyt et al.31 (Table 2.2). They divided the popu-
protein and a1-acid glycoprotein (orosomucoid) lation into four groups by ages: 12 years,
compared with cellulose and agarose gels. The high 34 years, 59 years, 1014 years. No differences
sialic acid content of orosomucoid interferes with were found between boys and girls for any of these
binding of protein dyes, whereas CZE detects pro- age groups in any of the fractions. The a2-globulin
teins via peptide bond absorbance that is not fraction values were lower in the older children
inuenced by this factor.28,29 (514 years) than in the younger children
In contrast to the Bossuyt et al.28 study, however, (14 years) because of higher values of a2-
Katzmann et al.23 noted a 21 per cent decrease in g- macroglobulin in the latter.31 Not surprisingly, the
globulin by CZE compared with their agarose g-globulin fraction was higher in the older groups
24 Techniques for protein electrophoresis
Table 2.2 Pediatric reference ranges for capillary zone electrophoresis (Paragon CZE 2000)
12 33 63.5 (54.770.4) 6.3 (4.28.5) 10.8 (715.6) 9.4 (7.511.6) 10.1 (4.716.0)
34 44 61.5 (53.970.4) 6.4 (4.88.1) 10.4 (7.615.2) 9.5 (7.411.6) 11.6 (7.117.8)
59 70 62.2 (52.666.3) 6.2 (4.27.6) 9.9 (7.413.5) 9.4 (7.911.3) 12.2 (8.518.7)
1014 48 61.2 (54.169.1) 5.9 (4.48.0) 8.9 (6.811.4) 9.7 (8.512.9) 13.7 (8.817.6)
a
Data from Bossuyt et al.31 with permission. Results expressed as fraction percentages as median, 95% condence limits in
parenthesis.
reported a similar occurrence with an IgM M- a2-region peak caused by sodium meglumine ioxi-
protein. In another case, 2-mercaptoethanol talamate (Telebrix, Guerbet Cedex, France) by
pretreatment was needed before the concentration desalting the sample (Fig. 2.14). For their pro-
of a b-migrating IgM M-protein could be accu- cedure, they used D-Salt Dextran plastic desalting
rately determined by CZE (Fig. 2.13).44 columns, 5 kDa cutoff from Pierce (Pierce
As with agarose and cellulose acetate elec- Biotechnology, Rockford, IL, USA). However,
trophoresis, the most challenging M-proteins are Arranz-Pea et al.32 reported obstruction in some
those that are of relatively low concentration that of their capillaries after trying that procedure.
also migrate in the b-region. Transferrin and C3 They removed the interference by adding 0.2 g of
may obscure these small M-proteins.43 The recent activated charcoal to 1 ml of serum, vortexing for
modication of the Paragon system to their version 20 s, and centrifuging at 200 g for 510 min at
1.6 system may help to detect some of these M- room temperature. Sometimes, two or three cen-
proteins. trifugations were needed to clear the serum. In
One of the most signicant ongoing problems Table 2.3 is the listing of radiocontrast dyes that
with CZE is the presence of false positive bands were tested by Arranz-Pea et al. complete with the
that occur when radiocontrast dyes are region where false positives are known to occur.32
present.32,33,45 With the increasing use of these dyes, In addition to radiocontrast dyes, other molecules
this problem occurs weekly in our laboratory. that absorb at 200215 nm will create unusual
Blessum et al.45 successfully removed an artifactual bands on CZE. For example, Bossuyt and co-
workers46,47 reported that the antibiotic
piperacillin-tazobactam (Tazocin; Wyeth Lederle)
(a)
a
(a)
(b)
b
(b)
Table 2.3 Location in the electropherogram and ultraviolet (UV) maxima for radio-opaque mediaa
a
Data from Arranz-Pea et al.,32 with permission.
will produce a small peak in the b-globulin region subtle a1-antitrypsin variants, small monoclonal
and the sulfamide sulfamethoxazole produces a gammopathies and oligoclonal patterns.5,9
small peak at the anodal edge of the albumin frac- Although the eye can detect variations in migra-
tion. tion more readily than most available clinical
In our laboratory, I have seen numerous small laboratory densitometers, differences in density of
deections that likely represent radiocontrast dyes staining may be objectively noted by the densito-
or other interferences. The most convenient way I meter. Therefore, the objective information
have found to rule out a monoclonal gammopathy obtained from densitometric scans of protein gels
is to perform an immunoxation with the Penta helps to draw the attention of the observer to a
(pentavalent) system (anti-G, A, M, K, and L in subtle quantitative abnormality that otherwise
the same reagent) on all suspicious cases (see might have been missed. Unfortunately, the preci-
Chapter 3). This simple semi-automated sion of densitometry is poor in examining the
immunoxation is negative with contrast dyes and smallest serum fraction (a1-globulin), although
other non-M-protein bands. Laboratories that use there may be considerable variation in other
CZE should encourage their clinicians to wait at fractions 48 (Table 2.4). Because of this, some
least a week to order serum protein electrophore- consider clinical densitometry of protein elec-
sis on patients that have received radiocontrast trophoretic fractions to be a semiquantitative
dyes.32 procedure.48
The values obtained by densitometric scanning
of electrophoretic gels will differ depending on the
DENSITOMETRIC SCANNING dilution of serum, the stain, the densitometer and
the electrophoretic system used for the analysis.
Densitometry provides objective information, The dilution of the sample used can be particu-
however, it should not be used without direct larly important when estimating the concentration
visual inspection of the gel because it may miss of an M-protein by densitometry.9,49 For example,
Densitometric scanning 27
aware that these measurements are not standard- to be overestimated compared with nephelometric
ized and that there is considerable variability in the techniques.56
measurement of an M-protein from one laboratory While densitometry suffers from the limitations
to another. of the dye used to stain the protein and the dilution
The correlation between polyclonal immunoglob- of the sample, nephelometric techniques also have
ulin concentration determined by nephelometry inherent problems due to antigen excess effects as
versus the concentration determined by densito- well as the characteristics of antisera used in the
metry is far from perfect.55 Schreiber et al.53 measurement.57,58 In their studies, Sinclair et al.57
compared the concentration of g-globulins with and Tichy59 found that electrophoresis followed by
nephelometric values of IgG + IgM + 1/2 IgA (they densitometry was superior to immunochemical
assumed that about half of the IgA migrated in the methods for following IgM monoclonal gammo-
b-region of their system, and hence would not be pathies.
included in the g-fraction). Although the correla- Stemerman et al.60 also recommended following
tion between the two techniques was very good (an monoclonal gammopathies (IgG in their study) by
average correlation coefcient of 0.95), the densit- densitometric scans of the serum protein electro-
ometric technique consistently gave lower results phoresis patterns. They noted that there was a loss
than did nephelometry. This discrepancy became of linearity above 6 g/dl requiring dilution for
more pronounced at higher immunoglobulin con- accurate results. Further, they presented a table to
centrations.53 Similar results were reported by aid the interpreter in determining if a change in the
Chang et al.49 who recommended diluting sera that quantity of a monoclonal gammopathy is signi-
contain total protein from 9.1 to 11.4 g/dl 1:10, cant (Table 2.5). They used Coomassie Brilliant
whereas total protein > 11.5 g/dl should be diluted Blue to improve the sensitivity and scanned the
1:20 to provide better linearity. Further, in esti- monoclonal band with the borders delineated
mating albumin concentrations, overstaining of automatically by their Pharmacia LKB 2220
albumin by some dyes may cause its concentration recording integrator.
Table 2.5 Minimal differences in paraprotein measurement that indicate true changes between seraa
0 1.1 1.8
10 1.6 2.5
20 2.0 3.3
30 2.4 4.0
40 2.9 4.7
50 3.3 5.4
60 3.7 6.1
70 7.2 6.8
80 4.6 7.5
a
Data modied from Stemerman et al.60
References 29
Table 2.6 Comparison of normal serum protein values by cellulose acetate versus high-resolution electrophoresis densitometry
a
Gelman ACD Densitometer used for scanning.
b
Beckman Appraise Densitometer used for scanning.
c
Results expressed in Gm/dl as 2 SD range.
d
Although the high-resolution scans can be separated into two b fractions, only one is used for comparison with the ve-band
pattern.9
The M-protein peak on densitometric scans is stained cellulose acetate (Table 2.6). However, the
measured in the same manner described above for superior band discrimination by Amido Black
CZE. The indicator is placed at the notch just made this difference in densitometric information
before and just after the M-protein peak. Clearly, insignicant for the purpose of clinical interpreta-
no currently available technique is able to tion.9
ompletely separate the M-protein concentration
from polyclonal immunoglobulins that may
migrate in the same region. As with capillary zone REFERENCES
electropherograms, the b-region is more problem-
atic because of the presence of other major proteins 1. Wieme R. Agar gel electrophoresis. Amsterdam:
(see Chapter 4). However, this technique provides Elsevier, 1965.
a reliable method to follow M-proteins in serum 2. Briere RO, Mull JD. Electrophoresis of serum
and urine. protein with cellulose acetate. A method for
We perform densitometric scanning on all gel- quantitation. Am J Clin Pathol 1964;34:547551.
based serum patterns. Even though gels with high- 3. Laurell CB. Electrophoresis, specic protein
resolution allow for recognition of up to twelve assays or both in measurement of plasma
proteins, our densitometric patterns of these gels proteins. Clin Chem 1973;19:99102.
use a standard ve serum-fraction pattern. This 4. Keren DF, Alexanian R, Goeken JA, Gorevic PD,
reects the variation in small bands such as b1- Kyle RA, Tomar RH. Guidelines for clinical and
lipoprotein and a1-antichymotrypsin, and the vari- laboratory evaluation patients with monoclonal
able proportions of C3 and C3 breakdown gammopathies. Arch Pathol Lab Med 1999;123:
product (which migrate in different regions see 106107.
Chapter 4), depending on the conditions of stor- 5. Aguzzi F, Kohn J, Petrini C, Whicher JT.
age of the sample and in vivo activation. A com- Densitometry of serum protein
parison with our former routine ve-band electrophoretograms. Clin Chem 1986;32:
cellulose acetate technique showed a lower level 20042005.
of al- and a2-globulins in the Amido Black-stained 6. Wild D. The immunoassay handbook. London:
agarose strips compared with the Ponceau S- Nature Publishing Group, 2001.
30 Techniques for protein electrophoresis
7. Johansson BG. Agarose gel electrophoresis. 20. Bienvenu J, Graziani MS, Arpin F, et al.
Scand J Clin Lab Invest Suppl 1972;124:719. Multicenter evaluation of the Paragon CZE 2000
8. Safety note. Clin Chem 1980;26:804. capillary zone electrophoresis system for serum
9. Keren DF, Di Sante AC, Bordine SL. protein electrophoresis and monoclonal
Densitometric scanning of high-resolution component typing. Clin Chem 1998;44:599605.
electrophoresis of serum: methodology and 21. Jenkins MA, Guerin MD. Optimization of serum
clinical application. Am J Clin Pathol 1986;85: protein separation by capillary electrophoresis.
348352. Clin Chem 1996;42:1886.
10. Kohn J. A cellulose acetate supporting medium 22. Jenkins MA, Guerin MD. Capillary
for zone electrophoresis. Clin Chim Acta electrophoresis as a clinical tool. J Chromatogr B
1957;2:297304. Biomed Appl 1996;682:2334.
11. Destro-Bisol G, Santini SA. Electrophoresis on 23. Katzmann JA, Clark R, Sanders E, Landers JP,
cellulose acetate and Cellogel: current status and Kyle RA. Prospective study of serum protein
perspectives. J Chromatogr A 1995;698:3340. capillary zone electrophoresis and immunotyping
12. Aguzzi F, Jayakar SD, Merlini G, Petrini C. of monoclonal proteins by immunosubtraction.
Electrophoresis: cellulose acetate vs agarose gel, Am J Clin Pathol 1998;110:503509.
visual inspection vs. densitometry. Clin Chem 24. Keren DF. Capillary zone electrophoresis in the
1981;27:19441945. evaluation of serum protein abnormalities. Am J
13. Merlini G, Piro P, Pavesi F, Epis R, Aguzzi F. Clin Pathol 1998;110:248252.
Detection and identication of monoclonal 25. Jolliff CR, Blessum CR. Comparison of serum
components: immunoelectrophoresis on agarose protein electrophoresis by agarose gel and
gel and immunoxation on cellulose acetate capillary zone electrophoresis in a clinical setting.
compared. Clin Chem 1981;27:18621865. Electrophoresis 1997;18:17811784.
14. Janik B, Dane RG. High-resolution 26. Klein GL, Jolliff CR. Chapter 16. Capillary
electrophoresis of serum proteins on cellulosic electrophoresis for the routine clinical
membranes and identication of individual laboratory. In: Landers JP, ed. Handbook of
components by immunoxation and capillary electrophoresis. Boca Raton: CRC
immunosubtraction. J Clin Chem Clin Biochem Press, 1993.
1981;19:712713. 27. Bossuyt X, Lissoir B, Marien G, et al. Automated
15. Ojala K, Weber TH. Some alternatives to the serum protein electrophoresis by Capillarys. Clin
proposed selected method for agarose gel Chem Lab Med 2003;41:704710.
electrophoresis. Clin Chem 1980;26: 28. Bossuyt X, Schiettekatte G, Bogaerts A,
17541755. Blanckaert N. Serum protein electrophoresis by
16. Destro-Bisol G. Reusing Cellogel strips after CZE 2000 clinical capillary electrophoresis
visualization of electrophoretically separated system. Clin Chem 1998;44:749759.
isozymes. Electrophoresis 1993;14:238239. 29. Dati F, Schumann G, Thomas L, et al. Consensus
17. Aguzzi F, Rezzani A. An advantageous but of a group of professional societies and
neglected technique for immunoxation of diagnostic companies on guidelines for interim
monoclonal components on cellulose acetate reference ranges for 14 proteins in serum based
membranes. Standardisation and evaluation. on the standardization against the
Giorn It Chim Clin 1986;11:293299. IFCC/BCR/CAP Reference Material (CRM 470).
18. Jenkins MA. Clinical applications of capillary International Federation of Clinical Chemistry.
electrophoresis. Status at the new millennium. Community Bureau of Reference of the
Mol Biotechnol 2000;15:201209. Commission of the European Communities.
19. Landers JP. Clinical capillary electrophoresis. College of American Pathologists. Eur J Clin
Clin Chem 1995;41:495509. Chem Clin Biochem 1996;34:517520.
References 31
30. Petrini C, Alessio MG, Scapellato L, Brambilla S, 39. Bossuyt X, Bogaerts A, Schiettekatte G,
Franzini C. Serum proteins by capillary zone Blanckaert N. Detection and classication of
electrophoresis: approaches to the denition of paraproteins by capillary immunoxation/
reference values. Clin Chem Lab Med subtraction. Clin Chem 1998;44:760764.
1999;37:975980. 40. Katzmann JA, Clark R, Wiegert E, et al.
31. Bossuyt X, Claeys R, Bogaert G, Said HI, Wouters Identication of monoclonal proteins in serum: a
C, Groven C, Sneyers L, Marien G, Gorus F. quantitative comparison of acetate, agarose gel,
Reference values for the ve electrophoretic serum and capillary electrophoresis. Electrophoresis
protein fractions in Caucasian children by 1997;18:17751780.
capillary zone electrophoresis. Clin Chem Lab 41. Smalley DL, Mayer RP, Bugg MF. Capillary zone
Med 2001;39:970972. electrophoresis compared with agarose gel and
32. Arranz-Pena ML, Gonzalez-Sagrado M, Olmos- immunoxation electrophoresis. Am J Clin
Linares AM, Fernandez-Garcia N, Martin-Gil FJ. Pathol 2000;114:487488.
Interference of iodinated contrast media in serum 42. Henskens Y, de Winter J, Pekelharing M, Ponjee
capillary zone electrophoresis. Clin Chem G. Detection and identication of monoclonal
2000;46:736737. gammopathies by capillary electrophoresis. Clin
33. Bossuyt X, Mewis A, Blanckaert N. Interference Chem 1998;44:11841190.
of radio-opaque agents in clinical capillary zone 43. Bossuyt X, Marien G. False-negative results in
electrophoresis. Clin Chem 1999;45:129131. detection of monoclonal proteins by capillary
34. Litwin CM, Anderson SK, Philipps G, Martins zone electrophoresis: a prospective study. Clin
TB, Jaskowski TD, Hill HR. Comparison of Chem 2001;47:14771479.
capillary zone and immunosubtraction with 44. Keren DF, Gulbranson R, Carey JL, Krauss JC.
agarose gel and immunoxation electrophoresis 2-Mercaptoethanol treatment improves
for detecting and identifying monoclonal measurement of an IgM kappa M-protein by
gammopathies. Am J Clin Pathol 1999;112: capillary electrophoresis. Clin Chem 2001;47:
411417. 13261327.
35. Clark R, Katzmann JA, Kyle RA, Fleisher M, 45. Blessum CR, Khatter N, Alter SC. Technique to
Landers JP. Differential diagnosis of remove interference caused by radio-opaque
gammopathies by capillary electrophoresis and agents in clinical capillary zone electrophoresis.
immunosubtraction: analysis of serum samples Clin Chem 1999;45:1313.
problematic by agarose gel electrophoresis. 46. Bossuyt X, Verhaegen J, Marien G, Blanckaert
Electrophoresis 1998;19:24792484. N. Effect of sulfamethoxazole on clinical
36. Jenkins MA, Kulinskaya E, Martin HD, Guerin capillary zone electrophoresis of serum proteins.
MD. Evaluation of serum protein separation by Clin Chem 2003;49:340341.
capillary electrophoresis: prospective analysis of 47. Bossuyt X, Peetermans WE. Effect of piperacillin-
1000 specimens. J Chromatogr B Biomed Appl tazobactam on clinical capillary zone
1995;672:241251. electrophoresis of serum proteins. Clin Chem
37. Jenkins MA, Guerin MD. Capillary 2002;48:204205.
electrophoresis procedures for serum protein 48. Kahn SN, Strony LP. Imprecision of
analysis: comparison with established techniques. quantication of serum protein fractions by
J Chromatogr B Biomed Sci Appl 1997;699: electrophoresis on cellulose acetate. Clin Chem
257268. 1986;32:356357.
38. Jenkins MA, Ratnaike S. Five unusual serum 49. Chang CY, Fritsche HA, Glassman AB, McClure
protein presentations found by capillary KC, Liu FJ. Underestimation of monoclonal
electrophoresis in the clinical laboratory. J proteins by agarose serum protein electrophoresis.
Biochem Biophys Methods 1999;41:3147. Ann Clin Lab Sci 1997;27:123129.
32 Techniques for protein electrophoresis
50. Carter PM, Slater L, Lee J, Perry D, Hobbs JR. 56. Mendler MH, Corbinais S, Sapey T, et al. In
Protein analyses in myelomatosis. J Clin Pathol patients with cirrhosis, serum albumin
Suppl 1975;6:4553. determination should be carried out by
51. Sun T. Interpretation of protein and isoenzyme immunonephelometry rather than by protein
patterns in body uids. New York/Tokyo: Igaku- electrophoresis. Eur J Gastroenterol Hepatol
Shoin, 1992. 1999;11:14051411.
52. Uriel L. Interpretation quantitative des resultats 57. Sinclair D, Ballantyne F, Shanley S, Caine E,
aprs electrophorese en gelose, 1. Considerations OReilly D, Shenkin A. Estimation of
gnrales, application a ltude de constituants paraproteins by immunoturbidimetry and
proteiques isols. Clin Chim Acta 1958;3: electrophoresis followed by scanning
234238. densitometry. Ann Clin Biochem 1990;27:
53. Schreiber WE, Chiang E, Tse SS. Electrophoresis 335337.
underestimates the concentration of polyclonal 58. Bush D, Keren DF. Over- and underestimation of
immunoglobulins in serum. Am J Clin Pathol monoclonal gammopathies by quantication of
1992;97:610613. kappa- and lambda-containing immunoglobulins
54. Sun T. High-resolution agarose electrophoresis. in serum. Clin Chem 1992;38:315316.
In: Ritzmann SE, ed. Protein abnormalities. Vol 59. Tichy M. A comparison of methods of
1. Physiology of immunoglobulins: diagnostic monoclonal immunoglobulin quantitation.
and clinical aspects. New York: Alan R. Liss, Neoplasma 1985;32:3136.
1982. 60. Stemerman D, Papadea C, Martino-Saltzman D,
55. Austin GE, Check IJ, Hunter RL. Analysis of the OConnell AC, Demaline B, Austin GE. Precision
reliability of serial paraprotein determinations in and reliability of paraprotein determinations by
patients with plasma cell dyscrasias. Am J Clin high-resolution agarose gel electrophoresis. Am J
Pathol 1983;79:227230. Clin Pathol 1989;91:435440.
3
Immunofixation, immunosubtraction, and
immunoselection techniques
Whereas serum protein electrophoresis can detect given sample. Immunoprecipitation involves the
restrictions that resemble monoclonal gammo- interaction of antibody molecules with antigen in
pathies, it cannot denitively identify a restriction either a gel or liquid matrix in which the molecules
as an M-protein. For that, immunochemical are free to diffuse. The key to precipitation is the
methods together with electrophoresis must be multivalent nature of antibodies and antigens (each
employed. In this chapter, I review the principles of has two or more sites with which they can inter-
the techniques employed for the identication of act).1,2
restrictions seen on protein electrophoresis as M- In most chemical reactions, when substance A is
proteins. Most laboratories now perform mixed with substance B to form a product C (lets
immunoxation electrophoresis (IFE) to identify call it a precipitate), the reaction can be expressed
the restriction as an M-protein. Laboratories using as
capillary zone electrophoresis (CZE) may be able
to perform immunosubtraction (ISUB), depending A+BC
on which CZE system they are using.
Immunoelectrophoresis (IEP) is still used by some As shown in Fig. 3.1, as one increases the amount
laboratories. It does have a couple of advantages of substance A while maintaining the amount of
over immunoxation, but it is slow, less sensitive
and more difcult to interpret than IFE. Lastly,
cases of heavy chain disease benet from perfor- A+B C
mance of immunoselection (ISEL).
Product of C
PRINCIPLES OF
IMMUNOPRECIPITATION
Whether one is performing IFE, ISUB, IEP, or ISEL
the basic principles of the precipitin reaction are Amount of A added
the same, and understanding them is critically Figure 3.1 As substance B is used up, no more product C is
important to making the correct interpretation of a formed by the addition of substance A.
34 Immunoxation, immunosubtraction, and immunoselection techniques
Ab + Ag Ag Ab D E F epitopes
B C G
A
Amount of precipitate
the need to consider the concentration of the anti- using dilutions that do not attempt to account for
body and antigen is critically important to optimize the approximate concentrations of the immuno-
IFE, ISUB, IEP, and ISEL techniques and to avoid globulins looked for. Why were broad precipitates
potential errors. Although the following concepts of IgG and k on the original IFE anodal to the
are quite basic, commercial products whose instruc- major band in the slow g-region? This represented
tions try to simplify the technique occasionally lower concentrations of the monoclonal protein
ignore them. Before launching into a discussion of that had migrated behind the major band. The
the chemistry involved, I will use the following case broad migration may have reected some self-
to illustrate the relevance of the antibody and anti- aggregation, heavy glycosylation, or the known
gen concentrations to the amount of protein present microheterogeneity of monoclonal proteins.3 It
in the gammopathy being studied. stained well, however, because the concentration
Although the protein electrophoresis tract looked of the monoclonal band in this region was consid-
like a monoclonal gammopathy serum protein erably lower than at the slow g-region where most
electrophoresis (SPE) for Fig. 3.4a the immuno- of the protein migrated. Because of its high con-
xation showed a very broad reactivity with IgG centration at the slow g-end, the complexes formed
and k that suggested a polyclonal reactivity. No were small and washed away during the wash steps
reaction was seen in the region where the slow g- (antigen excess effect), resulting in confusion, or
band was seen. Serum protein electrophoresis was worse, a false negative. Hopefully, this digression
performed on the sample. As shown in Fig. 3.4b, on a real case will whet your appetite for the rather
this gave a slow g-band similar to the one that dry, but straightforward, discussion of these
detected in the SPE lane of the referred gel. We important basics.
measured IgG, IgA, IgM, k and l by nephelometry. In the immune precipitin reaction, the multi-
From the data shown in Fig. 3.4c, it was concluded valency of the antibody and antigen allow for the
that the serum contained an IgG k monoclonal formation of a lattice (Fig. 3.5). In the zone of anti-
gammopathy. gen excess (Fig. 3.2), the large amount of antigen
The problem with the referred IFE related to the present makes it likely that only relatively small
dilutions of patient serum applied to the gel. We immune complexes are formed, with the formula
performed IFE, using the same commercial kit as AB(1)AG(2). These molecules are too small to pre-
the outside laboratory, but adjusting the dilution cipitate. As more antibody is added to the system,
of the serum to approximate equivalence for the a precipitate (large antibodyantigen latticework)
antibodyantigen interaction (see below). As begins to form until a maximal precipitate is noted.
shown in Fig. 3.4d, our IFE demonstrates that the This maximal precipitate occurs in the zone of
slow g-band is an IgG k monoclonal gammopathy. equivalence where the formula is AB(l)AG(1) (Fig.
We ran the k at two concentrations, 1:10 and 1:20, 3.5). With the addition of still more antibody, the
both of which gave interpretable results. This epitopes on the surface of the antigen are saturated
demonstrates that one must adjust the concentra- with antibody molecules and therefore are not
tion to be near the equivalence range to obtain an available for reaction with other crosslinking anti-
interpretable result. bodies. Here the formula is AB(x)AG(I) where x is
The dilutions of patients serum on the rst IFE the number of epitopes expressed on the surface of
gel resulted in antigen excess effect in the region of a particular antigen. For optimal IFE, a large
the monoclonal band. Note also that no precipitate crosslinked precipitate is needed. Smaller immune
at all is seen in the IgA or l lanes. Did the technol- complexes usually formed at antigen excess will
ogist forget to place the sample or the antiserum in wash away. In the case shown in Fig. 3.4, there was
this tract, or was the dilution used too large, result- too much antigen present at the cathodal end of the
ing in an undetectable precipitate (in effect, anti- gel resulting in the band being washed away (anti-
body excess)? This is another key problem with gen excess effect).
36 Immunoxation, immunosubtraction, and immunoselection techniques
(b)
Nonidentity Identity
Ag Anti-A, Anti-B Anti-A, Anti-B
Ag
Ag Ag
Ag
Ag Ag
Ag
Ag Ag
Ag Ag
Ag Ag Ag
Ag Ag Ag
Ag
Ag Ag
Ag Ag A B A A
A Spur A'
DOUBLE DIFFUSION IN TWO
DIRECTIONS (OUCHTERLONY Figure 3.7 Ouchterlony plates using specic antibodies can be
used to determine the antigen content of unknown solutions.
TECHNIQUE) When different antigens (A and B) are reacted with antibodies to
A and B, two lines form, which cross one another (nonidentity).
Ouchterlony devised a simple way to use the When the antigen in both wells is the same, the two lines meet
immunoprecipitation reaction to determine anti- (identity) because the antibodies to the antigen are absorbed out
body reactivity and to identify unknown antigens.4 in the precipitation and do not pass beyond the precipitin line to
react with the other antigen. When the antigen is similar (A), only
He cut wells in an agarose gel and put antibody in
some of the antibodies to the antigen (A) will be removed.
one well and antigen in another. These molecules Antibodies to antigen A that are not expressed on A will pass
diffuse radially out of the wells (Fig. 3.6). As they through the precipitin band formed by anti-A and A, and a second
diffuse away from the center of the well, their con- line (spur) will form with antigen A. This is the partial identity
centration decreases logarithmically. A precipitin pattern.
band forms somewhere between the two wells at the
point at which their concentrations are equivalent. of the antibodyantigen interactions when they dif-
If the precipitate is closer to the antigen well, it indi- fuse through agarose gel for a distance is that the
cates that the antibody was more concentrated than concentration of the reactants is automatically
the antigen because it had to diffuse further than the adjusted to form the precipitate.
antigen (thereby becoming more dilute) before a Ouchterlony also found that by using a known
precipitate could be seen. Other factors such as the antigen and antibody he could determine whether
size of the molecules and interactions with the gel an unknown antigen was structurally the same,
also affect this reaction but, in general, the beauty similar, or dissimilar. For example, immunoprecip-
itation of an antiserum with activity against both
antigens A and B is shown in Fig. 3.7. If antigen A
Antibody Antigen is in one test well and antigen B is in the other, a
pattern of non-identity occurs where the two lines
cross. The precipitin line does not resemble a solid
barrier, rather a grossly visible latticework (Fig.
3.5). Antibodies and antigen that lack specicity
for the epitopes on this immune complex readily
pass through this latticework.
If antigen A is placed in both wells, a pattern of
Figure 3.6 As antibody and antigen diffuse from the well, their identity occurs. Here, the lines meet but do not cross
concentrations decrease logarithmically. The precipitin line forms one another, because all the antibodies to antigen A
where the concentrations are equivalent. react with substance A they do not pass through the
38 Immunoxation, immunosubtraction, and immunoselection techniques
lattice. If antigen A is placed in one well and a chem- with the patients sample on the IEP strip. These
ically similar antigen A' (which may lack one antisera slowly diffuse from the trough into the gel
epitope that antigen A possesses) is placed in the while the protein components slowly diffuse in a
other test well, a pattern of partial identity results. radial fashion. Precipitin lines form where the anti-
This is a difcult pattern to identify and to under- sera and the specic antigens are at equivalence.
stand. Here, all of the molecules that react with anti- Owing to the geometry of the application wells and
gen A' are precipitated onto the lattice when they radial diffusion, the precipitin lines are in the form
meet antibody to A. However, since antigen A' lacks of arcs. Large quantities of monoclonal proteins are
one epitope that is present on antigen A, antibodies readily identied by IEP by comparing the migra-
to this unique epitope pass through the lattice tion of the control to the patients serum across a
formed by anti-A + antigen A' and are available to trough containing a specic antiserum (Fig. 3.9).
interact with antigen A. Because there are only
relatively few of these molecules and most of the
antibodies that react with anti-A have been removed
by the interaction with A', antigen A must diffuse
slightly further (to become more dilute in concen-
tration) before it is in equivalence to form a pre-
cipitate with the fewer remaining anti-A molecules +
that react with the unique epitope. This explains the
spur of antigen A, which is the classic denition of origin
IMMUNOELECTROPHORESIS
A logical development of the immune precipitin
reaction was to combine it with electrophoresis to
achieve separation by charge and then identication
of the molecules by immunological techniques. For anti-lgG in trough
many years, IEP was the mainstay for combining
wash
these two techniques in both research and clinical
laboratories.5 It is performed by placing the
patients serum into a series of wells in an agarose
gel. The sample is electrophoresed to permit sepa-
ration of the major serum proteins (Fig. 3.8). After
electrophoresis, the gel is removed from the elec-
trophoretic apparatus and the troughs are lled
Figure 3.8 Immunoelectrophoresis (IEP) takes advantage of the
with antisera to various specic components of
principles of electrophoresis, gel diffusion (which adjusts
interest, usually including antipentavalent human concentrations during diffusion), and antibodyantigen precipitin
immunoglobulin (which reacts with the three major reactions to identify specic proteins. In the example shown, only
heavy and light chain classes), and the remaining IgG is precipitated and present after the wash step. The closer the
troughs are lled with monospecic antisera against IgG is to the antiserum trough, the greater is its concentration.
Large molecules, such as pentameric IgM, can be difcult to
IgG, IgA, IgM, k-, and l-chains. For comparison of
examine by this technique because they diffuse slowly through the
electrophoretic migration patterns, and to be sure agarose. Small monoclonal proteins are also difcult to identify
that the correct antisera have been placed into the because this technique does not offer good resolution of individual
appropriate troughs, a control serum is alternated proteins.
Immunoelectrophoresis 39
3
Figure 3.9 Immunoelectrophoresis (IEP) shows a large IgG l
monoclonal gammopathy. Note that control (c) serum alternates
with patient (p) serum. Antisera to pentavalent (PV), IgG (G), IgA
(A), IgM (M), k (K), and l (L) were placed in the troughs. A large 4
arc with excessively anodal migration is seen in the polyvalent, IgG,
and l areas of the patients sample (indicated in all three
locations). The l arc stains weakly, which is the prozone
phenomenon often seen with l reagents. Diagnosis of such large 5
monoclonal gammopathies is relatively easy by IEP.
6
1
2
Figure 3.11 Immunoelectrophoresis from the serum of a patient
with both a monoclonal IgG l protein and a free l monoclonal
free light chain (MFLC) in the serum. The free l light chain
reactivity can be seen with the pentavalent reagent, where the 3
extra arc due to the anti-free l is indicated. The same area is
indicated in the anti-l reaction with the patients serum in the
bottom trough. This area is not seen with the anti-IgG reagent.
4
different migration than the intact monoclonal
gammopathy. Free light chains also express certain
epitopes on their surfaces that are not usually 5
expressed on the surfaces of a light chain attached
to intact immunoglobulin molecules. Commercial
antisera are able to detect these epitopes (Fig. 3.11). 6
Consequently, as with the Ouchterlony patterns of
partial identity discussed above, when the intact
immunoglobulin molecule reacts with the antiserum
7
it cannot remove antibodies to these hidden deter-
minants. The latter antibodies continue through the Figure 3.12 Nondiagnostic immunoelectrophoresis in a patient
precipitin band to react with the free light chain. with a fourfold increase in IgM and an obvious spike on serum
protein electrophoresis (shown on the top of Fig. 3.13). The
patients IgM is pentameric and is barely able to migrate out of the
well. The diffuse hazy area indicated near the well for each that
Limitations of IEP contains patients serum is where the monoclonal protein has
deposited. The normal and arcs represent the patients normal
Unfortunately, there are several problems with IEP serum IgG and do not reect his monoclonal protein (the so-called
that encouraged the development of newer umbrella effect).
Immunoelectrophoresis 41
detect biclonal gammopathies, that are more band may not enter the trough, and thus appear
readily detected with IFE, which allows greater dis- normal (Fig. 3.14).
crimination than does IEP between two different
monoclonal proteins that have relatively close pIs
(see Chapter 1). IMMUNOFIXATION
Immunoelectrophoresis lacks sensitivity to detect ELECTROPHORESIS
small M-proteins. Some believe that this insensitiv-
ity of IEP is an advantage because the larger mono- Immunoxation electrophoresis has been a dra-
clonal gammopathies detected are more likely to be matic development for the clinical labora-
of clinical signicance than smaller monoclonal tory.7,13,3538 This technique can be performed in
gammopathies. Small M-proteins, however, may about 34 h, is not subject to the umbrella effect,
have clinical importance. As discussed in Chapter can readily detect the small monoclonal gammo-
6, Kyle has coined the term monoclonal gammo- pathies, is easier to interpret than IEP, uses half as
pathy of undetermined signicance for patients much antisera as IEP, and requires the same equip-
with relatively small amounts of monoclonal ment as that needed for many of the commercially
gammopathy who lack clinical evidence of available protein electrophoresis methods. Further,
multiple myeloma.2224 Patients with monoclonal
gammopathy of undetermined signicance
(MGUS) need to be followed indenitely, because C
the gammopathy may develop into a malignant
process (myeloma or Waldenstrm macroglo- PV
bulinemia). Alternatively, small quantities of M-
proteins may represent the presence of a malignant P
B-cell neoplasm. Such small M-proteins may be
anti-G
detected in light chain myeloma, amyloid (AL),
chronic lymphocytic leukemia, Burkitt lymphoma, C
and in well-differentiated lymphocytic lym-
phoma.23,2527 It has become clear that some patients anti-A
with neurological complaints have monoclonal P
gammopathies that are related to their clinical anti-M
symptoms.2834
Another problem with IEP is over-interpretation C
of polyclonal increases, resulting in misdiagnosis of
inammatory conditions as monoclonal gammo- anti-K
pathies. This results when relatively massive poly-
P
clonal increases occur in IgG. Often, this produces
a distortion of the IgG band (as it requires a larger anti-L
space to diffuse to reach equivalence than is avail-
able). This can produce an appearance similar to C
the restriction seen in true monoclonal patterns Figure 3.14 Immunoelectrophoresis which was misdiagnosed as
(Fig. 3.14). Since there is twice as much total k an IgG k monoclonal gammopathy. It is a polyclonal pattern where
(light chain bound to intact immunoglobulin mole- there is so much IgG and k that they have spread into the trough
and their pattern falsely suggests a restriction (arrows). This
cules as well as the small amount of free light
should have been caught. The l actually also spills into the trough,
chains that normally circulate; see Chapter 7) as l but since it was broader and did not line up with the IgG, it was
under normal circumstances, and since the usual dismissed. Also, the IgA arc is markedly increased because of a
k/l ratio is 2.0 in most polyclonal increases, the l- corresponding polyclonal increase in IgA.
Immunoxation electrophoresis 43
semi-automated versions of IFE are currently avail- cases. They cautioned, however, that if equivocal
able that allow for examination of several samples results are obtained using these standard dilutions
with minimal technologist time. Sensitive new of serum, the analysis should be repeated using a
methods even allow for detection of oligoclonal different antiserum.40 Fortunately, there is usually a
bands in cerebrospinal uid by IFE technique on a relatively large leeway for this dilution, although
few microliters of unconcentrated sample (see this varies from one monoclonal protein to another
Chapter 8).39 and from one commercial antiserum to another. As
shown in Fig. 3.15, the monoclonal gammopathy
could be detected by a variety of dilutions around
the equivalence region. This would not have hap-
Selection of the dilution of patients pened, however, with all monoclonal proteins, or
serum with all commercial antisera (see Fig. 3.16). When
possible, it is a great advantage to know the con-
One advantage that IEP had over IFE was the slow centrations of the major immunoglobulin classes
diffusion step that allowed concentration of the before one sets up the IFE. A study by Hadler et
M-protein to adjust to the concentration of the al.41 recommended that the optimum concentration
antibody. This avoided the problem of the antigen for detection of M-proteins in the system they were
excess effect that may occur with IFE if the sample using was 2835 mg/dl. In our laboratory, we have
is not properly diluted. In IFE, the reaction occurs found that with most commercial antisera, the
quickly because there is no diffusion step. This was antigen (IgG, IgA, IgM, or k) needs to be present at
the step in IEP that allowed the antigen and anti- a concentration of about 100 mg/dl for best
body to adjust to equivalence for maximal precipi- results. Antisera to l tend to give weaker reactions
tation. and one may wish to dilute the patients serum to
To achieve optimum sensitivity and to avoid the about 50 mg/dl for l.
antigen excess effect, it is important to use a dilu- An example of the weakness of the anti-l
tion of the serum that places it close to the equiva- reagents is shown in Fig. 3.16. The serum protein
lence range for the immunoprecipitation reaction. electrophoresis gel shows a prominent slow b-
There is also considerable variation in the strength migrating monoclonal gammopathy. By nephelo-
and specicity of commercial antisera. Monos et metry, the IgA measured 14 400. Yet, with a 1:10
al.40 compared several commercial sera with the dilution of IgA, one did not have an antigen
same M-proteins and found considerable variation excess problem. The monoclonal band stained
in detection of the monoclonal protein. They also very strongly, perhaps obscuring the possible sec-
noted that an M-protein present in a concentration ond IgA band (likely a multimer of the major
of 700 mg/dl could be missed if the incorrect dilu- band). At 1:100, both the major band and the sec-
tion of the patients serum was used with a com- ond band are seen to advantage with the anti-IgA
mercial antisera that had poor reactivity for that reagent. However, with antisera to l at 1:10, a
particular protein. problem with antigen excess effect is seen. The
There are several ways to determine which dilu- center of the major band precipitate has washed
tion of serum to use. The rst is to use a standard away owing to inadequate size of the immune
dilution (suggested by commercial IFE kits). While complexes formed by the anti-l reagent; neverthe-
this is an easy approach, it presents problems in less, at 1:10 the same monoclonal protein reacts
cases with very high or very low concentrations of well with the anti-IgA reagent. Even at 1:50, the
M-proteins. Monos et al.40 found that a dilution of center of the l band has begun to wash away.
1:10 for IgG, 1:5 for IgA, IgM, and k-chains and With most antigen excess effects, one can still
1:2 for l-chains allowed them to make the proper make the diagnosis of the type of monoclonal
diagnosis of a monoclonal gammopathy in most gammopathy, however, in some cases, it can make
44 Immunoxation, immunosubtraction, and immunoselection techniques
Figure 3.15 Immunoxation of serum from patient with an IgG k monoclonal gammopathy. The concentration of the IgG was
1740 mg/dl. The monoclonal band is readily seen at the 1:20, 1:40, and 1:80 dilutions. At 1:10 the band is visible, but is considerably
obscured by the amount of polyclonal IgG present. At 1:5, there is a slight indentation at the junction of the band with the cathodal end of
the polyclonal IgG (arrow). However, I believe that this would have been missed. With the IgG undiluted (left lane), the band is not
detectable. (Paragon system stained with Paragon Violet; anode at the top.)
the correct interpretation difcult. When in doubt, samples for monoclonal gammopathies by rst
one should always rerun the sample at other dilu- performing protein electrophoresis and quantica-
tions or with other antisera. tion of IgG, IgA, IgM, k and l by standard
In the past few years, manufacturers have rec- immunological methods (nephelometry, turbidi-
ognized the need to adjust the concentration of metry, or radial immunodiffusion). Using this
serum to account for extremely high and low method, one can characterize large M-proteins
levels of antigen. For example, the Sebia without resorting to either IEP or IFE (see Chapter
Hydragel system (Sebia, Issy-les-Moulineaux, 8). In those cases where IFE is necessary (identify-
France) offers several levels of dilutions depend- ing small bands, abnormal k:l ratio, clinical
ing on the total immunoglobulin concentration picture compatible with monoclonal gammopathy
(Table 3.1). In our laboratory, strict adherence to process despite normal protein electrophoresis and
these instructions has provided excellent results quantications), one already knows the
with this kit. immunoglobulin concentrations and this allows
When using home brew IFE, or kits where no one to estimate readily the dilution to obtain
suggestion is provided for diluting the serum, optimal results. For example, with the
adjustment for concentration of a monoclonal immunoglobulin values shown in Table 3.2, results
protein may be accomplished in one of several are expressed as mg/dl. To determine the appro-
ways. One method involves evaluating serum priate dilution to use for a given sample, divide the
Immunoxation electrophoresis 45
(a)
XS
Figure 3.16 (a) Serum protein electrophoresis gel where the bottom lane contains a massive monoclonal gammopathy in the slow b-
region. The serum immediately above this sample has a small monoclonal band in the bg and another tiny restriction in the slow g-region.
Since the monoclonal band in the bottom lane stains as dense as the albumin bands on this gel and is considerably broader, one can
estimate that it will have about three times as much protein as albumin (about 12 g). (Paragon SPE2 system stained with Paragon Violet.)
(b) Immunoxation of the sample from the bottom lane of (a) using the indicated dilutions of patients serum. Note that the serum protein
electrophoresis (SPE) lane did not x the albumin band. Therefore, one cannot use it to estimate the protein concentration of the
monoclonal band. A large IgA restriction is easily seen in the 1:10 and 1:100 dilution, but a second IgA band is somewhat obscured by the
adjacent large IgA band at 1:10, but stands out nicely (arrow) at the 1:100 dilution. The center of the l precipitate has washed away at the
1:10 (XS arrow) dilution and even to some extent at the 1:50 dilution. A diffuse k band is seen using the 1:20 dilution. It indicates the
presence of this patients normal k. (Paragon system stained with Paragon Violet; anode at the top.)
46 Immunoxation, immunosubtraction, and immunoselection techniques
Table 3.1 Manufacturers suggested dilutions related to total a crude estimate of IgG. This is similar to the
immunoglobulin concentrationa
Sebia scheme above, since, in most sera, IgG
makes up most of the total immunoglobulin. With
Track Ratio of serum most g-migrating IgG monoclonal gammopathies,
to dilutent there is a reasonable correlation between the den-
sitometric scan of the g-region and the IgG deter-
IgG mined by immunochemical methods.42 Although
(total immunoglobulin > 0.5 < 2.0 g/dl) 1:5 these estimates can be used as approximations of
IgA, IgM, k, l the IgG concentration, they are far from per-
(total immunoglobulin > 0.5 < 2.0 g/dl) 1:2
fect.4244 This method only approximates the IgG
values, and one must estimate to the concentra-
IgG
tions for the other chains. Usually, however, two-
(total immunoglobulin > 2.0 g/dl) 1:10 thirds of the IgG is k and one-third is l. Also, IgA
IgA, IgM, k, l and IgM tend to be present in relatively low con-
(total immunoglobulin > 2.0 g/dl) 1:4 centrations (less than 300 mg/dl for most individ-
IgG uals). When using this method to estimate
(total immunoglobulin < 0.5 g/dl) 1:2.5 dilution, I prefer to start with a 1:2 dilution for
IgA and IgM. It is unusual that greater than a 1:4
IgA, IgM, k, l
dilution is required to avoid extreme antigen
(total immunoglobulin < 0.5 g/dl) 1:1
excess for IgA and IgM under normal circum-
a
Sebia Hydragel instructions.
stances. Since IFE is performed after serum pro-
tein electrophoresis, the presence of a suspicious
band, and its size, will alert us to the occasional
need to increase the dilution. When there is a and l), an M-protein can be readily ruled out
gross discrepancy between nephelometry and den- (Fig. 3.18).
sitometry, we usually favor the densitometric Following electrophoresis, the lane for compari-
value because, unlike nephelometry, it is not inu- son protein electrophoresis is xed with acid. For
enced by the vagaries of the different commercial the pentavalent or specic immunoglobulin lanes,
antisera and antigen excess effects.45,46 the patients sample is overlaid with specic anti-
One may estimate the g-globulin region of the serum. In some systems, commercial antiserum
serum protein electrophoresis without densito- against a specic isotype is coated onto a strip of
metry. Because serum protein electrophoresis is cellulose acetate. This strip is placed directly on top
always performed before IFE in our laboratory, of the electrophoresed sample. In other systems,
our technologists become adept in estimating the using a template, the commercial antiserum is
concentration of the g-region by examining the directly layered onto the gel. Diffusion of the anti-
stained gel itself. They base their dilutions on their sera directly into the thin gel beneath is quite rapid,
experience examining the gels. This method has requiring from as little as 530 min depending on
been adequate to estimate the dilution that should the system used.
be used to perform IFE and avoid the problems Following removal of the antisera, gels are
associated with the extreme antigen excess effect. washed and stained with Coomassie Brilliant Blue,
Whether one uses a standard dilution or one of the Amido Black or a specic commercial protein dye
above methods for performing dilutions of the provided by the manufacturers kit. Some workers
patients serum for IFE, it is recommend that a advocate silver staining for enhanced sensitivity. I
serum protein electrophoresis be performed at the nd silver stains clumsy and messy to use. In addi-
same time and compared with the IFE. Many of the tion, the background that may result makes the
commercial products provide a serum protein elec- interpretation difcult. Sensitivity is not usually a
trophoresis lane that can be used to detect the M- problem with IFE. Indeed, one of the more
protein for comparison purposes. Any band that is common complaints about the technique is that
not explained by the IFE should result in a repeat with the standard dyes, IFE shows too many
IFE with other dilutions, or occasionally other bands, not too few.
antisera.
INTERPRETATION OF IMMUNOFIXATION
Before looking at the IFE results, one should
Performance of immunoxation review the comparison protein electrophoresis gel
on the sample. Any suspicious band located any-
When performing IFE, a separate sample of where from the a- to the g-region should be
serum, appropriately diluted, is applied to the gel noted and the IFE must resolve its identity. If the
for each immunoglobulin or other antigen to be antisera used do not identify a band seen in the
assayed (Fig. 3.17). For example, although the comparison protein electrophoresis lane, one
typical IFE uses antisera against IgG, IgA, IgM, k, must consider additional steps: altered dilution of
and l, antisera against other antigens such as IgE, the antisera (addressing the possibility of antigen
IgD, brinogen, or C-reactive protein may be use- excess) or use of antisera against other proteins.
ful for identifying unusual bands in specic Other reagents used include antisera against IgE
samples (see Chapter 6). A sample of the patients or IgD, brinogen, C3, C-reactive protein
serum should also be placed in one lane for a depending on the position of the restriction and
comparison protein electrophoresis. Some manu- the clinical situation. Another approach that is
facturers have come up with ways to streamline sometimes used involves pretreatment of the
the initial screening. By using a pentavalent sample with 2-mercaptoethanol. Before IFE
(Penta) antisera (reactive with IgG, IgA, IgM, k became so well standardized, I would occasion-
48 Immunoxation, immunosubtraction, and immunoselection techniques
Anode +
Albumin
Origin
Gamma
SPE G A M K L
5. Wash and stain to 4. Fix first lane,
demonstrate overlay others
specific bands with antisera
Figure 3.17 Schematic overview of immunoxation electrophoresis. The dilution of the patients serum for each lane must be selected
(Step 1) by using one of four common methods for doing this listed AD and discussed in the text. After applying the serum to the
origin (Step 2), the proteins are separated by electrophoresis (Step 3) (in this example, the anode is at the top) and then (Step 4), acid
(SPE lane), or specic antisera (anti-G, A, M, k or l) are placed on each lane to precipitate the protein. In Step 5, the gels are washed and
stained to reveal, in this case, an IgG l monoclonal protein. Note that ideally, one should see diffusely staining bands at the appropriate
locations for the other analytes. This is achieved by appropriate dilutions of the patients serum. If a lane is empty, one would not be
certain if the correct antisera (or any antisera) was added, if the patients sample was added, or if there is just too little protein because
too large a dilution of the patients serum was used. HRE, high-resolution electrophoresis.
ally purify IgG and IgM with charge or sizing that the monoclonal protein lines up with the sus-
columns. With currently available IFE and pected band because other proteins, such as b-
immunosubtraction (ISUB; see below) techniques, rinogen, genetic variants (C3, transferrin) or such
I have not needed to use column purication for as C-reactive protein can produce a suspicious
several years. band in protein electrophoresis.47 When two or
The M-protein must have the same electro- more small bands are detected in the g-region, one
phoretic mobility in the IFE gel as it does in the is usually dealing with an oligoclonal expansion in
protein electrophoresis lane. Compare the migra- a patient with an infectious disease, an auto-
tion across the gel as this identies the position of immune disease, or some lymphoproliferative
the suspected monoclonal protein. It is important processes (see Chapters 4 and 6) with polyclonal
Immunoxation electrophoresis 49
Penta
(a)
(b)
Figure 3.18 (a) This capillary zone electrophoresis pattern contains a large g-region spike. The insert with the result of a Penta
(pentavalent) immunoxation is shown. On the left is the patients serum xed with acid. On the right is the result of the immunoxation
with antipentavalent serum. This immunoxation demonstrates that the spike is caused by an immunoglobulin. To characterize the
immunoglobulin, further studies need to be performed. With such a large g-spike, I usually go straight to immunosubtraction or
immunoxation. However, with subtle b-region restrictions, the conrmation that it is due to an immunoglobulin can save us from
performing immunoxation on a radiocontrast dye. (b) Overview of an entire Penta Gel. Each number represents the serum protein
electrophoresis pattern and the number prime to its right is the Penta immunoxation on that serum. For this gel, M-proteins that require
a more complete characterization are found in cases 2, 7, 8, 9 and 11. Case number 3 is arguable, but I would also perform a
characterization of that one. (Sebia Penta gel.)
expansion of immunoglobulins (Fig. 3.19). Small and B-cell neoplasms as opposed to the diffuse
bands on IFE can occur as artifacts. Cryoglobulins increase of immunoglobulins that accompanies
can produce small, artifactual bands at the origin polyclonal processes in many infections and
(see Chapter 6). Some commercial antibodies autoimmune diseases.
against human immunoglobulins contain antibod-
ies that react with normal b- or g-region compo-
nents. We have seen reactivity against several such
bands masquerading as monoclonal proteins Table 3.3 Reactivities of anti-immunoglobulin reagents that can
(Table 3.3).47 produce discrete bands resembling monoclonal gammopathies
Biclonal gammopathies occur uncommonly and
can be readily diagnosed by IFE (Figs. 3.20 and Additional reactanta
3.21). They are usually easily distinguished from
the polyclonal process, which has tiny oligoclonal Fibrinogen
bands shown in Fig. 3.19. The monoclonal pro- Transferrin
teins resulting from myeloma or B-cell lympho- C3
proliferative disorders are usually present in a
C4
greater concentration than oligoclonal bands
found in polyclonal processes. Furthermore there is a
Reactivities of commercial antisera with specicity for an
often an accompanying decrease in concentration immunoglobulin.
of the polyclonal immunoglobulins in myeloma
50 Immunoxation, immunosubtraction, and immunoselection techniques
L
Figure 3.19 Immunoxation of serum from a patient with
pneumonia who had a few (oligoclonal) bands in the g-region. The
serum protein electrophoresis lane at the top shows increased a1-
antitrypsin band, increased a1a2 interzone (I ), hemoglobin- Figure 3.20 Serum protein electrophoresis (top lane) of this
haptoglobin complex (H ) due to hemolysis during sample sample shows lightly staining a1- and a2-regions indicating that an
preparation and several tiny g-region bands. Immunoxation shows inammatory response is not likely. Two distinct g-bands (arrows)
that the bands are both k (K) and l (L), therefore polyclonal. are identied as IgG k by immunoxation electrophoresis. The
Some of the small, round, clear areas seen best in the anti-IgG and two bands may reect a monoclonal protein that forms monomers
anti-IgM reaction are caused by air bubbles that prevent the and dimers, post-translational modication of a monoclonal
precipitin reaction from occurring. Dilutions used: IgG, 1:15; IgA, protein or may be a true double (biclonal) gammopathy. In this
1:3; IgM, 1:2; k, 1:10; l, 1:4. (Panagel system stained with case, subclass determination showed that the two bands were of
Coomassie Blue; anode at the left.) different subclasses of IgG (a double gammopathy). Unlike Fig.
3.19, which shows a diffuse increase in the IgG proteins caused by
an infectious process, there is a relative hypogammaglobulinemia in
this IgG lane. The origin artifact (indicated) in the IgA and IgM
Limitations of immunoxation reactions is seen in some monoclonal proteins, which tend to self-
aggregate, but may also be seen with cryoglobulins or in normal
As mentioned above, a key problem to avoid is the samples with a polyclonal increase in g-globulins. The origin artifact
antigen excess effect. When a large amount of is most obvious in the IgA and IgM reactions here because the
monoclonal protein is present, an antigen excess serum samples were applied undiluted to the gel (since the IgA and
IgM concentrations were low). A slight antigen excess effect (x) is
effect can be seen (Fig. 3.22). In antigen excess, the
seen in one of the k bands. Note also that no precipitate was seen
immune complexes formed are small and wash with the anti-l reagent, likely due to too large a dilution of the
away during the washing steps. patients serum being used in this lane. Dilutions used: IgG, 1:20;
The lack of within-gel internal controls is a IgA, neat; IgM, neat; k (K), 1:12; l (L), 1:10. (Panagel system
potential source of error. Immunoelectrophoresis stained with Coomassie Blue; anode at the left.)
Immunoxation electrophoresis 51
G
M
K A
G
M
IMMUNOSUBTRACTION
Immunosubtraction was rst used as an adjunct to
K
gel-based protein electrophoresis. In 1977, Aguzzi
and Poggi48 reported a method to precipitate spe-
cic individual proteins by the use of monospecic
antisera prior to performing protein electro-
L
phoresis on cellulose acetate. They noted that by
removing specic bands or zones (immunosub-
tracting them) from the electrophoretic pattern it
improved their ability to interpret the nal
product. A few years later Merlini et al.49 passed
plasma samples through a layer of monospecic
Figure 3.25 Serum protein electrophoresis lane (top) shows a
faint brinogen band (F) just anodal to the origin. The anti-IgA
and anti-IgM reactions show a faint but distinct band (indicated)
due to reactivity of these commercial antisera with a b-region
migrating protein (possibly C4 or a complement breakdown
product). It was not identied. Minor reactivities such as these
usually were too small to be noticed with immunoelectrophoresis.
They can be controlled for by testing reagents with a normal
serum prior to use with patient samples. Dilutions used: IgG, 1:10;
IgA, 1:2; IgM, neat; k (K), 1:6; l (K), 1:3. (Panagel system stained
with Coomassie Blue; anode at the left.)
antisera to subtract out certain proteins. They and light chain components. As currently per-
applied this technique to prealbumin, albumin, formed on the Paragon CZE 2000, it is an entirely
a-lipoprotein, a1-antitrypsin, Gc-globulins, hands-free operation. A prepackaged container
a2-macroglobulin, haptoglobin, bronectin, with the antisera is placed on the instrument along
transferrin, b-lipoprotein, C3 and brinogen to with the patient sample to be tested. The instru-
demonstrate protein polymorphisms. The tech- ment mixes the serum with the reagents and fol-
nique was used to detect M-proteins by White and lowing an incubation period aspirates the sample
Attwood50 who found that it may be more sensitive into the capillary for protein electrophoresis.
than IEP. Removal of the M-protein peak identies its
In the past decade, the high-quality resolution of immunoglobulin composition52 (Fig. 3.28).
capillary zone immunosubtraction together with Because the analysis is performed in all six capil-
its ability for automation has found its way into laries simultaneously, the entire procedure takes
several high-volume clinical laboratories. under 10 min.
Currently, it is available on the Paragon CZE 2000 How well does ISUB compare with IFE for
technique as an automated module. For this, the accuracy and sensitivity? ISUB is convenient and
patients serum containing an M-protein is pre- when the M-protein is readily identiable in the
incubated in individual wells containing Sepharose CZE, ISUB should demonstrate the same heavy
beads coated with antibodies against one of the and light chain types as IFE. However, there are
following: anti-IgG, anti-IgA, anti-IgM, anti-k or some problems. Immunoxation electrophoresis is
anti-l. After a short incubation period, the beads slightly more sensitive than ISUB because ISUB
settle and the supernatant of each well is sampled can only deal with a band that one can remove.
for electrophoresis.51 As shown in Fig. 3.27, the M- Henskens et al.53 found that classication of four
protein is removed by the antisera against its heavy IgG monoclonal components was more readily
Immunosubtraction
A K
G K A L + A L G K
G K A K
G K G K G K G K
G K
M L G K Anti-K M L G K Anti-K
G K
G K M K G K
G K G L G K G K
G L
M K G K
G L G K G L
Figure 3.27 Schematic representation of one immunoxation lane. On the top, an arrow points to the prominent g-region spike in the
capillary zone electrophoresis pattern. Directly below is a diagram using block gures for the immunoglobulin molecules in this patients
serum. Beads are added to this serum that are coated with specic antisera. In this case the bead is coated with anti-k. On the top right,
after immunosubtraction with anti-k, the spike is gone. Below, the k-containing molecules have all bound to the beads (one shown),
leaving on the l-containing immunoglobulins in the serum. Note that the beads bind not only the monoclonal k immunoglobulins, but all
k-containing immunoglobulins.
Immunosubtraction 55
Modified immunoselection
Electro
Anti-g wash
stain
g Heavy chain
Normal IgG
Anti-k
& l 30 Pt sample
Figure 3.29 A modied immunoselection technique requires that specic antisera against both k and l light chain (anti-k & l) are
applied to the agarose just anodal to the origin 30 min prior to application of the patients serum (left). Just before application of the
patients serum, antiserum to the specic heavy chain is placed more anodal on the gel (center). After electrophoresis, any intact
immunoglobulins will precipitate near the well because it will react with the anti-k & l present in the gel (Normal IgG). Free heavy chains
can diffuse through this area containing anti-k & l and eventually precipitates with the anti-IgG (g Heavy Chain).
heavy chain of interest is applied to a more anodal immunoglobulin containing either k or l chains
portion of the gel. The patients sample is applied would precipitate near the origin. Any free heavy
in a well at the origin and the serum is electro- chains are able to migrate toward the anode where
phoresed. As with ISEL in the IEP format, any they precipitate as a second arc (Fig. 3.30).
1 1
2 2
3 3
4 4
5 5
6 6
7 7
Anti- Buffer Anti- Buffer Anti- Anti-
k&l k&l k&l k&l
Control Control Control Patient Patient Myeloma
Figure 3.30 Immunoselection on a case of g-heavy chain disease. On the bottom are listed the samples placed in each well (control
serum, patient with g-heavy chain disease and a myeloma patient that had an IgG k monoclonal gammopathy). Just above the contents of
the well are listed the reagents applied near the origin (anti-k & l, or buffer). At the top of the gel are listed the reagents applied at the
more anodal end of the gel (anti-k & l , or anti-g). Note only the lane that had the patients serum treated around the origin with anti-k &
l and treated toward the anode with anti-g has the band characteristic of g-heavy chain disease (arrow).
58 Immunoxation, immunosubtraction, and immunoselection techniques
13. Sun T, Lien YY, Degnan T. Study of 24. Kyle RA, Rajkumar SV. Monoclonal
gammopathies with immunoxation gammopathies of undetermined signicance.
electrophoresis. Am J Clin Pathol 1979;72: Hematol Oncol Clin North Am.
511. 1999;13:11811202.
14. Kahn SN, Bina M. Sensitivity of 25. Bajetta E, Gasparini G, Facchetti G, Ferrari L,
immunoxation electrophoresis for detecting Giardini R, Delia D. Monoclonal gammopathy
IgM paraproteins in serum. Clin Chem 1988;34: (IgM-k) in a patient with Burkitts type
16331635. lymphoblastic lymphoma. Tumori
15. Normansell DE. Comparison of ve methods for 1984;70:403407.
the analysis of the light chain type of monoclonal 26. Braunstein AH, Keren DF. Monoclonal
serum IgM proteins. Am J Clin Pathol gammopathy (IgM-kappa) in a patient with
1985;84:469475. Burkitts lymphoma. Case report and literature
16. Lane JR, Bowles KJ, Normansell DE. Detection review. Arch Pathol Lab Med 1983;107:235238.
of IgM monoclonal proteins in serum enhanced 27. Cesana C, Klersy C, Barbarano L, et al.
by removal of IgG. Lab Med 1985;16: Prognostic factors for malignant transformation
676678. in monoclonal gammopathy of undetermined
17. Prokesova L. Study of properties of structural signicance and smoldering multiple myeloma.
subunits of IgM immunoglobulin obtained by J Clin Oncol 2002;20:16251634.
reduction with 2-mercaptoethanol or by 28. Steck AJ, Murray N, Meier C, Page N,
oxidative sulphitolysis. Folia Microbiol Perruisseau G. Demyelinating neuropathy and
1969;14:8288. monoclonal IgM antibody to myelin-associated
18. Herrlinger JD, Kriegel W. Effect of glycoprotein. Neurology 1983;33:1923.
D-penicillamine and 2-mercaptoethanol on 29. Dalakas MC, Engel WK. Polyneuropathy with
human IgM in normal serum. Z Rheumatol monoclonal gammopathy: studies of 11 patients.
1976;35:108112. Ann Neurol 1981;10:4552.
19. Orr KB. Use of 2-mercaptoethanol to facilitate 30. Driedger H, Pruzanski W. Plasma cell neoplasia
detection and classication of IgM abnormalities with peripheral polyneuropathy. A study of ve
by immunoelectrophoresis. J Immunol Methods cases and a review of the literature. Medicine
1979;30:339347. (Baltimore) 1980;59:301310.
20. Capel PJ, Gerlag PG, Hagemann JF, Koene RA. 31. Lee YC, Came N, Schwarer A, Day B.
The effect of 2-mercaptoethanol on IgM and IgG Autologous peripheral blood stem cell
antibody activity. J Immunol Methods transplantation for peripheral neuropathy
1980;36:7780. secondary to monoclonal gammopathy of
21. Sorensen S. Monoclonal IgM kappa with unknown signicance. Bone Marrow Transplant
rheumatoid factor activity and 2002;30:5356.
cryoprecipitability identied only by 32. Gorson KC, Ropper AH, Weinberg DH,
immunoxation electrophoresis after Weinstein R. Efcacy of intravenous
2-mercaptoethanol treatment. Clin Chim Acta immunoglobulin in patients with IgG
1988;173:217224. monoclonal gammopathy and polyneuropathy.
22. Kyle RA. Benign monoclonal gammopathy Arch Neurol 2002;59:766772.
after 2035 years of follow-up. Mayo Clin Proc 33. Kvarnstrom M, Sidorova E, Nilsson J, et al.
1993;68:2636. Myelin protein P0-specic IgM producing
23. Kyle RA, Therneau TM, Rajkumar SV, et al. A monoclonal B cell lines were established from
long-term study of prognosis in monoclonal polyneuropathy patients with monoclonal
gammopathy of undetermined signicance. gammopathy of undetermined signicance
N Engl J Med 2002;346:564569. (MGUS). Clin Exp Immunol 2002;127:255262.
60 Immunoxation, immunosubtraction, and immunoselection techniques
34. Fisher MA, Wilson JR. Characterizing 44. Schreiber WE, Chiang E, Tse SS. Electrophoresis
neuropathies associated with monoclonal underestimates the concentration of polyclonal
gammopathy of undetermined signicance immunoglobulins in serum. Am J Clin Pathol
(MGUS): a framework consistent with classifying 1992;97:610613.
injuries according to ber size. Neurol Clin 45. Bush D, Keren DF. Over- and underestimation
Neurophysiol 2002;3:27. of monoclonal gammopathies by quantication
35. Janik B. Identication of monoclonal proteins of kappa- and lambda-containing
by immunoxation. Electrophor Today immunoglobulins in serum. Clin Chem
1981;2:14. 1992;38:315316.
36. Cawley LP, Minard BJ, Tourtellotte WW, Ma BI, 46. Su L, Keren DF, Warren JS. Failure of anti-
Chelle C. Immunoxation electrophoretic lambda immunoxation reagent mimics alpha
techniques applied to identication of proteins in heavy-chain disease. Clin Chem 1995;41:
serum and cerebrospinal uid. Clin Chem 121123.
1976;22:12621268. 47. Register LJ, Keren DF. Hazard of commercial
37. Ritchie RF, Smith R. Immunoxation. I. antiserum cross-reactivity in monoclonal
General principles and application to agarose gel gammopathy evaluation. Clin Chem
electrophoresis. Clin Chem 1976;22: 1989;35:20162017.
497499. 48. Aguzzi F, Poggi N. Immunosubtraction
38. Ritchie RF, Smith R. Immunoxation. II. electrophoresis: a simple method for identifying
Application to typing of alpha1-antitrypsin at specic proteins producing the cellulose acetate
acid pH. Clin Chem 1976;22:17351737. electrophoretogram. Boll Ist Sieroter Milan
39. Richard S, Miossec V, Moreau JF, Taupin JL. 1977;56:212216.
Detection of oligoclonal immunoglobulins in 49. Merlini G, Pavesi F, Carini A, Zorzoli I,
cerebrospinal uid by an immunoxation- Valentini O, Aguzzi F. Identication of specic
peroxidase method. Clin Chem 2002;48: plasma proteins determining the agarose gel
167173. electrophoresis by the immunosubtraction
40. Monos DS, Bina M, Kahn SN. Evaluation and technique. J Clin Chem Clin Biochem
optimization of variables in immunoxation 1983;21:841844.
electrophoresis for the detection of IgG 50. White WA, Attwood EC. Immunoxation and
paraproteins. Clin Biochem 1989;22: immunosubtraction on agarose gel: an aid in the
369371. typing of monoclonal gammopathies. Ann Clin
41. Hadler MB, Kimura EY, Leser PG, Kerbauy J. BioChem 1984;21:467470.
Standardization of immunoxation technique for 51. Bienvenu J, Graziani MS, Arpin F, et al.
the detection and identication of serum Multicenter evaluation of the Paragon CZE 2000
paraproteins. Rev Assoc Med Bras capillary zone electrophoresis system for serum
1995;41:119124. protein electrophoresis and monoclonal
42. Keren DF, Di Sante AC, Bordine SL. component typing. Clin Chem 1998;44:599605.
Densitometric scanning of high-resolution 52. Klein GL, Jolliff CR. Chapter 16. Capillary
electrophoresis of serum: methodology and electrophoresis for the routine clinical
clinical application. Am J Clin Pathol laboratory. In: Landers JP, ed. Handbook of
1986;85:348352. capillary electrophoresis. Boca Raton: CRC
43. Uriel L. Interpretation quantitative des resultats Press, 1993.
aprs electrophorese en gelose, 1. Considerations 53. Henskens Y, de Winter J, Pekelharing M, Ponjee
gnrales, application a ltude de constituants G. Detection and identication of monoclonal
proteiques isols. Clin Chim Acta gammopathies by capillary electrophoresis. Clin
1958;3:234238. Chem 1998;44:11841190.
References 61
54. Oda RP, Clark R, Katzmann JA, Landers JP. 64. Sun T, Peng S, Narurkar L. Modied
Capillary electrophoresis as a clinical tool for immunoselection technique for denitive diagnosis
the analysis of protein in serum and other of heavy-chain disease. Clin Chem 1994;40:664.
body uids. Electrophoresis 1997;18: 65. Radl J, Wels J, Hoogeveen CM. Immunoblotting
17151723. with (sub)class-specic antibodies reveals a high
55. Litwin CM, Anderson SK, Philipps G, Martins frequency of monoclonal gammopathies in
TB, Jaskowski TD, Hill HR. Comparison of persons thought to be immunodecient. Clin
capillary zone and immunosubtraction with Chem 1988;34:18391842.
agarose gel and immunoxation electrophoresis 66. Gerritsen E, Vossen J, van Tol M, Jol-van der
for detecting and identifying monoclonal Zijde C, van der Weijden-Ragas R, Radl J.
gammopathies. Am J Clin Pathol Monoclonal gammopathies in children. J Clin
1999;112:411417. Immunol 1989;9:296305.
56. Katzmann JA, Clark R, Sanders E, Landers JP, 67. Radl J. Monoclonal gammapathies. An attempt
Kyle RA. Prospective study of serum protein at a new classication. Neth J Med
capillary zone electrophoresis and immunotyping 1985;28:134137.
of monoclonal proteins by immunosubtraction. 68. Radl J, Liu M, Hoogeveen CM, et al.
Am J Clin Pathol 1998;110:503509. Monoclonal gammapathies in long-term
57. Bossuyt X, Bogaerts A, Schiettekatte G, surviving rhesus monkeys after lethal irradiation
Blanckaert N. Detection and classication of and bone marrow transplantation. Clin Immunol
paraproteins by capillary Immunopathol 1991;60:305309.
immunoxation/subtraction. Clin Chem 69. Radl J, Valentijn RM, Haaijman JJ, Paul LC.
1998;44:760764. Monoclonal gammapathies in patients
58. Keren DF. Detection and characterization of undergoing immunosuppressive treatment after
monoclonal components in serum and urine. Clin renal transplantation. Clin Immunol
Chem 1998;44:11431145. Immunopathol 1985;37:98102.
59. Vrethem M, Larsson B, von Schenck H, 70. Norden AG, Fulcher LM, Heys AD. Rapid
Ernerudh J. Immunoxation superior to plasma typing of serum paraproteins by immunoblotting
agarose electrophoresis in detecting small M- without antigen-excess artifacts. Clin Chem
components in patients with polyneuropathy. 1987;33:14331436.
J Neurol Sci 1993;120:9398. 71. Nooij FJ, van der Sluijs-Gelling AJ, Jol-Van der
60. Seligmann M, Mihaesco E, Hurez D, Mihaesco Zijde CM, van Tol MJ, Haas H, Radl J.
C, Preudhomme JL, Rambaud JC. Immunoblotting techniques for the detection of
Immunochemical studies in four cases of alpha low level homogeneous immunoglobulin
chain disease. J Clin Invest 1969;48: components in serum. J Immunol Methods
23742389. 1990;134:273281.
61. Radl J. Light chain typing of immunoglobulins in 72. Withold W, Reinauer H. An immunoblotting
small samples of biological material. procedure following agarose gel electrophoresis
Immunology 1970;19:137149. for detection of Bence Jones proteinuria
62. Al-Saleem TI, Qadiry WA, Issa FS, King J. The compared with immunoxation and quantitative
immunoselection technic in laboratory diagnosis light chain determination. Eur J Clin Chem Clin
of alpha heavy-chain disease. Am J Clin Pathol Biochem 1995;33:135138.
1979;72:132133. 73. Withold W, Rick W. An immunoblotting
63. Seligmann M, Mihaesco E, Preudhomme JL, procedure following agarose gel electrophoresis
Danon F, Brouet JC. Heavy chain diseases: for subclass typing of IgG paraproteins in human
current ndings and concepts. Immunol Rev sera. Eur J Clin Chem Clin Biochem
1979;48:145167. 1993;31:1721.
62 Immunoxation, immunosubtraction, and immunoselection techniques
74. Beaume A, Brizard A, Dreyfus B, Preudhomme 75. Musset L, Diemert MC, Taibi F, et al.
JL. High incidence of serum monoclonal Igs Characterization of cryoglobulins by
detected by a sensitive immunoblotting technique immunoblotting. Clin Chem 1992;38:798802.
in B-cell chronic lymphocytic leukemia. Blood
1994;84:12161219.
4
Proteins identified by serum
protein electrophoresis
The protein bands identied by serum protein clinical situations. The exact position of some of
electrophoresis may be divided into major and the bands will vary slightly with the methodology
minor protein bands (Tables 4.1 and 4.2). The employed.
major bands constitute those that are virtually
always seen in normal serum (brinogen is
included because it is visualized in various cir- TRANSTHYRETIN (PREALBUMIN)
cumstances). Minor bands are those that stain
weakly or not at all in normal serum but may Transthyretin, a 55 kDa protein, is the rst band
affect the electrophoretic pattern in a variety of encountered from the anodal side of the
a
Concentrations are for adults. They are expressed as g/dl for albumin and in mg/dl for all other proteins.
b
Concentration in plasma.
64 Proteins identied by serum protein electrophoresis
electrophoresis strip.1 Its concentration in serum is because it is synthesized locally in the choroid
2040 mg/dl. Structurally, transthyretin is a sym- plexus.5
metrical tetramer composed of four monomers Transthyretin has become a mainstay in assessing
each 13.7 kDa and 127 amino acids long.2 It is syn- the nutritional status of patients. When using
thesized mainly in the liver and it provides transthyretin levels to follow nutritional status,
transport for about 20 per cent of serum thyroxin quantication should be performed by nephelome-
(T4) (each molecule of transthyretin combines with try or radial immunodiffusion, not electrophoresis.
one molecule of T4).3 Together with retinol- It has advantages over albumin as a monitor for
binding protein, transthyretin also acts as a carrier protein-calorie malnutrition because albumin has a
for vitamin A.2 Its older name, prealbumin, merely relatively long half-life (3 weeks, compared with
referred to its position just anodal to albumin. the 2-day half-life of transthyretin). Therefore,
Transthyretin is the new name for this protein that transthyretin is a more sensitive indicator of
reects its roles transporting thyroxin and retinol- change in protein-calorie status.6,7 It is not,
binding protein.4 however, a perfect test for malnutrition because
By most electrophoretic procedures on serum, transthyretin levels are low in patients with severe
transthyretin produces only a weak, diffuse band liver disease.
just proximal to albumin (Fig. 4.1). The best reso- Rarely, one can note a marked increase in the
lution of transthyretin in serum from currently transthyretin band. This has been reported in
available techniques is that provided by capillary patients with inammatory bowel disease. In an
zone electrophoresis (CZE) (Fig. 4.2). However, extreme case of a patient with a long history of
even with CZE, the transthyretin band is too small diarrhea complicated with hepatitis B, Jolliff8
for useful monitoring of any clinical condition. noted that it could be confused with bisalbumine-
Although it is present in lower concentration in mia (a matter that can be cleared up by
CSF than it is in serum (Fig. 4.1), transthyretin is a immunoxation) (Fig. 4.3).
more substantial fraction in CSF than serum Structurally, transthyretin has a high content of
Albumin 65
Figure 4.1 Cerebrospinal uid (C) and serum (S) samples are try, these variants may be characterized in cere-
alternated on this gel. The cerebrospinal uid has been
brospinal uid.12
concentrated 80-fold, whereas the serum has been diluted 1:3.
Note the prominence of the transthyretin band in the
cerebrospinal uid samples. The transthyretin band in the serum is
barely visible in the top and bottom sera (arrows), but is not ALBUMIN
visible (although it is present) in the middle serum. (Paragon SPE2
system stained with Paragon Violet).
Albumin is a 69 kDa protein with a concentration
of 3.55.0 g/dl (3550 g/l) in adults 1860 years
old. It is the most prominent protein in normal
serum. Similar to transthyretin, albumin is synthe-
b-sheets with only one small a-helix by X-ray sized in the liver and functions as a transport
crystallography.1 The predominance of b-sheet protein. Albumin accounts for much of the osmotic
structure may be related to the propensity of effect of serum proteins and transports a variety of
familial amyloidosis to occur with minor struc- endogenous and exogenous molecules, including
tural changes to transthyretin. Transthyretin bilirubin, enzymes, hormones, lipid, metallic ions,
variants are associated with two types of amyloi- and drugs. Many of these molecules are poorly
dosis: senile systemic amyloidosis (SSA) and soluble in aqueous solution alone. The breadth of
familial amyloidotic polyneuropathy (FAP).9,10 the normal albumin band is partly due to its great
Familial amyloidotic polyneuropathy is the most serum concentration and partly due to the micro-
common form of inherited amyloidosis with an heterogeneity resulting from the charges and size of
incidence of 1 in 100 000.10 it may also be various molecules transported by albumin.13 The
caused by a genetic variant of apolipoprotein tendency for albumin to transport a variety of sub-
A1.11 Genetic variants of transthyretin associated stances accounts for some of the abnormal patterns
with FAP cannot be identied by serum protein of electrophoresis in patients receiving albumin-
electrophoresis. However, using mass spectrome- binding drugs (such as antibiotics, especially
66 Proteins identied by serum protein electrophoresis
(a)
1
Decreased albumin
2 Decreased concentration of serum albumin indi-
cates signicant pathology either in the produc-
tion of albumin by the liver or its leakage through
3 a damaged surface (glomerular disease, gastro-
intestinal loss, or thermal injury) (Table 4.3). In
Western countries, a decrease in the production of
albumin most commonly reects severe liver
4 injury. Because of the large reserve capacity of the
liver, hypoalbuminemia resulting from liver dam-
age occurs after most of the hepatocytes have been
5 damaged or destroyed. Such a decrease may be
accompanied by clotting abnormalities and
decreased synthesis of other hepatocyte products,
including haptoglobin and transferrin.19 Mendler
6 et al.20 recommend that serum albumin concentra-
tions should be performed by nephelometric tech-
niques rather than serum protein electrophoresis
7 in these patients.
In underdeveloped countries, severe protein mal-
(b)
nutrition (kwashiorkor) is the leading cause of
Figure 4.3 (a) Serum from patient with prominent transthyretin decreased synthesis of albumin. Patients with neo-
band (arrow) is on top and a normal control is on the bottom. (b)
plasia or other chronic diseases also develop a
Serum protein electrophoresis (SPE) and immunoxation of serum
from patient in (a) demonstrates that the fast-moving band reacts
nutritionally related hypoalbuminemia. Despite
with anti-transthyretin (aT). Anode at the top. Figures contributed the decreased serum albumin band on serum
by Carl R. Jolliff. protein electrophoresis in most cases of protein-
Albumin 67
Bisalbuminemia or alloalbuminemia
Another inherited abnormality of albumin is bis-
albuminemia, in which two types of albumin that
have slightly different electrophoretic mobility are
(a) produced; this results in two distinct and equal
peaks if the variant produces albumin at the same
rate as the normal gene. Occasionally, two distinct
peaks of different heights are seen when the variant
gene produces less albumin35 (Fig. 4.6). Variant
albumins may also alter binding of drugs such as
warfarin.36 Capillary zone electrophoresis is supe-
rior to gel-based techniques in demonstrating this
nding (Fig. 4.7).37 Some variants of albumin will
migrate anodally to the normal position and others
will migrate cathodally; the albumin variant
A depicted in Fig. 4.6 migrates cathodally to the nor-
mal albumin. Clinically, bisalbuminemia (alloalbu-
minemia) has no pathological consequence.38 The
Fraction Rel %
Albumin 4.2 incidence varies depending on the population with
Alpha 1 9.8 +++ a range of 1:10001:10 000 in Caucasians and
Alpha 2 33.6 +++
Beta 16.9 +++ Japanese, but with a frequency as high as about 1
Gamma 35.6 +++ per cent in some American Indian tribes.39,40
A/G: 0.04
(b)
Figure 4.5 (a) Serum from a normal individual (top) is compared
Increased albumin
with serum from a patient with analbuminemia (bottom). (b)
Densitometric scan from analbuminemic patient. Only minor An elevated albumin indicates acute dehydration
background distortions are seen in the albumin area (a). Figures and is accompanied by an increase in the other
contributed by Carl R. Jolliff. serum proteins. Albumin has been used for many
a-Region 69
(a) (b)
Figure 4.6 (a) The third sample from the top is a case of bisalbuminemia. Note that the two bands did not separate completely on this
gel. The fact that there are two bands present can be noted by the indentation (arrow) between the two bands. (Paragon SPE2 system
stained with Paragon Violet.) (b) Densitometric scan of the case of bisalbuminemia from (a). Note the two peaks atop the broadened
albumin band.
Reference ranges
Rel % g/dl
ALBUMIN 52.6 68.9 3.80 5.20
ALPHA 1 3.6 8.1 0.30 0.60
ALPHA 2 5.3 12.2 0.40 0.90
BETA 8.3 14.3 0.60 1.10
GAMMA 8.2 18.6 0.60 1.40
Figure 4.8 Normal albumin (but increased relative per cent of albumin) with decrease in all other bands. This serum is from a patient
that received plasmapheresis for a neuropathy and had albumin replacement. (Paragon CZE 2000.)
out an a1-antitrypsin deciency. In those cases, one fatty acids from triglycerides.48 The free fatty acids
must use an alternative electrophoretic technique bind to a1-lipoprotein and result in the anodal
(see below), or at least measure the amount of a1- migration of this band. The clinical laboratory can
antitrypsin by immunoassay to detect any de- take advantage of the effect of free fatty acids to
ciency. Even with quantication, however, an alter the migration of a1-lipoprotein in cases where
a1-antitrypsin deciency may be missed. Some the density of this region obscures the view of a1-
patients with a1-antitrypsin deciency can have antitrypsin. For this method, my laboratory
borderline low normal amounts of a1-antitrypsin prepares a 17.2 mmol/l solution of lauric acid in
during an acute-phase reaction. Therefore, a C- the standard electrophoresis buffer.50 The serum is
reactive protein assay should be run simultaneously diluted 1:4 in this lauric acid solution and routine
to rule out that possibility. electrophoresis is performed on the sample. This
The laboratory can remove the effect of a1- displaces the a1-lipoprotein to the prealbumin
lipoprotein by taking a lesson from the heparin region thus allowing a clear view of the a1-anti-
effect. Heparin alters the migration of the lipopro- trypsin band.
tein bands, but its effect depends on whether the
heparin was administered to the patient, or if the
tube to obtain the blood sample was heparinized. a1-Antitrypsin
When the patient is receiving heparin, it affects the
migration of both a1-lipoprotein and b1-lipoprotein Second in importance only to detection of mono-
(Figs 4.9 and 4.11),48 whereas, when heparin is clonal gammopathies is the detection of decreased
added to the blood sample from a non-heparinized and/or abnormal a1-antitrypsin. a1-Antitrypsin is
patient, only the migration of b1-lipoprotein is the major protein accounting for the discrete band
affected.49 in the cathodal end of the a1-region. Although it is
These different effects of heparin reect two represented by a relatively small band, in this text
processes. When heparin is added to a sample in it is classied as a major protein band because of
vitro, it binds directly to b1-lipoprotein, resulting in the clinical importance of abnormalities associated
a more diffuse and anodal migration.49 However, with this band. This 52 kDa, single-chain glyco-
in vivo, heparin activates lipoprotein lipase in protein has a normal concentration of
endothelial cells, resulting in the liberation of free 90200 mg/dl (0.92.0 g/l) in adults. It is produced
in the liver and is classied as a member of a family
of protease inhibitors called serpins (serine pro-
tease inhibitors) that react with proteolytic
enzymes containing a serine group at their active
site.51
Although its name implies that its functions are
limited to inhibiting the activity of trypsin, a1-
antitrypsin interferes with the enzymatic activity of
a variety of enzymes including trypsin, chymo-
trypsin, pancreatic elastase, skin collagenase,
renin, urokinase, Hageman-factor cofactor, and
leukocyte neutral proteases.52 It is by far the most
signicant protease inhibitor in serum, accounting
for the vast majority of the trypsin-inhibiting
capacity of human serum.53 a1-Antitrypsin works
Figure 4.10 Prominent a1-lipoprotein is present in this capillary like a mousetrap to inhibit proteases.54 It docks
zone electropherogram. (Paragon CZE 2000.) with the target protease, which cleaves the reactive
72 Proteins identied by serum protein electrophoresis
Figure 4.11 Immunoxation of serum from a patient receiving heparin. The anode is at the top of the gel. There is a faint diffuse band (b)
anodal to albumin (a). The extreme anodal edge of this band reacts with antisera against apolipoprotein A (a1-lipoprotein component)
(Apo A). Transthyretin (prealbumin) is more cathodal to this band (Pre-Alb). For comparison, antisera against apolipoprotein B (Apo B)
which reacts with a b1-lipoprotein component, a1-antichymotrypsin (A-1 CT) and a1-antitrypsin (A-1 AT) are shown. Anode at the top.
(Paragon Immunoxation stained with Paragon Violet.)
bands. We reported a coefcient variation of 10 M2, and M3, which produce normal levels of
per cent in this region.56 a1-antitrypsin.67 Combined, the M alleles have a
Capillary zone electrophoresis offers excellent gene frequency of about 0.94. The most cathodal
resolution in the a1-region. Because it detects molecule denoted by the letter Z (it is the zlowest)
peptide bonds, it is superior to gel-based tech- is the most common variant associated with severe
niques in detecting a1-acid glycoprotein (a heavily deciency. It results in a change from Glu to Lys at
glycated molecule that does not stain well with the position 342 of the M allele.68 About 3 per cent of
dyes used in gel-based methods). In addition, a1- the population in the USA are phenotype PiMZ,
lipoprotein and, in some CZE systems, which usually does not result in clinical deciency
triglycerides are also included in the a1-region. (because of the presence of the single PiM allele).
Because of this, Gonzalez-Sagrado et al.57 reported Fortunately, only one in 3630 individuals has the
that neither CZE nor gel-based methods provide severe decient PiZZ phenotype.69,70
an adequate method to rule out deciencies of a1- The second most common decient phenotype,
antitrypsin. They recommended performing referred to as S, migrates between M and Z on
nephelometry and phenotyping when the total a1- serum protein electrophoresis. The S gene has an
region (which includes all the proteins in this allelic frequency of 0.020.04. Individuals with
region) on CZE is less than 400 mg/dl (4 g/l). PiSZ also have clinically signicant deciency; this
Detection of a decreased a1-antitrypsin band, or occurs in about one in 500 individuals in the
one with altered migration, must be followed up by USA.71 The S mutation is a single base-pair substi-
genetic studies to provide prognostic information tution from Glu to Val at codon 264.72 Several
to the family.58,59 Because of interference with neph- other decient genes have been described.
elometric quantication of a1-antitrypsin levels by However, these are very rare and often difcult to
anticoagulants, serum should be used to measure detect by serum protein electrophoresis. Some have
the functional activity or the concentration by extraordinary symptoms such as in the case of
immune precipitation assays. However, either antitrypsin Pittsburgh (single base-pair substitu-
serum or plasma is suitable for phenotypic analysis tion from Met to Arg at codon 358), wherein the
by reference electrophoretic methods (isoelectric mutated protein serves as an inhibitor of thrombin
focus).60,61 Molecular techniques to detect both the (Fig. 4.12).73,74 One of the two cases of antitrypsin
Z and S mutations are now widely available.6266 Pittsburgh was associated with severe hemorrhage.
A complete discussion of the genetic polymor- Both cases were associated with a minor slow
phism (involving about 75 possible alleles) of
a1-antitrypsin is beyond the scope of this volume,
but some details are relevant to interpreting and
Table 4.4 Common a1-antitrypsin phenotypes
understanding the serum protein electrophoresis
patterns that one will come across in patients with
Genotype Electrophoretic migration
a1-antitrypsin deciency. Expression of the two
alleles of any individual is codominant, that is,
MMa Mid-a1
each allele controls production of a specic
a1-antitrypsin molecule unaffected by the other MS Mid- and slow a1
allele. The alleles are referred to as Pi (protease SS Slow a1
inhibitor) followed by a third letter characterizing MZ Mid- and slowest a1
the particular allele. For example, the a1-antit- SZ Slow and slowest a1
rypsin from the allele with the lowest ZZ zlowest a1
concentration (PiZ) has the slowest electrophoretic
mobility. The most common allele is PiM (Table a
Includes M1, M2, M3, M4 (types of M with normal activity).
4.4). There are several variants of this gene, M1,
74 Proteins identied by serum protein electrophoresis
albumin component (called proalbumin), which PiZZ product. The variant molecule of PiZZ is
migrated more toward the cathode than normal predisposed to polymerize resulting in disease due
(Fig. 4.12). to conformational change of the molecule.52 The
Genetic deciency of a1-antitrypsin is associated PiZZ and PiSZ genotypes may have hepatic dys-
with severe lung and, less commonly, liver disease. function as early as the rst three months of life.79
Some affected individuals develop neonatal hepati- However, the presentation even in individuals with
tis (those who do are highly likely to develop PiZZ is quite variable. Some cases of PiZZ do not
serious liver disease), others develop liver disease present until their seventh decade (at that time with
later as children, or in adult life they develop a both cirrhosis and emphysema).80 Others have
form of cirrhosis (about 20 per cent of patients), noted cases with no clinical symptoms until their
which is characterized by large globules of amor- eighth decade.81 As discussed above, this variability
phous periodic acidSchiff-positive material that implies that factors other than the genotype are
occurs within the cytoplasm of hepatocytes in the involved. Personal and environmental factors such
periportal areas (Fig. 4.13).7577 These globules are as smoking or living in an area with considerable
distinguished from glycogen because they are not air pollution have a negative effect on the lung
digested by treatment with diastase. Immuno- process.82 However, since only 1015 per cent of
histochemical studies have shown this material to individuals with PiZZ develop clinically signicant
be an a1-antitrypsin precursor that is not liver disease, a second factor relating to the ability
excreted.52,78 They represent aggregation of the of an individual to degrade these aggregates in
1 2 3 4 5 6 7 8 9 10 11
Figure 4.12 Agarose gel with the anode at the top. In lane 1, 2.5 ml of normal serum is shown to compare migration of albumin (note
location of the cathodal end) and a1-antitrypsin band (arrow). In lanes 2 and 7, 2.5 ml of plasma is shown from the original case of a1-
antitrypsin (Pittsburgh). Note the broader cathodal migration of the albumin band and the double a1-antitrypsin bands. In lanes 3, 6, and 8
are 2.5, 0.6, and 5.0 ml, respectively, of plasma samples from a second case of a1-antitrypsin (Pittsburgh). The cathodal migration of
albumin is apparent on all three, but the double a1-antitrypsin band is only seen to advantage in lane 8 (indicated). The abnormal albumin
was puried by diethylaminoethyl Sephadex column (lane 9) and trypsin treatment restored the normal migration of this band (lane 10). In
lane 11, a normal albumin is shown for comparison. (Figure contributed by Stephen O. Brennan.)74
a-Region 75
Figure 4.13 Liver from patient with a1-antitrypsin deciency (PiZZ) is depicted. The periodic acidSchiff (PAS), diastase stain, and
digestion disclose the large granules (arrows), which are especially prominent in the periportal hepatocytes.
hepatocytes may determine which patients develop disease (COPD) who also have PiMZ have a
cirrhosis.68 In addition to the more obvious associ- decreased forced expiratory volume (FEV1) com-
ation of PiZZ with cirrhosis, occasional cases of pared with PiMM individuals with COPD.
chronic liver failure have been reported in patients In examining electrophoretic patterns, the most
that are PiMZ or PiMS.83,84 common genetic variant seen in the laboratory is
The lung injury, resulting in early emphysema, the PiMS banding (Figs 4.14 and 4.15). As with the
has been hypothesized to result from the PiMZ heterozygotes, these patients have no clinical
unchecked endogenous activity of a variety of pro- disease, although family studies should be recom-
teolytic enzymes that are liberated by minor mended because siblings or children may carry two
inammatory events in these tissues.8587 Damage is defective genes. Another genetic variant likely to be
most severe at the base of the lung and is com- detected by examination of serum is reected by
pounded by environmental factors, especially the protein product that migrates anodal to PiM
smoking.68 However, even in non-smoking individ- and is termed PiF (fast). This genotype is not
uals homozygous for PiZZ, lung function will known to be associated with antitrypsin deciency.
decline, especially after 50 years of age.82,88 While serum protein electrophoresis cannot
Although heterozygotes are usually clinically well, detect some of the more unusual variants, it can be
Dahl et al.89 reported that individuals with clini- helpful in detecting the more common ones. When
cally established chronic obstructive pulmonary examining the electrophoretic pattern, one should
76 Proteins identied by serum protein electrophoresis
I
be especially aware of samples with decreased a1-
antitrypsin band and/or altered electrophoretic
migration. By noting that the weak a1-antitrypsin
band is located in the slow a1-region, it suggests the a
possibility of either PiZZ, PiSS or PiSZ phenotypes
(Figs 4.16 and 4.17).
Some clinically signicant variants such as
PiMMalton and PiMHeerlen require molecular techniques
to detect because they have a normal electro-
phoretic migration, but have a substantial decrease
in the a1-antitrypsin level and can cause liver and
lung disease.90,91 Some rare individuals are unable
to produce serum a1-antitrypsin because of dele- Figure 4.16 The second sample from the top is from a patient
tions and stop codons of a1-antitrypsin coding null with a1-antitrypsin deciency (PiZZ). No band can be seen in the
a1-antitrypsin area. A very faint a1-lipoprotein band (I) extends
variants.92 These patients are at high risk for the
from the cathodal end of the albumin band to the area where the
early development of emphysema even if they normal a1-antitrypsin band would normally be seen. For
do not smoke.93 Therefore, when there is altered comparison, the a1-antitrypsin PiMM band in the third sample is
electrophoretic mobility, or decrease or absence labeled a. (SPE2 system stained with Paragon Violet.)
a-Region 77
greatly elevated because its high sialic acid content protein (see below), increased a1-acid glycoprotein
interferes with binding of most stains such as levels are associated with an increased risk for
Amido Black or Coomassie Brilliant Blue. With myocardial infarction.110
gel-based techniques, when a1-acid glycoprotein is
increased to greater than 200 mg/dl it may show
up with a fuzzy appearance on the anodal side of a1-Antichymotrypsin
a1-antitrypsin.104 However, since CZE measures
peptide bond absorbance at from 200 to 214 nm, a1-Antichymotrypsin is a tiny band that may be
a1-acid glycoprotein more commonly is seen as a found in the a1a2 interregion (Fig. 4.19). It is
prominent anodal shoulder to a1-antitrypsin (Fig. about 1/10 as prominent as the a1-antitrypsin band
4.18). with a serum concentration of 3060 mg/dl, and a
a1-Acid glycoprotein is commonly seen with the molecular mass of 69 kDa. It is a serine protease
acute-phase reaction pattern and in uremic patients inhibitor genetically linked with a1-antitrypsin.111
receiving hemodialysis (this protein is normally a1-Antichymotrypsin reacts with the potent neu-
lost through the glomerulus).108 Quantication of trophil protease cathepsin G, mast cell chymase,
this protein may be useful in detecting neonatal and prostate specic antigen.112,113 Recent data for
infections because the serum levels of a1-acid the a1-antichymotrypsinprostate-specic antigen
glycoprotein are much lower at birth than by (PSA) complexes have been used to evaluate
1 year of age. Because of this, levels of modestly elevated PSA levels.114 However, this does
6080 mg/dl (68 g/l) have been associated with not affect the pattern visible on protein elec-
neonatal sepsis.109 These levels are too low, how- trophoresis. While its name implies a role in
ever, to be measured reliably by either CZE or gel- chymotrypsin inhibition, it provides less signicant
based techniques and should be determined by inhibition for chymotrypsin than does a1-anti-
immunochemical methods.109 Similar to C-reactive trypsin.115
In serum, its concentration increases rapidly after
acute injury, perhaps acting to inhibit some of the
enzymes liberated during this process, and it may
be seen in hepatocytes of individuals with hepatitis
C.116 It is increased in the serum and cerebrospinal
Figure 4.19 Small bands are often seen in the inter-a1a2 region.
In this photograph, they are best seen in the middle sample
(arrow). The serum in the bottom lane contains a slow-migrating,
Figure 4.18 Normal capillary zone electropherogram with a weakly staining a1-antitrypsin band (indicated) which should be
relatively prominent a1-acid glycoprotein (orosomucoid) shoulder further evaluated as a possible PiSS (see text). (Panagel System
(arrow) just anodal to a1-antitrypsin. (Sebia Capillarys.) stained with Amido Black.)
a-Region 79
uid of patients with Alzheimers disease, and the shown to bind to various cytokines, including
levels may be related to its heightened secretion by interleukin-2, interleukin-8, and heat shock pro-
astrocytes in those patients.117 By serum protein tein receptor CD91, indicating the diverse reper-
electrophoresis, one only sees a variable, deeper toire of this molecule.122124
staining in the a1a2 interregion. Obviously, a a2-Macroglobulin has a unique mechanism for
decreased a1-antichymotrypsin concentration trapping the proteinase in the bait region of the
would not be detectable by routine electrophoretic disulde-linked dimers. After the proteinase has
techniques. Congenital deciencies of this inhibitor cleaved a peptide bond in the middle of each of the
can be detected by immunoassay methods and may four a2-macroglobulin subunits (bait region), a2-
predispose individuals to liver and lung disease.113 macroglobulin rearranges itself to trap the pro-
teinase within the a2-macroglobulin molecule.125,126
The a2-macroglobulin bound to the proteinase
a2-Macroglobulin creates a conformational change that results in a
more compact a2-macroglobulin molecule that
a2-Macroglobulin is one of the two major proteins migrates faster during electrophoresis.125,126 Despite
in the a2-region. Synthesized mainly by the liver, it its effects on proteinases, a2-macroglobulin is not
is a tetramer composed of four identical subunits. an acute-phase reactant and does not increase with
The subunits are held together as dimers by di- inammation.
sulde bonds. Each a2-macroglobulin molecule The function of a2-macroglobulin as a pan-
consists of two sets of these dimers.118 With a serum proteinase inhibitor, together with its ability to
concentration of 130300 mg/dl (1.33.0 g/l) in bind with b-amyloid have raised the suggestion
adults, it accounts for about 3 per cent of the total that the a2-macroglobulin gene is a candidate gene
protein in serum.106 Because of the variable migra- in relation to Alzheimers disease.127129 However,
tion of the haptoglobin types (see below), this area of investigation is controversial because
a2-macroglobulin is often adjacent to, or comigrat- recent investigations have not conrmed this asso-
ing with, haptoglobin and is therefore not seen as a ciation.129131
discrete band. An important exception is in the
neonate where little haptoglobin is present and a2- INCREASED a2-MACROGLOBULIN
macroglobulin levels are much higher than in a2-Macroglobulin is elevated in neonates, the
adults. a2-Macroglobulin is an enormous molecule elderly, patients with elevated estrogen levels, and
(the tetramer has a molecular mass of 720 kDa) especially as part of the nephrotic pattern in
that functions as a protease inhibitor. patients with selective glomerular leakage. In
It is part of the family called thiol ester plasma nephrotic syndrome, there is a compensatory
proteins. This name refers to the presence of a increase in the synthesis of a2-macroglobulin.
unique cyclic thiol ester bond that is involved with Further, its large size prevents its passage into the
the covalent binding of proteases.118 This family of urine.
proteins includes a2-macroglobulin, pregnancy Serum levels of a2-macroglobulin are moderately
zone protein, C3, C4, and C5.118 While it is partic- higher in cord blood and during the rst week of
ularly effective at inhibiting plasmin activity, a2- life than in adults, averaging about 300 mg/dl.118,132
macroglobulin is also able to inhibit the enzymes Elevated levels of a2-macroglobulin are commonly
trypsin, chymotrypsin, thrombin, and elastase.119 seen in diabetics, especially in longstanding
a2-Macroglobulin binds to b2-microglobulin, sug- cases.133 In patients with a1-antitrypsin deciency
gesting that it may modify the metabolism of the and emphysema an increased concentration of a2-
latter.120 It is also the major serum collagenase macroglobulin is often also present.134
inhibitor, which may be important in wound heal- On serum protein electrophoresis, elevated a2-
ing.121 In addition, a2-macroglobulin has been macroglobulin may be seen as a sharp band in the
80 Proteins identied by serum protein electrophoresis
a
Data adapted in part from Langlois and Delanghe,135 Janik,137 and in part from van Lente.138
b
High-resolution electrophoresis.
DECREASED HAPTOGLOBIN
There are several important causes of decreased
haptoglobin levels that can be suggested by the
serum protein electrophoresis pattern (Table 4.6);
the most clinically signicant causes relate to
hemolysis. For the clinician studying a patient with
a hemolytic process, the laboratory interpretation
of marked decrease in haptoglobin consistent with
intravascular hemolysis is useful both in suggest-
ing the process (extravascular versus intravascular)
and in following the response of the patient to
Figure 4.21 The top sample shows a haptoglobin 1-1 variant therapy. As usual, there are some caveats. An
which migrates anodally from a2-macroglobulin. The second and
uncommon phenotypic variant of haptoglobin,
fourth sera contain a haptoglobin 2-2 variant that migrates
cathodally to a2-macroglobulin (from which it cannot be
referred to as type 0-0 (seen most frequently in
distinguished on this gel). The third serum contains a haptoglobin black patients), produces no or very low levels of
1-2 variant which migrates directly over a2-macroglobulin. The haptoglobin.145 Further, infants have little hapto-
slow g-region of the third serum also contains two dark-staining globin; the concentration goes from virtually zero
bands which may represent a double gammopathy (see Chapter 7). in the cord blood to adult levels by around the age
(Paragon SPE2 system stained with Paragon Violet.)
of 4 months.139 Ineffective erythropoiesis such as
occurs during folate and vitamin B12 deciency may
electrophoresis agarose gels and CZE, Hp2 produce a decrease in the haptoglobin region on
homozygotes produce a broad band at the catho- serum protein electrophoresis.
dal end of a2-macroglobulin (Fig. 4.21). When hemoglobin binds to haptoglobin, the
Heterozygotes, Hp1/Hp2, on starch gels produce complex has a much slower migration than that of
one weak band in the 1-1 position and several haptoglobin alone (Figs 4.22 and 4.23). The pat-
slower-migrating bands. By routine serum protein terns shown in these gures usually indicate that
electrophoretic techniques, these appear as a broad the sample has been handled poorly, with in vitro
band just anodal to and directly over a2- hemolysis producing the haptoglobinhemoglobin
macroglobulin. Molecular studies on the genotype band indicated. However, in cases of recent
82 Proteins identied by serum protein electrophoresis
Decreased haptoglobin
Hemolysis
In vivo: Intravascular hemolysis
Extravascular hemolysis
In vitro: Poor blood sampling technique; (hemoglobinhaptoglobin migrates
between a2-macroglobulin and transferrin)
Ineffective erythropoiesis Vitamin B12 deciency
Folate deciency
Congenital Hp O phenotype
Neonates Normally have low haptoglobin
Figure 4.23 Capillary zone electropherograms display a prominent slow b-region band (arrow) due to hemoglobinhaptoglobin complex.
Note, more recent buffer solution on the Sebia Capillarys may result in a different migration of the hemoglobinhaptoglobin complex.
broblasts and monocytes at sites of tissue rin from C3 can demonstrate relatively common
damage.152,153 It is increased in pregnancy, and is genetic variants of transferrin (Fig. 4.24). Both
further increased in patients with pre-eclampsia electrophoretically slow and fast variants of
caused by vascular injury, increased production or transferrin have been identied. The major fac-
enzymatic degradation.154 It is also increased in the tors that determine the migration of transferrin
vessel walls in active central nervous system on serum protein electrophoresis are the amino
plaques from patients with multiple sclerosis, pos- acid sequence and its sialic acid content (see
sibly enhancing myelin phagocytosis in these below). The migration of transferrin in the vast
lesions.155 During the rst 48 h of the acute inam- majority of people (98 per cent) is due to the
matory response, bronectin is rapidly deposited in common type of transferrin TfC.160,161 Variants
the damaged tissue. Simultaneously, its concentra- that move toward the anode are TfB and those
tion in the serum decreases.152 which move toward the cathode are TfD.156 Black
patients have a gene frequency of about 10 per
cent for TfCD.157,162 Homozygotes of fast or slow
b-REGION variants are rarely detected, perhaps because they
would be difcult to recognize by commonly used
Transferrin clinical laboratory electrophoretic techniques.
However, heterozygotes are easily seen as two
Transferrin is the major band at the anodal end of equal staining bands in the b1 region (Fig. 4.24).
the b1-region. Transferrin is a single polypeptide gly- These represent the codominant expression of
coprotein with a molecular mass of 76.5 kDa that two alleles, one normal and one variant.13 Despite
functions to transport non-heme ferric iron from the characteristic appearance, I recommend per-
the gastrointestinal tract and from the breakdown forming an immunoxation, usually a Penta
of hemoglobin to the bone marrow.156,157 Each trans- (pentavalent, see Chapter 3), to rule out a small
ferrin molecule can bind two molecules of free iron, monoclonal gammopathy. While earlier studies
but normally only about one-third of the transfer- suggested that these variants had no functional
rin molecules are saturated with iron. The total effect on patients, recent studies by Kasvosve et
iron-binding capacity of serum is a reection of the al.157 demonstrate that the amount of iron bound
amount of transferrin present; its normal concen- by transferrin is lower in TFCD individuals than
tration is 240480 mg/dl.158 In patients with iron- with the other common variants. This may have
deciency anemia, the levels of transferrin are functional signicance to partly protect these
considerably increased. Determinations of the individuals from increased iron accumulation in
transferrin levels are useful in distinguishing iron overload situations.157 While one cannot sub-
between iron deciency anemia (inadequate intake type the variants by serum protein electrophore-
or chronic hemorrhage with loss of iron stores) sis, it is important to understand the possible
from iron-refractory anemias. In iron deciency, the double bands associated with the variants. These
concentration of serum transferrin goes up, but as may be mistaken for small M-proteins. This is
less iron is available for transport, the saturation of one reason why bands suspected of being M-
the transferrin falls, often to less than 15 per cent proteins must always be characterized to prove
compared with 33 per cent normally. Transferrin that they are immunoglobulins.
levels are also increased in patients that are preg- Serum transferrin is normally glycosylated by two
nant or in patients receiving estrogens.159 complex carbohydrate chains that contain a
charged component, sialic acid. The number of the
TRANSFERRIN VARIANTS sialic acids attached varies from none to six. Each
Serum protein electrophoresis using either CZE sialic acid molecule present has been estimated to
or gels that provide a crisp separation of transfer- alter the pI of transferrin by 0.1 pH unit.163165 The
b-Region 85
Figure 4.24 The middle serum is from a patient with genetic variant of transferrin. There are two discrete bands (indicated) of
equivalent intensity. (Panagel System stained with Amido Black.)
most common isoform of transferrin has two mol- serum and a substantial transferrin band in the
ecules of sialic acid attached.163 Patients that b2-region (t fraction of transferrin) that lacks the
consume large quantities of alcohol contain serum sialic acid residues. This form of transferrin is
transferrin with few sialic acid molecules attached also present in human aqueous humor.174
(none, monosialated and disialated transferrin) Immunoxation of cerebrospinal uid with anti-
with a correspondingly higher pI than normal.166,167 bodies against transferrin will demonstrate two
This is referred to as carbohydrate-decient trans- transferrin bands (discussed in Chapter 8): one due
ferrin. Although, using iron specic stains, a to the normal sialated transferrin seen in the b1-
difference in migration on serum protein elec- region of serum and the other due to the t fraction,
trophoresis has been reported, it is not possible to (a serum control from the same patient must be
distinguish reliably this difference in migration examined in tract next to the nasal or ear uid to
using routine protein stains.168 A variety of be sure the patient does not normally have
methods, including CZE have been used to asialated transferrin in his or her serum).172 Altered
measure carbohydrate-decient transferrin.166 migration of genetic variants and changes of sialic
However, those capillary zone studies were per- acid in alcoholics could result in false positives if a
formed on a P/ACE 5000 electropherograph control serum was not studied.172,175
(Beckman), not the currently available clinical
equipment (Paragon CZE 2000 or Sebia INCREASED TRANSFERRIN
Capillarys).169 The transferrin band is increased in patients with
We can detect cerebrospinal uid leakage in the iron deciency anemia.176 Occasionally, the trans-
ear or nose following head trauma or secondary to ferrin band may be obscured by the presence of a
a neoplasm by taking advantage of the fact that large amount of preb or b1-lipoprotein (see below).
serum transferrin usually contains sialic acid mole- Also, when using CZE, some radiocontrast dyes
cules.170173 Cerebrospinal uid normally contains will migrate in the b-region and may result in a
both the b1-migrating glycated transferrin seen in falsely elevated transferrin band (see Chapter 2).177
86 Proteins identied by serum protein electrophoresis
In addition to radiocontrast dyes, the antibiotic 4.25).49,50 Another strategy to rule out a mono-
piperacillin-tazobactam (Tazocin; Wyeth Lederle, clonal gammopathy is to perform an
Collegeville, PA, USA) was reported to produce a immunoxation (such as a Penta evaluation) on all
small peak just anodal to the transferrin band on suspicious bands (see Chapter 3).
the Paragon CZE 2000, using software version
2.21).178 Such a band could be mistaken either for a DECREASED TRANSFERRIN
transferrin variant or small M-protein. It is recom- Transferrin is usually decreased in alcoholic cirrho-
mended that all suspicious unidentied bands be sis.179 During acute inammation, the synthesis of
evaluated by performing an immunoxation. A transferrin by the liver is largely shut down, result-
Penta screen may be the most efcient manner in ing in a faint b1-region band. Transferrin is also
which to do this (see Chapter 3). decreased during renal disease and thermal injuries
Lower resolution electrophoretic methods that do because of loss through the glomeruli and damaged
not provide a crisp separation of the b1 and b2 skin, respectively.180 Congenital atransferrinemia
bands do not offer a clear view of the transferrin has been reported, but is quite rare. These indivi-
band. This may make it more difcult to detect the duals suffer from a microcytic, hypochromic
occasional M-protein that causes a distortion of anemia, despite the presence of normal serum iron
the transferrin band. When b1-lipoprotein obscures levels; they have a very low iron-binding capacity.
this region, electrophoresis could be repeated with Their poor ability to transport iron also predisposes
heparin added to the sample. This pulls the b1- them to develop hemochromatosis unless detected
lipoprotein toward the anode and gives the and treated early.181,182 Therefore, the nding of an
interpreter a better view of the b1-region (Fig. undetectable b1-region band on serum protein
1000 160 80 40 20 10 0 0
Figure 4.25 Serum treated with heparin (concentrations indicated units/ml). Note that the b1-lipoprotein band becomes weaker and
moves toward the anode, making the transferrin region easier to inspect in difcult cases. (Photograph from work by Jeffrey Pearson).
b-Region 87
b-Migrating monoclonal
gammopathies (M-proteins) and
complement activation products
The major proteins that need to be distinguished
from the double transferrin band (heterozygous vari-
ants), are b-migrating monoclonal gammopathies
and complement activation products. It is often erro-
neously believed that immunoglobulin molecules
only migrate in the g-region. Although IgG usually
resides in the g-region, IgA is mainly a b-migrating Figure 4.27 The bottom sample has a very prominent b1-region
molecule while IgM tends to migrate between these (transferrin region) band (arrow) because of an IgA k monoclonal
gammopathy which migrated on top of the usual transferrin
two (Fig. 4.26). Further, the rare IgE, uncommon
location. Even though the migration is correct for transferrin, the
IgD and the more common monoclonal free light band is far too dense in staining to be caused by iron deciency.
chains (Bence Jones protein) will frequently occur in This sample, and those above it, also suffer from a slight distortion,
the b- or even in the a-globulin regions. Monoclonal possibly owing to insufcient blotting. (Paragon SPE2 system
gammopathies involving any of these b-migrating stained with Paragon Violet.)
immunoglobulins may only produce a subtle distor-
tion of the transferrin or C3 band.
directly over transferrin or C3 (see later) (Fig. 4.27).
The presence of a second b1-region band (other
Therefore, when there is a large transferrin band,
than transferrin) that is due to an M-protein of the
but no clinical history of an iron-decient anemia,
IgA class usually will be denser and more diffuse
one should rule out the possibility of a monoclonal
than the equal transferrin lines seen in the hetero-
gammopathy by an immunoxation (such as the
zygous variants. Performing an immunoxation to
studies discussed in Chapter 3).
rule out the possibility of an M-protein is recom-
The other major protein band of clinical signi-
mended when there is distortion of the transferrin
cance that is often confused with the transferrin
or C3 band. Occasionally, monoclonal bands lie
variant is the C3c product of complement (see
later). This may be an in vitro activation due to
Origin poor specimen handling or may reect in vivo acti-
vation caused by autoimmune disease or ongoing
inammation. As discussed later, when the band
+ anodal to transferrin is C3c, the usual C3 band (b2-
region) is decreased or absent. To be certain,
however, performing an immunoxation is recom-
IgG mended to rule out an M-protein.
Albumin
IgM
IgA
b1-lipoprotein
Figure 4.26 The usual locations of IgG, IgA and IgM molecules
are indicated by lines below the schematic representation of the b1-lipoprotein (low density lipoprotein) is an
gel. unusual molecule in two respects: its position on
88 Proteins identied by serum protein electrophoresis
gel electrophoresis varies considerably depending cal eld and endosmotic ow (see Chapter 1) in the
on its concentration and it has an irregular anodal immediate vicinity of the band. The partial inter-
front (Fig. 4.28). The molecular mass is enormous ruption of the regular electrical eld and
(2750 kDa) and its concentration spans a wide endosmotic ow results in an irregular band or
range between 57 mg/dl and 206 mg/dl in adults.183 small parallel bands (striae) with a crescent-type
b1-Lipoprotein transports lipids, cholesterol and shape (the middle lagging toward the cathode).184
hormones. Because the migration of b1-lipoprotein Such irregularity is especially pronounced in gels
on gels decreases with increasing concentration, it that have narrower pores with a greater molecular
may be found anodal to the transferrin band or sieve capability. While these gels often give excel-
cathodal to the C3 band. The band often has an lent resolution of g-globulins, the b1-lipoprotein
irregular anodal front on gel-based systems. This band can obscure other important beta proteins.
irregularity may also occur with a few IgM and IgA This tendency will vary from one manufacturer to
monoclonal gammopathies. It is related to the ten- another. Therefore, in some systems, b1-lipoprotein
dency of these molecules to aggregate when present will produce a diffuse band that does not signi-
in high concentration. At the anodal edge, the cantly interfere with the interpretation of the major
movement is faster and the concentration of the b-region bands or b-migrating monoclonal gam-
molecules is lower than at the center of the band, mopathies; other systems can give irregular,
where aggregation tends to occur. The aggregation dark-staining bands that may obscure transferrin
of these large molecules interferes with the electri- variants or small monoclonal gammopathies. In
CZE systems, b1-lipoprotein may migrate in the
slow a2-region (Fig. 4.29). Because of a relative
constancy of position of b1-lipoprotein on the
available CZE systems, I have found less confusion
with an M-protein than in gel-based systems.
Elevated levels of b1-lipoprotein are seen in condi-
tions with increased cholesterol, such as the
nephrotic syndrome and Type II hypercholes-
terolemia. Preincubating a patients serum with
40 units of heparin/ml serum will cause the b1-
a lipoprotein band to migrate more anodally as a
faint diffuse band and may provide the interpreter
with a better view of the b1-region 49,50 (Fig. 4.25).
Figure 4.30 The second sample from the top is a C3 heterozygote with two identical staining bands (indicated) in the b2-region.
(Cellogel high-resolution acetate system.) This photograph was contributed by Francesco Aguzzi.
90 Proteins identied by serum protein electrophoresis
complex has functional signicance. Most phago- The concentration of C3 is elevated late during
cytic cells have receptors for C3b on their surface the acute phase response. It is upregulated by inter-
and, therefore, the C3b on the surface of the initi- leukin-1 (IL-1), interleukin-6 (IL-6) and TNF.189192
ating complex will be more readily attached to Because C3 and haptoglobin are often elevated
neutrophils, monocytes and eosinophils. after other acute-phase reactants, a1-acid glycopro-
During inactivation of C3b, Factor I rst cleaves tein (orosomucoid), a1-antitrypsin and C-reactive
the a-chain into the inactive C3bi. Next, in a kinet- protein have declined to the normal range, I refer
ically slower reaction, Factor I cleaves the terminal to the combined elevation of a2-region and C3 (b2-
end of the a-chain, resulting in the formation of a band) together with a decrease in transferrin as a
small C3d fragment (30 000) and a larger C3c subacute reaction.
fragment (160 000) that is often seen in serum The use of a CZE system may be more sensitive
protein electrophoresis of poorly processed blood than gel-based methods for detecting C3 even on
samples or aged specimens. In the latter reaction, specimens that have been stored for up to 30 days
Factor I collaborates with serum proteases. at 4C.193 In addition to providing crisp separa-
Usually, the presence of the C3c band indicates tion of the b1- and b2-regions, it has been sug-
that the specimen has been processed poorly, gested that CZE may provide useful estimates of
stored at room temperature, lyophilized (most both C3 and transferrin that did not require
lyophilized commercial standard serum prepara- further testing.194 However, I recommend specic
tions will show this C3c band), or stored for nephelometric-determined concentrations for best
several days at 4C. Owing to the short biological accuracy.
half-life of C3c, it is unlikely that the C3c seen by
serum protein electrophoresis relates to active
inammation in vivo. C4
When C3 breaks down, the C3 (b2-band) band
declines and smaller, smudgy bands may appear C4 is a large, 206 kDa glycoprotein with an
anodal to transferrin, cathodal to transferrin or adult serum concentration of from 10 to
even in the g-region, depending on which electro- 40 mg/dl (0.10.4 g/l).106 It is visible on serum
phoretic system is used. On most gel-based systems protein electrophoresis only when present in the
and on the Paragon CZE 2000, using Version 1.5, higher normal range or when increased such as
the C3c breakdown product appears anodal to in patients with an acute inammatory reaction.
transferrin (Figs. 4.31 and 4.32). It may be seen just anodal to C3 on gel-based
Figure 4.31 C3 activation is seen in the middle sample. A C3c band is now present (indicated), while the usual position of intact C3 in
the b2-region shows only a faint slur toward the anode. (Panagel system stained with Amido Black.)
g-REGION Immunoglobulins
a1 a2 t
Figure 4.35 The top lane is an immunoxation with anti-C-reactive protein. It indicates the location of this prominent protein. The
second lane is the serum protein electrophoresis from the same case. This serum was from a patient with an acute phase reaction and
shows increased a1 (a1), a2 (a2), and decreased transferrin (t). The markedly elevated C-reactive protein is demonstrated by the discrete
band (indicated) in the mid-g-region. The C3 band is missing because the sample had been stored for several days before the
immunoxation was performed for this demonstration. (Panagel stained with Coomassie Blue.)
a
Subclasses, serum half life (days).215
b
Half-life varies with different subclasses.
c
All of the immunoglobulins appear on the surface of some B cells, however, only IgD and IgM are present on early B-lymphocytes.
g-Region 95
concentration is variable. Neonates will have the Fc portion, but also to the physical proximity
almost normal adult levels, because this molecule is of the ve Fc groups of the pentamer. The pI and
transported across the placenta through specic bulk of IgM are such that it does not stray far from
Fc-receptors on the placenta. However, the the origin on gel-based electrophoresis. Depending
newborn child does not begin to synthesize IgG for on the character of the support medium for the
several months. The maternal IgG that crossed the particular electrophoretic system, it may migrate
placenta is gradually catabolized so that at about slightly cathodally or slightly anodally. By CZE,
6 months a nadir is reached where extremely low IgM migrates mainly in the bg region. Marked
levels of IgG are present in the serum. Children elevations may occur with infections or malig-
with congenital immunodeciencies are asympto- nancy and are reviewed further in Chapter 7.
matic until after 6 months of age because they are
protected by the maternal IgG that crossed the IMMUNOGLOBULIN A
placenta.220 Immunoglobulin A is the other major isotype of
Many abnormalities of IgG occur. Elevations immunoglobulin in the serum. In the serum, IgA
with infections, autoimmune diseases, liver disease occurs mainly as a 160 kDa 7S monomer that has
and multiple myeloma are commonly seen. It is the a concentration of 70400 mg/dl in adults. It is
immunoglobulin most often found in multiple thought to function mainly along mucosal surfaces
myeloma.221 Two or three discrete bands (oligo- where it is the most prominent immunoglobulin
clonal bands) occur in the g-region of serum from isotype.215 Along the mucosa of the gastrointestinal
patients with circulating immune complexes, tract, respiratory tract, mammary or other glands,
acquired immune deciency syndrome, hepatitis or IgA is secreted from plasma cells as a dimer
polyclonal increases in response to other infec- attached by a small 12 kDa polypeptide called
tions.222 The bands may be especially prominent joining (J) chain. It then migrates toward the
during infectious diseases.223 epithelium where it combines with a 60 kDa glyco-
An isolated decrease in IgG demonstrated by protein called secretory component, which is made
nding faint staining of the g-region with other by the surface epithelium. This is secreted into the
fractions being normal is a highly signicant mucosal lumen where it prevents pathogenic
pattern that may indicate a variety of conditions microorganisms and their toxic products from
including: humoral immunodeciency, B-cell lym- entering and attaching to the surface epithelium. It
phoproliferative disorders, light chain disease, is a poor activator of complement and does not
nonsecretory myeloma, or amyloid AL (amyloid opsonize for phagocytosis. Its role in the serum is
associated with immunoglobulin light chain; dis- unclear, although some studies suggest it may be
cussed in Chapter 6). able to collaborate with killer cells in cytotoxic
reactions.224,225
IMMUNOGLOBULIN M Isolated IgA deciency is the most common of all
Immunoglobulin M usually occurs as a pentamer congenital immunodeciencies, occurring in 1 of
in the serum. As such, it has a molecular mass of 133 blood donors.220 Although there is usually a
about 900 kDa. The normal concentration is compensatory increase of IgM along the mucosal
40230 mg/dl.106 It is the most effective of all iso- surfaces, these patients have a signicantly increased
types in activating complement. Whereas IgG incidence of autoimmune disease and allergy. They
may require 100 or more molecules in an may also be subject to severe life-threatening ana-
antibodyantigen complex to effectively activate phylactic reactions to blood products if they have
complement, a single IgM molecule will sufce for developed a hypersensitivity to the IgA from previ-
complete activation through the terminal lytic ous transfusions. When these patients require intra-
sequence on a cell surface.215 This property prob- venous immunoglobulin therapy, preparations of
ably relates not only to the amino acid makeup of IgA-depleted immunoglobulin have proved to be
References 97
12. Bergquist J, Andersen O, Westman A. Rapid 22. Levinson SS. Urine protein electrophoresis and
method to characterize mutations in immunoxation electrophoresis supplement one
transthyretin in cerebrospinal uid from familial another in characterizing proteinuria. Ann Clin
amyloidotic polyneuropathy patients by use of Lab Sci 2000;30:7984.
matrix-assisted laser desorption/ionization time- 23. Lessard F, Bannon P, Lepage R, Joly JG. Light
of-ight mass spectrometry. Clin Chem chain disease and massive proteinuria. Clin
2000;46:12931300. Chem 1985;31:475477.
13. Laurell CB. Composition and variation of the gel 24. Cacoub P, Sbai A, Toan SV, Bellanger J, Hoang
electrophoretic fractions of plasma, cerebrosinal C, Godeau P, Piette JC. Collagenous colitis. A
uid and urine. Scand J Clin Lab Invest Suppl study of 11 cases. Ann Med Interne (Paris)
1972;124:7182. 2001;152:299303.
14. Yeung CY, Fung YS, Sun DX. Capillary 25. Schmidt PN, Blirup-Jensen S, Svendsen PJ,
electrophoresis for the determination of Wandall JH. Characterization and quantication
albumin binding capacity and free bilirubin in of plasma proteins excreted in faeces from
jaundiced neonates. Semin Perinatol healthy humans. Scand J Clin Lab Invest
2001;25:5054. 1995;55:3545.
15. Fung YS, Sun DX, Yeung CY. Capillary 26. Bologa RM, Levine DM, Parker TS, Cheigh JS,
electrophoresis for determination of free and Serur D, Stenzel KH, Rubin AL. Interleukin-6
albumin-bound bilirubin and the investigation of predicts hypoalbuminemia, hypocholesterolemia,
drug interaction with bilirubin-bound albumin. and mortality in hemodialysis patients. Am J
Electrophoresis 2000;21:403410. Kidney Dis 1998;32:107114.
16. Zhang B, Fung Y, Lau K, Lin B. Bilirubinhuman 27. Mackiewicz A, Schooltink H, Heinrich PC, Rose-
serum albumin interaction monitored by John S. Complex of soluble human IL-6-
capillary zone electrophoresis. Biomed receptor/IL-6 up-regulates expression of
Chromatogr 1999;13:267271. acute-phase proteins. J Immunol 1992;149:
17. Tarnoky AL. Genetic and drug-induced variation 20212027.
in serum albumin. Adv Clin Chem 28. Odamaki M, Kato A, Takita T, et al. Role of
1980;21:101146. soluble receptors for tumor necrosis factor alpha
18. Arvan DA, Blumberg BS, Melartin L. Transient in the development of hypoalbuminemia in
bisalmuminemia induced by drugs. Clin Chim hemodialysis patients. Am J Nephrol 2002;22:
Acta 1968;22:211218. 7380.
19. Daniels JC. Carrier protein abnormalities. In: 29. Holowachuk EW. Register of Analbuminemia
Ritzmann SE, Daniels JC, eds. Serum protein Cases 2001. www.albumin.org
abnormalities, diagnostic and clinical aspects. 30. Gossi B, Kleinert D, Gossi U. A further case of
Boston: Little, Brown & Co., 1975. analbuminemia. Schweiz Med Wochenschr
20. Mendler MH, Corbinais S, Sapey T, et al. In 2000;130:583589.
patients with cirrhosis, serum albumin 31. Benhold H, Klaus D, Schluren PG. Volumen
determination should be carried out by regulation and renal function in analbuminemia.
immunonephelometry rather than by protein Lancet 1960;ii:11691170.
electrophoresis. Eur J Gastroenterol Hepatol 32. Platt HS, Barron N, Giles AF, Midgley JE,
1999;11:14051411. Wilkins TA. Thyroid-function indices in an
21. Chew ST, Fitzwilliam J, Indridason OS, Kovalik analbuminemic subject being treated with
EC. Role of urine and serum protein thyroxin for hypothyroidism. Clin Chem
electrophoresis in evaluation of nephrotic-range 1985;31:341342.
proteinuria. Am J Kidney Dis 1999;34: 33. Dammacco F, Miglietta A, DAddabbo A,
135139. Fratello A, Moschetta R, Bonomo L.
References 99
Analbuminemia: report of a case and review of Gil A, Pineiro A. Characterization of the acute
the literature. Vox Sang 1980;39:153161. phase serum protein response in pigs.
34. Kallee E. Bennholds analbuminemia: a follow-up Electrophoresis 1994;15:672676.
study of the rst two cases (19531992). J Lab 45. Feher J, Romics L, Jakab L, Feher E, Szilvasi I,
Clin Med 1996;127:470480. Papp G. Serum lipids and lipoproteins in chronic
35. Minchiotti L, Campagnoli M, Rossi A, et al. A liver disease. Acta Med Acad Sci Hung
nucleotide insertion and frameshift cause 1976;33:217223.
albumin Kenitra, an extended and O- 46. Lopez D, Sanchez MD, Shea-Eaton W, McLean
glycosylated mutant of human serum albumin MP. Estrogen activates the high-density
with two additional disulde bridges. Eur J lipoprotein receptor gene via binding to estrogen
Biochem 2001;268:344352. response elements and interaction with sterol
36. Petersen CE, Ha CE, Harohalli K, Park DS, regulatory element binding protein1A.
Bhagavan NV. Familial dysalbuminemic Endocrinology 2002;143:21552168.
byperthyroxinemia may result in altered warfarin 47. Jula A, Marniemi J, Huupponen R, Virtanen A,
pharmacokinetics. Chem Biol Interact Rastas M, Ronnemaa T. Effects of diet and
2000;124:161172. simvastatin on serum lipids, insulin, and
37. Jaeggi-Groisman SE, Byland C, Gerber H. antioxidants in hypercholesterolemic men: a
Improved sensitivity of capillary electrophoresis randomized controlled trial. JAMA
for detection of bisalbuminemia. Clin Chem 2002;287:598605.
2000;46:882883. 48. Muckle TJ. The post-heparin pattern in routine
38. Huss K, Putnam FW, Takahashi N, Takahashi Y, serum protein immunoelectrophoresis. Am J Clin
Weaver GA, Peters T Jr. Albumin Cooperstown: Pathol 1977;67:450454.
a serum albumin variant with the same (313 49. Pearson JP, Keren DF. The effects of heparin on
LysAsn) mutation found in albumins in Italy lipoproteins in high-resolution electrophoresis of
and New Zealand. Clin Chem 1988;34:183187. serum. Am J Clin Pathol 1995;104:468471.
39. Madison J, Arai K, Sakamoto Y, et al. Genetic 50. Su LD, Keren DF. The effects of exogenous free
variants of serum albumin in Americans and fatty acids on lipoprotein migration in serum
Japanese. Proc Natl Acad Sci USA high-resolution electrophoresis: addition of free
1991;88:98539857. fatty acids improves visualization of normal and
40. Takahashi N, Takahashi Y, Isobe T, et al. Amino abnormal alpha1-antitrypsin. Am J Clin Pathol
acid substitutions in inherited albumin variants 1998;109:262267.
from Amerindian and Japanese populations. Proc 51. Khan H, Salman KA, Ahmed S. Alpha-1
Natl Acad Sci USA 1987;84:80018005. antitrypsin deciency in emphysema. J Assoc
41. Wolf M, Kronenberg H, Dodds A, et al. A safety Physicians India 2002;50:579582.
study of Albumex 5, a human albumin solution 52. Carrell RW, Lomas DA. Alpha1-antitrypsin
produced by ion exchange chromatography. Vox deciency a model for conformational diseases.
Sang 1996;70:198202. N Engl J Med 2002;346:4553.
42. Rifai N, Bachorik PS, Albers JJ. Lipids, 53. Hibbetts K, Hines B, Williams D. An overview of
lipoproteins, and apolipoproteins. In: Burtis CA, proteinase inhibitors. J Vet Intern Med
Ashwood ER, eds. Tietz textbook of clinical 1999;13:302308.
chemistry. Philadelphia: WB Saunders, 1999. 54. Lomas DA, Lourbakos A, Cumming SA,
43. Oram JF. Molecular basis of cholesterol Belorgey D. Hypersensitive mousetraps, alpha1-
homeostasis: lessons from Tangier disease and antitrypsin deciency and dementia. Biochem Soc
ABCA1. Trends Mol Med 2002;8:168173. Trans 2002;30:8992.
44. Lampreave F, Gonzalez-Ramon N, Martinez- 55. Weinberger SE. Recent advances in pulmonary
Ayensa S, Hernandez MA, Lorenzo HK, Garcia- medicine (1). N Engl J Med 1993;328:13891397.
100 Proteins identied by serum protein electrophoresis
56. Keren DF, Di Sante AC, Bordine SL. antitrypsin deciency alleles Pi*Z and Pi*S by
Densitometric scanning of high-resolution real-time uorescence PCR and melting curves.
electrophoresis of serum: methodology and Clin Chem 1999;45:18721875.
clinical application. Am J Clin Pathol 66. Hammerberg G, Keren DF. Polymerase chain
1986;85:348352. reaction-mediated site-directed mutagenesis
57. Gonzalez-Sagrado M, Lopez-Hernandez S, detection of Z and S alpha-1-antitrypsin alleles in
Martin-Gil FJ, Tasende J, Banuelos MC, family members. J Clin Lab Anal 1996;10:
Fernandez-Garcia N, Arranz-Pena ML. Alpha1- 384388.
antitrypsin deciencies masked by a clinical 67. Pierce JA, Eradio B, Dew TA. Antitrypsin
capillary electrophoresis system (CZE 2000). phenotypes in St Louis. JAMA 1975;231:
Clin Biochem 2000;33:7980. 609612.
58. Malfait R, Gorus F, Sevens C. Electrophoresis of 68. Teckman JH, Burrows J, Hidvegi T, Schmidt B,
serum protein to detect alpha1-antitrypsin Hale PD, Perlmutter DH. The proteasome
deciency: ve illustrative cases. Clin Chem participates in degradation of mutant alpha-1-
1985;31:13971399. antitrypsin Z in the endoplasmic reticulum of
59. Mullins RE, Miller RL, Hunter RL, Bennett B. hepatoma-derived hepatocytes. J Biol Chem
Standardized automated assay for functional 2001;276:4486544872.
alpha 1-antitrypsin. Clin Chem 69. Morse JO. Alpha1-antitrypsin deciency (rst
1984;30:18571860. of two parts). N Engl J Med 1978;299:
60. Jeppsson JO, Einarsson R. Genetic variants of 10451048.
alpha 1-antitrypsin and hemoglobin typed by 70. Morse JO. Alpha1-antitrypsin deciency (second
isoelectric focusing in preselected narrow pH of two parts). N Engl J Med 1978;299:
gradients with PhastSystem. Clin Chem 10991105.
1992;38:577580. 71. Morse JO, Lebowitz MD, Knudson RJ, Burrows
61. Berninger RW, Teixeira MF. Alpha 1-antitrypsin: B. Relation of protease inhibitor phenotypes to
the effect of anticoagulants on the trypsin obstructive lung diseases in a community. N Engl
inhibitory capacity, concentration and J Med 1977;296:11901194.
phenotype. J Clin Chem Clin Biochem 72. Elliott PR, Stein PE, Bilton D, Carrell RW,
1985;23:277281. Lomas DA. Structural explanation for the
62. von Ahsen N, Oellerich M, Schutz E. Use of two deciency of S alpha 1-antitrypsin. Nat Struct
reporter dyes without interference in a single- Biol 1996;3:910911.
tube rapid-cycle PCR: alpha(1)-antitrypsin 73. Lewis JH, Iammarino RM, Spero JA, Hasiba U.
genotyping by multiplex real-time uorescence Antithrombin Pittsburgh: an alpha1-antitrypsin
PCR with the LightCycler. Clin Chem variant causing hemorrhagic disease. Blood
2000;46:156161. 1978;51:129137.
63. Ortiz-Pallardo ME, Zhou H, Fischer HP, 74. Brennan SO, Sheat JM, Aiach M. Circulating
Neuhaus T, Sachinidis A, Vetter H, Bruning T, proalbumin associated with a second case of
Ko Y. Rapid analysis of alpha1-antitrypsin PiZ antitrypsin Pittsburgh. Clin Chim Acta
genotype by a real-time PCR approach. J Mol 1993;214:123128.
Med 2000;78:212216. 75. Ghishan FK, Greene HL. Liver disease in
64. Rieger S, Riemer H, Mannhalter C. Multiplex children with PiZZ alpha 1-antitrypsin
PCR assay for the detection of genetic variants of deciency. Hepatology 1988;8:307310.
alpha1-antitrypsin. Clin Chem 1999;45: 76. Iezzoni JC, Gaffey MJ, Stacy EK, Normansell
688690. DE. Hepatocytic globules in end-stage hepatic
65. Aslanidis C, Nauck M, Schmitz G. High-speed disease: relationship to alpha1-antitrypsin
detection of the two common alpha(1)- phenotype. Am J Clin Pathol 1997;107:692697.
References 101
77. Lima LC, Matte U, Leistner S, et al. Molecular 88. Piitulainen E, Tornling G, Eriksson S. Effect of
analysis of the Pi*Z allele in patients with liver age and occupational exposure to airway
disease. Am J Med Genet 2001;104:287290. irritants on lung function in non-smoking
78. Lomas DA, Evans DL, Finch JT, Carrell RW. individuals with alpha 1-antitrypsin deciency
The mechanism of Z alpha 1-antitrypsin (PiZZ). Thorax. 1997;52:244248.
accumulation in the liver. Nature 1992;357: 89. Dahl M, Nordestgaard BG, Lange P, Vestbo J,
605607. Tybjaerg-Hansen A. Molecular diagnosis of
79. Sveger T. Liver disease in alpha1-antitrypsin intermediate and severe alpha(1)-antitrypsin
deciency detected by screening of 200,000 deciency: MZ individuals with chronic
infants. N Engl J Med 1976;294:13161321. obstructive pulmonary disease may have lower
80. Lilley GL, Keren DF. A case of alpha1- lung function than MM individuals. Clin Chem
antitrypsin deciency. Tech Sample Clin 2001;47:5662.
Immunol 1986:C13. 90. Klaassen CH, de Metz M, van Aarssen Y,
81. Jack C, Evens CC. Three cases of alpha Janssen J. Alpha(1)-antitrypsin deciency as a
1-antitrypsin deciency in the elderly. Postgrad result of compound heterozygosity for the Z and
Med J 1991;67:840842. M(Heerlen) alleles. Clin Chem 2001;47:
82. Seersholm N. Epidemiology of emphysema in 978979.
subjects with severe alpha 1-antitrypsin 91. Canva V, Piotte S, Aubert JP, et al. Heterozygous
deciency. Dan Med Bull 2002;49:145158. M3Mmalton alpha1-antitrypsin deciency
83. Lohr HF, Schlaak JF, Dienes HP, Lorenz J, associated with end-stage liver disease: case
Meyer zum Buschenfelde KH, Gerken G. Liver report and review. Clin Chem 2001;47:
cirrhosis associated with heterozygous alpha-1- 14901496.
antitrypsin deciency type Pi MS and 92. Lee JH, Brantly M. Molecular mechanisms of
autoimmune features. Digestion 1995;56: alpha1-antitrypsin null alleles. Respir Med
4145. 2000;94(Suppl C):S711.
84. Graziadei IW, Joseph JJ, Wiesner RH, Therneau 93. Cook L, Janus ED, Brenton S, Tai E, Burdon J.
TM, Batts KP, Porayko MK. Increased risk of Absence of alpha-1-antitrypsin (Pi Null
chronic liver failure in adults with heterozygous Bellingham) and the early onset of emphysema.
alpha1-antitrypsin deciency. Hepatology. Aust N Z J Med 1994;24:263269.
1998;28:10581063. 94. Sandhaus RA. Alpha(1)-antitrypsin deciency
85. Hutchison DC, Tobin MJ, Cook PJ. Alpha therapy : pieces of the puzzle. Chest 2001;119:
1-antitrypsin deciency: clinical and 676678.
physiological features in heterozygotes of Pi type 95. Wencker M, Fuhrmann B, Banik N, Konietzko
SZ. A survey by the British Thoracic Association. N. Longitudinal follow-up of patients with
Br J Dis Chest 1983;77:2834. alpha(1)-protease inhibitor deciency before and
86. Tobin MJ, Cook PJ, Hutchison DC. Alpha during therapy with IV alpha(1)-protease
1-antitrypsin deciency: the clinical and inhibitor. Chest 2001;119:737744.
physiological features of pulmonary emphysema 96. Crystal RG, Brantly ML, Hubbard RC, Curiel
in subjects homozygous for Pi type Z. A survey DT, States DJ, Holmes MD. The alpha 1-
by the British Thoracic Association. Br J Dis antitrypsin gene and its mutations. Clinical
Chest 1983;77:1427. consequences and strategies for therapy. Chest
87. Larsson C, Dirksen H, Sundstrom G, Eriksson S. 1989;95:196208.
Lung function studies in asymptomatic 97. Sveger T, Thelin T. A future for neonatal alpha1-
individuals with moderately (Pi SZ) and severely antitrypsin screening? Acta Paediatr 2000;89:
(Pi Z) reduced levels of alpha1-antitrypsin. Scand 628631.
J Respir Dis 1976;57:267280. 98. Abboud RT, Ford GT, Chapman KR. Alpha1-
102 Proteins identied by serum protein electrophoresis
antitrypsin deciency: a position statement of the 108. Henriksen HJ, Petersen MU, Pedersen FB. Serum
Canadian Thoracic Society. Can Respir J alpha-1-acid glycoprotein (orosomucoid) in
2001;8:8188. uremic patients on hemodialysis. Nephron
99. Mullins CD, Huang X, Merchant S, Stoller JK. 1982;31:2426.
The direct medical costs of alpha(1)-antitrypsin 109. Philip AG, Hewitt JR. Alpha 1-acid glycoprotein
deciency. Chest 2001;119:745752. in the neonate with and without infection. Biol
100. Alkins SA, OMalley P. Should health-care Neonate 1983;43:118124.
systems pay for replacement therapy in patients 110. Mariotti R, Musumeci G, De Carlo M, et al.
with alpha(1)-antitrypsin deciency? A critical Acute-phase reactants in acute myocardial
review and cost-effectiveness analysis. Chest infarction: impact on 5-year prognosis. Ital Heart
2000;117:875880. J 2001;2:294300.
101. Ritchie RF, Palomaki GE, Neveux LM, 111. Billingsley GD, Walter MA, Hammond GL, Cox
Navolotskaia O, Ledue TB, Craig WY. Reference DW. Physical mapping of four serpin genes:
distributions for the positive acute phase serum alpha 1-antitrypsin, alpha 1-antichymotrypsin,
proteins, alpha1-acid glycoprotein corticosteroid-binding globulin, and protein C
(orosomucoid), alpha1-antitrypsin, and inhibitor, within a 280-kb region on
haptoglobin: a practical, simple, and clinically chromosome I4q32.1. Am J Hum Genet
relevant approach in a large cohort. J Clin Lab 1993;52:343353.
Anal 2000;14:284292. 112. Lilja H, Cockett AT, Abrahamsson PA. Prostate
102. Ritchie RF, Palomaki GE, Neveux LM, specic antigen predominantly forms a complex
Navolotskaia O. Reference distributions for the with alpha 1-antichymotrypsin in blood.
positive acute phase proteins, alpha1-acid Implications for procedures to measure prostate
glycoprotein (orosomucoid), alpha1-antitrypsin, specic antigen in serum. Cancer 1992;70:
and haptoglobin: a comparison of a large cohort 230234.
to the worlds literature. J Clin Lab Anal 113. Eriksson S, Lindmark B, Lilja H. Familial alpha
2000;14:265270. 1-antichymotrypsin deciency. Acta Med Scand
103. Lundholt BK, Madsen MW, Lykkesfeldt AE, 1986;220:447453.
Petersen OW, Briand P. Characterization of a 114. Miyake H, Hara S, Nomi M, Arakawa S,
nontumorigenic human breast epithelial cell line Kamidono S, Hara I. Value of prostate specic
stably transfected with the human estrogen antigen alpha1-antichymotrypsin complex for the
receptor (ER) cDNA. Mol Cell Endocrinol detection of prostate cancer in patients with a
1996;119:4759. PSA level of 4.110.0 ng/mL: comparison with
104. Jeppson JO, Laurell CB, Franzen B. Agarose gel PSA-related parameters. Int J Urol 2001;8:
electrophoresis. Clin Chem 1979;25:629638. 589593.
105. Leppert PC, Yu SY. Effect of parturition on 115. Burtin P, Grabar P. Nomenclature and
serum alpha 1-antiprotease (alpha 1-antitrypsin). identication of the normal human serum
Clin Chem 1993;39:905906. proteins. In: Bier M, ed. Electrophoresis theory,
106. Johnson AM, Rohlfs EM, Silverman LM. methods and applications. New York and
Chapter 20. Proteins. In: Burtis CA, Ashwood, London: Academic Press, 1989.
ER, eds. Tietz textbook of clinical chemistry. 116. Thomas RM, Schiano TD, Kueppers F, Black M.
Philadelphia: WB Saunders, 1999. Alpha1-antichymotrypsin globules within
107. Chen RJ, Chen CK, Chang DY, et al. hepatocytes in patients with chronic hepatitis C
Immunoelectrophoretic differentiation of alpha- and cirrhosis. Hum Pathol 2000;31:575577.
fetoprotein in disorders with elevated serum 117. Abraham CR. Reactive astrocytes and alpha1-
alpha-fetoprotein levels or during pregnancy. antichymotrypsin in Alzheimers disease.
Acta Oncol 1995;34:931935. Neurobiol Aging 2001;22:931936.
References 103
118. Petersen CM. Alpha 2-macroglobulin and receptor and low density lipoprotein receptor-
pregnancy zone protein. Serum levels, alpha 2- related protein suggests that this molecule is a
macroglobulin receptors, cellular synthesis and multifunctional receptor. J Biol Chem 1990;265:
aspects of function in relation to immunology. 1740117404.
Dan Med Bull 1993;40:409446. 128. Borth W. Alpha 2-macroglobulin, a
119. Sottrup-Jensen L. Role of internal thiol esters in multifunctional binding protein with targeting
the alpha-macroglobulin-proteinase binding characteristics. FASEB J 1992;6:33453353.
mechanism. Ann N Y Acad Sci 1994;737: 129. Koster MN, Dermaut B, Cruts M, et al. The
172187. alpha2-macroglobulin gene in AD: a population-
120. Gouin-Charnet A, Laune D, Granier C, et al. based study and meta-analysis. Neurology
alpha2-Macroglobulin, the main serum 2000;55:678684.
antiprotease, binds beta2-microglobulin, the light 130. Zill P, Burger K, Behrens S, et al. Polymorphisms
chain of the class I major histocompatibility in the alpha-2 macroglobulin gene in
complex, which is involved in human disease. psychogeriatric patients. Neurosci Lett.
Clin Sci (Lond) 2000;98:427433. 2000;294:6972.
121. Grinnell F, Zhu M, Parks WC. Collagenase-1 131. Wang X, Luedecking EK, Minster RL, Ganguli
complexes with alpha2-macroglobulin in the M, DeKosky ST, Kamboh MI. Lack of
acute and chronic wound environments. J Invest association between alpha2-macroglobulin
Dermatol 1998;110:771776. polymorphisms and Alzheimers disease. Hum
122. Kurdowska AK, Geiser TK, Alden SM, et al. Genet 2001;108:105108.
Activity of pulmonary edema uid interleukin-8 132. Cliver SP, Goldenberg RL, Neel NR, Tamura T,
bound to alpha(2)-macroglobulin in patients Johnston KE, Hoffman HJ. Neonatal cord serum
with acute lung injury. Am J Physiol Lung Cell alpha 2-macroglobulin and fetal size at birth.
Mol Physiol 2002;282:L10921098. Early Hum Dev 1993;33:201206.
123. Legres LG, Pochon F, Barray M, Gay F, Chouaib 133. James K, Merriman J, Gray RS, Duncan LJ, Herd
S, Delain E. Evidence for the binding of a R. Serum alpha 2-macroglobulin levels in
biologically active interleukin-2 to human alpha diabetes. J Clin Pathol 1980;33:163166.
2-macroglobulin. J Biol Chem 1995;270: 134. Brissenden JE, Cox DW. alpha 2-Macroglobulin
83818384. in patients with obstructive lung disease, with
124. Binder RJ, Karimeddini D, Srivastava PK. and without alpha 1-antitrypsin deciency. Clin
Adjuvanticity of alpha 2-macroglobulin, an Chim Acta 1983;128:241248.
independent ligand for the heat shock protein 135. Langlois MR, Delanghe JR. Biological and
receptor CD91. J Immunol 2001;166: clinical signicance of haptoglobin polymorphism
49684972. in humans. Clin Chem 1996;42:15891600.
125. DeStefano A, Hoffman M. The effect of alpha 2 136. Giblett ER. Recent advances in heptoglobin and
macroglobulinproteinase complexes on transferrin genetics. Bibl Haematol
macrophage Ia expression in vivo. Immunol 1968;29:1020.
Invest 1991;20:3343. 137. Janik B. High resolution electrophoresis and
126. Osada T, Ookata K, Athauda SB, Takahashi K, immunoxation of serum proteins on cellulosic
Ikai A. The active site titration of proteinases by media. Ann Arbor: Gelman Sciences, 1985.
using alpha 2-macroglobulin and high- 138. van Lente F. The diagnostic utility of haptoglobin.
performance liquid chromatography. Anal Clin Immunol Newsl 1991;11:135138.
Biochem 1992;207:7679. 139. Daniels JC, Larson DL, Abston S, Ritzmann SE.
127. Strickland DK, Ashcom JD, Williams S, Burgess Serum protein proles in thermal burns. II.
WH, Migliorini M, Argraves WS. Sequence Protease inhibitors, complement factors, and c-
identity between the alpha 2-macroglobulin reactive protein. J Trauma 1974;14:153162.
104 Proteins identied by serum protein electrophoresis
140. Daniels JC, Larson DL, Abston S, Ritzmann SE. 151. Blanco A, Solis P, Guisasola JA, Arranz E,
Serum protein proles in thermal burns. I. Serum Telleria JJ, Blanco C. Absence of bronectin in a
electrophoretic patterns, immunoglobulins, and 40-day-old child who died as a result of
transport proteins. J Trauma 1974;14:137152. septicemia with disseminated intravascular
141. Bowman BH. Haptoglobin. In: Bowman BH, ed. coagulation. Sangre (Barc) 1989;34:5962.
Hepatic plasma proteins. San Diego: Academic 152. Bevilacqua MP, Amrani D, Mosesson MW,
Press, 1993. Bianco C. Receptors for cold-insoluble globulin
142. Smithies O. Zone electrophoresis in starch gels: (plasma bronectin) on human monocytes. J Exp
group variations in the serum proteins of normal Med 1981;153:4260.
human adults. Biochem J 1955;61:629641. 153. Makogonenko E, Tsurupa G, Ingham K, Medved
143. Nakhoul FM, Marsh S, Hochberg I, Leibu R, L. Interaction of brin(ogen) with bronectin:
Miller BP, Levy AP. Haptoglobin genotype as a further characterization and localization of the
risk factor for diabetic retinopathy. JAMA bronectin-binding site. Biochemistry
2000;284:12441245. 2002;41:79077913.
144. Koch W, Latz W, Eichinger M, et al. Genotyping 154. Chavarria ME, Lara-Gonzalez L, Gonzalez-
of the common haptoglobin Hp 1/2 Gleason A, Sojo I, Reyes A. Maternal plasma
polymorphism based on PCR. Clin Chem cellular bronectin concentrations in normal and
2002;48:13771382. preeclamptic pregnancies: a longitudinal study
145. Kasvosve I, Gomo ZA, Gangaidzo IT, et al. for early prediction of preeclampsia. Am J Obstet
Reference range of serum haptoglobin is Gynecol 2002;187:595601.
haptoglobin phenotype-dependent in blacks. Clin 155. Sobel RA, Mitchell ME. Fibronectin in multiple
Chim Acta 2000;296:163170. sclerosis lesions. Am J Pathol 1989;135:
146. Fitzmaurice M, Valenzuela R, Winkelman EI. 161168.
Serum protein electrophoresis pattern associated 156. Kamboh MI, Ferrell RE. Human transferrin
with decreased haptoglobin as a poor prognostic polymorphism. Hum Hered 1987;37:6581.
indicator in severe liver disease. Am J Clin Pathol 157. Kasvosve I, Delanghe JR, Gomo ZA, et al.
1989;91:365(Abstr). Transferrin polymorphism inuences iron status
147. Baumann H, Onorato V, Gauldie J, Jahreis GP. in blacks. Clin Chem 2000;46:15351539.
Distinct sets of acute phase plasma proteins are 158. Wallach J. Interpretation of diagnostic tests.
stimulated by separate human hepatocyte- Philadelphia: Lippincott Williams & Wilkins,
stimulating factors and monokines in rat hepatoma 2000.
cells. J Biol Chem 1987;262:97569768. 159. Matsubara M, Odagaki E, Morioka T,
148. Baumann H, Richards C, Gauldie J. Interaction Nakagawa K. The clinical signicance of the
among hepatocyte-stimulating factors, measurement of plasma transferrin as a growth
interleukin 1, and glucocorticoids for regulation factor. II. The changes in various endocrine
of acute phase plasma proteins in human status. Nippon Naibunpi Gakkai Zasshi
hepatoma (HepG2) cells. J Immunol 1987;63:669674.
1987;139:41224128. 160. Smithies O. Variations in serum beta-globulins.
149. Ritzmann SE, Daniels JC. Serum electrophoresis Nature 1957;180:14821483.
and total serum proteins. In: Ritzmann SE, 161. Beckman G, Beckman L, Sikstrom C. Transferrin
Daniels JC, eds. Serum protein abnormalities: C subtypes in different ethnic groups. Hereditas
diagnostic and clinical aspects. New York: Alan 1980;92:189192.
R Liss, 1982. 162. Giblett E, Hickman C, Smithies O. Serum
150. Wallach J. Interpretation of diagnostic tests. transferrins. Nature 1959;183:15891590.
Philadelphia: Lippincott Williams & Wilkins, 163. Stibler H. The normal cerebrospinal uid
2000. proteins identied by means of thin-layer
References 105
185. Barnum SR, Fey G, Tack BF. Biosynthesis and 196. Fan Q, Uring-Lambert B, Weill B, Gautreau C,
genetics of C3. Curr Top Microbiol Immunol Menkes CJ, Delpech M. Complement component
1990;153:2343. C4 deciencies and gene alterations in patients
186. Tack BF, Morris SC, Prahl JW. Third component with systemic lupus erythematosus. Eur J
of human complement: structural analysis of the Immunogenet 1993;20:1121.
polypeptide chains of C3 and C3b. Biochemistry 197. Painter PC, Cope JY, Smith JL. Chapter 50.
1979;18:14971503. Reference information for the clinical laboratory.
187. McLean RH, Bryan RK, Winkelstein J. In: Burtis CA, Ashwood ER, ed. Tietz textbook
Hypomorphic variant of the slow allele of C3 of clinical chemistry. Philadelphia: WB Saunders,
associated with hypocomplementemia and 1999.
hematuria. Am J Med 1985;78:865868. 198. Handin RI, Laposata M. Case 2 1993
188. Taschini P, Addison N. Pseudomonoclonal A 72-year-old woman with a coagulopathy and
band caused by genetic polymorphism of the bilateral thigh masses. N Engl J Med 1993;328:
third component of complement. ASCP Check 121128.
Sample Immunopathol 1988:1. 199. Register LJ, Keren DF. Hazard of commercial
189. Platel D, Bernard A, Mack G, Guiguet M. antiserum cross-reactivity in monoclonal
Interleukin 6 upregulates TNF-alpha-dependent gammopathy evaluation. Clin Chem 1989;35:
C3-stimulating activity through enhancement of 20162017.
TNF-alpha specic binding on rat liver epithelial 200. Pepys MB. C-reactive protein fty years on.
cells. Cytokine 1996;8:895899. Lancet 1981;1:653657.
190. Tateda K, Matsumoto T, Yamaguchi K. Acute 201. Miettinen AK, Heinonen PK, Laippala P,
induction of interleukin-6 and biphasic changes Paavonen J. Test performance of erythrocyte
of serum complement C3 by carrageenan in mice. sedimentation rate and C-reactive protein in
Mediators Inamm 1998;7:221223. assessing the severity of acute pelvic
191. Platel D, Guiguet M, Briere F, Bernard A, Mack G. inammatory disease. Am J Obstet Gynecol
Human interleukin-6 acts as a co-factor for the up- 1993;169:11431149.
regulation of C3 production by rat liver epithelial 202. Teisala K, Heinonen PK. C-reactive protein in
cells. Eur Cytokine Netw 1994;5:405410. assessing antimicrobial treatment of acute pelvic
192. Katz Y, Revel M, Strunk RC. Interleukin 6 inammatory disease. J Reprod Med 1990;35:
stimulates synthesis of complement proteins 955958.
factor B and C3 in human skin broblasts. Eur J 203. Abdelmouttaleb I, Danchin N, Ilardo C, et al.
Immunol 1989;19:983988. C-Reactive protein and coronary artery disease:
193. Hirumi E, Safarian Z, Blessum C. Effect of additional evidence of the implication of an
sample age on electrophoretic resolution of C3 inammatory process in acute coronary
complement. Clin Chem 1995;41:S150(Abstr). syndromes. Am Heart J 1999;137:346351.
194. Smalley DL, Mayer R, Gardner C. Evaluation of 204. Lehtinen M, Laine S, Heinonen PK, et al. Serum
capillary zone electrophoresis assessment of beta C-reactive protein determination in acute pelvic
proteins. Clin Lab Sci 1999;12:262265. inammatory disease. Am J Obstet Gynecol
195. Blanchong CA, Zhou B, Rupert KL, et al. 1986;154:158159.
Deciencies of human complement component 205. Lacour AG, Gervaix A, Zamora SA, et al.
C4A and C4B and heterozygosity in length Procalcitonin, IL-6, IL-8, IL-1 receptor
variants of RP-C4-CYP21-TNX (RCCX) antagonist and C-reactive protein as
modules in caucasians. The load of RCCX identicators of serious bacterial infections in
genetic diversity on major histocompatibility children with fever without localising signs. Eur J
complex-associated disease. J Exp Med Pediatr 2001;160:95100.
2000;191:21832196. 206. Gendrel D, Raymond J, Coste J, et al.
References 107
Comparison of procalcitonin with C-reactive light chains: relative sensitivity for detection of
protein, interleukin 6 and interferon-alpha for monoclonal light chains. Clin Chem 2002;48:
differentiation of bacterial vs. viral infections. 14371444.
Pediatr Infect Dis J 1999;18:875881. 218. Taschini PA, Addison NM. An atypical
207. Grutzmeier S, Sandstrom E. C-reactive protein electrophoretic pattern due to a selective
levels in HIV complicated by opportunistic polyclonal increase of IgG4 subclass. ASCP
infections and infections with common bacterial Check Sample Immunopathol 1991:IP9195.
pathogens. Scand J Infect Dis 1999;31: 219. Batuman V, Sastrasinh M, Sastrasinh S. Light
229234. chain effects on alanine and glucose uptake by
208. Aufweber E, Jorup-Ronstrom C, Edner A, renal brush border membranes. Kidney Int
Hansson LO. C-reactive protein sufcient as 1986;30:662665.
screening test in bacterial vs. viral infections. 220. Schiff R, Schiff SE. Chapter 11. Flow cytometry
J Infect 1991;23:216220. for primary immunodeciency diseases. In:
209. Sheldon J, Riches P, Gooding R, Soni N, Hobbs Keren DF, McCoy JJP, Carey JL, eds. Flow
JR. C-reactive protein and its cytokine mediators cytometry in clinical diagnosis. Chicago: ASCP
in intensive-care patients. Clin Chem 1993;39: Press, 2001.
147150. 221. Kyle RA. Classication and diagnosis of
210. Magadle R, Weiner P, Sotzkover A, Berar-Yanay monoclonal gammopathies. In: Rose NR, ed.
N. C-reactive protein and vascular disease. Isr Manual of clinical laboratory immunology.
Med Assoc J 2001;3:5052. Washington, DC: American Society for
211. Canova CR, Courtin C, Reinhart WH. C-reactive Microbiology, 1986.
protein (CRP) in cerebro-vascular events. 222. Levinson SS, Keren DF. Immunoglobulins from
Atherosclerosis. 1999;147:4953. the sera of immunologically activated persons
212. Ridker PM, Cushman M, Stampfer MJ, Tracy with pairs of electrophoretic restricted bands
RP, Hennekens CH. Plasma concentration of C- show a greater tendency to aggregate than
reactive protein and risk of developing peripheral normal. Clin Chim Acta 1989;182:2130.
vascular disease. Circulation 1998;97:425428. 223. Keren DF, Morrison N, Gulbranson R. Evolution
213. Jacquiaud C, Bastard JP, Delattre J, Jardel C. C of a monoclonal gammopathy (MG) documented
reactive protein: a band mistaken as monoclonal by high-resolution electrophoresis (HRE) and
gammapathy in acute inammatory disease. Ann immunoxation (IFE). Lab Med 1994;
Med Interne (Paris) 1999;150:357358. 25:313317.
214. Yu A, Pira U. False increase in serum C-reactive 224. Tagliabue A, Boraschi D, Villa L, et al. IgA-
protein caused by monoclonal IgM-lambda: a dependent cell-mediated activity against
case report. Clin Chem Lab Med 2001;39: enteropathogenic bacteria: distribution,
983987. specicity, and characterization of the effector
215. Frazer JK, Capra JD. Chapter 3. cells. J Immunol 1984;133:988992.
Immunoglobulins: structure and function. In: 225. Tagliabue A, Nencioni L, Villa L, Keren DF,
Paul WE, ed. Fundamental immunology. Lowell GH, Boraschi D. Antibody-dependent
Philadelphia: Lippincott-Raven, 1999. cell-mediated antibacterial activity of intestinal
216. Bradwell AR, Carr-Smith HD, Mead GP, et al. lymphocytes with secretory IgA. Nature
Highly sensitive, automated immunoassay for 1983;306:184186.
immunoglobulin free light chains in serum and 226. Cunningham-Rundles C, Zhou Z, Mankarious S,
urine. Clin Chem 2001;47:673680. Courter S. Long-term use of IgA-depleted
217. Katzmann JA, Clark RJ, Abraham RS, et al. intravenous immunoglobulin in immunodecient
Serum reference intervals and diagnostic ranges subjects with anti-IgA antibodies. J Clin
for free kappa and free lambda immunoglobulin Immunol 1993;13:272278.
108 Proteins identied by serum protein electrophoresis
227. Yamaguchi M, Sayama K, Yano K, et al. IgE involved in defence against parasites. Nature
enhances Fc epsilon receptor I expression and 1994;367:183186.
IgE-dependent release of histamine and lipid 234. Corte G, Tonda P, Cosulich E, et al.
mediators from human umbilical cord blood- Characterization of IgD. I. Isolation of two
derived mast cells: synergistic effect of IL-4 and molecular forms from human serum. Scand J
IgE on human mast cell Fc epsilon receptor I Immunol 1979;9:141149.
expression and mediator release. J Immunol 235. Sinclair D, Craneld T. IgD myeloma: a potential
1999;162:54555465. missed diagnosis. Ann Clin Biochem 2001;38:
228. Pawankar R, Ra C. IgE-Fc epsilonRI-mast cell 564565.
axis in the allergic cycle. Clin Exp Allergy 236. de Dios Garcia-Diaz J, Alvarez-Blanco MJ. High
1998;28(Suppl 3):614. IgD could be a nonpathogenetic diagnostic
229. Sandor M, Lynch RG. The biology and marker of the hyper-IgD and periodic fever
pathology of Fc receptors. J Clin Immunol syndrome. Ann Allergy Asthma Immunol
1993;13:237246. 2001;86:587.
230. MacGlashan D Jr. Anti-IgE antibody therapy. 237. Summaries for patients. Can genetics help
Clin Allergy Immunol 2002;16:519532. diagnose the hyper-IgD and periodic fever
231. Imai S, Tezuka H, Fujita K. A factor of inducing syndrome. Ann Intern Med 2001;135:S36.
IgE from a larial parasite prevents insulin- 238. Picco P, Gattorno M, Di Rocco M,
dependent diabetes mellitus in nonobese diabetic Buoncompagni A. Non-steroidal anti-
mice. Biochem Biophys Res Commun 2001;286: inammatory drugs in the treatment of hyper-
10511058. IgD syndrome. Ann Rheum Dis 2001;60:904.
232. Selassie FG, Stevens RH, Cullinan P, et al. Total 239. de Hullu JA, Drenth JP, Struyk AP, van der Meer
and specic IgE (house dust mite and intestinal JW. Hyper-IgD syndrome and pregnancy. Eur J
helminths) in asthmatics and controls from Obstet Gynecol Reprod Biol 1996;68:223225.
Gondar, Ethiopia. Clin Exp Allergy 2000;30: 240. Haraldsson A, Weemaes CM, de Boer AW,
356358. Bakkeren JA. Clinical and immunological follow-
233. Gounni AS, Lamkhioued B, Ochiai K, et al. up in children with hyper-IgD syndrome.
High-afnity IgE receptor on eosinophils is Immunodeciency 1993;4:6365.
5
Approach to pattern interpretation
in serum
Initial processing of the sample 109 Protein loss through thermal injury 123
Overview of the electrophoretic strip 110 Acute-phase reaction pattern 124
Interpretation of the individual patients sample 112 Protein abnormalities in autoimmune disease 127
Serum pattern diagnosis 113 Protein patterns in hyperestrogenism 127
Liver disease patterns 114 g-Globulin patterns 128
Renal disease pattern 121 References 137
Gastrointestinal protein loss 123
As with any clinical laboratory method, control effects, improper storage, radiocontrast dye
serum samples must be run daily. Commercially effect on capillary zone electrophoresis (CZE),
available lyophilized controls are convenient and improperly labeled specimens, and
ensure adequate migration of the sample. Note that contaminated specimens.
C3 may not provide a discrete band in some 5. With gel-based methods use a densitometric
lyophilized preparations, altering the appearance scan for adjunctive information, but not as the
of the b-region. Densitometric scanning or electro- sole means to evaluate a specimen.
pherograms (for capillary zone electrophoresis) of 6. When encountering a pattern you do not
control serum provide objective criteria to deter- understand, call the clinician for more clinical
mine whether the percentage of the major bands is information.
within an acceptable range. Although some labora-
tories interpret gel-based serum protein electro-
phoresis without densitometric scanning, the
INITIAL PROCESSING OF THE
information available from densitometry is useful
in quality control, conrming impressions from
SAMPLE
direct visual examination of the gel and following
When the patients blood sample is received in the
patients with monoclonal gammopathies (M-
laboratory with a request for serum protein elec-
proteins).1 Below is a brief list of the approaches I
trophoresis, the serum should be allowed to clot,
recommend when interpreting protein electro-
separated from the clot and then refrigerated. The
phoretic patterns.
longer the sample sits at room temperature (with
1. Use all the information available, both clinical the many neutrophil proteases being released), the
and laboratory. more some key analytes (such as complement) dete-
2. Use a consistent logical approach. riorate. The sample should be checked to be sure the
3. Know the basics (Chapter 4). correct procedures have been followed in special
4. Know the artifacts: application problems, circumstances. For example, if the clinician has
hemolysis, incomplete clotting, drug-binding requested that the sample be examined for a cryo-
110 Approach to pattern interpretation in serum
globulin, be sure the sample has been drawn into a protein electrophoretic gels helps to avoid missing
prewarmed syringe, transported to the laboratory subtle alterations. Obvious ndings should be
at 37C and then centrifuged at 37C. If the sample noted and discussed (Fig. 5.1). Migration should
was not handled properly, redraw the patient.2 be consistent from one sample to the next. By
Previous electrophoretic results are of great reading down the strip, comparing the same
importance in following patients with M-proteins. protein band from sample-to-sample, the exam-
Therefore, as part of our initial specimen process- iner, in effect, uses adjacent samples as visual con-
ing, we check the laboratory le for any previous trols. This allows one to more easily note changes
electrophoretic studies performed on the patient. in concentration or subtle migration differences,
These les, together with the present material, are such as anodal slurring of albumin caused by drug
reviewed by the pathologist at the time of sign-out. or bilirubin binding. After reviewing albumin, I
It is recommended that clinicians provide pertinent proceed cathodally, observing the al-region for
information on the requisition slip. Even in a hos- increase or decrease in concentration and alter-
pital laboratory situation, however, it is often dif- ation of electrophoretic mobility of the a1-anti-
cult to obtain history; in a reference laboratory trypsin band (Fig. 5.2). By reading down the gel,
(off-site) the lack of history can be frustrating. This one is more likely to detect a double al-band or a
is why the person reviewing the electrophoretic shift in the migration of that band as will be seen
material should be aggressive about calling clini- with a1-antitrypsin Pittsburgh (see Chapter 4) than
cians when an unusual pattern occurs, or when the if one reads across a specimen on the gel.
present material is inconsistent with previous In the a2- and b-regions, monoclonal gammo-
reports. For example, if a patient has had an M-
protein documented by serum protein electro-
phoresis and immunological studies (characterizing
the M-protein) and the present sample shows no
evidence of a monoclonal process, one must be very
skeptical as to the correct identication of one of
the samples (how sure are you that the original was
the correct sample?). This requires contacting the
clinician and obtaining another sample, as well as a
little detective work to nd out where the sample
mix-up occurred.
OVERVIEW OF THE
ELECTROPHORETIC STRIP
On gel-based techniques, the basic information is
Figure 5.1 When comparing the top sample to the others on this
in the gel itself. The densitometric scan provides
gel, there is an obvious problem. The albumin band (1) migrates
useful objective measurable data. Therefore, when much too far toward the anode. There is a hazy zone between it
examining agarose or acetate gels, I prefer to and a second large band (2). Further, there is a densely staining
examine the gel itself rst to review the overall haze between the other bands and the g-region stains very lightly.
migration and staining of the bands. A study of the This is due to a double application to this gel. One was at the
normal location, and the other was anodal to it. The rst albumin
densitometric values for the control serum ensures
band (1) results from the anodal application and the second (2)
reliability for these values for the patient samples results from the correct application. Bisalbuminemia would not
of that run. have this great a separation and would not have the haze between
A consistent approach in examining serum the bands. (Paragon SPE2 system stained with Paragon Violet.)
Overview of the electrophoretic strip 111
(a)
Figure 5.2 By reading down the bands of this gel, one is struck by
the a1-antitrypsin band in the bottom lane (arrow). It stains more
weakly than those above it. Further, it has a much greater anodal
migration than the others. This is due to a homozygous PiFF (fast)
(typed by isoelectric focusing; clinically, they do not develop liver
or respiratory disease). Note that there is a slight difference in the
migrations of the a1-antitrypsin bands in the three top samples
because of the different M isoforms which have no clinical
signicance. (Paragon SPE2 system stained with Paragon Violet.)
laboratory that demonstrated the IgA l mono- a1-antitrypsin bands can be detected by paying
clonal gammopathy. Both high-resolution gel- close attention to this region of the electrophero-
based methods and CZE methods lend themselves gram.
to quantify the b1- and b2-regions thereby providing With regard to detecting subtle b-region mono-
objective evidence that an abnormality is present. clonal gammopathies, CZE offers high-resolution
After I have compared each band on the strip, I and crisp transferrin and C3 bands. By paying close
briey turn the gel 90 and look at them again. attention to the sharpness of these two analytes,
This only takes a minute (since I have already one can pick up small monoclonal gammopathies
formed an impression about any alterations seen). by detecting changes in and between these two
In addition to picking up several b-region mono- peaks. Figure 5.4 demonstrates the same subtle
clonal gammopathies, this examination technique monoclonal gammopathy by CZE that was shown
improves detection of a1-antitrypsin variants, in Fig. 5.3. Although one cannot read down the gel,
transferrin variants, or oligoclonal banding. I both currently available CZE systems provide a
know of no objective justication for this; perhaps control overlay that allows one to compare the
an analogy is when one turns one's head to the side crispness of the bands. Even immunosubtraction on
when looking at some paintings to try to achieve a such a subtle case is able to identify the monoclonal
different perspective. Try it. gammopathy as an IgA l (Fig. 5.5).
With CZE, there are no readily available methods
that compare with reading down the electro-
phoretic gel. Since each sample is processed INTERPRETATION OF THE
individually through a small capillary, there will be INDIVIDUAL PATIENTS SAMPLE
subtle variations in migration from one sample to
the next. However, CZE has the advantage of not After the entire strip has been examined and
having to rely upon a protein stain like Amido abnormalities noted, the pattern for each patient is
Black or Ponceau S. These dyes are taken up to reviewed and compared with the densitometric
varying extents depending on the degree of glyco- scan. When an abnormality such as an increase in
sylation of the protein. Therefore, the specic a1- the al-region is suggested by the initial review, it is
antitrypsin band can be distinguished from the useful to have objective, quantitative corrobora-
a1-acid glycoprotein (orosomucoid) band with this tion by the densitometric scan. However, with gel-
technique. Alterations in migration, or duplicate based methods I rely more on the visual inspection
Figure 5.4 The same College of American Pathologists (CAP) Survey sample from Fig. 5.3 is demonstrated on this capillary zone
electropherogram (right) with a control serum on the left. Here, I indicate the location of the monoclonal gammopathy (arrow). (Paragon
CZE 2000.)
Serum pattern diagnosis 113
(a)
Figure 5.6 (a) The top sample shows two faintly stained bands in
the a1-region (arrows) compared to the normal a1-region in the
other serum. (Panagel stained with Amido Black.) (b) A
densitometric scan indicates that the a1-region abnormality
observed in (a) can be seen (arrows) by enhancing the sensitivity of
(b) the densitometer.
Normal pattern
There is a decrease in all parameters suggesting hemodilution.
One measured value is slightly outside the normal range. The pattern is otherwise within normal limits.
Decreased albumin.
There is blurring of the anodal margin of albumin. This change may be caused by binding of a drug or bilirubin to
albumin. This may also be seen during heparin therapy where a1-lipoprotein may migrate anodally to albumin.
Markedly reduced a1-antitrypsin band. Recommend: a1-antitrypsin measurement and phenotype studies to exclude
congenital deciency.
Abnormal a1-antitrypsin band. Recommend: a1-antitrypsin measurement and phenotype studies to exclude congenital
deciency.
Relatively low a1-antitrypsin band. Recommend: a1-antitrypsin measurement to exclude congenital deciency.
Increased a1- and a2-globulins, decreased albumin and transferrin, consistent with acute inammation.
Relative elevation of acute phase proteins suggesting a reactive condition.
Increased a2-globulin and C3 band suggest subacute inammation.
Increased acute-phase reactants and polyclonal increase in immunoglobulins suggest active, chronic
inammation.
Decreased haptoglobin with a slow a2 band is likely due to hemoglobinhaptoglobin complex. This pattern suggests
hemolysis. Hemolysis may occur as a result of venepuncture technique.
Haptoglobin is decreased. This is consistent with a hemolytic process or hereditary deciency.
Increase in a2-globulin consistent with estrogen effect.a
Increase in a2- and b-globulins consistent with estrogen effect.a
The transferrin band is increased. If this is not due to estrogen effect, an evaluation for iron deciency is
warranted.a
Increased b1-band may be due to increased transferrin in iron deciency, or to the presence of an exogenous agent
such as a radiocontrast dye. Recommend: correlate with history and iron studies if warranted.b
The pattern reveals bg bridging. This suggests an increase in IgA. This may occur in liver disease, intestinal or
respiratory infections, and rheumatoid arthritis.
Decreased albumin and haptoglobin with a polyclonal increase in -globulins and bg bridging. This is consistent with
liver disease.
Decreased total protein with low albumin and -globulin. This is a mild protein loss pattern.
Very low albumin, transferrin and g-globulin with increased a2-globulin. This protein loss pattern suggests renal
disease (nephrotic syndrome) or gastrointestinal protein loss.
A small restriction is present at the application point on the gel. This may indicate the presence of a cryoglobulin.
Recommend: evaluate serum for cryoglobulin on a new sample drawn in a 37C tube and transported to the
laboratory at 37C.
Polyclonal increase in immunoglobulins is consistent with chronic inammation.
Borderline polyclonal increase in g-globulins suggesting a chronic reactive condition.
116 Approach to pattern interpretation in serum
There is a polyclonal increase in g-globulins with oligoclonal banding. This pattern may be seen in a variety of
conditions, most often chronic infections, occasionally autoimmune diseases and less commonly lymphoproliferative
processes.
A few tiny oligoclonal bands are seen in the g-region. These bands can be seen with infections and autoimmune
diseases. Occasionally, they have been noted as part of lymphoproliferative processes.
One of the oligoclonal bands is more prominent than the rest. Recommend: urine protein electrophoresis now and
repeat serum studies in 36 months.
Borderline hypogammaglobulinemia. This may be seen in some chronic lymphoproliferative conditions.
Hypogammaglobulinemia is present. In adults, this may reect the presence of immune deciency, chemotherapy or
B-cell neoplasm. Recommendation: immunoxation of urine or quantication of serum free k and l chains to
detect monoclonal free light chains (or Bence Jones protein). In addition, immunoxation of the serum and
measurement of IgG, IgA, and IgM will help to rule out a monoclonal process.
A tiny restriction is seen in the -region. This may indicate the presence of a lymphoproliferative process.
Recommend: immunoxation of urine or quantication of serum free k and l chains to detect monoclonal free
light chains (or Bence Jones protein) and repeat serum electrophoresis in 612 months to see if the process
regresses.
A monoclonal gammopathy is present. This indicates the presence of a lymphoproliferative process. Recommend:
serum and urine immunoxation.
Recommend: immunoxation of urine or quantication of serum free k and l chains to detect monoclonal free light
chains (or Bence Jones protein) if light chain disease is part of the differential diagnosis.
The monoclonal gammopathy is identied as ____ (specify type) and measures ____ (specify quantity measured by
densitometry or electropherograms measurement).
The monoclonal gammopathy persists with no signicant change from the previous study.
The monoclonal gammopathy persists signicantly greater than on the previous study.
The monoclonal gammopathy persists signicantly less than on the previous sample.
The normal -globulins appear suppressed.
To identify the monoclonal protein, please call (telephone number) to add an immunoxation to this sample. No
redraw is necessary.
Unusual electrophoretic pattern indicates that the sample is denatured. Please resubmit.
Repeat immunoxation is not recommended to follow patients with previously characterized M-proteins.
Serum and 24-h urine electrophoresis with measurement of the M-protein will provide quantitative information to
follow this patient. Alternatively, a serum measurement of serum free k and l chains will provide a useful
assessment of free light chains associated with the M-protein.
a
Use for women suspected of having estrogen effect.
b
Use only for capillary zone electrophoresis.
These suggested sign-outs should be adjusted for the needs and communication style of each institution. They are not all inclusive.
Often, I will combine different sign-outs as needed. Not infrequently, a narrative report is required when a sample does not t into a
particular category.
Liver disease patterns 117
even low) a1-antitrypsin, polyclonal hypergamma- to perform serum protein electrophoresis on ascites
globulinemia, beta-gamma bridging, and often, uid.13
decreased haptoglobin (Figs 5.7 and 5.8).7 The complex pathophysiology of hypoalbumine-
This relatively complex pattern is the result of mia prevents the interpreter from equating the
several pathophysiological alterations. Synthesis of level of albumin with prognosis.14 When hyper-
albumin and other proteins is affected by the bilirubinemia accompanies the liver disease, the
number of hepatocytes remaining, their current bilirubin binds to albumin, causing an anodal
state of health (damage by ethanol, toxins, or bio- slurring of this band (Figs 5.7 and 5.8). In addi-
logical agents), and the nutritional and metabolic tion to albumin, other hepatocyte-derived proteins
status of the individual resulting from diet and including al-acid glycoprotein (orosomucoid) and
hormonal changes.811 Clearly, when sufcient haptoglobin are often decreased.7 Whereas
hepatocytes are damaged, synthesis of albumin will transthyretin (prealbumin) is decreased consis-
be decreased, but this is not necessarily the cause of tently in cirrhosis and serves as a sensitive indica-
hypoalbuminemia in most cases of cirrhosis. tor of the level of hepatocyte function,15 serum
Although considerable loss of hepatocytes has protein electrophoresis is too insensitive for mean-
occurred in patients with cirrhosis and ascites, the ingful quantication of this band, even with the
large reserve capacity of the liver can result in a excellent resolution found with newer CZE tech-
normal or even elevated synthesis of serum niques.
albumin in some of these individuals.12 Therefore, Although transferrin is often decreased in cirrho-
hypoalbuminemia in many individuals results from sis, this is difcult to detect on serum protein
the altered distribution of albumin due to the pres- electrophoresis because of increased IgA. The bg
ence of ascites. Attempts to remove the ascites bridging in cirrhosis results from a marked poly-
result in an absolute loss of albumin. It is not useful clonal increase in the level of IgA that migrates in
Figure 5.7 Center serum has a classic cirrhosis pattern. Albumin is slightly decreased and moved anodally (indicated) due to bilirubin
binding. The a2-region is quite low. Prominent bg bridging (B) is seen together with a polyclonal increase in g-globulin. Note that the
top sample has a diffuse polyclonal increase in the slow g-region. The sample in the bottom lane is normal. (Panagel stained with Amido
Black.)
118 Approach to pattern interpretation in serum
Reference ranges
Rel % g/dl
ALBUMIN 52.6 68.9 3.80 5.20
ALPHA 1 3.6 8.1 0.30 0.60
ALPHA 2 5.3 12.2 0.40 0.90
BETA 8.3 14.3 0.60 1.10
GAMMA 8.2 18.6 0.60 1.40
Figure 5.8 This capillary zone electropherogram demonstrates the features of cirrhosis. Albumin is decreased and the band itself is
broadened toward the anode, likely because of bilirubin binding. Although the a2-region is in the normal range, it is at the lower end.
There is a prominent bg bridging and a polyclonal increase in g-globulin. (Paragon CZE 2000.)
albumin levels. However, an anodal slurring of clonal gammopathy. These bands reect the greater
albumin results from attachment of bilirubin (Figs expansion of those particular B-cell clones and are
5.9 and 5.10). As in cirrhosis, the a1-lipoproteins further evidence of the polyclonal nature of the
are usually decreased, but unlike cirrhosis the b1- process. When immunoxation is performed on
lipoproteins show no characteristic changes during such a sample, both k- and l-bands will usually be
hepatitis. identied. The polyclonal gammopathy is partly in
Examination of the g-globulin region can supply response to the inciting agent (typically hepatitis B
useful information about hepatitis patients. A poly- or C). The better prognosis, found in individuals
clonal gammopathy, occasionally with oligoclonal whose polyclonal and oligoclonal gammopathy
bands suggesting the presence of particular expan- regresses, parallels recovery of normal hepatocyte
sion of a few of the B-cell clones is often seen during structure with restoration of more normal intra-
clinically active hepatitis (Figs 5.9 and 5.10). The hepatic blood ow.25,26 Some clonal expansions in
polyclonal increase in patients with chronic active the g-region may be especially prominent, produc-
hepatitis tends to be especially prominent in the slow ing areas of restricted mobility in the slow g-region
g-region. Patients with acute hepatitis usually have that may be mistaken for monoclonal gammo-
a polyclonal gammopathy initially that may become pathies.27 Indeed, monoclonal gammopathies do
more pronounced if the disease progresses but usu- occur with increased frequency in patients with
ally regresses with clinical improvement.2224 hepatitis (see below). However, a small monoclonal
The polyclonal expansion may have two, three, or gammopathy superimposed on a polyclonal pattern
more small discrete bands that reect an oligo- may reect a transient process (Fig. 5.11).
Figure 5.9 Bottom serum is from a patient with hepatitis. The anodal slurring is caused by bilirubin binding to albumin. In the slow g-
region, several prominent bands (indicated) are superimposed on a polyclonal (diffuse) increase in g-globulin. There is no bg bridging. The
top two lanes contain normal serum. (Panagel stained with Amido Black.)
120 Approach to pattern interpretation in serum
Figure 5.10 Capillary zone electropherogram from a patient with active hepatitis. The a1-region is at the upper limit of normal and the
concentration of the a2-region is slightly increased. There is a prominent bg bridging with polyclonal and oligoclonal increase in the g-
region. The irregularity and slightly angled peaks indicate the oligoclonal banding. (Sebia Capillarys.)
No particular pattern will allow the interpreter to Chapter 6).28,29 Some of these patients may go on to
identify the causative agent of the hepatitis. This have one clone predominate and switch to a Type
should be accomplished by appropriate serological II cryoglobulin.30 In one study, 11 per cent of
and molecular investigations. The currently avail- patients with hepatitis C virus infection developed
able cadre of assays for specic antibodies, anti- an M-protein.31 The likelihood of this occurring
gens and virus-specic nucleic acid allows increased when the disease was caused by hepatitis
determination of the probable etiological agent, as C genotype 2a/c.31 If a large g band develops, I
well as often assisting the clinician in determining recommend characterizing it immediately by
the infectivity of the patient. The most serious error immunoxation, testing the serum for the presence
I have seen in interpretation of these patterns has of a cryoglobulin and evaluating the urine for the
been to overcall the polyclonal slow g increase as presence of monoclonal free light chains (MFLC;
a monoclonal gammopathy. Sometimes, aberrant see Chapter 7).
bound (i.e. intact immunoglobulin) k/l ratios can
develop in these patients, although the vast major- BILIARY OBSTRUCTION
ity have a ratio close to 2:0. Biliary obstruction is an end result of a wide variety
Patients with hepatitis C virus are particularly of disease processes, which may involve the hepatic
prone to develop Type III cryoglobulinemia (see parenchyma proper, such as in primary biliary
Renal disease pattern 121
Reference Ranges
Rel % g/dl
ALBUMIN 52.6 68.9 3.80 5.20
ALPHA 1 3.6 8.1 0.30 0.60
ALPHA 2 5.3 12.2 0.40 0.90
BETA 8.3 14.3 0.60 1.10
GAMMA 8.2 18.6 0.60 1.40
Figure 5.11 Capillary zone electropherogram from a patient with hepatitis C infection. In addition to decreased albumin and a broad
polyclonal increase in g-globulin, a modest-sized monoclonal band is present (arrow). On cases with these features I note the presence of
the monoclonal gammopathy superimposed on a polyclonal pattern. I recommend performing a urine test to rule out the presence of
monoclonal free light chains, but caution the clinician that monoclonal gammopathies in the presence of a polyclonal pattern could indicate
the presence of a transient, reactive process. Because of this, I recommend a repeat electrophoresis of serum in 36 months to see if the
process resolves. (Paragon CZE 2000.)
cirrhosis, or which may result from external obstruc- demonstrates a characteristic pattern consisting of
tion of the biliary tree by stones, inammation, or a low serum albumin, a diffuse slurring of a1-
neoplasm. The main effect on the serum protein elec- lipoprotein anodal to albumin (possibly due to
trophoresis pattern is anodal slurring of albumin heparins activation of lipoprotein lipase; see
because of its association with bilirubin. Additional Chapter 4), decreased or low normal al globulin,
irregularities may include acute-phase reaction, ele- and decreased g-globulin. The a2- and b-regions are
vated C3, increased b1-lipoprotein, decreased a- often elevated (Figs 5.12 and 5.13).3638 Examining
lipoprotein, and occasionally elevated g-globulin, the selectivity index, however, provides better clin-
especially in patients with primary biliary cirrhosis ical information about ultimate outcome for the
(who usually have an elevated IgM level).3235 patients than the electrophoretic pattern of their
However these ndings do not allow a specic diag- serum proteins (see Chapter 7).39
nosis. Although the albumin band is broadened The nephrotic pattern results from loss of serum
(slurred) toward the anode because of the binding of proteins through the damaged nephron and the
bilirubin, it is normal or only slightly decreased in bodys attempts to restore the oncotic pressure by
concentration and IgA is only slightly elevated overproduction of large proteins that do not pass
(therefore, no bg bridge is seen). These features are through the damaged glomeruli. Nephrotic syn-
helpful in the differential diagnosis from the cirrho- drome is dened as proteinuria (> 3 g/24 h.1.73 m2
sis and hepatitis patterns. of body surface area), hypoproteinemia, edema, and
hyperlipidemia. Considerably less proteinuria can
produce a clinical picture of nephrosis, especially in
RENAL DISEASE PATTERN children having a relapse of renal disease.40 Electro-
phoresis of serum will only detect the severe cases
When renal disease is severe enough to produce the of renal damage and, as mentioned above, is of little
nephrotic syndrome, serum protein electrophoresis help in dening the specic site of damage in the
122 Approach to pattern interpretation in serum
Reference ranges
Rel % g/dl
ALBUMIN 52.6 68.9 3.80 5.20
ALPHA 1 3.6 8.1 0.30 0.60
ALPHA 2 5.3 12.2 0.40 0.90
BETA 8.3 14.3 0.60 1.10
GAMMA 8.2 18.6 0.60 1.40
Figure 5.13 Capillary zone electropherogram demonstrates more clearly some of the quantitative features of a nephrotic pattern.
Albumin and g-globulin are decreased with a prominent increase in a2-globulin. (Paragon CZE 2000.)
Protein loss through thermal injury 123
Elevated cholesterol in patients with nephrotic nephrotic pattern but, in contrast, it is unusual to
syndrome is directly related to the increased b1- have an elevated b1-lipoprotein in patients with
lipoprotein.41,42 Because of the self-aggregation that protein-losing enteropathies. Whicher42 suggested
occurs with b1-lipoprotein at elevated concentra- that the amount of protein loss into the gastro-
tions (Chapter 4), its electrophoretic mobility is intestinal tract can be estimated by measuring the
slowed and it appears as an irregular band catho- a1-antitrypsin excretion in stool. Obviously, the
dal to its usual location using gel-based methods. features of protein-losing enteropathy are far too
Samples from patients receiving hemodialysis non-specic for serum protein electrophoresis pat-
may be incompletely clotted by the heparin that terns to suggest a specic diagnosis. An absolute
they are given during their treatment. In addition distinction between a nephrotic and a protein-
to a band in the brinogen region, patients losing enteropathy pattern cannot be made reliably
receiving heparin have the a1-lipoprotein band just from looking at the electrophoresis pattern.
migrating anodal to albumin and the b1-lipoprotein Clinical history and ancillary laboratory informa-
band migrates in the a2 region owing to activation tion are needed for the diagnostic process.
of lipoprotein lipase and the binding of heparin to
b1-lipoprotein respectively (see Chapter 4).43,44
PROTEIN LOSS THROUGH
THERMAL INJURY
GASTROINTESTINAL PROTEIN
LOSS Thermal injury to the skin produces a large surface
area through which serum protein is lost. In the
Damage to the gastrointestinal tract will affect the rst few days following thermal injury, there is an
serum protein electrophoresis pattern in a variety increase in acute-phase reactants, including C-reac-
of ways. Acute damage from invasive bacteria or tive protein and a1-antitrypsin along with a2-
acute exacerbation of inammatory bowel disease macroglobulin (the last is not typically an
produces an acute-phase reaction pattern. acute-phase reactant its rise may reect renal loss
Similarly, dehydration due to agents such as Vibrio of other proteins see below), while albumin is
cholera or staphylococcal enterotoxin B will result decreased.46 At the same time, patients with
in an increase in the concentration of all serum thermal injury experience a rapid fall in the levels
proteins. Patients with inammatory bowel disease of transthyretin (prealbumin) that reach their nadir
usually have a polyclonal increase in the g-globulin at 6 days post-burn.47 The administration of low
region reecting the chronic inammatory nature dose insulin-like growth factor-1/binding protein-3
of that condition. Also, depending on the extent intravenously to children with burns covering > 40
and severity of the process, absorption of nutrients, per cent of their skin has been shown to increase
including amino acids, may be impaired and serum the levels of constitutive serum proteins such as
protein may be lost into the lumen of the bowel. transthyretin and transferrin, while attenuating the
One cannot distinguish between ulcerative colitis increases in a1-antitrypsin, a1-acid glycoprotein,
and Crohns disease by this technique. and haptoglobin.48
Protein-losing enteropathy may occur at any age The early thermal injury pattern differs from the
as the result of a wide variety of pathological nephrotic and protein-losing enteropathy patterns
processes including: gluten-sensitive enteropathy, in that the last two usually show decreased al-
acute enteritis, Whipples disease, and globulin.46 After burn injury, protein is also lost
HenochSchnlein purpura. The serum protein into the urine. Initially the pattern of loss resembles
electrophoresis pattern usually displays hypoalbu- a mild glomerular proteinuria that may transform
minemia, occasionally with decreased -globulins.45 to a tubular proteinuria (see Chapter 7).49
a2-Macroglobulin may be increased as in the Studies of serum from burn patients with sodium
124 Approach to pattern interpretation in serum
has a diffuse increase in density caused by the rapid concentration of C-reactive protein can be quite
increase of al-acid glycoprotein. However, when dramatic after some inammatory stimuli (as much
using CZE, the al-acid glycoprotein band can be as 1000-fold).60 Quantication of the C-reactive
distinguished as a shoulder to the al-antitrypsin protein (by immunochemical means) has been rec-
band (Fig. 5.16). C-Reactive protein is not ommended to detect early infection in patients with
detectable by serum protein electrophoresis in leukemia who often do not demonstrate typical
control specimens (normal concentration granulocyte responses to the infection.61 I have not
< 2 mg/dl), but with a vigorous acute-phase reac- observed a C-reactive protein band in CZE
tion, it may be seen as a small band in the mid-g- samples of acute-phase reaction.
region on gel-based systems (Fig. 5.14). As such, it By 1224 h after the onset of inammation, a1-
may be confused with a small monoclonal gam- antitrypsin and haptoglobin levels have usually
mopathy. By observing the presence of the other increased (Table 5.2). Because a1-antitrypsin
elements of the acute-phase pattern, noting other levels increase during acute tissue damage even in
laboratory values and appropriate clinical history, patients with a1-antitrypsin deciency, the pres-
one will not be led astray. If there is any doubt, a ence of acute inammation may obscure the
serum immunoxation reaction will quickly presence of this deciency (see Chapter 4). In
demonstrate that the band is not an immuno- such individuals it is important to perform a con-
globulin (see Chapter 3). The increase in serum comitant measurement of C-reactive protein to be
126 Approach to pattern interpretation in serum
Table 5.2 Time coursea of positive acute-phase reactants which usually occur in an acute-phase reaction (iron
affect the high-resolution electrophoresis pattern deciency being an important exception).
Haptoglobin is variable in the acute-phase reac-
Protein Earliest Peak tion. Although it is usually increased 24 h after
elevation elevation (h) an acute episode of tissue injury, if hemolysis
accompanies the acute tissue injury then hapto-
C-reactive protein 612 h 4872 globin will bind hemoglobin and this product
a1-acid glycoprotein 1224 h 4872 will be removed by the reticuloendothelial
a1-antitrypsin 24 h 7296 system, resulting in a decreased haptoglobin
a1-antichymotrypsin 24 h 7296
level. Fibrinogen is increased within 24 h of the
injury, but one needs to quantify brinogen in
Haptoglobin 24 h 7296
plasma to measure this. Less consistently, eleva-
Fibrinogen 24 h 7296 tions in C3, hemopexin, ceruloplasmin, and Gc
C3 47 days globulin are seen within a week after acute tissue
a
damage.52,53,6264
Following a single acute episode of tissue damage. Note that
many diseases have ongoing tissue injury, which results in
As noted in Table 5.2, the elevation in C3 occurs
more complicated, overlapping elevations of several later during the inammatory response and I refer
components. Data modied from Ritchie and Whicher,62 and to it as an indicator of a later or subacute stage of
Fischer et al.63
the inammation. C3, however, is an inconsistent
marker because both synthesis and catabolism may
be increased. During the inammatory process,
certain that the patient does not have evidence of complement is activated by the alternative and/or
an acute-phase pattern. Pregnancy or the use of classical pathway (depending on the initial mode of
birth control pills is usually accompanied by an stimulation see Chapter 4), therefore, the
elevation of a1-antitrypsin, but these also produce increased C3 produced may be used up in the
an elevation in transferrin, which does not process.65
Reference ranges
Rel % g/dl
ALBUMIN 52.6 68.9 3.80 5.20
ALPHA 1 3.6 8.1 0.30 0.60
ALPHA 2 5.3 12.2 0.40 0.90
BETA 8.3 14.3 0.60 1.10
GAMMA 8.2 18.6 0.60 1.40
Figure 5.16 Capillary zone electropherogram from a patient with an acute-phase reaction, In addition to the features mentioned in
Figure 5.15, this serum has a prominent anodal shoulder to a1-antitrypsin due to an increase in a1-acid glycoprotein (orosomucoid)
(arrow). (Paragon CZE 2000.)
Protein patterns in hyperestrogenism 127
Several proteins consistently decrease in concen- where C-reactive protein usually does not correlate
tration following an episode of acute tissue injury. with inammation. An increasing C-reactive
These negative acute-phase reactants are useful protein in patients with systemic lupus erythe-
guides in distinguishing an acute-phase pattern matosus may indicate a coexistent bacterial infec-
from estrogen effect. Typically, albumin and trans- tion.67
ferrin decrease within a few days following injury
(Table 5.3). However, the concurrent presence of
an iron deciency anemia can produce an elevation
PROTEIN ABNORMALITIES IN
of transferrin which would produce an atypical
pattern.65 Although transthyretin (prealbumin) and
AUTOIMMUNE DISEASE
a1-lipoprotein also decline during an acute-phase
While early studies held some promise for the use
reaction, such a decrease is not reliably detected by
of electrophoresis in the diagnosis or prognosis of
serum protein electrophoresis because of the small
autoimmune diseases,68,69 it has become clear that
quantity of transthyretin and the diffuse staining
the electrophoretic ndings in these conditions are
and wide variation of a1-lipoprotein in the general
too non-specic and varied to provide much useful
population.
information. There is often a polyclonal increase in
The acute-phase pattern is non-specic in that it
g-globulin. Further, during active autoimmune
can be seen following a wide variety of tissue
disease, circulating immune complexes are often
injuries; however, when some clinical history is
accompanied by an oligoclonal pattern. An acute-
provided, its detection can be useful in conrming
phase reaction is seen during acute exacerbations;
clinical impressions, in timing an internal injury
the a2-globulin fraction may be altered with hapto-
event, and in predicting infections in patients with
globin binding to hemoglobin during episodes of
impaired leukocyte responses such as individuals
intravascular hemolysis.70 Overall, screening for
with leukemia or others receiving chemotherapy.66
autoimmune diseases and for following particular
In patients with autoimmune diseases, C-reactive
patients, specic autoantibody titers and evidence
protein levels have been useful as a more objective
of in vivo complement activation such as CH50, C3
indication of disease activity than other tests such
and C4 levels are more useful than serum protein
as erythrocyte sedimentation rate.54 Interestingly,
electrophoresis.
this is not true in systemic lupus erythematosus
polyclonal nature of this increase can be demon- bridge.87 About half of the patients have
strated. splenomegaly and lymphoid aggregates in bone
Polyclonal increases in IgM can present a difcult marrow express FMC7.86 Immunophenotyping in
diagnostic problem (Fig. 5.22). They must be dis- the case reported by Woessner et al.87 demon-
tinguished from monoclonal lesions by either strated these cells to be CD19+, HLA-DR+, IgM+,
immunoxation or immunosubtraction. The case CD3-, CD5-, CD23-, k+, and l+ (the last two
in Fig. 5.22 was polyclonal. Recently, a benign markers prove the polyclonal nature of the pro-
lymphoproliferation has been demonstrated with a liferation).87 The lymphocytosis persists for years,
polyclonal increase in IgM. Persistent polyclonal B- and although it is still unclear whether this is a pre-
cell lymphocytosis (PPBL) presents with a poly- malignant or benign condition, aggressive treat-
clonal increase in IgM, low to normal levels of IgA ment has been discouraged.85,88
and IgG in association with a polyclonal B-cell Polyclonal increase in IgM has also been reported
lymphocytosis.85 The condition occurs more often in the X-linked hyper-IgM immunodeciency syn-
in women, and especially in those that are cigarette drome.89 In addition to polyclonal increases in
smokers.85 PPBL has been reported in identical IgM, these patients have hypogammaglobulinemia.
twins, is associated with HLA-DR7 phenotype and As its name implies, these cases present during
in three-quarters of the patients, a +i(3Q) chromo- infancy with recurrent infections, diarrhea and
somal abnormality.85,86 The peripheral blood of often require parenteral nutrition to combat failure
these patients contains bilobed lymphocytes where to thrive.9097 The defect is caused by a mutation
the two lobes are joined by a delicate chromatin in the CD40 ligand (CD40L), which results in
130 Approach to pattern interpretation in serum
Figure 5.21 This capillary zone electropherogram (Paragon CZE Figure 5.22 This capillary zone electropherogram (Paragon CZE
2000) demonstrates a rounded area of restriction in the b-g region 2000) demonstrates a rounded area of restriction in the mid-g-
(arrow). Whereas most monoclonal gammopathies have a more region (arrow). I recommend performing an immunoxation or
narrow restriction, I have seen cases of IgA and IgD monoclonal immunosubtraction on such restrictions. They are unusual, could
gammopathies with similar appearances. In this case, the be monoclonal gammopathies, and should have the polyclonal
restriction was caused by a polyclonal increase in IgG4 subclass. nature documented. This is a polyclonal increase of IgM.
-Globulin patterns 131
disturbed antibody production and poor T-cell increases in g-globulin, often accompanied by
function.98 oligoclonal, or occasionally monoclonal gammo-
pathies (Figs 5.24 and 5.25).100106 Some of these
DIFFUSE OLIGOCLONAL bands have been shown to have specicity for
The prex oligo is Greek and means having few. human immunodeciency virus antigens.107
In a variety of benign and malignant conditions, a Taichman et al.108 reported a positive correlation
few clones of B cells proliferate in excess. between the presence of oligoclonal bands and the
Sometimes these are in response to an obvious degree of anti-human immunodeciency virus
stimulus such as in viral hepatitis or in bacterial (HIV) antibody positivity. In their study, they
endocarditis. In those cases, the responses are tran- found that 43 per cent of patients that were HIV
sient, although they may persist for months. Other antibody positive had oligoclonal bands by serum
oligoclonal expansions are the result of signi- protein electrophoresis. These patterns are the
cantly altered immunoregulatory mechanisms, result of immune suppression due to decreased
such as angioimmunoblastic lymphadenopathy CD4 lymphocytes and extraordinary antigenic
where aberrant T cell clones (usually CD4+) result stimulation by the many infections they contract.
in massive polyclonal B-cell proliferation.99 These In other instances, patients with autoimmune
oligoclonal responses were once thought to repre- diseases can have complex serum protein electro-
sent immune complexes, but it is now clear that phoresis patterns because they may have aberrant
they reect the immunoglobulin products of a few immunoregulation with T-cell dysfunction, be
clones of B cells/plasma cells. Patients with chronic receiving a variety of immunosuppressive regi-
lymphocytic leukemia and chronic B-cell lym- mens, suffer from infections related to the
phoma may have hypogammaglobulinemia and immunosuppression, and may be receiving plasma-
oligoclonal bands (Fig. 5.23) pheresis or g-globulin replacement therapy.
Occasionally, a combination of several factors is Oligoclonal gammopathies have been described
responsible for an oligoclonal pattern. For in the serum of patients with circulating immune
example, transplant patients have profound complexes.109111 Kelly et al.109,110 pointed out that
immunosuppression from the required therapy and during an oligoclonal B-cell expansion due to an
they suffer from a variety of infections because of infection, individual clones do not always achieve
this suppression. Also profoundly immunosup- equivalent serum concentrations. Furthermore, the
pressed by their disease, acquired immune-de- peak response of one clone may not occur synchro-
ciency syndrome (AIDS) patients have a complex nously with those of others. Thus, a particular
pattern with occasionally massive polyclonal sample from such a patient may show only the
larger peak that could be mistaken for a mono-
clonal gammopathy that was part of a lymphopro-
liferative process.109,112
Oligoclonal bands may be particularly prominent
in chronic hepatitis B and C.25,113 These are other
instances of infectious diseases where M-proteins
have been reported.114118 One must be cautious,
however, when describing an M-protein in a poly-
clonal background, or a background with oligo-
clonal expansion. In some infectious diseases, the
monoclonal gammopathy is transient. The M-
Figure 5.23 The bottom sample has a hypogammaglobulinemia
with three or four oligoclonal bands (three are indicated). The
protein may persist for as long as 6 months.119
sample above has a normal g-globulin region for comparison. Monoclonal gammopathies may be especially
(Paragon SPE2 system stained with Paragon Violet.) prominent in bacterial endocarditis caused by
132 Approach to pattern interpretation in serum
Figure 5.24 The middle sample shows a massive polyclonal expansion of g-globulins in a patient with acquired immunodeciency
syndrome (AIDS). This massive expansion is also seen in patients with hepatitis and occasionally in angioimmunoblastic lymphadenopathy.
This pattern is not diagnostic of AIDS, but this, or a similar polyclonal g increase with or without oligoclonal bands should cause one to
include these conditions in the differential diagnosis. (Paragon SPE2 system stained with Paragon Violet.)
Figure 5.26 (a) The top serum is from a patient who had
angioimmunoblastic lymphadenopathy that evolved to
immunoblastic lymphoma. The prominent bg bridging together
with the marked diffuse increase in g-globulins, and the slightly
decreased albumin, give this serum the appearance of a cirrhotic
pattern. There are also several small areas of restriction
(oligoclonal bands) in this g-region (indicated on the
immunoxation in (b). SPE, serum protein electrophoresis.
(Paragon SPE1 system stained with Paragon Violet.) (b)
Immunoxation of the top serum from (a) identies several of the
oligoclonal bands (indicated) as IgG, IgM, k, and l. This emphasizes
the altered polyclonal B-cell proliferation that occurs in some
cases of immunoblastic lymphoma. SPE, serum protein
electrophoresis. (Paragon immunoxation stained with Paragon
Violet; anode at the top.) Photographs contributed by Dorothy
Wilkins.
(a)
regress when the clinician decreases the immuno- The type of immunosuppression has a major
suppressive therapy. Recently, the use of anti- effect on when the disorder begins. For example,
CD20 (rituximab) has been shown to be effective patients receiving either cyclosporine A or mono-
in treating post-transplantation lymphoprolifera- clonal anti-T3 (OKT3) can develop PTLD rela-
tive disorder (PTLD).143 tively early (1 month after transplantation),
-Globulin patterns 135
whereas individuals receiving neither of these pattern, either renal or gastrointestinal loss (see
drugs may take years before PTLD develops.144 The above). Isolated hypogammaglobulinemia, how-
disease is often regional, but may be widespread. ever, with other serum protein fractions in the
The latter has the worse prognosis. normal range, implies an immunological abnor-
Both monoclonal and oligoclonal gammopathies mality. A low g-globulin in an individual older
have been found in these individuals. Myara et than 2 years of age may indicate an abnormal
al.105 found that 35 per cent of their 76 heart trans- immune system, and should be pursued. The
plant patients had oligoclonal bands; 15 per cent immunodeciency may be congenital or acquired
had monoclonal components. The most common owing to suppression by a neoplasm or suppres-
protein band found was IgG and it was seven times sion by chemotherapy. Even individuals in their
more common than IgM; IgA was not found in eighth and ninth decades should have normal g-
their series.105 Oligoclonal bands in these patients regions (assessed by densitometric scan) (Fig.
are usually present in low concentration, and are 5.27). Knowledge of the age and sex of the patient
usually transient. Monoclonal gammopathies in are of great importance in evaluating a patient with
these patients may precede the demonstration of hypogammaglobulinemia.
PTLD.145,146
Monoclonal gammopathies are discussed in BRUTONS X-LINKED AGAMMAGLOBULINEMIA
Chapter 6. Brutons X-linked agammaglobulinemia (XLA) is a
congenital deciency of B cells in males which
occurs once in each 100 000 live male births.147,148 It
Decreased g-globulin is caused by mutations in the gene coding for
Brutons agammaglobulinemia tyrosine kinase.149 patients vary from one to another. This indicates
Their bone marrow and peripheral blood B-lym- that CVID is really a cluster of diseases which
phocytes are markedly decreased.150 The few circu- share the end-stage characteristic of extreme
lating B lymphocytes they do possess may express hypogammaglobulinemia.155 Many cases of CVID
the CD20 surface marker, but not CD21.147 A mat- demonstrate benign lymphoid lesions, including
urational arrest occurs between the cytoplasmic atypical lymphoid hyperplasia, reactive lymphoid
negative and positive stages.150 Because maternal hyperplasia and chronic granulomatous inamma-
IgG crosses the placenta, the newborn with XLA tion.156
has a normal g-globulin until about 6 months of
age. IgG has a half-life of about 21 days, therefore, LYMPHOPROLIFERATIVE DISEASE
after about 6 months the maternal IgG has been As discussed in Chapter 6, light chain disease,
catabolized. By that age, most normal children will chronic lymphocytic lymphoma and well-differen-
begin to synthesize their own IgG. Patients with tiated lymphocytic lymphoma are commonly asso-
XLA will not be able to synthesize IgG and begin ciated with isolated hypogammaglobulinemia
to suffer from recurrent pyogenic, bacterial infec- (Table 5.4). This is usually a much more subtle
tions because of the lack of opsonins and comple- decrease than the hypogammaglobulinemia associ-
ment-xing antibody.89 ated with the immunodeciency diseases men-
tioned above. When the deciency in g-globulin is
TRANSIENT HYPOGAMMAGLOBULINEMIA OF profound, intravenous g-globulin therapy has been
INFANCY (THI) reported to decrease bacterial infections.157
THI must be considered in the differential diagno- Therefore, in adults, any decrease in the g-globulin
sis of hypogammaglobulinemia pattern in an infant. region below our two standard deviation limits
This condition has been associated with a delayed causes us to note its existence in a narrative sign-
maturation of helper CD4 lymphocytes.151153 They out. We also recommend performing immunoxa-
possess normal numbers of peripheral blood B lym- tion on a urine sample to rule out the possibility of
phocytes. Some patients with transient hypo- a monoclonal free light chain (Bence Jones
gammaglobulinemia of infancy develop mucosal protein).
infections (usually respiratory), however, they do
not suffer from the pattern of recurrent pyogenic IATROGENIC HYPOGAMMAGLOBULINEMIA
infections seen in patients with XLA. Some forms of therapy suppress the immune
system. Chemotherapy may be given for a variety
COMMON VARIABLE IMMUNODEFICIENCY of autoimmune and neoplastic diseases.
DISEASE Depending on dose and duration, this may
Common variable immunodeciency disease decrease the g-globulin levels. Plasma exchange is
(CVID) was the most common form of acquired now commonly used for autoimmune diseases and
immunodeciency disease in adults prior to the for hyperviscosity syndromes. Immediately after
AIDS epidemic. It usually manifests with diarrhea plasma exchange, most plasma components are
or respiratory tract infections in young adults. The decreased, but they usually recover within a day
disease is uncommon in childhood or in the elderly. or two. It takes about 2 weeks for IgG to return
Despite the similarity in name to acquired immune to its former value, whereas IgA and IgM often
deciency, the pathogenesis is not related to HIV. return to their reference values within a week of
The electrophoretic pattern shows extreme plasma exchange.158 Therefore, when hypogamma-
hypogammaglobulinemia (usually less than globulinemia is detected on protein electrophore-
250 mg/dl of IgG). These patients respond well to sis, clinical correlation is important to help avoid
-globulin replacement and judicious antibiotic confusion as to the signicance of this nding for
therapy.154 The cell surface marker patterns in these the patient.
References 137
24. Hirayama C, Tominaga K, Irisa T, Nakamura 35. Teramoto T, Kato H, Hashimoto Y, Kinoshita
M. Serum gamma-globulins and hepatitis- M, Toda G, Oka H. Abnormal high density
associated antigen in blood donors, chronic liver lipoprotein of primary biliary cirrhosis analyzed
disease and primary hepatoma. Digestion by high performance liquid chromatography.
1972;7:257265. Clin Chim Acta 1985;149:135148.
25. Tsianos EV, Di Bisceglie AM, Papadopoulos 36. Jensen H. Plasma protein and lipid pattern in the
NM, Costello R, Hoofnagle JH. Oligoclonal nephrotic syndrome. Acta Med Scand
immunoglobulin bands in serum in association 1967;182:465473.
with chronic viral hepatitis. Am J Gastroenterol 37. Markiewicz K. Serum lipoprotein pattern in the
1990;85:10051008. nephrotic syndrome. Pol Med J
26. Perez Losada J, Losa Garcia JE, Alonso Ralero L, 1966;5:737743.
Lerma Marquez J, Sanchez Sanchez R. Oligoclonal 38. Ghai CL, Chugh KS, Kumar S, Chhuttani PN.
gammapathy and infection by hepatitis B virus. An Electrophoretic pattern of serum proteins in
Med Interna 1998;15:397398. secondary renal amyloidosis with nephrotic
27. Demenlenaere L, Wieme RJ. Special syndrome. J Assoc Physicians India
electrophoretic anomalies in the serum of liver 1966;14:141144.
patients: a report of 1145 cases. Am J Dig Dis 39. Bazzi C, Petrini C, Rizza V, Arrigo G, DAmico
1961;6:661675. G. A modern approach to selectivity of
28. Staak JO, Glossmann JP, Diehl V, Josting A. proteinuria and tubulointerstitial damage in
Hepatitis-C-virus-associated cryoglobulinemia. nephrotic syndrome. Kidney Int
Pathogenesis, diagnosis and treatment. Med Klin 2000;58:17321741.
2002;97:601608. 40. Robson AM, Loney LC. Nephrotic syndrome. In:
29. Kayali Z, Buckwold VE, Zimmerman B, Schmidt Conn RB, ed. Current diagnosis. Philadelphia:
WN. Hepatitis C, cryoglobulinemia, and WB Saunders, 1985.
cirrhosis: a meta-analysis. Hepatology 41. Appel GB, Blum CB, Chien S, Kunis CL, Appel
2002;36:978985. AS. The hyperlipidemia of the nephrotic
30. Agnello V. The etiology and pathophysiology of syndrome. Relation to plasma albumin
mixed cryoglobulinemia secondary to hepatitis C concentration, oncotic pressure, and viscosity.
virus infection. Springer Semin Immunopathol N Engl J Med 1985;312:15441548.
1997;19:111129. 42. Whicher JT. The interpretation of
31. Andreone P, Zignego AL, Cursaro C, et al. electrophoresis. Br J Hosp Med 1980;24:
Prevalence of monoclonal gammopathies in 348356.
patients with hepatitis C virus infection. Ann 43. Su LD, Keren DF. The effects of exogenous free
Intern Med 1998;129:294298. fatty acids on lipoprotein migration in serum
32. Hickman PE, Dwyer KP, Masarei JR. high-resolution electrophoresis: addition of free
Pseudohyponatraemia, hypercholesterolaemia, fatty acids improves visualization of normal and
and primary biliary cirrhosis. J Clin Pathol abnormal alpha1-antitrypsin. Am J Clin Pathol
1989;42:167171. 1998;109:262267.
33. Lindgren S. Accelerated nonproteolytic cleavage 44. Pearson JP, Keren DF. The effects of heparin
of C3 in plasma from patients with primary on lipoproteins in high-resolution
biliary cirrhosis. J Lab Clin Med electrophoresis of serum. Am J Clin Pathol
1989;114:655661. 1995;104:468471.
34. Loginov AS, Reshetniak VI, Chebanov SM, 45. Reif S, Jain A, Santiago J, Rossi T. Protein losing
Matiushin BN. Clinical signicance of gel enteropathy as a manifestation of
electrophoresis of blood lipoproteins in primary HenochSchonlein purpura. Acta Paediatr Scand
biliary cirrhosis. Klin Lab Diagn 1996:3335. 1991;80:482485.
References 139
46. Latha B, Ramakrishnan KM, Jayaraman V, Babu 55. Nakayama T, Sonoda S, Urano T, Yamada T,
M. Action of trypsin:chymotrypsin (Chymoral Okada M. Monitoring both serum amyloid protein
forte DS) preparation on acute-phase proteins A and C-reactive protein as inammatory markers
following burn injury in humans. Burns in infectious diseases. Clin Chem 1993;39:293297.
1997;23(Suppl 1):S37. 56. Mariotti R, Musumeci G, De Carlo M, et al.
47. Moody BJ. Changes in the serum concentrations Acute-phase reactants in acute myocardial
of thyroxine-binding prealbumin and retinol- infarction: impact on 5-year prognosis. Ital Heart
binding protein following burn injury. Clin Chim J 2001;2:294300.
Acta 1982;118:8792. 57. Ikeda T, Kawakami K, Fujita J, Bandoh S,
48. Spies M, Wolf SE, Barrow RE, Jeschke MG, Yamadori I, Takahara J. Thymic carcinoma
Herndon DN. Modulation of types I and II acute associated with a high serum level of interleukin
phase reactants with insulin-like growth factor- 6 diagnosed through the evaluation for
1/binding protein-3 complex in severely burned asymptomatic elevation of acute-phase reactants.
children. Crit Care Med 2002;30:8388. Intern Med 1998;37:414416.
49. Shakespeare PG, Coombes EJ, Hambleton J, 58. Kushner I. Regulation of the acute phase response
Furness D. Proteinuria after burn injury. Ann by cytokines. Perspect Biol Med 1993;36:611622.
Clin BioChem 1981;18:353360. 59. Jiang SL, Samols D, Sipe J, Kushner I. The acute
50. Huang WH. Study of the abnormal plasma phase response: overview and evidence of roles
proteins after burn injury. Zhonghua Zheng Xing for both transcriptional and post-transcriptional
Shao Shang Wai Ke Za Zhi 1992;8:8992, 163. mechanisms. Folia Histochem Cytobiol
51. Ritchie RF, Palomaki GE, Neveux LM, 1992;30:133135.
Navolotskaia O, Ledue TB, Craig WY. Reference 60. Kushner I. C-reactive protein in rheumatology.
distributions for the positive acute phase serum Arthritis Rheum 1991;34:10651068.
proteins, alpha1-acid glycoprotein 61. Schoeld KP, Voulgari F, Gozzard DI, Leyland
(orosomucoid), alpha1-antitrypsin, and MJ, Beeching NJ, Stuart J. C-reactive protein
haptoglobin: a practical, simple, and clinically concentration as a guide to antibiotic therapy in
relevant approach in a large cohort. J Clin Lab acute leukaemia. J Clin Pathol 1982;35:866869.
Anal 2000;14:284292. 62. Ritchie RF, Whicher JT. Acute phase proteins:
52. Ritchie RF, Palomaki GE, Neveux LM, markers of inammation. Clin Chem News
Navolotskaia O. Reference distributions for the 1993;19(Suppl):7.
positive acute phase proteins, alpha1-acid 63. Fischer CL, Gill C, Forrester MG, Nakamura R.
glycoprotein (orosomucoid), alpha1-antitrypsin, Quantitation of acute-phase proteins
and haptoglobin: a comparison of a large cohort postoperatively. Value in detection and
to the worlds literature. J Clin Lab Anal monitoring of complications. Am J Clin Pathol
2000;14:265270. 1976;66:840846.
53. Ritchie RF, Palomaki GE, Neveux LM, 64. Killingsworth LM. Plasma protein patterns in
Navolotskaia O, Ledue TB, Craig WY. Reference health and disease. CRC Crit Rev Clin Lab Sci
distributions for the negative acute-phase serum 1979;3:130.
proteins, albumin, transferrin and transthyretin: 65. Jolliff C. Agarose gel electrophoresis in the acute
a practical, simple and clinically relevant and chronic inammatory states. Clin Immunol
approach in a large cohort. J Clin Lab Anal Newsl 1991;11:132135.
1999;13:273279. 66. Smith SJ, Bos G, Esseveld MR, van Eijk HG,
54. Thompson D, Milford-Ward A, Whicher JT. The Gerbrandy J. Acute-phase proteins from the liver
value of acute phase protein measurements in and enzymes from myocardial infarction; a
clinical practice. Ann Clin BioChem 1992;29(Pt quantitative relationship. Clin Chim Acta
2):123131. 1977;81:7585.
140 Approach to pattern interpretation in serum
67. Becker GJ, Waldburger M, Hughes GR, Pepys MB. 79. Lindstrom FD, Eriksson P, Tejle K, Skogh T. IgG
Value of serum C-reactive protein measurement in subclasses of anti-SS-A/Ro in patients with
the investigation of fever in systemic lupus primary Sjogrens syndrome. Clin Immunol
erythematosus. Ann Rheum Dis 1980;39:5052. Immunopathol 1994;73:358361.
68. Labac E, Tirzonalis A, Gossart J. Electrophoresis 80. Shakib F, Pritchard DI, Walsh EA, et al. The
of blood proteins in collagen diseases. Acta detection of autoantibodies to IgE in plasma of
Gastroenterol Belg 1959;22:544549. individuals infected with hookworm (Necator
69. Stava Z. Serum proteins in scleroderma. americanus) and the demonstration of a
Dermatologica 1958;117:147153. predominant IgG1 anti-IgE autoantibody
70. Ganrot PO. Variation of the concentrations of response. Parasite Immunol 1993;15:4753.
some plasma proteins in normal adults, in 81. Shiddo SA, Huldt G, Nilsson LA, Ouchterlony
pregnant women and in newborns. Scand J Clin O, Thorstensson R. Visceral leishmaniasis in
Lab Invest Suppl 1972;124:8388. Somalia. Signicance of IgG subclasses and of
71. Horne CH, Howie PW, Goudie RB. Serum- IgE response. Immunol Lett 1996;50:8793.
alpha2-macroglobulin, transferrin, albumin, and 82. Suzuki S, Kida S, Ohira Y, et al. A case of
IgG levels in preeclampsia. J Clin Pathol Sjogrens syndrome accompanied by
1970;23:514516. lymphadenopathy and IgG4
72. Hubel CA, Kozlov AV, Kagan VE, et al. hypergammaglobulinemia. Ryumachi
Decreased transferrin and increased transferrin 1993;33:249254.
saturation in sera of women with preeclampsia: 83. van Nieuwkoop JA, Brand A, Radl J, Skvaril F.
implications for oxidative stress. Am J Obstet Increased levels of IgG4 subclass in 5 patients
Gynecol 1996;175:692700. with acquired respiratory disease. Int Arch
73. Vladutiu AO, Roach BM, Farahmand SM. Allergy Appl Immunol 1982;67:6165.
Polyclonal gammopathy with marked increase in 84. Taschini PA, Addison NM. An atypical
serum viscosity. Clin Chem 1991;37:17881793. electrophoretic pattern due to a selective
74. Aalberse RC, Schuurman J. IgG4 breaking the polyclonal increase of IgG4 subclass. ASCP
rules. Immunology 2002;105:919. Check Sample Immunopathol 1991:IP9195.
75. Akdis CA, Blesken T, Akdis M, Alkan SS, 85. Troussard X, Mossafa H, Flandrin G. Persistent
Heusser CH, Blaser K. Glucocorticoids inhibit polyclonal B-cell lymphocytosis. Ann Biol Clin
human antigen-specic and enhance total IgE (Paris) 2002;60:273280.
and IgG4 production due to differential effects 86. Carr R, Fishlock K, Matutes E. Persistent
on T and B cells in vitro. Eur J Immunol polyclonal B-cell lymphocytosis in identical
1997;27:23512357. twins. Br J Haematol 1997;96:272274.
76. Brouwer E, Tervaert JW, Horst G, et al. 87. Woessner S, Florensa L, Espinet B. Bilobulated
Predominance of IgG1 and IgG4 subclasses of circulating lymphocytes in persistent polyclonal
anti-neutrophil cytoplasmic autoantibodies B-cell lymphocytosis. Haematologica
(ANCA) in patients with Wegeners 1999;84:749.
granulomatosis and clinically related disorders. 88. Mossafa H, Malaure H, Maynadie M, et al.
Clin Exp Immunol 1991;83:379386. Persistent polyclonal B lymphocytosis with
77. Segelmark M, Wieslander J. ANCA and IgG binucleated lymphocytes: a study of 25 cases.
subclasses. Adv Exp Med Biol 1993;336:7175. Groupe Franais dHematologie Cellulaire. Br J
78. Mellbye OJ, Mollnes TE, Steen LS. IgG subclass Haematol 1999;104:486493.
distribution and complement activation ability of 89. Schiff R, Schiff SE. Chapter 11. Flow cytometry
autoantibodies to neutrophil cytoplasmic for primary immunodeciency diseases. In: Keren
antigens (ANCA). Clin Immunol Immunopathol DF, McCoy JJP, Carey JL, eds. Flow cytometry
1994;70:3239. in clinical diagnosis. Chicago: ASCP Press, 2001.
References 141
90. Schneider LC. X-linked hyper IgM syndrome. 101. Zeman AZ, Keir G, Luxton R, Thompson EJ.
Clin Rev Allergy Immunol 2000;19:205215. Serum oligoclonal IgG is a common and
91. Tsuge I, Matsuoka H, Nakagawa A, et al. persistent nding in multiple sclerosis, and has a
Necrotizing toxoplasmic encephalitis in a child systemic source. QJM 1996;89:187193.
with the X-linked hyper-IgM syndrome. Eur J 102. Frankel EB, Greenberg ML, Makuku S, Kochwa
Pediatr 1998;157:735737. S. Oligoclonal banding in AIDS and hemophilia.
92. Leiva LE, Junprasert J, Hollenbaugh D, Sorensen Mt Sinai J Med 1993;60:232237.
RU. Central nervous system toxoplasmosis with 103. Zeman A, McLean B, Keir G, Luxton R, Sharief
an increased proportion of circulating gamma M, Thompson E. The signicance of serum
delta T cells in a patient with hyper-IgM oligoclonal bands in neurological diseases.
syndrome. J Clin Immunol 1998;18:283290. J Neurol Neurosurg Psychiatry 1993;56:3235.
93. Miller ML, Algayed IA, Yogev R, Chou PM, Scholl 104. Pisa P, Cannon MJ, Pisa EK, Cooper NR, Fox
PR, Pachman LM. Atypical Pneumocystis carinii RI. EpsteinBarr virus induced
pneumonia in a child with hyper-IgM syndrome. lymphoproliferative tumors in severe combined
Pediatr Pathol Lab Med 1998;18:7178. immunodecient mice are oligoclonal. Blood
94. Chang MW, Romero R, Scholl PR, Paller AS. 1992;79:173179.
Mucocutaneous manifestations of the hyper-IgM 105. Myara I, Quenum G, Storogenko M, Tenenhaus
immunodeciency syndrome. J Am Acad D, Guillemain R, Moatti N. Monoclonal and
Dermatol 1998;38:191196. oligoclonal gammopathies in heart-transplant
95. Hostoffer RW, Berger M, Clark HT, Schreiber recipients. Clin Chem 1991;37:13341337.
JR. Disseminated Histoplasma capsulatum in a 106. Amadori A, Gallo P, Zamarchi R, et al. IgG
patient with hyper IgM immunodeciency. oligoclonal bands in sera of HIV-1 infected
Pediatrics 1994;94:234236. patients are mainly directed against HIV-1
96. Iseki M, Anzo M, Yamashita N, Matsuo N. determinants. AIDS Res Hum Retrovirus
Hyper-IgM immunodeciency with disseminated 1990;6:581586.
cryptococcosis. Acta Paediatr 1994;83:780782. 107. Papadopoulos NM, Costello R, Ceroni M,
97. Di Santo JP, de Saint Basile G, Durandy A, Moutsopoulos HM. Identication of HIV-
Fischer A. Hyper-IgM syndrome. Res Immunol specic oligoclonal immunoglobulins in serum of
1994;145:205209. carriers of HIV antibody. Clin Chem
98. Lobo FM, Scholl PR, Fuleihan RL. CD40 ligand- 1988;34:973975.
decient T cells from X-linked hyper-IgM 108. Taichman DB, Bayer K, Senior M, Goodman DB,
syndrome carriers have intrinsic priming Kricka LJ. Oligoclonal immunoglobulins in HIV-
capability. J Immunol 2002;168:14731478. antibody-positive serum. Clin Chem
99. Willenbrock K, Roers A, Seidl C, Wacker HH, 1988;34:2377.
Kuppers R, Hansmann ML. Analysis of T-cell 109. Kelly RH, Hardy TJ, Shah PM. Benign
subpopulations in T-cell non-Hodgkins monoclonal gammopathy: a reassessment of the
lymphoma of angioimmunoblastic problem. Immunol Invest 1985;14:183197.
lymphadenopathy with dysproteinemia type by 110. Kelly RH, Scholl MA, Harvey VS, Devenyi AG.
single target gene amplication of T cell receptor- Qualitative testing for circulating immune
beta gene rearrangements. Am J Pathol complexes by use of zone electrophoresis on
2001;158:18511857. agarose. Clin Chem 1980;26:396402.
100. Franciotta D, Zardini E, Bono G, Brustia R, 111. Levinson SS, Keren DF. Immunoglobulins from
Minoli L, Cosi V. Antigen-specic oligoclonal the sera of immunologically activated persons
IgG in AIDS-related cytomegalovirus and with pairs of electrophoretic restricted bands
toxoplasma encephalitis. Acta Neurol Scand show a greater tendency to aggregate than
1996;94:215218. normal. Clin Chim Acta 1989;182:2130.
142 Approach to pattern interpretation in serum
112. Keren DF, Morrison N, Gulbranson R. Evolution 123. Keshgegian AA. Oligoclonal banding in sera of
of a monoclonal gammopathy (MG) documented hospitalized patients. Clin Chem 1992;38:169.
by high-resolution electrophoresis (HRE) and 124. Chang KL, Flaris N, Hickey WF, Johnson RM,
immunoxation (IFE). Lab Med 1994; Meyer JS, Weiss LM. Brain lymphomas of
25:313317. immunocompetent and immunocompromised
113. Papadopoulos NM, Tsianos EV, Costello R. patients: study of the association with
Oligoclonal immunoglobulins in serum of EpsteinBarr virus. Mod Pathol
patients with chronic viral hepatitis. J Clin Lab 1993;6:427432.
Anal 1990;4:180182. 125. Surico G, Muggeo P, Muggeo V, Lucarelli A,
114. Perrone A, Deramo MT, Spaccavento F, Novielli C, Conti V, Rigillo N. Polyclonal
Santarcangelo P, Favoino B, Antonaci S. hypergammaglobulinemia at the onset of acute
Hepatitis C virus (HCV) genotypes, human myeloid leukemia in children. Ann Hematol
leucocyte antigen expression and monoclonal 1999;78:445448.
gammopathy prevalence during chronic HCV 126. Mouzaki A, Matthes T, Miescher PA, Beris P.
infection. Cytobios 2001;106(Suppl 1):125134. Polyclonal hypergammaglobulinaemia in a case
115. Schott P, Pott C, Ramadori G, Hartmann H. of B-cell chronic lymphocytic leukaemia: the
Hepatitis C virus infection-associated non- result of IL-2 production by the proliferating
cryoglobulinaemic monoclonal IgMkappa monoclonal B cells? Br J Haematol
gammopathy responsive to interferon-alpha 1995;91:345349.
treatment. J Hepatol 1998;29:310315. 127. Hirokawa M, Lee M, Kitabayashi A, et al.
116. Andreone P, Gramenzi A, Cursaro C, Bernardi Autoreactive T cell-dependent polyclonal
M, Zignego AL. Monoclonal gammopathy in hypergammaglobulinemia in mantle cell
patients with chronic hepatitis C virus infection. lymphoma. Leuk Lymphoma 1994;14:509513.
Blood 1996;88:1122. 128. Machii T, Yamaguchi M, Inoue R, et al.
117. Gattoni A, Cecere A, Romano C, Di Martino P, Polyclonal B-cell lymphocytosis with features
Caiazzo R, Rippa A. Case report of a resembling hairy cell leukemia-Japanese variant.
monoclonal gammopathy in a patient with Blood 1997;89:20082014.
chronic hepatitis: effects of beta-IFN treatment. 129. McCarty MJ, Vukelja SJ, Banks PM, Weiss RB.
Panminerva Med 1996;38:175178. Angiofollicular lymph node hyperplasia
118. Heer M, Joller-Jemelka H, Fontana A, Seefeld U, (Castlemans disease). Cancer Treat Rev
Schmid M, Ammann R. Monoclonal 1995;21:291310.
gammopathy in chronic active hepatitis. Liver 130. Menke DM, Camoriano JK, Banks PM.
1984;4:255263. Angiofollicular lymph node hyperplasia: a
119. Kanoh T. Fluctuating M-component level in comparison of unicentric, multicentric, hyaline
relation to infection. Eur J Haematol vascular, and plasma cell types of disease by
1989;42:503504. morphometric and clinical analysis. Mod Pathol
120. Struve J, Weiland O, Nord CE. Lactobacillus 1992;5:525530.
plantarum endocarditis in a patient with benign 131. Ohsaka A, Saito K, Sakai T, Mori S, Kobayashi
monoclonal gammopathy. J Infect Y, Amemiya Y, Sakamoto S, Miura Y.
1988;17:127130. Clinicopathologic and therapeutic aspects of
121. Larrain C. Transient monoclonal gammopathies angioimmunoblastic lymphadenopathy-related
associated with infectious endocarditis. Rev Med lesions. Cancer 1992;69:12591267.
Chil 1986;114:771776. 132. Attygalle A, Al-Jehani R, Diss TC, et al.
122. Keshgegian AA. Prevalence of small monoclonal Neoplastic T cells in angioimmunoblastic T-cell
proteins in the serum of hospitalized patients. lymphoma express CD10. Blood
Am J Clin Pathol 1982;77:436442. 2002;99:627633.
References 143
133. Strupp C, Aivado M, Germing U, Gattermann N, 143. Berney T, Delis S, Kato T, et al. Successful
Haas R. Angioimmunoblastic lymphadenopathy treatment of posttransplant lymphoproliferative
(AILD) may respond to thalidomide treatment: disease with prolonged rituximab treatment in
two case reports. Leuk Lymphoma 2002;43: intestinal transplant recipients. Transplantation
133137. 2002;74:10001006.
134. Schultz DR, Yunis AA. Immunoblastic 144. Craig FE, Gulley ML, Banks PM.
lymphadenopathy with mixed cryoglobulinemia. Posttransplantation lymphoproliferative
A detailed case study. N Engl J Med 1975;292: disorders. Am J Clin Pathol 1993;99:265276.
812. 145. Badley AD, Portela DF, Patel R, et al.
135. Chodirker WB, Komar RR. Angioimmunoblastic Development of monoclonal gammopathy
lymphadenopathy in a child with unusual clinical precedes the development of EpsteinBarr virus-
and immunologic features. J Allergy Clin induced posttransplant lymphoproliferative
Immunol 1985;76:745752. disorder. Liver Transpl Surg 1996;2:375382.
136. Hsu SM, Waldron JA Jr, Fink L, et al. 146. Stevens SJ, Verschuuren EA, Pronk I, et al.
Pathogenic signicance of interleukin-6 in Frequent monitoring of EpsteinBarr virus DNA
angioimmunoblastic lymphadenopathy-type T- load in unfractionated whole blood is essential
cell lymphoma. Hum Pathol 1993;24: for early detection of posttransplant
126131. lymphoproliferative disease in high-risk patients.
137. Steinberg AD, Seldin MF, Jaffe ES, et al. NIH Blood 2001;97:11651171.
conference. Angioimmunoblastic 147. Conley ME. B cells in patients with X-linked
lymphadenopathy with dysproteinemia. Ann agammaglobulinemia. J Immunol 1985;134:
Intern Med 1988;108:575584. 30703074.
138. Whitten RO, Zutter M, Iaci-Hall J, Odell M, 148. Conley ME, Rohrer J, Minegishi Y. X-linked
Kidd P. Oligoclonal immunoglobulin heavy chain agammaglobulinemia. Clin Rev Allergy Immunol
gene rearrangement in a childhood 2000;19:183204.
immunoblastic lymphoma. Presentation as a 149. Vihinen M, Kwan SP, Lester T, et al. Mutations
polyphenotypic atypical lymphoproliferative of the human BTK gene coding for bruton
reaction. Am J Clin Pathol 1990;93:286293. tyrosine kinase in X-linked
139. Tienhaara A, Irjala K, Rajamaki A, Pulkki K. agammaglobulinemia. Hum Mutat 1999;13:
Four monoclonal immunoglobulins in a patient 280285.
with chronic lymphocytic leukemia. Clin Chem 150. Noordzij JG, de Bruin-Versteeg S, Comans-Bitter
1986;32:703705. WM, et al. Composition of precursor B-cell
140. Jain A, Nalesnik M, Reyes J, et al. Posttransplant compartment in bone marrow from patients with
lymphoproliferative disorders in liver X-linked agammaglobulinemia compared with
transplantation: a 20-year experience. Ann Surg. healthy children. Pediatr Res 2002;51:159168.
2002;236:429436. 151. Siegel RL, Issekutz T, Schwaber J, Rosen FS,
141. Nepomuceno RR, Snow AL, Robert Beatty P, Geha RS. Deciency of T helper cells in transient
Krams SM, Martinez OM. Constitutive hypogammaglobulinemia of infancy. N Engl J
activation of Jak/STAT proteins in EpsteinBarr Med 1981;305:13071313.
virus-infected B-cell lines from patients with 152. Kilic SS, Tezcan I, Sanal O, Metin A, Ersoy F.
posttransplant lymphoproliferative disorder. Transient hypogammaglobulinemia of infancy:
Transplantation 2002;74:396402. clinical and immunologic features of 40 new
142. Dunphy CH, Gardner LJ, Grosso LE, Evans HL. cases. Pediatr Int 2000;42:647650.
Flow cytometric immunophenotyping in 153. Rosefsky JB. Transient hypogammaglobulinemia
posttransplant lymphoproliferative disorders. Am of infancy. Acta Paediatr Scand 1990;79:
J Clin Pathol 2002;117:2428. 962963.
144 Approach to pattern interpretation in serum
A monoclonal gammopathy is dened as the has recommended the use of the term M-protein
electrophoretically and antigenically homogeneous and I will use it in this chapter interchangeably
protein product of a single clone of B lymphocytes with monoclonal gammopathy.3
and/or plasma cells that has proliferated beyond
the constraints of normal control mechanisms.
Monoclonal gammopathies are detected in serum
and/or urine from individuals with a wide variety DIFFERENTIATION OF B
of neoplastic, potentially neoplastic, neurological LYMPHOCYTES
and infectious conditions. The monoclonal
gammopathy is produced by a single clone of The discovery that there are major subpopulations
plasma cells or B lymphocytes. Several terms have of lymphocytes resulted from careful observations
been used to refer to monoclonal gammopathies: of immune deciency in humans and a serendipi-
paraprotein, dysproteinemia, monoclonal gammo- tous discovery in bursectomized chickens. In
pathy, monoclonal component, and M-protein.1,2 1952, Bruton4 described a child that suffered from
The guidelines for clinical and laboratory evalua- recurrent infections with pyogenic bacteria. Using
tion of patients with monoclonal gammopathies serum protein electrophoresis, then a relatively
146 Conditions associated with monoclonal gammopathies
new clinical laboratory test, he demonstrated that poietic stem cells to plasma cells is a continuous
this childs serum lacked a g-globulin region. The process, however, it is useful to divide the process
patient was treated successfully by administering into maturation before and after activation by
g-globulin parenterally. Such patients are now antigen. The earliest stage of B-lymphocyte devel-
known to have Brutons X-linked agammaglobu- opment occurs in the fetal liver and continues in
linemia (XAG), which is caused by decient matu- adult life in the bone marrow (Table 6.1).7 The
ration of the B lymphocytes as a result of the availability of cluster designation (CD) mono-
absence of Brutons tyrosine kinase (BTK).5 A few clonal antibody markers to lymphocyte surface
years after Brutons report, Glick et al.6 discovered antigens has dramatically improved the delineation
that removal of the cloacal Bursa of Fabricius of the stages of B-lymphocyte maturation and ne-
early in the life of chickens produced agamma- tuned the characterization of lymphoproliferative
globulinemia similar to that found in Brutons disorders. At the stem cell stage CD34 is
patients. Both bursectomized chickens and indivi- expressed.8 Following this, the earliest B-
duals with XAG lack plasma cells in their tissues, lymphocytes (pro-B cells) express CD19 (a marker
and neither has germinal centers in their lymphoid present at all stages of B-cell maturation), and
tissues. Nonetheless, normal numbers of peri- HLA-DR (an HLA Class II antigen prominently
pheral blood lymphocytes are present, and the expressed on B-cells) in addition to CD34. These
patients do not have problems with viral, fungal, early B- lymphocytes also contain the enzyme
or intracellular bacterial infections. We now know terminal deoxyribonucleotidyl transferase (TdT) in
that the remaining lymphocytes are the normal the nucleus. At this time, the B cells have already
peripheral blood T-lymphocytes, which play the begun the process of immunoglobulin gene
main role in host defense against viral, fungal, and rearrangement that will result in the production of
intracellular bacterial infections. The B-lympho- antibody directed against a single epitope. Even
cytes are so named because in the chicken they are before the pre-B cell stage, the cell has rearranged
derived from the Bursa of Fabricius. In humans, its m chain gene, although the m heavy chain cannot
B- and T-cells originate in the bone marrow. yet be detected in the cytoplasm.9,10 These pre-
However, T-cells must be processed subsequently pre-B cells usually express the major histocom-
in the thymus gland, while B-lymphocytes are sub- patability antigen HLA-DR, and have surface
sequently processed in the fetal liver and bone markers that are recognized by CD10 (common
marrow. acute lymphoblastic leukemia antigen, CALLA),
Differentiation of B-lymphocytes from hemato- CD19 and CD20 (dim).11
a
cm, cytoplasmic m-heavy chain.
b
sIgD and/or sIgM, surface IgD and/or surface IgM.
Differentiation of B-lymphocytes 147
Around the eighth week of human gestation, CD19, CD20 and, when they pass into the circula-
large lymphoid cells with a small amount of tion, CD21.7
detectable cytoplasmic m heavy chain but no The genes that direct the expression of the heavy
detectable light chains are present in the fetal chain undergo rearrangement as the B-lymphocytes
liver. These are termed pre-B cells, they do not mature to plasma cells. Whereas the light chain is
show surface expression of the m chain at this always the same, these mature B cells may express
stage.12,13 Pre-B cells already have selected the more than one isotype on their surface; under-
variable region that will be part of the standing this helps to explain the occurrence of
immunoglobulin heavy chain that their plasma some biclonal gammopathies. In most biclonal
cell progeny will eventually produce. Pre-B cells gammopathies, both monoclonal proteins have the
divide at a rapid rate (generation time is about same light chain type. When the light chains are the
12 h), leading to the production of small pre-B same, it is likely that we are seeing expression of
lymphocytes that still contain cytoplasmic m. At two heavy chain genes by the same clone. This reca-
this stage, allelic exclusion occurs wherein either pitulates the events seen during development.
the k or l gene is selected for production. When the light chain types differ, the monoclonal
Although there will be subsequent switches in components of a biclonal gammopathy truly arise
heavy chain expression, the light chain remains from different clones.14
constant for this clone. On the surface of the cell, Following contact between B-lymphocytes and
one nds CD10, HLA-DR, CD19, and CD20. A antigen with the appropriate costimulation by
small subset of pre-B cells also express CD5 macrophages and helper T cells, the B-lymphocytes
(usually a T-cell marker).7 become activated. At this level of maturity, B-
The next stage of development can be recognized lymphocytes migrate to secondary lymphoid
in the fetus by the tenth to twelfth week of gesta- tissues where further differentiation occurs in the
tion.12 These mature B cells contain surface whole lymphoid follicles.7,15 Here, the surface antigen
IgM and/or IgD molecules with the selected light expression varies depending upon the location of
chain, but they no longer contain the cytoplasmic m the B-lymphocytes within the lymph node (Table
chain. The variable regions of both the light and 6.2).
heavy chain are the same as those in the The nal differentiation of activated B cells to
immunoglobulins ultimately produced by the plasma cells results in loss of expression of surface
plasma cell progeny of these B cells. Although IgD immunoglobulin, HLA-DR, CD20, and CD40.
and/or IgM are the major surface isotypes at this Normal bone marrow plasma cells are strongly
stage, they will not usually be the nal isotype pro- positive for CD38 and CD79a (an immunoglobulin-
duced by this clone. At this point they lose their associated antigen), express CD19 and CD45 weakly,
CD10 marker while they continue to express but lack expression of CD56.1619
Mantle zone and marginal zone CD21, CD22, CD23, CD24, CD39, CDw76,
Follicular center CD10, CD38(dim), CD77, CD22(dim), CD23(dim), CD24(dim), CD79a(dim)
Pan lymph node CD19, CD20, CD37, CD40, CD45, CD72, CD74, CD275, CDw78
a
Data from Pirruccello and Aoun.7
148 Conditions associated with monoclonal gammopathies
I. Multiple myeloma
IgG Large g spike
IgM Broad spike at origin
IgA Broad b spike
IgD Small g or b spike
IgE Small b spike
Light chain disease (k or l) Hypogammaglobulinemia, occasional b spike
Heavy chain disease (a, m, g) Broad b band
Biclonal (double) gammopathy Two g spikes
Nonsecretory Low g region, normal
II. Waldenstrms macroglobulinemia
IgM Broad band near origin
IgA Broad band near origin or slightly b
IgG Broad band near origin or slightly g
III. B-lymphoproliferative disorders
Chronic lymphocytic leukemia Low g, small band
Well-differentiated lymphocytic lymphoma Low g, small band
IV. Amyloidosis (AL)
With multiple myeloma g Spike
Not associated with myeloma Normal, low g
V. Monoclonal gammopathy with other clinical correlation
Autoimmune diseases Small g spike
Neuropathy Small usually IgM bg restriction
Infectious disease Small g spike usually with a diffuse or oligoclonal
increase in g-globulin
VI. Monoclonal gammopathy of undetermined Small g spike
signicance
Conditions associated with monoclonal gammopathies 149
a
MIDD, monoclonal immunoglobulin deposition disease. This includes light chain deposition disease (LCDD), heavy chain deposition
disease (HCDD), and a combination of the two.
150 Conditions associated with monoclonal gammopathies
(a)
monoclonal gammopathies to be about 1.5 per cent SPE2 system (Beckman Coulter, Fullerton, CA,
of the population over 50 years of age and about 3 USA) and reported that of 1732 individuals, 8.4 per
per cent of the population over 70 years of age. cent of blacks and 3.8 per cent of whites had mon-
Looking at different age groups, Axelsson et al.25 oclonal gammopathies. This nding is notable
found monoclonal gammopathies in 2 per cent of because multiple myeloma also occurs approxi-
Swedish subjects between the ages of 70 years and mately twice as often in blacks than in whites.29,30
79 years and 5.7 per cent in the subjects older than Lastly, using a high-resolution acetate method,
80 years. Cohen et al.28 evaluated samples from indi- Aguzzi detected monoclonal gammopathies in 78
viduals older than 70 years of age using the Paragon per cent of patients over 55 years of age.31
Multiple myeloma 151
Nonetheless, these studies all point out the common cases of multiple myeloma make only MFLC and
occurrence of M-proteins and agree that the occur- may be mistaken for non-secretory myeloma if the
rence of monoclonal gammopathies progressively urine is not examined because monomeric MFLC
increases with age. are more readily detected in urine than in serum by
electrophoretic techniques. However, a new quan-
titative immunoassay can measure free light chains
MULTIPLE MYELOMA (unbound to heavy chains) in both serum and urine.
As discussed below, this technique will improve
Clinical picture in myeloma detection of cases of light chain disease and should
provide an improvement over 24-h urine samples to
Multiple myeloma is a malignant neoplasm, follow the tumor burden of these patients.
expressed mainly as a proliferation of plasma cells, Patients typically present with fatigue, bone pain
that usually presents with bone marrow involve- and pathological fractures. The prominent bone
ment.3234 While the plasma cells form the vast marrow involvement in this disease is associated
majority of the malignant cells, remnants of the with lytic lesions in the ribs, vertebrae, skull, and
parent B-cell clone and other B-clonal precursors long bones. The bone involvement results in hyper-
persist and play a role in the evolution of the neo- calcemia and a normochromic, normocytic anemia.
plastic process.3537 Typically, these malignant In addition to its effects on the bone marrow, mon-
plasma cells synthesize and secrete considerable oclonal free light chains damage the kidneys and
amounts of monoclonal whole immunoglobulin or other organs occasionally with deposition of MFLC
fragments of immunoglobulin and/or monoclonal or amyloid AL (amyloid associated with immuno-
free light chains (MFLC; Bence Jones proteins). globulin light chain; see chapter 7).4247 Suppression
Depending on their structure and concentration, of normal g-globulins in advanced cases is associ-
these M-proteins may increase the viscosity of ated with recurrent bacterial infections.48 Ong et al.49
blood or precipitate in various tissues or form a looked at the medical histories of 127 patients diag-
cryoglobulins.3841 Rare cases of non-secretory nosed with multiple myeloma to see which symp-
myeloma produce no detectable immunoglobulin toms were associated with patients who were
in serum or urine when studied by conventional diagnosed quickly and which with those in whom
electrophoretic techniques. About 15 per cent of the diagnosis was delayed. As shown in Table 6.5,
a
Data from Ong et al. 49 ESR, erythrocyte sedimentation rate.
b
Chi-square signicance P 0.001.
152 Conditions associated with monoclonal gammopathies
those with delayed diagnosis were signicantly less peripheral blood stem cell transplantation. He notes
likely to have bone pain and/or fractures, anemia or that the main issues at the present time for this
hypercalcemia.49 therapy are the ability of chemotherapeutic agents
Multiple myeloma accounts for about 1 per cent to destroy the tumor cells in the patient and the need
of all malignancies and 10 per cent of hematopoi- to remove myeloma cells and their precursors from
etic malignancies.50 Although the incidence the harvested stem cells that are given back to the
increases dramatically with age, multiple myeloma patient to reconstitute the bone marrow.64 This
has been reported (rarely) in a few children.5153 treatment has been highly successful in some cases
However, the clinical picture is often unusual, (Fig. 6.2). In addition to more conventional chemo-
lacking key features such as anemia, elevated therapeutic agents, thalidomide and new immuno-
calcium (in about one-third of the patients), and modulatory agents show promise.65 Nonetheless,
lytic lesions. It has been better documented in a few despite these new techniques and aggressive chemo-
young adults, but even this is exceptionally rare.54 therapy, the prognosis is poor for most cases.50
Individuals under the age of 40 years account for
only 2 per cent of myeloma cases and those under
30 years comprise just 0.3 per cent.55 Staging
Nonetheless, the possibility of this disease should
not be ignored when appropriate symptoms are Prognosis in myeloma is highly dependent on clini-
present.5658 A comparison of the clinical presenta- cal and laboratory features. The system devised by
tions and laboratory features of the disease Durie and Salmon66 is still largely used to predict
revealed no signicant differences between those prognosis (Table 6.6). Based on their criteria,
that present with multiple myeloma prior to individuals with stage I, II and III disease had median
50 years of age and those who present after age survivals of 38, 35 and 13 months, respectively.67
50 years.59 Although prognosis for younger Using a multivariate analysis of 265 patients
patients does not generally differ from that of older with multiple myeloma, Cherng and coworkers68
individuals, young patients with good renal func- found that the three most important factors were
tion and low b2-microglobulin levels have a longer plasmacytosis (> 30 per cent), hypercalcemia
survival than the older patients and may benet (> 11.5 mg/dl), and hypoalbuminemia (< 3.5 g/dl).69
from aggressive therapy.55,59 The overwhelming Other factors predicting an unfavorable course
majority of children with monoclonal gammo- include elevated alkaline phosphatase, hyper-
pathies do not have multiple myeloma. They usu- uricemia, renal insufciency, and male gender.70
ally have B-cell lymphoproliferative lesions that are Flow cytometry studies are also improving our
sometimes related to the presence of primary and ability to stage patients with multiple myeloma.
secondary immunodeciency diseases, occasionally Patients with plasmablasts positive for CD56 had a
to autoimmune diseases and also to hematological 63-month overall survival compared with 22 months
malignancies other than myeloma.6063 for those that had CD56-negative plasmablasts.71
There have been dramatic improvements in the Others have shown that hyperdiploidy on initial
therapy offered for myeloma patients in the past few bone marrow examination has a signicantly better
years. The therapy offered depends on the stage of overall survival than diploid or hypodiploid tumor
the disease (see below). Depending on their perfor- cells.72 This area is still somewhat controversial, how-
mance status, newly diagnosed cases of multiple ever, because although an Eastern Cooperative
myeloma may now be treated with stem cell trans- Oncology Group (ECOG) study was able to conrm
plantation rather than the previous chemo- the survival advantage of hyperdiploid versus
therapeutic approach with melphalan and diploid, it was unable to conrm the worse prognosis
prednisone.50 Kyle64 recommends that patients under of the hypodiploid cases.73 Clearly, the issue of stag-
the age of 70 years be considered for autologous ing will be undergoing changes.
Multiple myeloma 153
Reference Ranges
Rel % g/dl
ALBUMIN 49.7 64.4 3.55 5.04
ALPHA 1 4.8 10.1 0.25 0.74
ALPHA 2 8.5 15.1 0.55 1.14
BETA 7.8 13.1 0.55 1.04
GAMMA 10.5 19.5 0.65 1.44
(c)
Figure 6.2 (a) Densitometric scan of serum protein electrophoresis from a patient with multiple myeloma prior to receiving a bone
marrow transplantation. The massive monoclonal gammopathy in the b-region measures 7.71 g/dl. (Scan from Paragon SPE2 gel stained
with Amido Black.) (b) Urine from pre-transplant analysis demonstrated a massive monoclonal free light chain (MFLC) (> 5 g/24 h). (c)
Capillary zone electropherogram of serum protein electrophoresis from same patient as in (a) 9 years after bone marrow transplant. The
patient is clinically well with no sign of multiple myeloma. The urine is negative for MFLC.
Table 6.6 Myeloma staging criteriaa Neither does there appear to be a risk of develop-
ing multiple myeloma associated with the small
Stage Criteria
doses of radiation one receives through diagnostic
radiology procedures.85
I Must have all of the following:
Hemoglobin > 10g/dl
Serum calcium < 12 mg/dl Genetics and cytogenetics
No lytic lesions or solitary plasmacytoma on
There are suggestions of a genetic predisposition to
radiographs
the development of monoclonal gammopathies.
Relatively low quantity of M-protein:
Many studies have recorded multiple myeloma in
IgG < 5g/dl rst degree relatives.8699 Individuals that report
IgA < 3g/dl having rst-degree relatives with multiple myeloma
Urine monoclonal free light chains < 4 g/24 h have a 3.7 times greater chance of developing mul-
II Lacks criteria for stage III, but exceeds tiple myeloma than those that do not share this
criteria for stage I
family history.86 Broader epidemiological evidence
that genetics plays a role derives from Bowden et al.
III Has criteria for stage I, and has any one or
who documented that monoclonal gammopathies
more of the following: are more often detected in Americans than in a
Hemoglobin < 8.5g/dl Japanese cohort.100 However, while genetic differ-
Serum calcium > 12 mg/dl ences may play a role, the American population had
Advanced lytic bone lesions a larger proportion of older individuals, which may
Relatively large quantity of M-protein: have accounted for some of their results.
Cytogenetic abnormalities have been found in
IgG > 7 g/dl
multiple myeloma.101 The most typical transloca-
IgA > 5 g/dl
tions occur at the site of the immunoglobulin heavy
Urine monoclonal free light chains = 12 g/24h chain locus in switch regions.102,103 Fluorescence in
Subclassication: situ hybridization (FISH) probes are available to
A. serum creatinine < 2.0 mg/dl detect the presence of translocations on chromo-
B. serum creatinine = 2.0 mg/dl some 11 at position q13.104 Cytogenetics may be
especially useful in cases of non-secretory myeloma
a
Table from Durie and Salmon.66 where M-proteins are not demonstrable in serum
or urine by traditional electrophoretic methods.105
Karyotypic instability has been detected at the ear-
67 years) than among whites (mean age of onset liest stages of multiple myeloma and increases
71 years) and is somewhat more common in men along with progression of the disease.106 These
than women.29,30,75 These differences may relate to cytogenetic abnormalities may eventually explain
socioeconomic factors including diet and dietary why myeloma plasma cells do not undergo their
supplement use (such as vitamin C).77,78 There is a usual programmed death by apoptosis.48
clear association of multiple myeloma with radia-
tion exposure in Japanese survivors of the atomic
bomb, radium dial workers, and even among radi- Cellular features of myeloma
ologists.7983 Although there has been a report that
individuals who work at nuclear power plants may Most neoplastic plasma cells in multiple myeloma
have an increased risk for myeloma, there does not lack surface immunoglobulins despite the secre-
appear to be a risk for those living near the plants.84 tions they produce. Bone marrow plasma cells
Multiple myeloma 155
from patients with multiple myeloma express There are some markers that can help to detect
CD45, CD38, and CD79a but, in contrast to the B-cell precursors of the neoplastic plasma cells
normal plasma cells, they have an aberrant expres- in multiple myeloma. Unlike normal activated B
sion of CD56 (the natural killer cell antigen) and cells, the circulating B cells in patients with
CD19 (CD19-/CD56+ or CD19-/CD56-).16,17 multiple myeloma express adhesion molecules
This surface marker information may be useful in including b1-integrins that may facilitate trans-
the differential diagnosis between multiple portation of these cells as they metastasize
myeloma and MGUS. Sezer et al.17 noted that all (localize) in the bone marrow and other sites.111
MGUS patients in their study contained both phe- Despite the fact that the neoplastic B cells circulate,
notypically normal (CD19+/CD56-) plasma cells a case of myeloma in a 27-year-old pregnant
and aberrant monoclonal populations (CD19- woman whose child did not show evidence of
/CD56+ and/or CD19-/CD56-). The ratio of the myeloma after a 2-year follow-up suggests that
normal to all bone marrow plasma cells was these B cells do not readily pass through the
always 20 per cent. In contrast, individuals with placenta.112 Although circulating CD19 lympho-
multiple myeloma had either no phenotypically cytes are increased in myeloma patients, circulating
normal plasma cells (CD19+/CD56-), or when mature plasma cells in these patients may be dif-
they were present always totaled less than 20 per cult to detect except in the extreme case of plasma
cent of the bone marrow plasma cells.17 cell leukemia (see below).
A histological correlate of this immaturity and Although T-lymphocytes in myeloma are not
malignant behavior is the presence of large atypical thought to be involved directly in the neoplastic
cells that may be seen, occasionally with binucleate process, there are notable alterations in their
cells, in the involved marrow. However, atypical peripheral blood subpopulations because they are
nuclei and multinucleate plasma cells are insuf- affected by the myeloma process. The production
cient criteria to absolutely distinguish between of transforming growth factor b (TGF-b) by the
proliferating polyclonal plasma cells in chronic myeloma cells suppresses the normal proliferation
osteomyelitis (for example) and monoclonal and blastogenic response of T-lymphocytes to
plasma cells of multiple myeloma. interleukin-2 (IL-2).113 This observation explains
Although multiple myeloma is mainly a disease of the decreased responsiveness of peripheral blood
malignant plasma cells, both circulating B-lympho- T-lymphocytes from patients with multiple
cytes and even pre-B cells in the bone marrow can myeloma to lectin stimulation and also the
be found that express the specic idiotype of the decrease in the number of CD4-positive cells.114,115
monoclonal protein being produced by the myeloma Individuals with multiple myeloma who have
cells from that patient.107110 The nding of such pre- normal levels of CD4+ and CD19+ cells at the time
cursor cells may seem surprising considering that of diagnosis have a better prognosis than those
the neoplasm is characterized by mature-appearing with decreased levels.116,117 This observation has
plasma cells. However, this is entirely consistent been extended by a recent ECOG study where the
with our understanding of the variable maturation peripheral blood lymphocyte levels were followed
by neoplastic cells. For example, a squamous cell from baseline to the end of a chemotherapy
carcinoma may consist mainly of mature keratinized regimen. They found that when patients undergo-
cells, yet one is not at all surprised to nd some ing chemotherapy maintained the baseline
immature non-keratinizing elements in the same numbers of CD3+, CD4+, CD8+ and CD19+ cells
neoplasm. Indeed, it would be difcult to explain in the peripheral blood, they had a better long-term
why a neoplasm of mature plasma cells would pro- survival than those that did not.118
duce such a rapidly fatal disease if there were not Cytokines play a major role in the growth and
progenitor cells that formed the silent proliferative drug responsiveness of myeloma cells.119,120 Both
compartment of the neoplasm. interleukin-6 (IL-6) and insulin-like growth factor
156 Conditions associated with monoclonal gammopathies
1 (IGF-1) are important in the proliferation and blood.129134 Two-thirds of cases of plasma cell
drug resistance of myeloma and may operate by leukemia do not have evidence of multiple
upregulating telomerase activity of the malignant myeloma at the time of presentation and are
cells.121 This interferes with dexamethasone- termed primary plasma cell leukemia.129,131 The
induced apoptosis in the myeloma cells.122 remainder of cases are usually the end-stage of a
Interestingly the plasma cells themselves can case of multiple myeloma.129,131 Perhaps because of
produce IL-6 along with the bone marrow cells. the late stage of presentation in the secondary
While not thought of as a currently used marker of form, plasma cell leukemia has a reputation for an
the disease, the level of IL-6 in serum has been aggressive course.21
shown to correlate inversely with survival.123
Patients with multiple myeloma also have higher
serum levels of TGF-b than do patients with
MGUS.124 Interleukin 1 (IL-1) and IL-6 have both Monoclonal gammopathies in multiple
been implicated either as a marker or as part of the myeloma
process resulting in conversion of MGUS to
myeloma.119 In addition to correlating with conver- IMMUNOGLOBULIN ISOTYPES IN MULTIPLE
sion of MGUS to myeloma, IL-6, tumor necrosis MYELOMA
factor-a (TNF-a) and soluble interleukin-2 recep- The prevalence of the major immunoglobulin
tor (sIL-2r) have been used in experimental studies isotypes among large studies of monoclonal
as markers of cases of known myeloma that have gammopathies roughly parallels the concentration
an aggressive advanced disease.125 Again, these of that immunoglobulin isotype in serum. For
markers are not currently recommended to distin- example, in his large series of patients with
guish between these conditions. myeloma, Kyle found an IgG monoclonal protein
in about 5560 per cent of cases, IgA in about 25
per cent, and k light chain was twice as frequent as
l (consistent with the normal 2:1 ratio of k to l in
Circulating plasma cells and plasma the serum).135 About 15 per cent of cases only
cell leukemia secrete MFLC, IgD accounts for about 12 per cent
of cases, IgM in 0.5 per cent, and IgE is exceedingly
Under normal circumstances, plasma cells are not rare.136
detected in the peripheral blood. However, The vast majority of IgM monoclonal gammo-
patients with multiple myeloma and patients with pathies associated with lymphoproliferative dis-
amyloid AL may have circulating monoclonal orders are considered separately as Waldenstrms
plasma cells.126,127 Their presence in concentrations in the Kyle series, but they typically comprise about
4 per cent of immunoglobulin positive cells or 2530 per cent of cases of monoclonal gammo-
4 106/l is a negative prognostic indicator and pathies.135 Similar ndings for the occurrence of
suggests the presence of active multiple major isotypes in multiple myeloma have been
myeloma.126,127 The mere presence of plasma cells in reported in smaller studies.137,138 There are, how-
the circulation of patients with multiple myeloma, ever, some notable exceptions to the generalization
however, does not equate with the diagnosis of that the occurrence of M-proteins in myeloma fol-
plasma cell leukemia. lows their concentration in the serum. Among the
Plasma cell leukemia is an uncommon occurrence subclasses of IgG, Schur et al.139 found signicantly
in myeloma (about 2 per cent of cases), making it a fewer cases of IgG2 than would be predicted by its
rare disease indeed.128130 To qualify, a patient needs concentration in the serum. Similar observations
to have an absolute plasma cell count of 2 109/l have been made about the infrequency of IgA2
and 20 per cent plasma cells in the peripheral monoclonal gammopathies.140
Multiple myeloma 157
In most cases of myeloma, electrophoretic nd- with multiple myeloma).142144 IgA myeloma pro-
ings are straightforward. The characteristic densely teins can occur as monomers or as polymers with
staining spike typically occurs in the g-region, near variable molecular weight (monomers, dimers,
the origin, and in the b-region, for IgG, IgM, and trimers and tetramers in the same myeloma
IgA monoclonal proteins, respectively (Fig. 6.3). serum).145,146 Because these molecules can self-
Note, however, that monoclonal gammopathies aggregate, they have been known to cause
may migrate in the a-region (uncommonly), and problems with hyperviscosity.145,147149 This poly-
they may bind to other serum proteins thereby merization may result in the presence of two or
altering their migration on electrophoretic gels.141 three M-protein peaks on the serum protein
electrophoresis (Fig. 6.4). Since most of the IgA
IGG AND IGA MYELOMA monoclonal gammopathies migrate in the b-region,
IgG myeloma proteins are almost always 160 kDa they may be masked by the C3, b1 lipoprotein and
monomers that only rarely produce clinical symp- transferrin bands on serum protein electrophoresis.
toms of hyperviscosity (although the measured Because of these concerns, the Guidelines for
viscosity of serum may be elevated in most patients Clinical and Laboratory Evaluation of Patients
Figure 6.3 Three samples with monoclonal proteins in typical positions for their heavy chain class. The top sample has an IgG
monoclonal protein migrating in the mid-g-region. Middle sample has an IgM monoclonal protein near the origin. Bottom sample has a
broad IgA monoclonal protein just cathodal to the C3 band. Although these are typical locations for monoclonal proteins of these
isotypes, they may migrate at a variety of locations from a2-region to the slow g-region. (Panagel system stained with Coomassie Blue.)
158 Conditions associated with monoclonal gammopathies
Reference Ranges
Rel % g/dl
ALBUMIN 49.7 64.4 3.55 5.04
ALPHA 1 4.8 10.1 0.25 0.74
ALPHA 2 8.5 15.1 0.55 1.14
BETA 7.8 13.1 0.55 1.04
GAMMA 10.5 19.5 0.65 1.44
Figure 6.4 Capillary zone electrophoresis demonstrates three peaks (arrows) of an IgA k monoclonal gammopathy. (Paragon CZE 2000.)
with Monoclonal Gammopathies recommends the immunophenotyping to conrm the presence of the
use of immunoxation in cases where the serum plasma cell surface antigen CD38 is helpful to dis-
protein electrophoresis is negative, but where there tinguish this condition from Waldenstrms
is a high index of suspicion.3 In practice, however, macroglobulinemia.153,154 Another useful marker for
in the laboratory we often do not know the clinical making this distinction is the translocation
history. Therefore, anytime a distortion in the b- t(11;14)(q13;q32), which is present in the vast
region protein bands is observed, I perform an majority of cases of IgM myeloma.155
immunoxation to rule out a subtle M-protein. By
using the pentavalent (Penta) antisera available on IGD MYELOMA
semiautomated immunoxation gels, this can be IgD myelomas are uncommon, accounting for only
done quickly and, using otherwise unused portions 12 per cent of cases, but they have some
of the Penta gels, at very little cost. By performing characteristics of which one must be aware to
this extra step I have detected cases of M-proteins avoid misdiagnosis.138,156,157 Most IgD monoclonal
where the major portion was found in the urine. gammopathies migrate in the g-region, 25 per cent
of cases have b-region spikes, and in one case the
IGM MYELOMA monoclonal component was in the a2-region (Fig.
Although unusual, IgM may be the main 6.5).157 IgD myeloma may be missed because the
immunoglobulin in cases of multiple myeloma, monoclonal gammopathy can be relatively small
accounting for about 0.5 per cent of cases.150,151 and it may be hidden among the normal a2- or b-
Clinically, these cases resemble the typical case of region proteins. A high index of suspicion about
multiple myeloma but may have an increased subtle distortions in the b- and g-region helps in
propensity to hyperviscosity.152 Typically, there is detecting them. Individuals with IgD myeloma tend
suppression of the uninvolved immunoglobulin to have a worse prognosis than most other isotypes
classes, IgG and IgA.153 IgM myeloma is associated because of the increased chance of renal involve-
with plasma cell proliferations as opposed to the ment due to the high proportion of cases with
more typical IgM monoclonal gammopathies (part monoclonal free light chains, amyloidosis and
of Waldenstrms macroglobulinemia) in which extramedullary plasmacytomas, although, rarely,
lymphoplasmacytoid cells are seen.152 The use of they may have good responses to chemo-
Multiple myeloma 159
(a) (b)
Figure 6.5 (a) Immunoxation of serum with mid-g M-protein in the ELP lane. The M-protein is matched by a broad restriction in the k
(K) lane, but in none of the other lanes. (Sebia Immunoxation). (b) Immunoxation of serum from case shown in (a). Here, the IgD and k
lanes both demonstrate this IgD k monoclonal gammopathy. The anti-free k demonstrates that the extra band in the k lane is due to k
monoclonal free light chain. (Sebia Immunoxation). This case was contributed by Chrissie Dyson.
therapy.138,156159 Cytoplasmic crystalline inclusions the case shown in Figure 6.5. Therefore, when send-
and amyloidosis have been described in patients ing a sample to a reference laboratory, inform them
with IgD myeloma.160,161 The latter may require that you suspect an elevated IgD and ask that they
immunohistochemical identication. Although an perform the assay at a 1:10 and 1:100 dilutions. IgD
IgD monoclonal gammopathy is usually associated myeloma should always be suspected when a light
with a lymphoproliferative process, IgD MGUS chain is identied in the serum by immunoxation
does occur.159 Cutaneous plasmacytomas have been or immunoelectrophoresis but there is no corre-
reported rarely in patients with IgD myeloma.162 sponding heavy chain (g, m, or a). One should also
Older literature reported the k/l ratio of cases with have a higher index of suspicion for an IgD mono-
IgD myeloma to be 1:9 (as mentioned above, it is clonal gammopathy in cases containing l than for
2:1 for most other monoclonal gammopathies, k spikes (reecting the disproportionate k/l ratio in
paralleling the usual k/l in the serum); however, a the occurrence of IgD myelomas).
more recent study of 53 cases at the Mayo Clinic
revealed a ratio of k/l of 1:2 in their cases.156 In any IGE MYELOMA
event, l clearly predominates. IgE myeloma is rare (38 cases were reported in the
To follow the tumor burden in IgD myelomas, most recent tally I found).136 The reported patients
measurement of IgD may be performed by a variety have been younger than is typical for myeloma,
of techniques, including nephelometry, radial and the disease pursues a relatively rapid course.
immunodiffusion and enzyme immunoassay.163,164 Patients with IgE myeloma are more likely to have
The IgD measurements may be more useful than the MFLC, anemia and plasma cell leukemia than
typical densitometric or electropherogram mea- patients with other isotypes (Table 6.7).135 The
surements of the M-protein because accuracy average survival time from time of diagnosis is
decreases with smaller quantities of the protein mea- 16 months compared with 30 months for non-IgE
sured. However, because the immunoassays for IgD multiple myeloma.135,165 Osteolytic lesions have
are standardized to the low concentrations of IgD been found in 14 of 28 patients where skeletal
normally found in serum, a falsely low value may information was available.165 Other boney lesions
be obtained unless the serum is further diluted prior include osteoporosis, vertebral collapse and one
to performing the assay. This is what happened in case with osteoblastic lesions.165,166 In addition, two
160 Conditions associated with monoclonal gammopathies
a
Data from Macro et al.165 and Kyle.135
cases with hyperviscosity have been reported.167,168 polyclonal increase in immunoglobulin produc-
Small, occasionally diffuse peaks have been tion, glomerular damage or tubular damage. In
described in the few cases where the IgE mono- those circumstances, by electrophoresis on concen-
clonal protein migrates in the b-region.169,170 Macro trated urine, PFLC migrate broadly in the b- and
et al.s summary of 29 published cases where g-regions. Yet, by immunoxation, they may
electrophoretic data was available is shown in produce a ladder pattern that could be confused
Table 6.7.165 As with IgD myelomas, when I detect with a monoclonal gammopathy (see Chapter 7 for
a monoclonal light chain in the serum without a discussion of ladder pattern) (Fig. 6.6). In contrast,
corresponding heavy chain (using the more usual when MFLC are present, they usually migrate as
IgG, IgA and IgM antisera) I perform an one, two or more discrete bands in the b- and g-
immunoxation to rule out an IgE M-protein. regions (Fig. 6.7). The specic isotype may be
determined by immunoxation. The Guidelines
MONOCLONAL FREE LIGHT CHAINS AND recommended that both urine protein electro-
LIGHT CHAIN MYELOMA phoresis and immunoxation be performed on
As discussed in Chapter 7, plasma cells produce concentrated urine as the screening test for MFLC.
heavy and light chains separately. Under normal Urine protein electrophoresis alone is too insensi-
circumstances, these chains combine and are tive to reliably detect some of the small MFLC that
secreted as intact immunoglobulin molecules may be associated with amyloid AL (Fig. 6.8).
bearing two identical heavy chains and two identi- For the past several years, the standard has been
cal light chains. Free light chains are produced in to perform electrophoresis and immunoxation on
excess by normal plasma cells. These light chains concentrated urine to detect and identify the
are secreted as monomers (mainly k) with a molec- MFLC. After identication, it is important to
ular mass of about 25 kDa, or as dimers (mainly l) measure the amount of MFLC to estimate the
with a molecular mass of 50 kDa.171173 These poly- tumor burden. For this, a 24-h urine specimen is
clonal free light chains (PFLC) are usually submitted for electrophoresis and protein determi-
reabsorbed by the proximal convoluted tubules nation. The MFLC peak is integrated by
where they are catabolysed. As a result, little PFLC densitometry and the nal result is expressed as
nds its way into urine under normal circum- mg/24 h (see Chapter 7). Using these electro-
stances. PFLC proteinuria results when there is a phoretic techniques, Kyle135 demonstrated that
Multiple myeloma 161
1 1
A A
2 2
3 3
4 4
F 5
5
X 6
6
P 7
7
Figure 6.7 Immunoxation of urine concentrated 100-fold. The
far left lane (protein electrophoresis xed in acid) shows a weakly
Figure 6.6 Immunoxation of a urine concentrated 100-fold with
staining albumin band (A) along with a few weakly staining a2- and
a mild protein loss pattern indicated by the presence of a small
b-region bands (small arrows) most consistent with a tubular
amount of albumin (A) in the far left lane (protein electrophoresis
proteinuria. The g-region shows one prominent band (P) and one
xed in acid). Although no other protein bands are visible in the
faint band (F). Immunoxation for k shows two strong bands in
left lane, there are multiple bands visible in the k immunoxation
with the same migration as the prominent and faint bands in the
electrophoresis lane (arrows), and a faintly staining diffuse haze in
adjacent lane. Note that the center of the prominent band shows a
the l lane. This is a classic ladder pattern that occurs in any urine
slight antigen excess effect (X). Slightly above the faint band and
containing polyclonal free light chains. The bands are more often
slightly below the prominent band are lightly staining areas (large
seen in k, although they are also seen, on occasion, with l. The
arrowheads), which may represent other forms of the monoclonal
bands with this polyclonal pattern are evenly spaced (like the rungs
proteins (breakdown products, post-translational modications, or
on a ladder). I interpret this pattern as Negative for Bence Jones
forms associated with intact heavy chains). Note that this
protein. Although the bands vary somewhat in their staining
immunoxation only shows k and l reactions, therefore, we
intensity from one to another, none shows the type of dense
needed to perform immunoxation for heavy chains to be certain
staining or antigen excess effect of the monoclonal free light chain
that these were not intact monoclonal proteins (in this case they
(MFLC) protein bands in Fig. 6.7. (Paragon system stained with
were not). Contrast this pattern of true k monoclonal free light
Paragon Violet; anode at the top.)
chain (Bence Jones protein) with a classic ladder pattern in Fig.
6.6. (Paragon system stained with Paragon Violet; anode at the
about 60 per cent of cases of multiple myeloma top.)
produce MFLC (formerly termed Bence Jones pro-
teins) in addition to the intact monoclonal urine. This is why urine must always be studied
immunoglobulin, or exclusively when present as (Figure 6.9). Weber et al.176 studied the course of
part of light chain myeloma. The presence of 101 asymptomatic patients with multiple myeloma
MFLC produces a more dire prognosis than their who were detected by chance while they lacked
absence.174180 Occasionally, patients with small or symptoms and found that the presence of MFLC
modest-sized monoclonal gammopathies in the excretion of > 50 mg/24 h in the urine was a signif-
serum will have massive amounts of MFLC in the icant independent variable that indicated treatment
162 Conditions associated with monoclonal gammopathies
SPE IgG K L
(a) (b)
Figure 6.8 (a) The top sample (1) is a lyophilized serum for reference. The bottom two samples of urine concentrated 100-fold (2, 3)
have no protein bands visible. (Paragon SPE2 system stained with Paragon Violet.) (b) Immunoxation of the urine sample labeled 2 from
(a). In the serum protein electrophoresis (SPE) lane a very faint band is barely discernable (arrow), yet the immunoxation using anti-k
shows an obvious k monoclonal free light chain. (Paragon system stained with Paragon Violet; anode at the top.)
even when the patients lacked lytic skeletal lesions. ate clinical picture: lytic lesions, anemia, hypercal-
Further, Knudsen et al.181 reported that myeloma cemia, and elevated creatinine. Urine must always
patients with renal failure who have relatively low be studied to rule out light chain multiple
amounts of MFLC in their urine have a better myeloma. Patients with multiple myeloma are
chance of achieving an improvement in their renal more predisposed to develop infections and the
function with chemotherapy than patients with antigenic stimulation from such infections may
high quantities of MFLC. result in a polyclonal increase in g-globulins.187
In most cases of light chain multiple myeloma, the While detection of MFLC by electrophoresis and
serum protein electrophoresis shows hypogamma- immunoxation has been useful, it will likely be
globulinemia and only a small M-protein spike supplemented by assays that allow quantication
because most of the MFLC passes through the of total (polyclonal and/or monoclonal) free light
glomerular basement membrane into the urine. chains (FLC) in serum or urine. Using sensitive
However, in cases of tetrameric light chain disease, nephelometric techniques to measure FLC,
or aggregates of light chains such as the mono- Bradwell et al.173 found that the vast majority of
clonal l hexameric aggregates (trimolecular patients with multiple myeloma produce excessive
complex of l dimers) reported by Abraham et al.,182 amounts of the FLC type that corresponds to the
a large M-protein is seen in the serum that contains intact M-protein (i.e. free k chains in an IgG k M-
only light chain and no heavy chain.183186 protein) that can be detected in serum (Freelite;
Nonetheless, in the vast majority of cases of The Binding Site Ltd, Birmingham, UK; see
MFLC, urine protein electrophoresis and Chapter 7). Katzmann et al.188 established
immunoxation lead to the correct diagnosis. 0.261.65 as the diagnostic interval of FLC k/l in
In some patients, the presence of a polyclonal serum. Their 95 per cent reference interval for k
increase in serum g-globulins does not rule out FLC in serum was 3.319.4 mg/l and for l was
multiple myeloma in the presence of an appropri- 5.726.3 mg/l. This method was more sensitive
Multiple myeloma 163
IgG SPE
IgA K
IgM L
(a)
Figure 6.9 (a) Immunosubtraction of serum from a patient with a modest-sized monoclonal gammopathy in the g-region (arrow in serum
protein electrophoresis (SPE) box, top right). When the serum was preincubated with antisera against the immunoglobulin indicated in
each box, only the antisera against IgA (arrow) and k (arrow) removed the monoclonal protein. (Paragon CZE 2000 immunosubtraction.)
G A K L
(b)
Figure 6.9 (contd) (b) Composite gure showing the protein electrophoresis on urine concentrated 50-fold (arrow indicates sample)
from the patient identied in (a) as having an IgA k monoclonal gammopathy. The pattern shows a small amount of albumin and several
small proteins in the b- and g-regions consistent with tubular damage. The densitometric scan demonstrates the massive amount of
monoclonal protein present. The immunoxation to the right of the densitometric scan has dense staining only in the k (K) lane with a
tiny amount of IgA staining. Although both the small amount of intact IgA k and the large amount of k monoclonal free light chain (MFLC)
migrate in the same location, the entire peak is integrated to measure the MFLC. (Sebia b1,2 gel and Sebia Immunoxation.)
than current immunoxation methods of serum use the free light chain assay to follow myeloma
and urine to detect free light chains in patients with patients. The main limitation of the FLC assay is its
amyloid AL.188 Marin et al.189 compared capillary inability to distinguish between MFLC and poly-
zone electrophoresis (CZE) with determination of clonal FLC of the same isotype. Electrophoresis
FLC in frozen sera from 54 patients conrmed to and immunoxation do this by the presence of
contain M-proteins by immunoxation. They restricted mobility. Therefore the demonstration of
found that sera from all 16 patients studied had probable monoclonality is accomplished by statis-
abnormal k/l and increased absolute values for the tical comparison of reference ranges. In cases
relevant FLC.189 where there is a polyclonal increase along with
The method to measure FLC in serum is a power- MFLC, a combination of electrophoresis,
ful new tool that will improve our ability to detect immunoxation and FLC measurement may prove
and follow many of these patients. It will be much to be ideal.
easier to obtain the serum sample than the 24-h
urine to follow the MFLC level in these patients. At NON-SECRETORY MYELOMA
the time of writing, however, 24-h urine is still the Non-secretory myeloma has been reported in from
standard, although our laboratory is beginning to 1 to 4 per cent of patients with multiple
Multiple myeloma 165
myeloma.190194 These are cases of monoclonal bone marrow biopsy and aspirate are obtained to
plasma cell proliferations in which no monoclonal explain the hypercalcemia and lytic skeletal
protein can be detected in the serum or urine by lesions. Similarly, because of the lack of a measur-
conventional electrophoretic techniques. The able M-protein, the response to therapy has been
plasma cells in this condition may contain a mono- also followed by bone marrow studies.191
clonal immunoglobulin, which may be detected by Occasional cases of acquired non-secretory
immunohistochemical analysis and occasionally myeloma secondary to chemotherapy have been
ultrastructural studies.195,196 A model of non-secre- reported. These can be a particular problem
tion by plasma cells is offered by Mott cells because the M-protein is an important measure of
(plasma cells with large intracytoplasmic inclu- tumor burden.202 The decreased, or absent synthe-
sions of immunoglobulin). They have a partial or sis in the natural or chemotherapy-acquired
complete block of the secretion of their non-secretory myeloma may relate to a defective B-
immunoglobulins, resulting in distended endoplas- cell response to differentiation signals.203
mic reticulum.197 Indeed, this phenomenon seems to Another unusual situation is that of pseudo-non-
be responsible for at least one report of non-secre- secretory myeloma described in two patients where
tory IgA k myeloma where immunohistochemistry no M-protein was found in serum or urine, but
was used to demonstrate the trapped intracellular where it was found by in vitro studies of the neo-
M-protein.198 plastic plasma cells.204 In those cases, the authors
Some cases of non-secretory myeloma result from concluded that the secreted M-protein was rapidly
synthesis of structurally abnormal molecules that catabolized and deposited in the kidneys and other
may be degraded intracellularly.196 Within the organs.204
plasma cells from these patients, one may detect The FLC nephelometric assay improves our
truncated or even intact M-proteins, but they are ability to detect most cases of non-secretory
not secreted in sufcient quantity to be found in myeloma. Drayson et al.205 studied serum from
the serum or urine by conventional electrophoresis patients with non-secretory myeloma using the
and immunoxation.199 Molecular evidence of FLC assay and detected abnormal ratios of free k/l
abnormal immunoglobulin formation in non- in 19 of 28 patients. This indicates that non-secre-
secretory myeloma was presented by Cogn and tory myeloma may really be a pauci-secretory
Guglielmi200 who demonstrated a truncated 42 kDa disease in most of the cases described where small
g1 heavy chain that lacked a variable region amounts of MFLC are secreted.
because of a 2-base pair (bp) deletion.
Cytogenetically, the vast majority of non-secretory BICLONAL GAMMOPATHIES
myeloma cases have the same t(11;14)(q13;q32) Biclonal gammopathies (or double gammopathies)
translocation seen with other types of multiple are uncommon, but not rare. They occur in about
myeloma.155 35 per cent of patients with conditions associated
Non-secretory multiple myeloma presents clini- with monoclonal gammopathies.27,135,206 Because of
cally with similar symptoms as cases of secretory their unusual appearance, biclonal gammopathies
myeloma, yet they lack the renal involvement (due can be somewhat confusing. Clinically, there is no
to the absence of MFLC production) and lack the difference between biclonal gammopathies and
convenient M-protein in serum and/or urine to monoclonal gammopathies in terms of outcome for
conrm the diagnosis.131 To avoid delay in diagno- the patient.206 There have been two cases reported
sis, special diligence is needed when confronted in the past few years where biclonal gammopathies
with symptoms of hypercalcemia in patients with (an uncommon lesion) were associated with the
radiographic and other features consistent with occurrence of angioimmunoblastic T-lymphomas
multiple myeloma but no M-protein in serum or (another unusual lesion).207210 This is probably a
urine.105,201 Typically, the diagnosis is made when a coincidence, but it deserves notice. However, in
166 Conditions associated with monoclonal gammopathies
patients with multiple myeloma and a biclonal classes but only one light chain type, there may be
gammopathy, it is important to measure both preferential switches to explain the frequency of the
clones during follow-up because the clinical two heavy chain types observed. Fortunately, other
response to the two clones may be asynchronous.211 than being confused by the initial electrophoretic
The term biclonal implies that the plasma cell pattern itself; there is no known clinical signicance
neoplasm arose from two separate clones of B lym- to the demonstration of double (or even true
phocytes. As such, they should always have a biclonal) gammopathies. Most of these patients
different variable region and may differ in light have monoclonal gammopathy of undetermined
chain class. There is no reason to look at variable signicance (see below), while others have B-cell
regions in clinical laboratory testing. Biclonal gam- lymphoproliferative disorders or multiple myeloma.
mopathies may originate from a single clone or Oligoclonal expansions due to infections, often
separate clones.212,213 seen as part of hepatitis C virus infection or acquired
The presence of different light chains is a priori immunodeciency syndrome may produce two,
evidence that the neoplasm represents the product three or more M-protein peaks.216 It is not clear
of two separate clones, which would be true where one of these peaks ceases to be a prominent
biclonal gammopathies (Fig. 6.10). However, oligoclonal band and becomes an M-protein. When
most reported cases and most that we have seen in one peak of an oligoclonal process appears appre-
our laboratory have had the same light chain type ciably larger than the rest, I measure it and recom-
with two different heavy or heavy chain subclass mend that the clinician perform a urine study to rule
types. Some heavy chain classes occur more fre- out MFLC, and follow the serum sample in
quently in double gammopathies with IgGIgM 36 months to see if the process regresses (Fig. 6.12).
occurring most frequently followed by IgGIgA,
IgGIgG (detected by their different electrophoretic
mobilities), and IgAIgM.214 These may represent
one clone of B-lymphocytes that is expressing two Immunosuppression in multiple
different heavy chain constant regions (similar to myeloma and B-cell
the stage in B-cell ontogeny where both IgM and lymphoproliferative disorders
another heavy chain class are present on the surface
of the B cell). Therefore, for cases with the same In multiple myeloma and chronic lymphocytic
light chain type one may use the term double gam- leukemia, concomitant suppression of the normal
mopathy, which recognizes the fact that two dis- immunoglobulin secretion is a key feature recog-
tinct proteins are seen but does not imply that they nized by examining the electrophoretic pattern
resulted from two separate clones. (Fig. 6.13).217,218 The decrease in production of
The reasons for the occurrence of double gammo- polyclonal immunoglobulins predisposes these
pathies are better understood by examining the patients to recurrent bacterial infections and may
structure of the immunoglobulin heavy chain con- be a cause of death.218
stant-region gene sequence (Fig. 6.11). When a cell The production of normal immunoglobulins is
switches from expressing one heavy chain gene to the result of a balanced interaction between B cells,
another, such as from m to a1, the intervening genes helper T-cells, cytokines, and antigen presenting
(in this case cmca1) are deleted and cannot be cells. Peripheral blood lymphocytes from patients
expressed later.215 The remaining segments, how- with multiple myeloma exhibit a mechanism
ever, could be selected and subsequently expressed. extrinsic to the B cells that mediates an arrest in
This occurs normally during B-lymphocyte matu- terminal B-lymphocyte maturation.114 These data
ration where B cells often bear both surface IgM and are also consistent with a profound decrease (as
another heavy chain isotype. Hammarstrom et al.140 much as 20- to 600-fold) of the normal polyclonal
noted that in gammopathies with two heavy chain B-lymphocytes in the circulation of patients with
Multiple myeloma 167
Figure 6.10 (a) The sample in the third lane has one relatively
small band in the g-region (arrow). However, by comparing the C3
regions of all of the samples, one may note that the C3 region of
this sample is considerably darker (C). Therefore, despite the less
than optimal separation of the different protein bands on this gel,
one should be suspicious that a second monoclonal band may also
be lurking in this area. (Paragon SPE2 system stained with Paragon
Violet.) (b) Immunoxation of the serum from (a) demonstrates a
true biclonal gammopathy (IgA k and IgM l). These small bands
were not associated with clinical symptoms and in this case are
considered biclonal gammopathies of undetermined signicance
(BiGUS?). (Paragon system stained with Paragon Violet; anode at
the top.)
(a)
(b)
168 Conditions associated with monoclonal gammopathies
5 3
V1 V2 Vn Cm Cd Cg3 Cg1 Ca1 Cg2 Cg4 Ce Ca2
Figure 6.11 Schematic representation of heavy chain gene rearrangement during B-cell maturation. In B lymphocytes, the C m chain is
usually selected to be the heavy chain isotype expressed on the lymphocyte surface membrane. During further maturation to a plasma cell,
another heavy chain gene is selected (in this case C a1. While intervening genes are deleted (C m, C d, C g3, and C g1 in this case) during
maturation of a particular clone, the remaining heavy chain genes are still available and may be selected for expression at a later time in
maturation. This could result in the double gammopathies; that is, two heavy chains that originated from a single clone. Note that the
variable region gene (actually a combination of genes) selected is the same in the B-cell surface membrane and in the eventual
immunoglobulin product secreted by the plasma cell.
multiple myeloma, implying the existence of a sup- suppression of the normal g-globulin content. This
pressive inuence on B-lymphocyte proliferation.219 number has not been derived through rigorous
Interestingly, the number of B cells does not seem investigation. It reects the cut off of 0.25 g/dl that
to correlate with disease status or the concentra- has been used as a marker for common variable
tion of the monoclonal protein.219 immune deciency (CVID). In patients with CVID,
The decrease in normal immunoglobulins that those with IgG levels over 0.3 g/dl have adequate
occurs in myeloma patients has variously been lymphocyte proliferation responses to S. aureus
hypothesized as reecting excessive suppressor-T and E. coli, whereas those with IgG levels < 0.125
cell activities, decient helper T-cell numbers and g/dl have markedly decreased responses.224
function, decreased numbers of pre-B cells,
unusual macrophage products or dysfunctional
natural killer (NK) cells.220223 In studies using
mitogen stimulation of peripheral blood mononu- HEAVY CHAIN DISEASE
clear cells from patients with multiple myeloma
and reduced serum immunoglobulin levels, a Heavy chain disease
Walchner and Wick218 identied CD8+, CD11b+,
Leu-8- T cells as playing a role in suppressing a Heavy chain disease is extremely uncommon,
polyclonal immunoglobulin secretion. especially in the Western world. It occurs most
In evaluating serum protein electrophoresis, after frequently in the Middle East and Mediterranean
I measure the M-protein in the g-region, I note the region.225,226 The disease has an onset in much
amount of non-M-protein g-globulin. If the younger patients than myeloma or most B-cell
number is less than 0.25 g/dl I note that there is lymphoproliferative diseases. It usually develops
Heavy chain disease 169
(a)
in the third decade, but has been reported as early the patients present with diarrhea (often with
as 9 years of age.227 Using immunoelectrophoresis steatorrhea), malabsorption and weight loss.
or immunoselection techniques (see Chapter 3), Usually, there is slow progression of their disease
free a heavy chains have been demonstrated in the during the early phase and, importantly, cures
serum, urine and intestinal uids.227229 Clinically, have been reported when treated at this stage.230
170 Conditions associated with monoclonal gammopathies
Fatigue 70
Lymphadenopathy 22
Splenomegaly 18
Hepatomegaly 13
Extranodal involvement 6
Bleeding tendency 17
Infection 17
Hyperviscosity syndrome 12
Cardiac failure 25
a
Data from Kyrtsonis et al.263
(a)
(b)
Waldenstrms macroglobulinemia 173
IgG SPE
IgA K
(c)
IgM L
Figure 6.14 (a) The top lane shows a serum with an acute-phase reaction and a polyclonal increase in g-globulins. The serum in the
second lane has a somewhat diffuse restriction (arrow) in the fast g-region. It stands out especially well, despite being rather small because
the remainder of the g-region is decreased. The third sample has a borderline low/normal g-region. The bottom lane shows a polyclonal
increase in g-globulin. SPE, serum protein electrophoresis. (SPE2 system stained with Paragon Violet.) (b) Immunoxation demonstrates
the diffuse band in the fast g-region is IgG (arrow), but at 1:5 dilution, no such band is seen in the K or L lanes. This is presumptive
evidence for g heavy chain disease, however, an immunoselection procedure which precipitates the intact IgG molecules and
demonstrates the free IgG would be needed to absolutely conrm this impression. (Paragon system stained with Paragon Violet.)
(c) Immunosubtraction of the same serum from (a) and (b) demonstrates that the large monoclonal gammopathy can only be removed by
the antisera against IgG, not by either the anti-k or anti-l antisera. Further supportive evidence of this case being g heavy chain disease.
(Paragon CZE 2000.) Case contributed by Dr Gary Assarian.
174 Conditions associated with monoclonal gammopathies
Monoclonal gammopathies in
lymphoma and leukemia
Protein electrophoresis of serum and urine may
provide useful information about B-cell lympho-
proliferative processes. As discussed above,
patients with multiple myeloma have clonal B-cell
precursors with surface immunoglobulin and even
pre-B cells with cytoplasmic m of the same idiotype
(roughly equivalent with reactivity), implying that
although the neoplasm manifests mainly as mono-
clonal plasma cells, this is a disease of the entire
lineage of that particular B-cell clone.3537,109,298 B-cell
chronic lymphocytic leukemia (CLL) and myeloma
represent cells from the same lineage at various
stages of maturation; although myeloma seems to
involve the transformation of a pluripotent stem
cell, whereas, CLL seems to involve a more mature,
terminally committed cell.299,300 The stage of a given
B-cell neoplasm is not irreversibly xed, and may
change during the course of an illness. Not surpris-
ingly then, B cell neoplasms such as chronic
lymphocytic leukemia may transform to a predom-
Figure 6.16 Composite gure showing original electropherogram inantly plasma cell neoplasm producing the same
on a patient with Waldenstrms macroglobulinemia but an heavy and light chain types found on the B cells,
unimpressive M-spike (arrow in top gure) (Paragon CZE 2000).
and some cases of myeloma evolve into an aggres-
The middle gure is the densitometric scan from the Sebia b1,2 gel
that demonstrates the prominent M-protein peak. The bottom sive lymphoproliferative phase, characterized by
gure is the same serum re-assayed by capillary zone rapidly enlarging soft-tissue masses.301,302
electrophoresis following treatment of the serum with 2- When less sensitive techniques were available to
mercaptoethanol. evaluate serum and urine from patients with B-cell
lymphoproliferative disorders, monoclonal gam-
mopathies were detected only in a small number of
study of 57 patients with Waldenstrms macro- cases.303 However, the current widespread use of
globulinemia, Jonsson et al.295 reported that 51 per immunoxation demonstrates that a large number
cent of them had clinical and/or serological of patients with chronic lymphocytic leukemia will
autoimmune ndings: autoimmune hemolytic have at least one and some will have a few (oligo-
anemia, rheumatoid arthritis, parietal cell anti- clonal) clonal immunoglobulin products demon-
bodies and IgM-anticardiolipin. Occasionally, the strable in serum and/or urine.304 Using
reactivity of these M-proteins explains specic immunoxation and high-resolution agarose gel
symptoms, such as IgM antibodies against myelin electrophoresis Berstein et al.303 detected mono-
basic protein in patients with neurological symp- clonal gammopathies in the serum from 36 of 111
toms.296 Another was associated with lupus patients with CLL. This is not a new nding. In
176 Conditions associated with monoclonal gammopathies
1909, Decastello detected monoclonal free light of T lymphocytes to facilitate proper maturation of
chains in the urine from a patient with CLL.305 Not uninvolved B lymphocytes.328 Another hypothesis
surprisingly, the monoclonal gammopathy in the notes that CD95 is upregulated on the uninvolved
serum and/or urine of patients with CLL corre- plasma cells of patients with CLL. Interaction of the
sponds to the molecules expressed on the cell sur- CD95 ligand (CD95L) produced by the neoplastic
faces.306308 B-lymphocytes from patients with CLL B cells with this receptor leads to death of normal
can be induced, by EpsteinBarr virus or mitogens plasma cells via apoptosis.328 Paradoxically, even
such as phorbol ester, to differentiate into with their hypogammaglobulinemia, patients with
immunoglobulin-secreting cells in vitro.309 Other CLL are more likely than the general population to
CLL cells have been shown to spontaneously secrete develop autoantibodies, and especially those
monoclonal light chain or monoclonal whole against hematologic cells (i.e., autoimmune
immunoglobulins.310 Further, the demonstration of hemolytic anemias and immune thrombocytope-
a monoclonal gammopathy in those individuals was nias).11,329,330
associated with a decrease in median survival from
103 months for those without the monoclonal gam-
mopathy to 63 months for those with one.303 MONOCLONAL GAMMOPATHY
Interestingly, that decrease was independent of the ASSOCIATED WITH TISSUE
clinical stage of the patients. DEPOSITION: AMYLOIDOSIS AND
Other B-cell lymphoproliferative processes have
NON-AMYLOID MONOCLONAL
also been reported to show monoclonal proteins in
serum or MFLC in urine. Nodular lymphoma,
IMMUNOGLOBULIN DEPOSITION
Burkitts lymphoma, lymphoplasmacytoid leu- DISEASE
kemia with hairy-cell morphology, and even
angioimmunoblastic T-cell lymphoma (AITL) Amyloidosis
evolving into an immunoblastic lymphoma have
had monoclonal proteins demonstrated by a com- Amyloidosis is another major clinical condition
bination immunoxation.311322 It should be noted, associated with monoclonal gammopathies.
however, that in most cases of AITL the serum con- Amyloid is a general term that literally means
tains a polyclonal increase in g-globulins.323 One starchlike. It is dened by the tinctorial quality of
extraordinary case of T-cell acute lymphoblastic the staining reaction in parafn-embedded tissue.
leukemia has been reported in a child with a mono- Amyloid stains with Congo Red and has a charac-
clonal gammopathy.324 Molecular studies on that teristic yellow-green birefringence under polarized
case demonstrated both rearrangement of the T cell light. Several different proteins may result in this
receptor b gene and the immunoglobulin heavy deposition; therefore, amyloidosis may be caused
chain genes.324 by a wide variety of conditions. On an ultrastruc-
As mentioned above, similar to patients with tural level, the amyloid brils of serum amyloid A
multiple myeloma, those with B-cell lymphopro- (see below) are 7.58.0 nm wide whereas those for
liferative disorders may also suffer hypogamma- transferrin-associated familial amyloidic poly-
globulinemia and have difculty synthesizing neuropathy are 13.0 nm wide.331 All forms of
immunoglobulins in response to infectious dis- amyloid have the b-pleated sheet structure that
eases.11,325,326 Intravenous g-globulin has been rec- accounts for the Congo Red birefringence.331
ommended to ameliorate this problem.327 As in Amyloid may be broadly classied as acquired or
multiple myeloma, mechanisms have been hypothe- inherited. The most common form of acquired
sized for the decreased immunoglobulin production amyloidosis is amyloid AL that has an incidence of
against normal stimuli. One mechanism involves 0.9 per 100 000 person years.332 Amyloid AL
impairment by the neoplastic B cells of the function results in the deposition of MFLC systemically as
Monoclonal gammopathy associated with tissue deposition 177
amyloid brils. Tissue deposition of AL preferen- k/l ratio in serum using the nephelometric FLC
tially involves the tongue, heart, gastrointestinal assay is more sensitive than immunoxation in
tract, blood vessels (including glomerular capillar- detecting patients with amyloid AL as well as in
ies), tendons, skin, and peripheral nerves. The patients with light chain deposition disease.
clinical picture in these patients parallels the sites The other form of acquired amyloidosis is reac-
of involvement, with macroglossia, congestive tive systemic or secondary amyloid A.340 This form
heart failure, carpal tunnel syndrome, purpura, of amyloid is found in patients with a wide variety
renal failure, and peripheral neuropathy as promi- of chronic inammatory processes and is com-
nent features. Further, the optimal sites (which posed of the an acute-phase reactant protein,
should be judged on the symptoms for the individ- serum amyloid A (SAA).341 Its synthesis in both
ual case) for biopsies of suspected cases reect hepatic and extrahepatic locations is stimulated by
distribution and availability of the site. These IL-1, IL-6 and TNF.341 The deposition of SAA is a
patients usually do not have bone pain or oste- dynamic process during the inammation such that
olytic lesions.333 Kyle and Greipp334 recorded 229 the use of anti-inammatory agents to keep SAA
patients with AL of whom 47 (20.5 per cent) had concentration below 10 mg/l may allow the
multiple myeloma; they found that the presence of amyloid deposition to regress and the organ
myeloma did not contribute to prediction of sur- involved to recover its function.340
vival at 1 year. Using a ve-band electrophoretic The hereditary form of amyloidosis may be
technique, Kyle and Greipp found a discrete band caused by deposition of a wide variety of struc-
in only 40 per cent of their patients with amyloido- turally abnormal forms of plasma proteins:
sis while demonstrating monoclonal protein in 68 transthyretin (prealbumin), apolipoprotein A-I,
per cent of the sera and MFLC in 70 per cent of the apolipoprotein A-II, brinogen, lysozyme, gelsolin,
urine by immunoelectrophoresis.334 Immuno- and cystatin C.342 Variants of transthyretin are the
xation is more sensitive and may detect a most common forms of autosomal dominant
monoclonal component in as many as 95 per cent systemic amyloidosis typically resulting in neuro-
of amyloid AL cases.335,336 Prognosis varies with the pathies (although some do involve the heart or
type of light chain involved. The presence of mono- kidneys).343 It is important to distinguish between
clonal l light chains in the urine of patients with hereditary forms of amyloid, amyloid AL, and
primary systemic amyloidosis was found to have amyloid AA because the treatment, prognosis and
an average 12-month survival compared with the genetic counseling differ. Although the presence of
30-month survival of patients with k chain excre- an M-protein in serum or urine is helpful, it does
tion and 35 months for those with no monoclonal not conrm that the patient has amyloid AL. As
protein in the urine.337 discussed below, monoclonal gammopathy of
Kyle338 cautions that amyloid AL needs to be ruled undetermined signicance is common and may be
out when a patient has unexplained renal failure, an innocent bystander in the serum of a patient
congestive heart disease, peripheral neuropathies, with a hereditary form of amyloid. Therefore, if
hepatomegaly, or malabsorption. Amyloid AL is immunohistochemistry of the tissue involved does
always associated with clonal plasma cell prolifer- not conrm the presence of immunoglobulin light
ation, however, in 15 per cent of them a chain, genetic studies should be performed.343
monoclonal gammopathy is not detected in either
the urine or serum even when immunoxation is
used to enhance sensitivity.20,135,339 Non-amyloid monoclonal
When the serum and urine are negative in an indi- immunoglobulin deposition disease
vidual suspected of having amyloid AL, one further
study should be performed. Katzmann et al.188 have Light chains are made in excess by plasma cells,
demonstrated that the presence of abnormal free pass readily through the glomerulus and are
178 Conditions associated with monoclonal gammopathies
reabsorbed mainly by the proximal convoluted such as restrictive cardiomyopathy, or more com-
tubules. About 12 g of protein can be reabsorbed monly, renal disease.345,347,350 Reports of tumoral
by normal tubules in a 24-h period. Since the non-amyloidotic monoclonal immunoglobulin
amount of polyclonal free light chains rarely light chain deposits in lymphoid and pulmonary
exceeds this, only trivial quantities of polyclonal tissue may be an early presentation of LCDD.351
free light chains nd their way into urine daily. Amyloidosis AL is more likely to occur in the
These can be visualized by performing immunox- absence of multiple myeloma than is MIDD;
ation only after concentrating the urine much however, the clinical overlap of the two conditions
greater than the typical 50- to 100-fold used in the is considerable. In Buxbaums series, 13 per cent of
clinical laboratory. patients with amyloidosis AL and 20 per cent of
Large amounts of MFLC overwhelm the low- patients with LCDD had neither serum nor urine
afnity receptors present on the brush border of monoclonal component (Fig. 6.17).352 Light chain
the proximal convoluted tubules that normally deposition disease is present much less often than
bind the polyclonal free light chains and begin either amyloidosis or BJP (Bence Jones protein)
them on their journey to lysosomal acid hydroly- cast nephropathy. In an autopsy series of 57
sis.344 In patients with large amounts of MFLC, patients with myeloma, Ivnyi353 found that 32 per
these proteins may deposit in glomeruli. cent had MFLC cast nephropathy, 11 per cent had
Absorption of the excess amount of MFLC occa- renal amyloidosis and only 5 per cent (three
sionally damages the proximal convoluted tubule patients) had light chain deposition nephropathy.
cells. Sanders and Booker344 note that the proximal The severity of the renal involvement varies widely
convoluted tubule damage and the myeloma cast with the survival related to the diagnosis of the
nephropathy are different events. underlying plasma cell process.345
Although amyloidosis has been classically associ- In those cases where there is no monoclonal com-
ated with systemic dysfunction due to deposition ponent in serum or urine the diagnosis of MIDD
of amyloid in the involved tissues, non-amyloid can be a challenge. Light chain deposition disease
light chains (and occasionally truncated heavy has been described in cases of non-secretory
chains) can also deposit in glomeruli and other myeloma, later called pseudo-nonsecretory
organs, resulting in disturbed function.345,346 In this myeloma because of the demonstration in vitro
situation, where amyloid cannot be demonstrated
(negative for Congo Red staining), but light chains
can, the term light chain deposition disease
(LCDD) has been used.345 Because we now recog- Proteinuria
nize that occasionally truncated heavy chains Kidney disease
may also be involved either alone or with light
Heart disease
chains the terms HCDD and LHCDD have been
added.46,345,347349 Since these terms become clumsy, I Liver disease
prefer the lumpers designation of monoclonal Neuropathy Amyloidosis
immunoglobulin deposition disease (MIDD).335 LCDD
Myeloma
Serum and urine protein electrophoresis and urine
immunoxation demonstrate the presence of a No-M
monoclonal gammopathy in 7085 per cent of 0 20 40 60 80 100
patients with MIDD.335 Often, hypogammaglo- Percentage of patients
bulinemia is notable in the serum.
Figure 6.17 Percentages of patients with the indicated clinical
Although many cases of MIDD occur in patients features are shown for those with amyloid AL (black bars) and
with multiple myeloma,345,347 it can be a part of those with light chain deposition disease (LCDD; white bars). Data
MGUS or may present with clinical manifestations, from Table 2 of Buxbaum.352
Monoclonal gammopathies not associated with B-lymphoproliferative disorders 179
that the bone marrow plasma cells produce a radiotherapy. When the monoclonal protein dis-
defective monoclonal protein.204 As mentioned appears after radiotherapy, a long-term disease-
above, in those cases with a high index of suspicion free survival can be anticipated.359,360 Liebross et
but a negative immunoxation of serum and urine, al.360 found that plasmacytoma patients with non-
measurement of the free k/l ratio in serum using secretory lesions and those with a monoclonal
the nephelometric assay specic for polyclonal free gammopathy that persisted both had about a 60
light chains may demonstrate the presence of an per cent chance of developing multiple myeloma
abnormal ratio.188,205 as opposed to only about 20 per cent of patients
with a monoclonal gammopathy that disappeared.
SOLITARY PLASMACYTOMA
MONOCLONAL GAMMOPATHIES
Solitary plasmacytomas may be located in bone in
extramedullary locations or, rarely, at both sites.354
NOT ASSOCIATED WITH B-
There is a major difference in outcome depending LYMPHOPROLIFERATIVE
on whether the bone is involved. Whereas all of DISORDERS
these patients are probably part of the same
spectrum of disease, patients with solitary plasma- Autoimmune disease
cytomas of bone have a worse prognosis than
individuals with extramedullary plasmacytomas Monoclonal gammopathies occasionally have been
(typically the head and neck, especially in the detected in patients with autoimmune diseases. In
upper respiratory tract).355 This may partly reect some cases, such as monoclonal anti-rheumatoid
the difculty in excluding involvement of other factor, anti-nuclear antibody, lupus anticoagulant,
bones at the time of diagnosis; since in multiple anti-platelet, anti-neutrophil, and anti-insulin, the
myeloma, the bone marrow involvement is often specicity of the autoantibody is known.361367 In
patchy.356 The best outcome seems to be among other cases, the relationship of the monoclonal
individuals with primary lymph node plasma- antibody to the autoimmune disease is not known.
cytomas, where none of the 25 cases progressed to However, removal of the monoclonal antibody by
multiple myeloma.357 plasmapheresis has been reported to result in clini-
Although solitary plasmacytoma involving bone cal improvement in patients with monoclonal
may be an early presentation for multiple gammopathies associated with polymyositis
myeloma, Frassica et al.358 reported that they had (where monoclonal antibodies have been detected
a much better 5-year survival (74 per cent) than in the sarcolemmal basement membrane).368 Of
did patients with multiple myeloma (18 per cent). course, other antibodies and non-immunoglobulin
They recommend the use of aggressive radiother- molecules with signicant biological activity are
apy for these lesions. Monoclonal proteins are also removed by this process. In some patients with
found in 56 per cent of serum screens (using a Sjgrens syndrome, IgM monoclonal proteins
low-resolution electrophoresis method) of these have been associated with plasma cell inltrates in
patients.358 The presence of a monoclonal the salivary glands.369
gammopathy, however, is not required for a diag- In addition to autoantibodies, monoclonal pro-
nosis of either solitary plasmacytoma of bone or teins have reactivity to other common antigens. A
extramedullary plasmacytoma. Typically, mono- wide variety of reactivities of monoclonal proteins
clonal gammopathies in both groups of patients that have been determined include bacterial pro-
are relatively small. If a serum or urine mono- teins, cardiolipin, polysaccharides, viral antigens,
clonal protein is present, follow-up electro- and other major serum proteins including iso-
phoresis is useful to gauge response to the enzymes, albumin, and al-antitrypsin.370373
180 Conditions associated with monoclonal gammopathies
Monoclonal gammopathies and Even though most of the recorded cases have an
neuropathies IgM monoclonal antibody, IgA and IgG mono-
clonal isotypes have also been described in patients
About 10 per cent of patients with idiopathic with polyneuropathy, especially those associated
peripheral neuropathies have monoclonal gammo- with osteosclerotic myeloma (POEMS syndrome;
pathies in their serum.374,375 Looking at it from see below).375,395 Histologically, nerve biopsy may
another perspective, almost half of the individuals reveal loss of both myelinated and unmyelinated
with macroglobulinemia have clinical evidence of nerve bers, and in cases associated with
neuropathy.376 The underlying process may be macroglobulinemia, diffuse inltration by lympho-
benign or malignant.375 Yet, because monoclonal plasmacytic B-cells.396 The monoclonal antibodies
gammopathies are relatively common, coincidental have been found in widened lamellae of myelinated
occurrence of this nding with a patient that has bers by immunoelectron microscopy.397,398
peripheral neuropathy from another cause needs to The quantity of these monoclonal proteins in
be ruled out.375 This can be accomplished by serum may be quite small, requiring electrophoretic
looking at the specic reactivity of the monoclonal techniques of high-resolution and immunoxation
protein. The neuropathies are usually sensori- for adequate demonstration; routine ve-band
motor, but may be limited to motor disturbances. electrophoresis may miss the monoclonal band.399
IgM k is the most common M-protein detected, Vrethem et al.400 have documented that immuno-
although other isotypes have been described. xation, overall, is superior to routine agarose
Whereas the relationship between the monoclonal electrophoresis in detecting small monoclonal
gammopathy and the peripheral neuropathy is gammopathies in these patients. The correct diag-
unclear in most cases, autoreactivity with myelin nosis is important because plasma exchange has
has been shown in some.377384 been shown to be helpful in treating some neuro-
Specic reactivity against myelin-associated glyco- pathies associated with MGUS.401 Surprisingly, con-
protein (MAG; 100 kDa) has been characterized in sidering that IgM is located mainly in the
many cases.385387 Demyelinating peripheral neuro- intravascular compartment and would be expected
pathy is associated with IgM monoclonal anti-MAG, to decline more rapidly with plasma exchange than
whereas antibody reactivity against ganglioside would IgG or IgA, plasma exchange has been more
antigens GM1 is most closely associated with motor effective with IgA and IgG monoclonal proteins than
neuropathies, and anti-GD1b, anti-sulfatide and with IgM monoclonals.401 This may reect a differ-
anti-chondroitin sulfate reactivity has been associ- ence in the pathogenesis of neuropathy associated
ated with sensory neuropathies.375,388,389 Further, the with IgM than that associated with IgG or IgA. IgM
monoclonal antibody HNK-1 has been shown to monoclonal proteins associated with polyneuropa-
react with an epitope similar to that recognized by thy can result in a complement mediated demyeli-
IgM anti-MAG.390 In vitro studies suggest that the nation.402 In contrast, a controlled study of
anti-GM1 antibodies are a key participant in facili- intravenous immunoglobulin in demyelinating
tating bloodnerve barrier dysfunction.391 neuropathy associated with IgM anti-MAG con-
While most cases of monoclonal gammopathy cluded that this form of therapy had only a modest
associated with peripheral neuropathies are spo- benet to less than 20 per cent of their patients.403
radic, familial occurrence of polyneuropathy has
been reported.392,393 It is important to distinguish
between hereditary neuropathies and those associ- POEMS SYNDROME
ated with monoclonal gammopathies directed
against MAG because the latter may respond to Another association of monoclonal gammopathies
interventional drug therapy while hereditary neu- and neuropathies is the POEMS syndrome, also
ropathies lacking these antibodies do not.394 called CrowFukase syndrome.404406 The acronym
Monoclonal gammopathy of undetermined signicance 181
POEMS stands for peripheral neuropathy, organo- increases with time such that 2 per cent of individ-
megaly (usually hepatosplenomegaly although uals over the age of 50 years and 3 per cent over 70
lymphadenopathy has been included as an alterna- will have one. Another study found 10 per cent of
tive), endocrine dysfunction (including diabetes the ambulatory elderly population have a demon-
mellitus, thyroid dysfunctions, impotence, viriliza- strable monoclonal gammopathy.20,31,412 Although
tion, gynecomastia and infertility), monoclonal the cause of this increased incidence of monoclonal
gammopathy and skin changes (hyperpigmenta- gammopathies with age is unknown, it is clear that
tion).407 The disease tends to occur in a younger age immunoregulatory capability also declines with
group than multiple myeloma, with a mean of age.413,414 Deciencies of regulatory T-suppressor
51 years reported in a study of 99 patients.407 activity could allow emergence of clonal prolifera-
Median survival was 165 months in that study. tions, resulting in monoclonal gammopathies.415
The presenting signs and/or symptoms in most of The risk of an MGUS to progress to either multiple
these cases relate to neuropathy or weakness.408 myeloma or a related B-cell lymphoproliferative
The POEMS syndrome is associated with the rare process was reported by Kyle et al. to be about 1
osteosclerotic variant of multiple myeloma. Most per cent per year.20 However, that may be a rela-
patients with myeloma develop lytic skeletal tively high estimate because the MGUS cases culled
lesions that result in bone pain and pathologic frac- early on in that data used electrophoretic tech-
tures. However, about 3 per cent of myeloma niques of low resolution that may not have
patients have osteosclerotic lesions and do not detected some of the small M-proteins detected by
manifest bone pain. These individuals usually have many laboratories today.
a single or multiple sites of osteosclerotic bone It is unclear how, or even if, a benign monoclonal
lesions, but do not have diffuse osteosclerotic bone gammopathy evolves into a malignant process;
involvement. This distinction is important because hence Kyle coined the term monoclonal gammo-
Lacy et al.409 report that patients with widespread pathy of undetermined signicance (MGUS) to
bone lesions have a more aggressive clinical course account for this phenomenon in order to avoid the
typical of multiple myeloma rather than the more term benign monoclonal gammopathy.416 An
indolent course of POEMS syndrome. The mono- MGUS is the presence of a monoclonal gammopathy
clonal proteins that occur with the POEMS in serum at a concentration of 3.0 g/dl (30 g/l) in
syndrome are usually small IgA or IgG l mono- patients with no or at most moderate amounts of
clonal gammopathies that do not have specicity MFLC in the urine and who do not have lytic lesions,
for myelin.409 However, there is no clinical value in anemia, hypercalcemia, or renal insufciency (sec-
distinguishing the more common osteosclerotic ondary to a monoclonal protein).20 Bone marrow
myeloma (with one or a few sites involved) from must contain less than 10 per cent plasma cells.20
POEMS syndrome because they have a similar clin- In a review of 241 patients thought to have
ical course.408 benign monoclonal gammopathy, 24 per cent
developed myeloma or related disorders after
2035 years of follow-up (Table 6.9).416 Because of
MONOCLONAL GAMMOPATHY this long-term potential problem, evaluation for a
OF UNDETERMINED monoclonal gammopathy should be part of the
SIGNIFICANCE workup of a potential bone marrow donor.417
Recently, Kyle et al.20 reported on the long-term
The incidence of both polyclonal and monoclonal follow-up of 1384 patients with MGUS diagnosed
gammopathies increases with advancing age.410 A between 1960 and 1994. With 11 009 person-years
monoclonal gammopathy is demonstrable in the of follow-up they found that 115 of the patients
serum of about 1 per cent of individuals over the progressed to develop either multiple myeloma, or
age of 25 years.25,411 The incidence of MGUS another B-cell lymphoproliferative disorder (Table
182 Conditions associated with monoclonal gammopathies
6.9). The greatest risk (compared with an age- and have been reported with varying degrees of suc-
sex-matched population) of progression was for cess.124,421423 None of these special markers has yet
Waldenstrms macroglobulinemia and multiple been embraced to evaluate MGUS patients.
myeloma, and the overall risk of progression to However, some more common clinical features of
some lymphoplasmacytic neoplasm was 7.3 times MGUS may be helpful, in identifying those indivi-
that of the controls.20 duals that might benet from stricter monitoring.
There have been many attempts to identify the indi- The presence of > 5 per cent bone marrow plasma-
viduals with MGUS who are most likely to progress. cytosis, detectable MFLC, hypogammaglobuline-
As mentioned above (see Multiple Myeloma mia, and increased sedimentation rate indicates an
section), the neoplastic cells from patients with increased chance of MGUS progressing to a neo-
multiple myeloma typically have translocations plastic lymphoplasmacytic process.424 However,
involving the immunoglobulin heavy chain locus in these factors are not agreed upon by all studies.
switch regions.102,103 Some workers have suggested Kyle et al.20 did not nd a clear association between
that 14q32 translocations and monosomy 13 sug- suppression of the uninvolved immunoglobulin
gest a transition phase from MGUS to multiple class and progression. They also did not nd that
myeloma.418,419 Unfortunately, translocations of the the presence of an MFLC in the urine indicated pro-
immunoglobulin heavy chain locus were found in gression. Yet, they did note that patients with IgM
46 per cent, l light chain translocations in 11 per or IgA MGUS were more likely to develop a malig-
cent, and t(11;14)(q13;q32) in 25 per cent of nant lymphoplasmacytic process than patients that
patients with MGUS and were not felt to be useful had an IgG MGUS.20 They also found that the
in early detection of progression.420 At present, cyto- higher the concentration of the M-protein was, the
genetics does not help to distinguish the subpopu- more likely it was that the disease would progress.20
lation of MGUS patients who are at highest risk for
progression of their disease. The clinical signicance
of these translocations for the prognosis of those Smoldering multiple myeloma
patients is not known.
A variety of non-invasive markers including bone To diagnose myeloma, one must document the pre-
turnover markers, cytokines, and labeling studies sence of increased plasma cells, tissue involvement,
Multiple myeloma 75 25
c
Lymphoma 19 3.9
Amyloid AL 10 8.4
Macroglobulinemia 7 46
d
Chronic lymphocytic leukemia 6 1.7
Plasmacytoma 1 8.5
a
Data from Kyle et al.20
b
Risk is compared with age- and sex-matched populations.
c
Includes patients with IgM, IgA or IgG monoclonal gammopathy and lymphoma.
d
Includes patients with IgM, IgA or IgG monoclonal gammopathy and chronic lymphocytic leukemia.
Monoclonal gammopathies in infectious diseases 183
and monoclonality. Kyle and Greipp334 noted that examination, skeletal X-rays, and examination of
some patients with these features did not undergo tissue lesions for the conditions discussed
progressive deterioration; they did not have anemia, above.426428
lytic bone lesions, hypercalcemia, or renal failure.
Even though the median initial monoclonal protein
was 3.1 g/l, overt symptoms of myeloma did not MONOCLONAL GAMMOPATHIES
develop for at least 5 years of follow-up. They IN INFECTIOUS DISEASES
termed the disease of these individuals smoldering
multiple myeloma and recommended following
them closely without therapy. Other investigators Monoclonal gammopathies have been reported in
also have reported a slow course. Kanoh425 reported association with infections.425,429432 Endocarditis is a
a case with 34 g/dl of IgG k monoclonal protein particularly frequent clinical diagnosis in infections
and 10 per cent plasma cells in the bone marrow. with monoclonal gammopathies.433435 Most of the
Although the patient was mildly anemic (hemoglo- monoclonal gammopathies associated with infec-
bin was 10.2 per cent), he was otherwise well and tious diseases are transient, although some persist
was followed with no disease progression for more for more than 6 months.24,436 More typically,
than two decades. The type of monoclonal protein patients with infections have oligoclonal gammo-
does not appear to be a distinguishing feature; even pathies.437 The reported monoclonal gammopathies
IgD MGUS cases have been described. The presence may reect the fact that during an oligoclonal
of > 10 per cent plasma cells in the bone marrow, expansion caused by an infection, individual clones
detectable MFLC in urine and a monoclonal pro- may not always produce the same serum concen-
tein of the IgA isotype are useful hints that closer trations of antibody directed against the infectious
monitoring of patients with smoldering multiple agent. Therefore, the peak response of one clone
myeloma may be indicated.424 may predominate to such an extent that it has the
It is not always possible to categorize patients as same electrophoretic appearance as that of a mono-
having myeloma or MGUS. There are many clonal gammopathy caused by a neoplastic lympho-
reported cases in the literature in which a patient proliferative process. I have seen many cases of
with a small monoclonal protein was followed for oligoclonal gammopathies and occasionally mono-
several years, sometimes longer than two decades, clonal gammopathies in patients with acquired
before the condition evolved into clear-cut immune deciency syndrome (AIDS). In patients
myeloma. I have seen a case in which a solitary with AIDS, relatively large M-proteins may be seen
plasmacytoma was removed, and 17 years later a along with plasma cell hyperplasia as part of this
monoclonal protein of the same isotype was infection.438 When I see one or more M-proteins
detected in the serum. Therefore, although the lym- superimposed on a polyclonal background, in my
phoproliferative condition in most patients with interpretation, I caution that this may be reecting
MGUS will not evolve, it is important to follow a reactive process. I recommend urine immuno-
these patients every 612 months with a serum and xation and a follow-up of the serum electro-
urine protein electrophoresis (depending on the phoresis to determine if the process regresses or
location of their gammopathy) to determine if the evolves into either a clearly infectious or denable
disease is evolving. When a monoclonal gammo- neoplastic condition. Since infections are occasion-
pathy is detected for the rst time, the patient needs ally a presenting feature in patients with multiple
to have a thorough physical examination, labora- myeloma, one cannot dismiss the potential impor-
tory evaluation for hemoglobin, hematocrit, white tance of a monoclonal gammopathy in patients with
blood cell count and differential, calcium, urine infections. I look for indications that the mono-
study for MFLC (note that the serum assay for clonal gammopathy is probably caused by an
free k and l may supersede this), bone marrow infection (Table 6.10). Despite the electrophoretic
184 Conditions associated with monoclonal gammopathies
a
If the isotype of the M-protein is IgG, the uninvolved isotypes would be IgA and IgM.
b
MFLC, monoclonal free light chain.
associated with the presence of monoclonal in large amounts (> 500 mg/dl) (Table 6.11). In the
immunoglobulins.439 Therefore, in addition to the original data from Brouet et al.,457 Type I cryoglo-
conditions described above, immune dysfunction bulinemia accounts for about 25 per cent of cases
may be the underlying cause of monoclonal and of cryoglobulinemia. However, the recent data
oligoclonal gammopathies. from Trejo et al.461 demonstrate that the extraordi-
nary number of HCV infections have dramatically
changed the percentage of cases that result from
CRYOGLOBULINS infectious diseases in recent years (Table 6.12) such
that in their series only 7 per cent of cases were
Cryoglobulins are immunoglobulins that aggregate associated with hematologic conditions. Those
and precipitate or gel at temperatures lower than numbers more likely reect differences between the
37C. Most are not monoclonal proteins. They current case mix and that of the original Brouet
have clinical importance because, in addition to data.
problems caused by the underlying condition (such Type II cryoglobulins are also associated with
as multiple myeloma, hepatitis C, autoimmune monoclonal proteins, but are different from those
disease, etc.), they can precipitate in blood vessels of Type I. As shown in Table 6.11, these are
with life-threatening consequences. They were rst usually present in smaller amounts than Type I
recognized in association with multiple myeloma cryoglobulin. Type II cryoglobulins are a mixture
in 1933 by Wintrobe and Buel, but the term cryo- of an IgM (rarely IgA and IgG) usually k mono-
globulin was introduced in 1947 by Lerner and clonal protein with rheumatoid factor activity
Watson.446448 Depending on their primary disease (reacts with the Fc portion of IgG).457,462,463 These
association, they have been treated by various proteins are present in lower concentration than
means, including ribaviron and a-interferon, cyto- Type I, and are most often found in patients with
toxic drugs, steroids, plasmapheresis, and even infectious conditions. They have occasionally been
colchicine.449454 Occasionally, the therapies seemed described in association with autoimmune or
to trigger adverse events.455,456 Cryoglobulins have lymphoproliferative diseases. In the past decade,
been classied by Brouet into Types I, II, and III several studies have pointed to an extraordinarily
(Table 6.11).457 When no underlying disease such high incidence of HCV infection in patients with
as multiple myeloma is known to be present, the Type II cryoglobulinemia.459,464474 Indeed, HCV
term essential is used to describe the cryoglobulin. was a major factor in the recent series by Trejo et
However, as we learn more about the etiology of al.461 where they evaluated sera from 7043
cryoglobulins, the use of essential has declined. samples sent to their immunology department
For example, as discussed below, most cases of between 1991 and 1999. Of the 443 patients with
Type II cryoglobulins used to be thought of as a cryocrit of 1 an extraordinary 321 (73 per
essential cryoglobulins, but since the discovery of cent) had HCV (Table 6.12). In contrast, hepatitis
hepatitis C virus (HCV) it has become clear that B infection accounted for only 15 (3 per cent)
HCV is a major cause of both Type II and Type III cases.
cryoglobulins.458,459 Type III cryoglobulins, the type most frequently
Type I cryoglobulins are most often seen in encountered and unrelated to monoclonal pro-
patients with lymphoproliferative diseases, espe- teins, consist of polyclonal rheumatoid factor,
cially multiple myeloma, Waldenstrms usually IgM, that reacts with polyclonal IgG.460
macroglobulinemia, and monoclonal gammopathy Since both Types II and III cryoglobulinemias
of undetermined signicance. IgM is the most fre- contain more than one type of immunoglobulin,
quent isotype encountered in this category they are termed mixed cryoglobulins. Currently,
followed by IgG, then IgA.460 In Type I cryoglobu- the vast majority of these patients are infected
linemia, the monoclonal protein is usually present with HCV, although the exact percentage varies
186 Conditions associated with monoclonal gammopathies
a
Data from Brouet et al.457
a
Data from Trejo et al.461
Cryoglobulins 187
widely from one study to the next. In addition to signicantly increases the likelihood that the
HCV, however, these cases may be associated individual will develop renal disease.485
with other infectious diseases or autoimmune In the clinical laboratory, a tiny amount of cryo-
diseases.475477 Interestingly, however, even the globulin may be detected in normal individuals.
autoimmune diseases may have more than a Levels up to 80 mg/ml may occur in controls.446
passing relationship with the HCV infection.478 When a cryoglobulin is detectable, but present in
Typically, Type III cryoglobulin is present in low amounts < 2 per cent, we cannot characterize
concentration (< 100 mg/dl). Because of the rela- them. I report them as trace of cryoglobulin
tively small amount of cryoglobulin in these present, but too small to quantify. Types II and III
cases, they prove the most difcult to character- (the mixed cryoglobulins) can present with a
ize. Immunoblotting and two-dimensional electro- variety of laboratory ndings, including rheuma-
phoresis studies have demonstrated that rather toid factor activity and often a low level of C1q
than a polyclonal IgM, the cryoglobulin is com- and C4 (although C3 is usually normal).446,478 A
posed of a few clones of (oligoclonal) IgM with denition of clinically symptomatic mixed cryo-
polyclonal IgG.460,463,479 globulinemia by Invernizzi et al.486 provided a
It is not clear why cryoglobulins precipitate. usable standard. Key laboratory features of their
Some have noted structural changes in the vari- denition of mixed cryoglobulins are: the presence
able portions of the immunoglobulin heavy and of a cryocrit > 1 per cent for at least 6 months, C4
light changes, an abundance of hydrophobic less than 8 mg/dl, and a positive rheumatoid
amino acids, unusual glycosylation, a decrease in factor.486 In addition, patients often have increased
galactose at the Fc portion of IgG, or a change erythrocyte sedimentation rates and a normo-
in the CH3 domain glycosylation sites.478,480483 chromic normocytic anemia.446
Unfortunately, there is no consensus on the Cryoglobulins may be missed by electrophoretic
mechanism, and it is likely there are several analysis, especially if they precipitate at relatively
mechanisms depending on the individual proteins high temperatures a feature with considerable
involved. clinical importance.484 If proper precautions are not
Cryoglobulins can have signicant clinical conse- taken in handling the specimen, the cryoglobulin
quences that occur secondary to the obstruction of will precipitate during the clotting process and will
blood vessels and/or to the vasculitis that results be missed by electrophoresis or when the sample is
from the inammatory effects of immune complex placed in the cold. To detect cryoglobulins, the
deposition.460 Prominent signs and symptoms specimen should be drawn in a prewarmed (37C)
include: purpura (virtually always), arthralgias, red top tube or syringe. To minimize transporta-
renal disease (often membranoproliferative tion problems, ideally the patient would have their
glomerulonephritis), peripheral neuropathy, venepuncture in the laboratory. However, because
hepatic involvement, abdominal pain (likely due to of the condition of the patient or the location of the
vessel involvement in the gastrointestinal tract), laboratory, this is often not practical. To maintain
Raynauds phenomenon, and leg ulcers.446 The clin- the temperature during transportation to the labo-
ical consequences of cryoglobulins may depend as ratory, some house ofcers place the sample close
much on the temperature at which they precipitate to the body such as under an armpit or in a pocket.
as on their amount. Letendre and Kyle484 described As a more reasonable alternative when the patient
two patients with relatively small amounts of Type cannot be readily transported to the laboratory, I
I cryoglobulin who had signicant clinical conse- recommend using a device to maintain that
quences because they precipitated in vitro at temperature. A commercial device is now available
temperatures higher than 25C. The presence of for this, facilitating temperature maintenance
cryoglobulinemia in patients with autoimmune during transportation. One inexpensive solution is
disease, such as systemic lupus erythematosus, to use a thermos lled with a material to preserve
188 Conditions associated with monoclonal gammopathies
the temperature at 37C. For years, I recom- a uffy, white occulant substance; however, some
mended using sand in the thermos kept in a 37C are crystalline or gelatinous (the latter are often
incubator. However, the sand can be messy, and a Type III cryoglobulins and may take up to 7 days
clever alternative is to put commercially available to form).488 One may visualize cryoglobulins as
gel-lled plastic bags in the thermos (suggested by small as 15 mg/ml.488 The cryocrit is measured by
Linda Thomas at the University of Michigan placing the warm serum sample into a Wintrobe or
Immunology Laboratory, Ann Arbor, MI, USA) other calibrated tube and incubate it at 4C until
instead of the sand. These gel packs are available as the precipitate forms. Centrifugation at
Gel-Ice (Pioneer Packaging Co., Kent, WA, USA) 10001200 g for 30 min in the cold provides the
and as Polar Pack (Mid-Lands Chemical Co., cryocrit.488 Unfortunately, there is not a world-
DesMoines, IA, USA). This provides excellent wide standard for the measurement of the
thermal stability when compared with 37C water cryoglobulin and even low cryocrits have been
in a styrofoam cup, and has resulted in improved associated with severe disease whereas some high
yield of cryoglobulins (Fig. 6.18).487 cryocrits are clinically asymptomatic.478 An alterna-
After separating the clot, the specimen is split into tive to quantify the cryoglobulin involves washing
two fractions. One is kept at 37C while the other the cryoprecipitate at least six times with ice-cold
is placed at 4C. Samples are examined daily for up saline by repeated centrifugations and vortexing.
to 7 days. When a precipitate is seen in the 4C Then one redissolves the cryoprecipitate in 37C
tube and not in the 37C tube, it is reported as pos- and performs a total protein determination on the
itive for cryoglobulin. The precipitate is most often redissolved precipitate. Unfortunately, consider-
able cryoglobulin may be lost in the washing
procedure. Therefore, I prefer to use the simpler
38 cryocrit.
After establishing the cryocrit, our laboratory
36
characterizes cryoglobulins by immunoxation on
the thoroughly washed cryoglobulin (as above).
One needs to wash thoroughly in order to remove
34
the polyclonal immunoglobulins that are in the
Temperature (C)
1 1
2 2
3 3
4 4
5 5
6 6
7 7
Figure 6.19 Immunoxation of thoroughly washed cryoprecipitate. Note that the serum protein electrophoresis (SPE) lane shows no
band in the albumin position, indicating that the sample has been washed adequately. There is a diffuse broad staining in the IgG and faintly
in the l lanes. There is a dense band near the origin in the IgM and the k lanes. In addition, there is some diffuse staining in the k lane both
anodal and cathodal to the band. This is a Type II cryoglobulin. It contains monoclonal IgM k and polyclonal IgG (and of course the light
chains bound to the polyclonal IgG). If this had been a Type I cryoglobulin, there should be no polyclonal antibody present. The wash step
that removes unbound polyclonal antibodies is critical to making this distinction. (Paragon Immunoxation stained with Paragon Violet.)
When a Type II or III cryoglobulin is present, I rec- In addition to the above patterns, on gel-based
ommend performing serology for HCV, rheuma- electrophoretic techniques, there may be a precipi-
toid factor and C4. tate at the origin because of the cooler tempera-
Serum protein electrophoresis in patients with ture of the gel allowing precipitation of some
Type I cryoglobulin usually shows the M-protein. cryoglobulins (Fig. 6.20). Finding such a precipi-
With Type II cryoglobulins, an M-protein may or tate should prompt one to investigate the patient
may not be seen by electrophoresis. It may be for cryoglobulin. Since the protein precipitates
present in an amount so small that an immunox- without any specic antisera being added, it may
ation will be needed to identify it. Type III also produce confusing patterns on immunoxa-
cryoglobulins will usually have a polyclonal tion of the serum. Therefore, when an immunox-
increase, and sometimes a polyclonal and oligo- ation of a serum sample shows an origin
clonal increase in the g-region. However, serum precipitate in several lanes, I repeat the
protein electrophoresis is not an appropriate immunoxation, replacing one of the antisera
screening test for cryoglobulins. Normal patterns with buffer or saline. When the precipitate recurs
and even hypogammaglobulinemia have been in this location, one is certain to be dealing with a
reported in patients with cryoglobulinemia.446,491493 spontaneously precipitating protein, usually a
190 Conditions associated with monoclonal gammopathies
Fibrinogen
One of our most common problems in dealing with
serum protein electrophoresis is the presence of the
Figure 6.20 Serum protein electrophoresis of four samples. The brinogen band. This should not be present in a
top and bottom samples contain obvious M-proteins in the g- properly clotted specimen. However, it may be
region and near the origin, respectively. The third lane contains a present owing to a variety of circumstances: a
large transferrin band that deserves an immunoxation to be blood sample may not have been allowed to clot
certain it is not a monoclonal gammopathy. The second lane
for a sufcient period of time; the patient may be
contains an origin artifact (arrow) that may indicate a cryoglobulin.
Just cathodal to this origin artifact is a lightly staining, but distinct on an anticoagulant which prevents complete clot-
band likely representing a monoclonal cryoglobulin. (Paragon SPE2 ting; the sample may have been collected in a tube
system stained with Paragon Violet.) containing an anticoagulant and then the brino-
gen band appears in the bg region of virtually all
gel-based electrophoretic systems (Fig. 6.22). With
the earlier Paragon CZE 2000 system 1.5
cryoglobulin. At that point, a fresh specimen (Beckman Coulter), a brinogen band was not
drawn and transported as recommended above visible. In the more recent version of this system
will provide the best results to characterize the 1.6 and in the Sebia Capillarys system, brinogen
cryoglobulin. also occurs at the bg region. There is no way to
Cryobrinogens may be confused with cryoglob- absolutely rule out a monoclonal gammopathy
ulins when the sample is drawn in a tube with an when detecting a band in that location by just
anticoagulant rather than the recommended red examining the serum protein electrophoresis. If the
top (no anticoagulant present) tube. They result tube containing the sample is examined, and a clot
from a precipitation of brinogen with brin in the is present at the bottom, this suggests that some b-
cold.488 Detection of cryobrinogen requires rinogen was present after the clot was removed.
drawing a sample of blood into a prewarmed cit- One may repeat the electrophoresis on the next
rated tube and allowing it to sit overnight at 4C. run, if the band is gone (because the brinogen has
One should not use ethylenediaminetetraacetic now clotted), it is due to brinogen. Care is needed
acid (EDTA) because this may inhibit cryobrino- here because, if the band is due to a cryoprecipitat-
gen formation, neither should one use heparin ing monoclonal gammopathy, this too would form
because it may enhance cryoprecipitation of a precipitate in the bottom of the tube (however,
bronectin.488 A washed cryobrinogen immuno- the precipitate from cryoglobulins usually looks
xation may be used to characterize the nature of different from the typical brin clot). When I
the precipitate with anti-brinogen antibody suspect a brinogen band is present, I perform a
(Figure 6.21). Penta immunoxation (see Chapter 3). An alterna-
Bands mistaken for monoclonal gammopathies 191
Lambda
Kappa
Fibrinogen
IgA
(a)
HEMOLYSIS
A hemoglobinhaptoglobin band usually migrates
in the a2- to b1- region. It can resemble a monoclonal
gammopathy. Usually, there is marked depletion of
the normal haptoglobin band and the serum is red.
If the serum is not red, or if the haptoglobin band
is not depleted, perform an immunoxation to rule
out a monoclonal gammopathy.
disease, a cryoglobulin that inhibited brin poly- cell differentiation: monoclonal antibodies and
merization, and disseminated intravascular cluster designation (CD)-dened hematopoietic
coagulation.500503 Unusual physiological manifesta- cell antigens. In: Keren DF, McCoy JJP, Carey
tions can occur, for example the monoclonal JL, eds. Flow cytometry in clinical diagnosis.
protein that bound to insulin and produced recur- Chicago: ASCP Press, 2001.
rent severe hypoglycemia,504,505 or the binding of an 8. Ciudad J, Orfao A, Vidriales B, et al.
IgM monoclonal gammopathy to calcium resulting Immunophenotypic analysis of CD19+
in a hypercalcemia while the ionized fraction of precursors in normal human adult bone marrow:
calcium was normal.506 Interaction between an ery- implications for minimal residual disease
throcyte lysing reagent and an IgM monoclonal detection. Haematologica 1998;83:10691075.
gammopathy resulted in inaccurate hemoglobin 9. Korsmeyer SJ, Greene WC, Cossman J, et al.
measurements in one patient with Waldenstrms Rearrangement and expression of
macroglobulinemia.507 Because of their unique immunoglobulin genes and expression of Tac
properties, cryoglobulins may be mistaken by elec- antigen in hairy cell leukemia. Proc Natl Acad
tronic cell counters when the size of the precipitates Sci USA 1983;80:45224526.
fall within the ranges the instrument interprets as 10. Caligaris-Cappio F, Janossy G. Surface markers
cellular components, resulting in elevated leuko- in chronic lymphoid leukemias of B cell type.
cyte or platelet counts.508512 Direct analysis of such Semin Hematol 1985;22:112.
blood smears will demonstrate suspicious aggre- 11. Winkelstein A, Jordan, PS. Immune deciencies
gates of amorphous particles.513 in chronic lymphocytic leukemia and multiple
myeloma. In: Ballow M, ed. Intravenous
immunoglobulin therapy today. Totowa: The
REFERENCES Humana Press Inc, 1992.
12. Hayward AR. Development of lymphocyte
1. Pick AI, Shoenfeld Y, Frohlichmann R, Weiss H, responses and interactions in the human fetus
Vana D, Schreibman S. Plasma cell dyscrasia. and newborn. Immunol Rev 1981;57:3960.
Analysis of 423 patients. JAMA 1979;241: 13. Hofman FM, Danilovs J, Husmann L, Taylor
22752278. CR. Ontogeny of B cell markers in the human
2. Katzin WE. Cancer (multiple myeloma and fetal liver. J Immunol 1984;133:11971201.
related disorders). Anal Chem 1993;65: 14. Jensen GS, Mant MJ, Pilarski LM. Sequential
382R387R. maturation stages of monoclonal B lineage cells
3. Keren DF, Alexanian R, Goeken JA, Gorevic PD, from blood, spleen, lymph node, and bone
Kyle RA, Tomar RH. Guidelines for clinical and marrow from a terminal myeloma patient. Am J
laboratory evaluation of patients with Hematol 1992;41:199208.
monoclonal gammopathies. Arch Pathol Lab 15. Rudin CM, Thompson CB. B-cell development
Med 1999;123:106107. and maturation. Semin Oncol 1998;25:435446.
4. Bruton OC. Agammaglobulinemia. Pediatrics 16. Grogan TM, van Camp B, Kyle RA. Plasma cell
1952;9:722727. neoplasms. In: Jaffe ES, Harris NL, Stein H,
5. Schiff R, Schiff SE. Chapter 11. Flow cytometry Vardiman JW, ed. Pathology and genetics of
for primary immunodeciency diseases. In: Keren tumours of haematopoietic and lymphoid tissues.
DF, McCoy JJP, Carey JL, eds. Flow cytometry Lyon: IARC Press, 2001.
in clinical diagnosis. Chicago: ASCP Press, 2001. 17. Sezer O, Heider U, Zavrski I, Possinger K.
6. Glick B, Chang TS, Jaap RG. The bursa of Differentiation of monoclonal gammopathy of
Fabricius and antibody production. Poult Sci undetermined signicance and multiple myeloma
1956;35:224232. using ow cytometric characteristics of plasma
7. Pirruccello SJ, Aoun P. Chapter 3. Hematopoietic cells. Haematologica 2001;86:837843.
194 Conditions associated with monoclonal gammopathies
18. van Riet I, De Waele M, Remels L, Lacor P, statistics review, 19731991: tables and graphs,
Schots R, van Camp B. Expression of National Cancer Institute. Bethesda: NIH,
cytoadhesion molecules (CD56, CD54, CD18 1994.
and CD29) by myeloma plasma cells. Br J 30. Parker SL, Davis KJ, Wingo PA, Ries LA, Heath
Haematol 1991;79:421427. CW Jr. Cancer statistics by race and ethnicity.
19. van Camp B, Durie BG, Spier C, et al. Plasma CA Cancer J Clin 1998;48:3148.
cells in multiple myeloma express a natural killer 31. Aguzzi F, Bergami MR, Gasparro C, Bellotti V,
cell-associated antigen: CD56 (NKH-1; Leu-19). Merlini G. Occurrence of monoclonal
Blood 1990;76:377382. components in general practice: clinical
20. Kyle RA, Therneau TM, Rajkumar SV, et al. A implications. Eur J Haematol 1992;48:
long-term study of prognosis in monoclonal 192195.
gammopathy of undetermined signicance. 32. Alexanian R, Weber D, Liu F. Differential
N Engl J Med 2002;346:564569. diagnosis of monoclonal gammopathies. Arch
21. Kyle RA. Benign monoclonal gammopathy Pathol Lab Med 1999;123:108113.
after 2035 years of follow-up. Mayo Clin Proc 33. Kyle RA. Multiple myeloma. Diagnostic
1993;68:2636. challenges and standard therapy. Semin Hematol
22. Kyle RA, Rajkumar SV. Monoclonal 2001;38:1114.
gammopathies of undetermined signicance. 34. Kyle RA. Sequence of testing for monoclonal
Hematol Oncol Clin North Am gammopathies. Arch Pathol Lab Med
1999;13:11811202. 1999;123:114118.
23. Clark KT. Monoclonal gammopathy of 35. Epstein J, Barlogie B, Katzmann J, Alexanian R.
undetermined signicance. J Insur Med 1997;29: Phenotypic heterogeneity in aneuploid multiple
136138. myeloma indicates pre-B cell involvement. Blood
24. Keren DF, Morrison N, Gulbranson R. Evolution 1988;71:861865.
of a monoclonal gammopathy (MG) 36. Barlogie B. Pathophysiology of human multiple
Documented by high-resolution electrophoresis myeloma recent advances and future directions.
(HRE) and immunoxation (IFE). Lab Med Curr Top Microbiol Immunol 1992;182:
1994; 25:313317. 245250.
25. Axelsson U, Bachmann R, Hallen J. Frequency of 37. Taylor BJ, Pittman JA, Seeberger K, et al.
pathological proteins (M-components) in 6,995 Intraclonal homogeneity of clonotypic
sera from an adult population. Acta Med Scand. immunoglobulin M and diversity of nonclinical
1966;179:235247. post-switch isotypes in multiple myeloma:
26. Malik AA, Ganti AK, Potti A, Levitt R, Hanley insights into the evolution of the myeloma clone.
JF. Role of Helicobacter pylori infection in the Clin Cancer Res 2002;8:502513.
incidence and clinical course of monoclonal 38. Dimopoulos MA, Papadimitriou C, Sakarellou
gammopathy of undetermined signicance. Am J N, Athanassiades P. Complications and
Gastroenterol 2002;97:13711374. supportive therapy of multiple myeloma.
27. Kyle RA. Diagnostic criteria of multiple Baillieres Clin Haematol 1995;8:845852.
myeloma. Hematol Oncol Clin N Am 1992;6: 39. Vaswani SK, Sprague R. Pseudohyponatremia in
347358. multiple myeloma. South Med J 1993;86:
28. Cohen HJ, Crawford J, Rao MK, Pieper CF, 251252.
Currie MS. Racial differences in the prevalence 40. Avashia JH, Walsh TD, Valenzuela R, Fernando
of monoclonal gammopathy in a community- Quevedo J, Clough J. Pseudohypoproteinemia
based sample of the elderly. Am J Med and multiple myeloma. Cleve Clin J Med
1998;104:439444. 1990;57:298300.
29. Ries LAG, Miller BA, Hankey BF. SEER cancer 41. van Dijk JM, Sonnenblick M, Weissberg N,
References 195
dexamethasone for newly diagnosed myeloma. and nutrition as risk factors for multiple
J Clin Oncol 2002;20:43194323. myeloma among blacks and whites in the United
66. Durie BG, Salmon SE. A clinical staging system States. Cancer Causes Control 2001;12:117125.
for multiple myeloma. Correlation of measured 79. Hayakawa N, Ohtaki M, Ueoka H, Matsuura
myeloma cell mass with presenting clinical M, Munaka M, Kurihara M. Mortality statistics
features, response to treatment, and survival. of major causes of death among atomic bomb
Cancer 1975;36:842854. survivors in Hiroshima Prefecture from 1968 to
67. Alexanian R, Balcerzak S, Bonnet JD, et al. 1982. Hiroshima J Med Sci 1989;38:5367.
Prognostic factors in multiple myeloma. Cancer. 80. Shimizu Y, Kato H, Schull WJ. Studies of the
1975;36:11921201. mortality of A-bomb survivors. 9. Mortality,
68. Cherng NC, Asal NR, Kuebler JP, Lee ET, 19501985: Part 2. Cancer mortality based on
Solanki D. Prognostic factors in multiple the recently revised doses (DS86). Radiat Res
myeloma. Cancer 1991;67:31503156. 1990;121:120141.
69. Chen YH, Magalhaes MC. Hypoalbuminemia in 81. Matanoski GM, Seltser R, Sartwell PE, Diamond
patients with multiple myeloma. Arch Intern EL, Elliott EA. The current mortality rates of
Med 1990;150:605610. radiologists and other physician specialists:
70. Kaneko M, Kanda Y, Oshima K, et al. Simple specic causes of death. Am J Epidemiol
prognostic model for patients with multiple 1975;101:199210.
myeloma: a single-center study in Japan. Ann 82. Stebbings JH, Lucas HF, Stehney AF. Mortality
Hematol 2002;81:3336. from cancers of major sites in female radium dial
71. Sahara N, Takeshita A, Shigeno K, et al. workers. Am J Ind Med 1984;5:435459.
Clinicopathological and prognostic 83. Stebbings JH. Health risks from radium in
characteristics of CD56-negative multiple workplaces: an unnished story. Occup Med
myeloma. Br J Haematol 2002;117:882885. 2001;16:259270.
72. Smadja NV, Bastard C, Brigaudeau C, Leroux D, 84. Jablon S, Hrubec Z, Boice JD. Cancer in
Fruchart C. Hypodiploidy is a major prognostic populations living near nuclear facilities.
factor in multiple myeloma. Blood 2001;98: Washington, DC: DHHSUS Government Print
22292238. Ofce, 1990.
73. Greipp PR, Trendle MC, Leong T, et al. Is ow 85. Hatcher JL, Baris D, Olshan AF, et al. Diagnostic
cytometric DNA content hypodiploidy radiation and the risk of multiple myeloma
prognostic in multiple myeloma? Leuk (United States). Cancer Causes Control
Lymphoma 1999;35:8389. 2001;12:755761.
74. Kyle RA, Beard CM, OFallon WM, Kurland LT. 86. Brown LM, Linet MS, Greenberg RS, et al.
Incidence of multiple myeloma in Olmsted Multiple myeloma and family history of cancer
County, Minnesota: 1978 through 1990, with a among blacks and whites in the US. Cancer
review of the trend since 1945. J Clin Oncol 1999;85:23852390.
1994;12:15771583. 87. Alexander LL, Benninghoff DL. Familial
75. Mundy GR. Myeloma bone disease. Eur J multiple myeloma. J Natl Med Assoc 1965;57:
Cancer 1998;34:246251. 471475.
76. American Cancer Society. Multiple Myeloma 88. Alexander LL, Benninghoff DL. Familial multiple
Cancer Resource Center; 2002. myeloma. II. Final pathological ndings in two
77. Baris D, Brown LM, Silverman DT, et al. brothers and a sister. J Natl Med Assoc 1967;59:
Socioeconomic status and multiple myeloma 278281.
among US blacks and whites. Am J Public Health 89. Robbins R. Familial multiple myeloma: the tenth
2000;90:12771281. reported occurrence. Am J Med Sci 1967;254:
78. Brown LM, Gridley G, Pottern LM, et al. Diet 848850.
References 197
90. Barbieri D, Grampa A. Familial multiple study of 24 patients with t(11;14)(q13;q32) or its
myeloma: beta myeloma in two sisters. variant. Cancer Genet Cytogenet. 1998;104:
Haematologica 1972;58:565566. 133138.
91. Wiedermann D, Urban P, Wiedermann B, Cidl K. 103. Fenton JA, Pratt G, Rawstron AC, Morgan GJ.
Multiple myeloma in two brothers. Isotype class switching and the pathogenesis of
Immunoglobulin levels among their relatives. multiple myeloma. Hematol Oncol 2002;20:
Neoplasma 1976;23:197207. 7585.
92. Law MI. Familial occurrence of multiple 104. Lynch HT, Sanger WG, Pirruccello S, Quinn-
myeloma. South Med J 1976;69:4648. Laquer B, Weisenburger DD. Familial multiple
93. Shoenfeld Y, Berliner S, Shaklai M, Gallant LA, myeloma: a family study and review of the
Pinkhas J. Familial multiple myeloma. A review literature. J Natl Cancer Inst 2001;93:
of thirty-seven families. Postgrad Med J 14791483.
1982;58:1216. 105. Abdalla IA, Tabbara IA. Nonsecretory multiple
94. Grosbois B, Gueguen M, Fauchet R, et al. myeloma. South Med J 2002;95:761764.
Multiple myeloma in two brothers. An 106. Pratt G. Molecular aspects of multiple myeloma.
immunochemical and immunogenetic familial Mol Pathol 2002;55:273283.
study. Cancer 1986;58:24172421. 107. Bast EJ, van Camp B, Reynaert P, Wiringa G,
95. Eriksson M, Hallberg B. Familial occurrence of Ballieux RE. Idiotypic peripheral blood
hematologic malignancies and other diseases in lymphocytes in monoclonal gammopathy. Clin
multiple myeloma: a case-control study. Cancer Exp Immunol 1982;47:677682.
Causes Control 1992;3:6367. 108. van Acker A, Conte F, Hulin N, Urbain J.
96. Crozes-Bony P, Palazzo E, Meyer O, De Bandt Idiotypic studies on myeloma B cells. Eur J
M, Kahn MF. Familial multiple myeloma. Report Cancer 1979;15:627635.
of a case in a father and daughter. Review of the 109. Kubagawa H, Vogler LB, Capra JD, Conrad ME,
literature. Rev Rheum Engl Ed 1995;62: Lawton AR, Cooper MD. Studies on the clonal
439445. origin of multiple myeloma. Use of individually
97. Roddie PH, Dang R, Parker AC. Multiple specic (idiotype) antibodies to trace the
myeloma in three siblings. Clin Lab Haematol oncogenic event to its earliest point of expression
1998;20:191193. in B-cell differentiation. J Exp Med 1979;150:
98. Grosbois B, Jego P, Attal M, et al. Familial 792807.
multiple myeloma: report of fteen families. Br J 110. Rasmussen T, Kastrup J, Knudsen LM, Johnsen
Haematol 1999;105:768770. HE. High numbers of clonal CD19+ cells in the
99. Deshpande HA, Hu XP, Marino P, Jan NA, peripheral blood of a patient with multiple
Wiernik PH. Anticipation in familial plasma cell myeloma. Br J Haematol 1999;105:265267.
dyscrasias. Br J Haematol 1998;103:696703. 111. Helfrich MH, Livingston E, Franklin IM, Soutar
100. Bowden M, Crawford J, Cohen HJ, Noyama O. RL. Expression of adhesion molecules in
A comparative study of monoclonal malignant plasma cells in multiple myeloma:
gammopathies and immunoglobulin levels in comparison with normal plasma cells and
Japanese and United States elderly. J Am Geriatr functional signicance. Blood Rev 1997;11:
Soc 1993;41:1114. 2838.
101. Zandecki M, Lai JL, Facon T. Multiple 112. Pajor A, Kelemen E, Mohos Z, Hambach J,
myeloma: almost all patients are cytogenetically Varadi G. Multiple myeloma in pregnancy. Int J
abnormal. Br J Haematol 1996;94:217227. Gynaecol Obstet 1991;35:341342.
102. Lai JL, Michaux L, Dastugue N, Vasseur F, 113. Cook G, Campbell JD, Carr CE, Boyd KS,
Daudignon A, Facon T, Bauters F, Zandecki M. Franklin IM. Transforming growth factor beta
Cytogenetics in multiple myeloma: a multicenter from multiple myeloma cells inhibits
198 Conditions associated with monoclonal gammopathies
135. Kyle RA. Multiple myeloma: review of 869 cases. 146. Vaerman JP, Langendries A, Vander Maelen C.
Mayo Clin Proc 1975;50:2940. Homogenous IgA monomers, dimers, trimers and
136. Jako JM, Gesztesi T, Kaszas I. IgE lambda tetramers from the same IgA myeloma serum.
monoclonal gammopathy and amyloidosis. Int Immunol Invest 1995;24:631641.
Arch Allergy Immunol 1997;112:415421. 147. Kimizu K, Hamada A, Haba T, et al. A case of
137. Qiu WX, Huang NH, Wu GH. Immunological IgA-kappa multiple myeloma with hyperviscosity
classication of 31 multiple myeloma patients. syndrome terminating in plasma cell leukemia.
Proc Chin Acad Med Sci Peking Union Med Coll Rinsho Ketsueki. 1983;24:572579.
1990;5:7983. 148. Tuddenham EG, Whittaker JA, Bradley J,
138. Ameis A, Ko HS, Pruzanski W. M components Lilleyman JS, James DR. Hyperviscosity
a review of 1242 cases. Can Med Assoc J 1976; syndrome in IgA multiple myeloma. Br J
114:889892, 895. Haematol 1974;27:6576.
139. Schur PH, Kyle RA, Bloch KJ, et al. IgG 149. Whittaker JA, Tuddenham EG, Bradley J.
subclasses: relationship to clinical aspects of Hyperviscosity syndrome in IgA multiple
multiple myeloma and frequency distribution myeloma. Lancet 1973;2:572.
among M-components. Scand J Haematol 150. De Gramont A, Grosbois B, Michaux JL, et al.
1974;12:6068. IgM myeloma: 6 cases and a review of the
140. Hammarstrom L, Mellstedt H, Persson MA, literature. Rev Med Interne 1990;11:1318.
Smith CI, Ahre A. IgA subclass distribution in 151. Takahashi K, Yamamura F, Motoyama H. IgM
paraproteinemias: suggestion of an IgG-IgA myeloma its distinction from Waldenstroms
subclass switch pattern. Acta Pathol Microbiol macroglobulinemia. Acta Pathol Jpn 1986;36:
Immunol Scand (C) 1984;92:207211. 15531563.
141. Vincent C, Bouic P, Revillard JP, Bataille R. 152. Zarrabi MH, Stark RS, Kane P, Dannaher CL,
Complexes of alpha 1-microglobulin and Chandor S. IgM myeloma, a distinct entity in the
monomeric IgA in multiple myeloma and spectrum of B-cell neoplasia. Am J Clin Pathol
normal human sera. Mol Immunol 1985;22: 1981;75:110.
663673. 153. Dierlamm T, Laack E, Dierlamm J, Fiedler W,
142. Pruzanski W, Watt JG. Serum viscosity and Hossfeld DK. IgM myeloma: a report of four
hyperviscosity syndrome in IgG multiple cases. Ann Hematol 2002;81:136139.
myeloma. Report on 10 patients and a review of 154. Haghighi B, Yanagihara R, Cornbleet PJ. IgM
the literature. Ann Intern Med 1972;77: myeloma: case report with immunophenotypic
853860. prole. Am J Hematol 1998;59:302308.
143. Pruzanski W, Russell ML. Serum viscosity and 155. Avet-Loiseau H, Garand R, Lode L, Harousseau
hyperviscosity syndrome in IGG multiple JL, Bataille R. Translocation t(11;14)(q13;q32) is
myeloma the relationship of Sia test and to the hallmark of IgM, IgE, and nonsecretory
concentration of M component. Am J Med Sci multiple myeloma variants. Blood 2003;101:
1976;271:145150. 15701571.
144. Seward CW, Osterland CK. Hyperviscosity in 156. Blade J, Lust JA, Kyle RA. Immunoglobulin D
IgG myeloma: detection and treatment in multiple myeloma: presenting features, response
community medical facilities. Tex Med 1974;70: to therapy, and survival in a series of 53 cases.
6365. J Clin Oncol 1994;12:23982404.
145. Chandy KG, Stockley RA, Leonard RC, 157. Jancelewicz Z, Takatsuki K, Sugai S, Pruzanski
Crockson RA, Burnett D, MacLennan IC. W. IgD multiple myeloma. Review of 133 cases.
Relationship between serum viscosity and Arch Intern Med 1975;135:8793.
intravascular IgA polymer concentration in IgA 158. Kyle RA. IgD multiple myeloma: a cure at 21
myeloma. Clin Exp Immunol 1981;46:653661. years. Am J Hematol 1988;29:4143.
200 Conditions associated with monoclonal gammopathies
159. Blade J, Kyle RA. IgD monoclonal gammopathy 173. Bradwell AR, Carr-Smith HD, Mead GP, et al.
with long-term follow-up. Br J Haematol Highly sensitive, automated immunoassay for
1994;88:395396. immunoglobulin free light chains in serum and
160. Jennette JC, Wilkman AS, Benson JD. IgD urine. Clin Chem 2001;47:673680.
myeloma with intracytoplasmic crystalline 174. Okada K, Oguchi N, Shinohara K, et al. BUN,
inclusions. Am J Clin Pathol 1981;75:231235. Bence Jones protein, and chromosomal
161. Schaldenbrand JD, Keren DF. IgD amyloid in aberrations predict survival in multiple myeloma.
IgD-lambda monoclonal conjunctival Rinsho Ketsueki 1997;38:12541262.
amyloidosis. A case report. Arch Pathol Lab Med 175. Sirohi B, Powles R, Mehta J, et al. Complete
1983;107:626628. remission rate and outcome after intensive
162. Patel K, Carrington PA, Bhatnagar S, Houghton treatment of 177 patients under 75 years of age
JB, Routledge RC. IgD myeloma with multiple with IgG myeloma dening a circumscribed
cutaneous plasmacytomas. Clin Lab Haematol disease entity with a new staging system. Br J
1998;20:5355. Haematol 1999;107:656666.
163. Vladutiu AO. Immunoglobulin D: properties, 176. Weber DM, Dimopoulos MA, Moulopoulos LA,
measurement, and clinical relevance. Clin Diagn Delasalle KB, Smith T, Alexanian R. Prognostic
Lab Immunol 2000;7:131140. features of asymptomatic multiple myeloma. Br J
164. Levan-Petit I, Cardonna J, Garcia M, et al. Haematol 1997;97:810814.
Sensitive ELISA for human immunoglobulin D 177. Baldini L, Guffanti A, Cesana BM, et al. Role of
measurement in neonate, infant, and adult sera. different hematologic variables in dening the
Clin Chem 2000;46:876878. risk of malignant transformation in monoclonal
165. Macro M, Andre I, Comby E, et al. IgE gammopathy. Blood 1996;87:912918.
multiple myeloma. Leuk Lymphoma 1999;32: 178. Peest D, Coldewey R, Deicher H, et al.
597603. Prognostic value of clinical, laboratory, and
166. Rogers JS 2nd, Spahr J, Judge DM, Varano LA, histological characteristics in multiple myeloma:
Eyster ME. IgE myeloma with osteoblastic improved denition of risk groups. Eur J Cancer
lesions. Blood 1977;49:295299. 1993;29A:978983.
167. Proctor SJ, Chawla SL, Bird AG, Stephenson J. 179. Dimopoulos MA, Moulopoulos A, Smith T,
Hyperviscosity syndrome in IgE myeloma. Br Delasalle KB, Alexanian R. Risk of disease
Med J (Clin Res Edn) 1984;289:1112. progression in asymptomatic multiple myeloma.
168. West N. Hyperviscosity syndrome in IgE Am J Med 1993;94:5761.
myeloma. Br Med J (Clin Res Edn) 1984;289: 180. Nagai M, Kitahara T, Minato K, et al.
1539. Prognostic factors and therapeutic results in
169. Alexander RL Jr, Roodman ST, Petruska PJ, Tsai multiple myeloma. Jpn J Clin Oncol 1985;15:
CC, Janney CG. A new case of IgE myeloma. 505515.
Clin Chem 1992;38:23282332. 181. Knudsen LM, Hjorth M, Hippe E. Renal failure
170. Rowan RM. Multiple myeloma: some recent in multiple myeloma: reversibility and impact on
developments. Clin Lab Haematol 1982;4: the prognosis Nordic Myeloma Study Group.
211230. Eur J Haematol 2000;65:175181.
171. Dul JL, Aviel S, Melnick J, Argon Y. Ig light 182. Abraham RS, Charlesworth MC, Owen BA,
chains are secreted predominantly as monomers. Benson LM, Katzmann JA, Reeder CB, Kyle RA.
J Immunol 1996;157:29692975. Trimolecular complexes of lambda light chain
172. Shapiro AL, Scharff MD, Maizel JV, Uhr JW. dimers in serum of a patient with multiple
Synthesis of excess light chains of gamma myeloma. Clin Chem 2002;48:18051811.
globulin by rabbit lymph node cells. Nature 183. Inoue N, Togawa A, Yawata Y. Tetrameric
1966;211:243245. Bence Jones protein case report and review of
References 201
the literature. Nippon Ketsueki Gakkai Zasshi Seligmann M. Intracytoplasmic and surface-
1984;47:14561459. bound immunoglobulins in nonsecretory and
184. Caggiano V, Dominguez C, Opfell RW, Kochwa Bence-Jones myeloma. Clin Exp Immunol
S, Wasserman LR. IgG myeloma with closed 1976;25:428436.
tetrameric Bence Jones proteinemia. Am J Med 197. Alanen A, Pira U, Lassila O, Roth J, Franklin
1969;47:978985. RM. Mott cells are plasma cells defective in
185. Hom BL. Polymeric (presumed tetrameric) immunoglobulin secretion. Eur J Immunol
lambda Bence Jones proteinemia without 1985;15:235242.
proteinuria in a patient with multiple myeloma. 198. Raubenheimer EJ, Dauth J, Senekal JC. Non-
Am J Clin Pathol 1984;82:627629. secretory IgA kappa myeloma with distended
186. Inoue S, Nagata H, Yozawa H, Terai T, endoplasmic reticulum: a case report.
Hasegawa H, Murao M. A case of IgG myeloma Histopathology 1991;19:380382.
with tetrameric Bence Jones proteinemia and 199. Doster DR, Folds J, Gabriel DA. Nonsecretory
abnormal brin polymerization (authors transl). multiple myeloma. Arch Pathol Lab Med
Rinsho Ketsueki 1980;21:200207. 1988;112:147150.
187. Duffy TP. The many pitfalls in the diagnosis of 200. Cogne M, Guglielmi P. Exon skipping without
myeloma. N Engl J Med 1992;326:394396. splice site mutation accounting for abnormal
188. Katzmann JA, Clark RJ, Abraham RS, et al. immunoglobulin chains in nonsecretory human
Serum reference intervals and diagnostic ranges myeloma. Eur J Immunol 1993;23:12891293.
for free kappa and free lambda immunoglobulin 201. Derk CT, Sandor N. Nonsecretory multiple
light chains: relative sensitivity for detection of myeloma masquerading as a new osteoporotic
monoclonal light chains. Clin Chem 2002;48: vertebral compression fracture in an older
14371444. female. J Am Geriatr Soc 2002;50:17471748.
189. Marien G, Oris E, Bradwell AR, Blanckaert N, 202. Kubota K, Kurabayashi H, Kawada E, Okamoto
Bossuyt X. Detection of monoclonal proteins in K, Tamura J, Shirakura T. Nonsecretion of
sera by capillary zone electrophoresis and free myeloma protein in spite of an increase in tumor
light chain measurements. Clin Chem 2002;48: burden by chemotherapy. Ann Hematol 1991;63:
16001601. 232233.
190. Kyle RA. The monoclonal gammopathies. Clin 203. Shustik C, Michel R, Karsh J. Nonsecretory
Chem 1994;40:21542161. myeloma: a study on hypoimmunoglobulinemia.
191. Bourantas K. Nonsecretory multiple myeloma. Acta Haematol 1988;80:153158.
Eur J Haematol 1996;56:109111. 204. Matsuzaki H, Yoshida M, Akahoshi Y,
192. Cavo M, Galieni P, Gobbi M, et al. Nonsecretory Kuwahara K, Satou T, Takatsuki K. Pseudo-
multiple myeloma. Presenting ndings, clinical nonsecretory multiple myeloma with light chain
course and prognosis. Acta Haematol deposition disease. Acta Haematol 1991;85:
1985;74:2730. 164168.
193. Dreicer R, Alexanian R. Nonsecretory multiple 205. Drayson M, Tang LX, Drew R, Mead GP, Carr-
myeloma. Am J Hematol 1982;13:313318. Smith H, Bradwell AR. Serum free light-chain
194. Rubio-Felix D, Giralt M, Giraldo MP, et al. measurements for identifying and monitoring
Nonsecretory multiple myeloma. Cancer. patients with nonsecretory multiple myeloma.
1987;59:18471852. Blood 2001;97:29002902.
195. Kawada E, Shinonome S, Saitoh T, et al. Primary 206. Kyle RA, Robinson RA, Katzmann JA. The
nonsecretory plasma cell leukemia: a rare variant clinical aspects of biclonal gammopathies. Review
of multiple myeloma. Ann Hematol 1999;78: of 57 cases. Am J Med 1981;71:9991008.
2527. 207. Tasaka T, Matsuhashi Y, Uehara E, Tamura T,
196. PreudHomme JL, Hurez D, Danon F, Brouet JC, Kuwajima M, Nagai M. Angioimmunoblastic
202 Conditions associated with monoclonal gammopathies
228. Guardia J, Rubies-Prat J, Gallart MT, et al. The C, Preudhomme JL, Rambaud JC.
evolution of alpha heavy chain disease. Am J Immunochemical studies in four cases of alpha
Med 1976;60:596602. chain disease. J Clin Invest 1969;48:23742389.
229. Martin IG, Aldoori MI. Immunoproliferative 241. Al-Saleem TI, Qadiry WA, Issa FS, King J. The
small intestinal disease. Mediterranean immunoselection technic in laboratory diagnosis
lymphoma and alpha heavy chain disease. Br J of alpha heavy-chain disease. Am J Clin Pathol
Surg 1994;81:2024. 1979;72:132133.
230. Novis BH, King HS, Gilinsky NH, Mee AS, 242. Fermand JP, Brouet JC, Danon F, Seligmann M.
Young G. Long survival in a patient with alpha- Gamma heavy chain disease: heterogeneity of
chain disease. Cancer 1984;53:970973. the clinicopathologic features. Report of 16 cases
231. Stoop JW, Ballieux RE, Hijmans W, Zegers BJ. and review of the literature. Medicine
Alpha-chain disease with involvement of the (Baltimore) 1989;68:321335.
respiratory tract in a Dutch child. Clin Exp 243. Franklin EC, Lowenstein J, Bigelow B. Heavy
Immunol 1971;9:625635. chain disease a new disorder of serum gamma-
232. Tracy RP, Kyle RA, Leitch JM. Alpha heavy- globulins: report of the rst case. Am J Med
chain disease presenting as goiter. Am J Clin 1964;37:332350.
Pathol 1984;82:336339. 244. Bloch KJ, Lee L, Mills JA, Haber E. Gamma
233. Florin-Christensen A, Doniach D, Newcomb PB. heavy chain disease an expanding clinical and
Alpha-chain disease with pulmonary laboratory spectrum. Am J Med 1973;55:6170.
manifestations. Br Med J 1974;2:413415. 245. Ellman LL, Bloch KJ. Heavy-chain disease.
234. Fine KD, Stone MJ. Alpha-heavy chain disease, Report of a seventh case. N Engl J Med
Mediterranean lymphoma, and 1968;278:11951201.
immunoproliferative small intestinal disease: a 246. Franklin EC, Kyle R, Seligmann M, Frangione B.
review of clinicopathological features, Correlation of protein structure and
pathogenesis, and differential diagnosis. Am J immunoglobulin gene organization in the light of
Gastroenterol 1999;94:11391152. two new deleted heavy chain disease proteins.
235. Lavergne A, Brocheriou I, Rambaud JC, et al. T- Mol Immunol 1979;16:919921.
cell rich alpha-chain disease mimicking T-cell 247. Alexander A, Steinmetz M, Barritault D, et al.
lymphoma. Histopathology 1997;30:394396. gamma Heavy chain disease in man: cDNA
236. Isaacson PG, Dogan A, Price SK, Spencer J. sequence supports partial gene deletion model.
Immunoproliferative small-intestinal disease. An Proc Natl Acad Sci USA 1982;79:32603264.
immunohistochemical study. Am J Surg Pathol 248. Alexander A, Barritault D, Buxbaum J. Gamma
1989;13:10231033. heavy chain disease in man: translation and
237. Su L, Keren DF, Warren JS. Failure of anti- partial purication of mRNA coding for the
lambda immunoxation reagent mimics alpha deleted protein. Proc Natl Acad Sci USA
heavy-chain disease. Clin Chem 1995;41: 1978;75:47744778.
121123. 249. Buxbaum JN, Alexander A, Olivier O. Gamma
238. Rambaud JC, Galian A, Danon FG, et al. Alpha- heavy chain disease in man: synthesis of a deleted
chain disease without qualitative serum IgA gamma3 immunoglobulin by lymphoid cells in
abnormality. Report of two cases, including a short and long term tissue culture. Clin Exp
nonsecretory form. Cancer 1983;51:686693. Immunol 1978;32:489497.
239. Cogne M, Preudhomme JL. Gene deletions force 250. Lennert K, Stein H, Kaiserling E. Cytological and
nonsecretory alpha-chain disease plasma cells to functional criteria for the classication of
produce membrane-form alpha-chain only. malignant lymphomata. Br J Cancer
J Immunol 1990;145:24552458. 1975;31(Suppl 2):2943.
240. Seligmann M, Mihaesco E, Hurez D, Mihaesco 251. Presti BC, Sciotto CG, Marsh SG. Lymphocytic
204 Conditions associated with monoclonal gammopathies
specic antigenic determinants among IgM, IgA, 303. Bernstein ZP, Fitzpatrick JE, ODonnell A, Han
and IgG. J Immunol 1976;117:944949. T, Foon KA, Bhargava A. Clinical signicance of
293. McNutt DR, Fudenberg HH. IgG myeloma and monoclonal proteins in chronic lymphocytic
Waldenstrom macroglobulinemia. Coexistence leukemia. Leukemia 1992;6:12431245.
and clinical manifestations in one patient. Arch 304. Tienhaara A, Irjala K, Rajamaki A, Pulkki K.
Intern Med 1973;131:731736. Four monoclonal immunoglobulins in a patient
294. Keren DF, Gulbranson R, Carey JL, Krauss JC. with chronic lymphocytic leukemia. Clin Chem
2-Mercaptoethanol treatment improves 1986;32:703705.
measurement of an IgMkappa M-protein by 305. Decastello AV. Beitrage zur Kenntnis der Bence-
capillary electrophoresis. Clin Chem 2001;47: Jonesschen Albuminurie. Z Klin Med 1909;67:
13261327. 319343.
295. Jonsson V, Kierkegaard A, Salling S, et al. 306. Deegan MJ, Abraham JP, Sawdyk M, van Slyck
Autoimmunity in Waldenstroms EJ. High incidence of monoclonal proteins in the
macroglobulinaemia. Leuk Lymphoma serum and urine of chronic lymphocytic leukemia
1999;34:373379. patients. Blood 1984;64:12071211.
296. Kira J, Inuzuka T, Hozumi I, et al. A novel 307. Fu SM, Winchester RJ, Feizi T, Walzer PD,
monoclonal antibody which reacts with a high Kunkel HG. Idiotypic specicity of surface
molecular weight neuronal cytoplasmic protein immunoglobulin and the maturation of leukemic
and myelin basic protein (MBP) in a patient with bone-marrow-derived lymphocytes. Proc Natl
macroglobulinemia. J Neurol Sci 1997;148: Acad Sci USA 1974;71:44874490.
4752. 308. Qian GX, Fu SM, Solanki DL, Rai KR.
297. Tait RC, Oogarah PK, Houghton JB, Farrand SE, Circulating monoclonal IgM proteins in B cell
Haeney MR. Waldenstroms chronic lymphocytic leukemia: their
macroglobulinaemia secreting a paraprotein with identication, characterization and relationship
lupus anticoagulant activity: possible association to membrane IgM. J Immunol 1984;133:
with gastrointestinal tract disease and 33963400.
malabsorption. J Clin Pathol 1993;46:678680. 309. Deegan MJ, Maeda K. Differentiation of chronic
298. Boccadoro M, Van Acker A, Pileri A, Urbain J. lymphocytic leukemia cells after in vitro
Idiotypic lymphocytes in human monoclonal treatment with EpsteinBarr virus or phorbol
gammopathies. Ann Immunol (Paris) 1981; ester. I. Immunologic and morphologic studies.
132C:919. Am J Hematol 1984;17:335347.
299. Labardini-Mendez J, Alexanian R, Barlogie B. 310. Hannam-Harris AC, Gordon J, Smith JL.
Multiple myeloma and chronic lymphocytic Immunoglobulin synthesis by neoplastic B
leukemia. Rev Invest Clin 1997;49(Suppl 1): lymphocytes: free light chain synthesis as a
2833. marker of B cell differentiation. J Immunol
300. Barlogie B, Gale RP. Multiple myeloma and 1980;125:21772181.
chronic lymphocytic leukemia: parallels and 311. Murata T, Fujita H, Harano H, et al. Triclonal
contrasts. Am J Med 1992;93:443450. gammopathy (IgA kappa, IgG kappa, and IgM
301. Pines A, Ben-Bassat I, Selzer G, Ramot B. kappa) in a patient with plasmacytoid lymphoma
Transformation of chronic lymphocytic derived from a monoclonal origin. Am J Hematol
leukemia to plasmacytoma. Cancer 1984;54: 1993;42:212216.
19041907. 312. Tursi A, Modeo ME. Monoclonal gammopathy
302. Suchman AL, Coleman M, Mouradian JA, Wolf of undetermined signicance predisposing to
DJ, Saletan S. Aggressive plasma cell myeloma. A Helicobacter pylori-related gastric mucosa-
terminal phase. Arch Intern Med 1981;141: associated lymphoid tissue lymphoma. J Clin
13151320. Gastroenterol 2002;34:147149.
References 207
313. Pascali E. Monoclonal gammopathy and cold 323. Attygalle A, Al-Jehani R, Diss TC, et al.
agglutinin disease in non-Hodgkins lymphoma. Neoplastic T cells in angioimmunoblastic T-cell
Eur J Haematol 1996;56:114115. lymphoma express CD10. Blood 2002;99:
314. Bajetta E, Gasparini G, Facchetti G, Ferrari L, 627633.
Giardini R, Delia D. Monoclonal gammopathy 324. Yetgin S, Olcay L, Yel L, et al. T-ALL with
(IgM-k) in a patient with Burkitts type monoclonal gammopathy and hairy cell features.
lymphoblastic lymphoma. Tumori 1984;70: Am J Hematol 2000;65:166170.
403407. 325. Fernandez LA, MacSween JM, Langley GR.
315. Braunstein AH, Keren DF. Monoclonal Immunoglobulin secretory function of B cells
gammopathy (IgM-kappa) in a patient with from untreated patients with chronic
Burkitts lymphoma. Case report and literature lymphocytic leukemia and
review. Arch Pathol Lab Med 1983;107: hypogammaglobulinemia: role of T cells. Blood
235238. 1983;62:767774.
316. Heinz R, Stacher A, Pralle H, et al. 326. Freeland HS, Scott PP. Recurrent pulmonary
Lymphoplasmacytic/lymphoplasmacytoid infections in patients with chronic lymphocytic
lymphoma: a clinical entity distinct from chronic leukemia and hypogammaglobulinemia. South
lymphocytic leukaemia? Blut 1981;43:183192. Med J 1986;79:13661369.
317. Orth T, Treichel U, Mayet WJ, Storkel S, Meyer 327. Pangalis GA, Vassilakopoulos TP, Dimopoulou
zum Buschenfelde KH. Reversible myelobrosis MN, Siakantaris MP, Kontopidou FN,
in angioimmunoblastic lymphadenopathy. Dtsch Angelopoulou MK. B-chronic lymphocytic
Med Wochenschr 1994;119:694698. leukemia: practical aspects. Hematol Oncol
318. Oft K, Macris NT, Finkbeiner JA. Monoclonal 2002;20:103146.
hypergammaglobulinemia without malignant 328. Sampalo A, Brieva JA. Humoral
transformation in angioimmunoblastic immunodeciency in chronic lymphocytic
lymphadenopathy with dysproteinemia. Am J leukemia: role of CD95/CD95L in tumoral
Med 1986;80:292294. damage and escape. Leuk Lymphoma
319. Toccanier MF, Kapanci Y. Lymphomatoid 2002;43:881884.
granulomatosis, polymorphic reticulosis and 329. Caligaris-Cappio F. Biology of chronic
angioimmunoblastic lymphadenopathy with lymphocytic leukemia. Rev Clin Exp Hematol
pulmonary involvement. Similar or different 2000;4:521.
entities? Ann Pathol 1983;3:2941. 330. Pritsch O, Maloum K, Dighiero G. Basic biology
320. Palutke M, McDonald JM. Monoclonal of autoimmune phenomena in chronic
gammopathies associated with malignant lymphocytic leukemia. Semin Oncol 1998;25:
lymphomas. Am J Clin Pathol 1973;60: 3441.
157165. 331. Serpell LC, Sunde M, Blake CC. The molecular
321. Pascali E, Pezzoli A, Melato M, Falconieri G. basis of amyloidosis Cell Mol Life Sci 1997;53:
Nodular lymphoma eventuating into 871887.
lymphoplasmocytic lymphoma with monoclonal 332. Kyle RA, Gertz MA. Primary systemic
IgM/lambda cold agglutinin and Bence-Jones amyloidosis: clinical and laboratory features in
proteinuria. Acta Haematol 1980;64:94102. 474 cases. Semin Hematol 1995;32:4559.
322. Yamasaki S, Matsushita H, Tanimura S, et al. B- 333. Pick AI, Frohlichmann R, Lavie G, Duczyminer
cell lymphoma of mucosa-associated lymphoid M, Skvaril F. Clinical and immunochemical
tissue of the thymus: a report of two cases with a studies of 20 patients with amyloidosis and
background of Sjgrens syndrome and plasma cell dyscrasia. Acta Haematol 1981;66:
monoclonal gammopathy. Hum Pathol 1998;29: 154167.
10211024. 334. Kyle RA, Greipp PR. Amyloidosis (AL). Clinical
208 Conditions associated with monoclonal gammopathies
and laboratory features in 229 cases. Mayo Clin 347. Buxbaum JN, Chuba JV, Hellman GC, Solomon
Proc 1983;58:665683. A, Gallo GR. Monoclonal immunoglobulin
335. Dhodapkar MV, Merlini G, Solomon A. Biology deposition disease: light chain and light and
and therapy of immunoglobulin deposition heavy chain deposition diseases and their relation
diseases. Hematol Oncol Clin North Am to light chain amyloidosis. Clinical features,
1997;11:89110. immunopathology, and molecular analysis. Ann
336. Perfetti V, Garini P, Vignarelli MC, et al. Intern Med 1990;112:455464.
Diagnostic approach to and follow-up of difcult 348. Gallo G, Picken M, Buxbaum J, Frangione B.
cases of AL amyloidosis. Haematologica The spectrum of monoclonal immunoglobulin
1995;80:409415. deposition disease associated with immunocytic
337. Gertz MA, Kyle RA. Prognostic value of urinary dyscrasias. Semin Hematol 1989;26:234245.
protein in primary systemic amyloidosis (AL). 349. Kambham N, Markowitz GS, Appel GB, Kleiner
Am J Clin Pathol 1990;94:313317. MJ, Aucouturier P, DAgati VD. Heavy chain
338. Kyle R. AL Amyloidosis: current diagnostic and deposition disease: the disease spectrum. Am J
therapeutic aspects. Amyloid: J Protein Folding Kidney Dis. 1999;33:954962.
Disord 2001;8:6769. 350. McAllister HA Jr, Seger J, Bossart M, Ferrans VJ.
339. Gertz MA, Lacy MQ, Dispenzieri A. Restrictive cardiomyopathy with kappa light
Immunoglobulin light chain amyloidosis and the chain deposits in myocardium as a complication
kidney. Kidney Int 2002;61:19. of multiple myeloma. Histochemical and electron
340. Gillmore JD, Lovat LB, Persey MR, Pepys MB, microscopic observations. Arch Pathol Lab Med
Hawkins PN. Amyloid load and clinical outcome 1988;112:11511154.
in AA amyloidosis in relation to circulating 351. Rostagno A, Frizzera G, Ylagan L, Kumar A,
concentration of serum amyloid A protein. Ghiso J, Gallo G. Tumoral non-amyloidotic
Lancet 2001;358:2429. monoclonal immunoglobulin light chain deposits
341. Marhaug G, Dowton SB. Serum amyloid A: an (aggregoma): presenting feature of B-cell
acute phase apolipoprotein and precursor of AA dyscrasia in three cases with
amyloid. Baillieres Clin Rheumatol 1994;8: immunohistochemical and biochemical analyses.
553573. Br J Haematol 2002;119:6269.
342. Saraiva MJ. Sporadic cases of hereditary systemic 352. Buxbaum J. Mechanisms of disease: monoclonal
amyloidosis. N Engl J Med 2002;346:18181819. immunoglobulin deposition. Amyloidosis, light
343. Lachmann HJ, Booth DR, Booth SE, et al. chain deposition disease, and light and heavy
Misdiagnosis of hereditary amyloidosis as AL chain deposition disease. Hematol Oncol Clin
(primary) amyloidosis. N Engl J Med 2002;346: North Am 1992;6:323346.
17861791. 353. Ivanyi B. Frequency of light chain deposition
344. Sanders PW, Booker BB. Pathobiology of cast nephropathy relative to renal amyloidosis and
nephropathy from human Bence Jones proteins. Bence Jones cast nephropathy in a necropsy
J Clin Invest 1992;89:630639. study of patients with myeloma. Arch Pathol Lab
345. Buxbaum J, Gallo G. Nonamyloidotic Med 1990;114:986987.
monoclonal immunoglobulin deposition disease. 354. Miwa T, Kimura Y, Nonomura A, Kamide M,
Light-chain, heavy-chain, and light- and heavy- Furukawa M. Unusual case of plasma cell tumor
chain deposition diseases. Hematol Oncol Clin with monoclonal gammopathy of the sino-nasal
North Am 1999;13:12351248. cavity and clavicular bone. ORL J
346. Rivest C, Turgeon PP, Senecal JL. Lambda light Otorhinolaryngol Relat Spec 1993;55:4548.
chain deposition disease presenting as an 355. Lasker JC, Bishop JO, Wilbanks JH, Lane M.
amyloid-like arthropathy. J Rheumatol 1993;20: Solitary myeloma of the talus bone. Cancer
880884. 1991;68:202205.
References 209
356. Woodruff RK, Malpas JS, White FE. Solitary Nippon Naika Gakkai Zasshi 1979;68:
plasmacytoma. II: Solitary plasmacytoma of 13061312.
bone. Cancer 1979;43:23442347. 367. Zlotnick A, Benbassat J. Monoclonal
357. Menke DM, Horny HP, Griesser H, et al. gammopathy associated with autoimmune
Primary lymph node plasmacytomas (plasmacytic disease. Harefuah 1976;90:424427.
lymphomas). Am J Clin Pathol 2001;115: 368. Kiprov DD, Miller RG. Polymyositis associated
119126. with monoclonal gammopathy. Lancet
358. Frassica DA, Frassica FJ, Schray MF, Sim FH, 1984;2:11831186.
Kyle RA. Solitary plasmacytoma of bone: Mayo 369. Moutsopoulos HM, Tzioufas AG, Bai MK,
Clinic experience. Int J Radiat Oncol Biol Phys. Papadopoulos NM, Papadimitriou CS.
1989;16:4348. Association of serum IgM kappa monoclonicity
359. Dimopoulos MA, Goldstein J, Fuller L, Delasalle in patients with Sjogrens syndrome with an
K, Alexanian R. Curability of solitary bone increased proportion of kappa positive plasma
plasmacytoma. J Clin Oncol 1992;10: cells inltrating the labial minor salivary glands.
587590. Ann Rheum Dis 1990;49:929931.
360. Liebross RH, Ha CS, Cox JD, Weber D, 370. Seligmann M, Brouet JC. Antibody activity of
Delasalle K, Alexanian R. Solitary bone human monoclonal immunoglobulins. Pathol
plasmacytoma: outcome and prognostic factors Biol (Paris) 1990;38:822823.
following radiotherapy. Int J Radiat Oncol Biol 371. Seligmann M, Brouet JC. Antibody activity of
Phys 1998;41:10631067. human myeloma globulins. Semin Hematol
361. Johnson TL, Keren DF, Thomas LR. Indirect 1973;10:163177.
immunouorescence technique to detect 372. Potter M. Myeloma proteins (M-components)
monoclonal antinuclear antibody. Arch Pathol with antibody-like activity. N Engl J Med
Lab Med 1987;111:560562. 1971;284:831838.
362. Jonsson V, Svendsen B, Vorstrup S, et al. 373. Laurell CB. Complexes formed in vivo between
Multiple autoimmune manifestations in immunoglobulin light chain kappa, prealbumin
monoclonal gammopathy of undetermined and-or alpha-1-antitrypsin in myeloma sera.
signicance and chronic lymphocytic leukemia. Immunochemistry 1970;7:461465.
Leukemia 1996;10:327332. 374. Vrethem M, Cruz M, Wen-Xin H, Malm C,
363. Nakase T, Matsuoka N, Iwasaki E, Ukyo S, Holmgren H, Ernerudh J. Clinical,
Shirakawa S. CML with autoimmune neurophysiological and immunological evidence
thrombocytopenia observed in the course of IgG of polyneuropathy in patients with monoclonal
(kappa) type monoclonal gammopathy. Rinsho gammopathies. J Neurol Sci 1993;114:193199.
Ketsueki 1992;33:706708. 375. Latov N. Pathogenesis and therapy of
364. Guardigni L, Rasi F, Nemni R, Lorenzetti I, neuropathies associated with monoclonal
Valzania F, Pretolani E. Autoimmune neuropathy gammopathies. Ann Neurol 1995;37(Suppl 1):
in monoclonal gammopathy. Recenti Prog Med S3242.
1991;82:669671. 376. Nobile-Orazio E, Marmiroli P, Baldini L, et al.
365. Wasada T, Eguchi Y, Takayama S, Yao K, Peripheral neuropathy in macroglobulinemia:
Hirata Y, Ishii S. Insulin autoimmune syndrome incidence and antigen-specicity of M proteins.
associated with benign monoclonal gammopathy. Neurology 1987;37:15061514.
Evidence for monoclonal insulin autoantibodies. 377. Freddo L, Ariga T, Saito M, et al. The
Diabetes Care 1989;12:147150. neuropathy of plasma cell dyscrasia: binding of
366. Kozuru M, Nakashima Y, Kurata T, Kaneko S, IgM M-proteins to peripheral nerve glycolipids.
Ibayashi H. Autoimmune neutropenia associated Neurology 1985;35:14201424.
with monoclonal gammopathy (authors transl). 378. Latov N, Sherman WH, Nemni R, et al. Plasma-
210 Conditions associated with monoclonal gammopathies
cell dyscrasia and peripheral neuropathy with a RH. Sensory neuropathy associated with
monoclonal antibody to peripheral-nerve myelin. monoclonal immunoglobulin M to GD1b
N Engl J Med 1980;303:618621. ganglioside. Ann Neurol 1992;31:683685.
379. Latov N, Braun PE, Gross RB, Sherman WH, 390. Burger D, Perruisseau G, Simon M, Steck AJ.
Penn AS, Chess L. Plasma cell dyscrasia and Comparison of the N-linked oligosaccharide
peripheral neuropathy: identication of the structures of the two major human myelin
myelin antigens that react with human glycoproteins MAG and P0: assessment of the
paraproteins. Proc Natl Acad Sci USA 1981;78: structures bearing the epitope for HNK-1 and
71397142. human monoclonal immunoglobulin M found in
380. Latov N. Plasma cell dyscrasia and motor neuron demyelinating neuropathy. J NeuroChem
disease. Adv Neurol 1982;36:273279. 1992;58:854861.
381. Shy ME, Rowland LP, Smith T, et al. Motor 391. Kanda T, Iwasaki T, Yamawaki M, Tai T,
neuron disease and plasma cell dyscrasia. Mizusawa H. Anti-GM1 antibody facilitates
Neurology 1986;36:14291436. leakage in an in vitro blood-nerve barrier model.
382. Driedger H, Pruzanski W. Plasma cell neoplasia Neurology 2000;55:585587.
with peripheral polyneuropathy. A study of ve 392. Manschot SM, Notermans NC, van den Berg
cases and a review of the literature. Medicine LH, Verschuuren JJ, Lokhorst HM. Three
(Baltimore) 1980;59:301310. families with polyneuropathy associated with
383. Dalakas MC, Engel WK. Polyneuropathy with monoclonal gammopathy. Arch Neurol
monoclonal gammopathy: studies of 11 patients. 2000;57:740742.
Ann Neurol 1981;10:4552. 393. Jensen TS, Schroder HD, Jonsson V, et al. IgM
384. Ilyas AA, Quarles RH, Dalakas MC, Fishman monoclonal gammopathy and neuropathy in two
PH, Brady RO. Monoclonal IgM in a patient siblings. J Neurol Neurosurg Psychiatry 1988;
with paraproteinemic polyneuropathy binds to 51:13081315.
gangliosides containing disialosyl groups. Ann 394. Latov N. Neuropathy, heredity, and monoclonal
Neurol 1985;18:655659. gammopathy. Arch Neurol 2000;57:641642.
385. Steck AJ, Murray N, Meier C, Page N, 395. Simmons Z, Bromberg MB, Feldman EL, Blaivas
Perruisseau G. Demyelinating neuropathy and M. Polyneuropathy associated with IgA
monoclonal IgM antibody to myelin-associated monoclonal gammopathy of undetermined
glycoprotein. Neurology 1983;33:1923. signicance. Muscle Nerve 1993;16:7783.
386. Meier C, Vandevelde M, Steck A, Zurbriggen A. 396. Kanda T, Usui S, Beppu H, Miyamoto K,
Demyelinating polyneuropathy associated with Yamawaki M, Oda M. Blood-nerve barrier in
monoclonal IgM-paraproteinaemia. Histological, IgM paraproteinemic neuropathy: a
ultrastructural and immunocytochemical studies. clinicopathologic assessment. Acta Neuropathol
J Neurol Sci 1984;63:353367. (Berl) 1998;95:184192.
387. Steck AJ, Murray N, Justafre JC, et al. Passive 397. Lach B, Rippstein P, Atack D, Afar DE, Gregor
transfer studies in demyelinating neuropathy A. Immunoelectron microscopic localization of
with IgM monoclonal antibodies to myelin- monoclonal IgM antibodies in gammopathy
associated glycoprotein. J Neurol Neurosurg associated with peripheral demyelinative
Psychiatry 1985;48:927929. neuropathy. Acta Neuropathol (Berl) 1993;85:
388. Latov N, Hays AP, Donofrio PD, et al. 298307.
Monoclonal IgM with unique specicity to 398. Vital C, Vital A, Deminiere C, Julien J, Lagueny
gangliosides GM1 and GD1b and to lacto-N- A, Steck AJ. Myelin modications in 8 cases of
tetraose associated with human motor neuron peripheral neuropathy with Waldenstroms
disease. Neurology 1988;38:763768. macroglobulinemia and anti-MAG activity.
389. Daune GC, Farrer RG, Dalakas MC, Quarles Ultrastruct Pathol 1997;21:509516.
References 211
399. Zuckerman SJ, Pesce MA, Rowland LP, et al. An on human serum immunoglobulin
alert for motor neuron diseases and peripheral concentrations. J Immunol 1970;105:964972.
neuropathy: monoclonal paraproteinemia may be 411. Fine JM, Lambin P, Leroux P. Frequency of
missed by routine electrophoresis. Arch Neurol monoclonal gammopathy (M components) in
1987;44:250251. 13,400 sera from blood donors. Vox Sang
400. Vrethem M, Larsson B, von Schenck H, 1972;23:336343.
Ernerudh J. Immunoxation superior to plasma 412. Crawford J, Eye MK, Cohen HJ. Evaluation of
agarose electrophoresis in detecting small M- monoclonal gammopathies in the well elderly.
components in patients with polyneuropathy. Am J Med 1987;82:3945.
J Neurol Sci 1993;120:9398. 413. Franceschi C, Cossarizza A. Introduction: the
401. Dyck PJ, Low PA, Windebank AJ, et al. Plasma reshaping of the immune system with age. Int
exchange in polyneuropathy associated with Rev Immunol 1995;12:14.
monoclonal gammopathy of undetermined 414. Makinodan T, Kay MM. Age inuence on the
signicance. N Engl J Med 1991;325: immune system. Adv Immunol 1980;29:
14821486. 287330.
402. Monaco S, Bonetti B, Ferrari S, et al. 415. Radl J. Differences among the three major
Complement-mediated demyelination in patients categories of paraproteinaemias in aging man
with IgM monoclonal gammopathy and and the mouse. A minireview. Mech Ageing Dev
polyneuropathy. N Engl J Med 1990;322: 1984;28:167170.
649652. 416. Kyle RA. Monoclonal gammopathy of
403. Dalakas MC, Quarles RH, Farrer RG, et al. A undetermined signicance. Natural history in
controlled study of intravenous immunoglobulin 241 cases. Am J Med 1978;64:814826.
in demyelinating neuropathy with IgM 417. Peters SO, Stockschlader M, Zeller W,
gammopathy. Ann Neurol 1996;40:792795. Mross K, Durken M, Kruger W, Zander AR.
404. Koike H, Sobue G. CrowFukase syndrome. Monoclonal gammopathy of unknown
Neuropathology 2000;20(Suppl):S6972. signicance in a bone marrow donor. Ann
405. Araki T, Konno T, Soma R, et al. CrowFukase Hematol 1993;66:9395.
syndrome associated with high-output heart 418. Avet-Loiseau H, Li JY, Morineau N, Facon T,
failure. Intern Med 2002;41:638641. Brigaudeau C, Harousseau JL, Grosbois B,
406. Nakanishi T, Sobue I, Toyokura Y, et al. The Bataille R. Monosomy 13 is associated with the
CrowFukase syndrome: a study of 102 cases in transition of monoclonal gammopathy of
Japan. Neurology 1984;34:712720. undetermined signicance to multiple myeloma.
407. Dispenzieri A, Kyle RA, Lacy MQ, et al. POEMS Intergroupe Francophone du Myelome. Blood
syndrome: denitions and long-term outcome. 1999;94:25832589.
Blood 2003;101:24962506. 419. Avet-Loiseau H, Facon T, Daviet A, et al. 14q32
408. Miralles GD, OFallon JR, Talley NJ. Plasma-cell translocations and monosomy 13 observed in
dyscrasia with polyneuropathy. The spectrum of monoclonal gammopathy of undetermined
POEMS syndrome. N Engl J Med 1992;327: signicance delineate a multistep process for the
19191923. oncogenesis of multiple myeloma. Intergroupe
409. Lacy MQ, Gertz MA, Hanson CA, Inwards DJ, Francophone du Myelome. Cancer Res 1999;59:
Kyle RA. Multiple myeloma associated with 45464550.
diffuse osteosclerotic bone lesions: a clinical 420. Fonseca R, Bailey RJ, Ahmann GJ, et al.
entity distinct from osteosclerotic myeloma Genomic abnormalities in monoclonal
(POEMS syndrome). Am J Hematol 1997;56: gammopathy of undetermined signicance.
288293. Blood 2002;100:14171424.
410. Buckley CE 3rd, Dorsey FC. The effect of aging 421. Vejlgaard T, Abildgaard N, Jans H, Nielsen JL,
212 Conditions associated with monoclonal gammopathies
Heickendorff L. Abnormal bone turnover in bacterial infection. Acta Med Okayama 1976;30:
monoclonal gammopathy of undetermined 209214.
signicance: analyses of type I collagen 431. Crapper RM, Deam DR, Mackay IR.
telopeptide, osteocalcin, bone-specic alkaline Paraproteinemias in homosexual men with HIV
phosphatase and propeptides of type I and type infection. Lack of association with abnormal
III procollagens. Eur J Haematol 1997;58: clinical or immunologic ndings. Am J Clin
104108. Pathol 1987;88:348351.
422. Pecherstorfer M, Seibel MJ, Woitge HW, et al. 432. Tsuji T, Tokuyama K, Naito K, Okazaki S,
Bone resorption in multiple myeloma and in Shinohara T. A case report of primary hepatic
monoclonal gammopathy of undetermined carcinoma with prolonged HB virus infection
signicance: quantication by urinary pyridinium and monoclonal gammopathy. Gastroenterol Jpn
cross-links of collagen. Blood 1997;90: 1977;12:6975.
37433750. 433. Godeau P, Herson S, De Treglode D, Herreman
423. Zheng C, Huang DR, Bergenbrant S, Sundblad G. Benign monoclonal immunoglobulins during
A, et al. Interleukin 6, tumour necrosis factor subacute infectious endocarditis. Coeur Med
alpha, interleukin 1beta and interleukin 1 Interne. 1979;18:312.
receptor antagonist promoter or coding gene 434. Herreman G, Godeau P, Cabane J, Digeon M,
polymorphisms in multiple myeloma. Br J Laver M, Bach JF. Immunologic study of
Haematol 2000;109:3945. subacute infectious endocarditis through the
424. Cesana C, Klersy C, Barbarano L, et al. search for circulating immune complexes.
Prognostic factors for malignant transformation Preliminary results apropos of 13 cases. Nouv
in monoclonal gammopathy of undetermined Presse Med 1975;4:23112314.
signicance and smoldering multiple myeloma. 435. Struve J, Weiland O, Nord CE. Lactobacillus
J Clin Oncol 2002;20:16251634. plantarum endocarditis in a patient with benign
425. Kanoh T. Fluctuating M-component level in monoclonal gammopathy. J Infect 1988;17:
relation to infection. Eur J Haematol 1989;42: 127130.
503504. 436. Covinsky M, Laterza O, Pfeifer JD, Farkas-
426. Pasqualetti P, Casale R. Risk of malignant Szallasi T, Scott MG. An IgM lambda antibody
transformation in patients with monoclonal to Escherichia coli produces false-positive results
gammopathy of undetermined signicance. in multiple immunometric assays. Clin Chem
Biomed Pharmacother 1997;51:7478. 2000;46:11571161.
427. Pasqualetti P, Festuccia V, Collacciani A, Casale 437. Kelly RH, Hardy TJ, Shah PM. Benign
R. The natural history of monoclonal monoclonal gammopathy: a reassessment of the
gammopathy of undetermined signicance. problem. Immunol Invest 1985;14:183197.
A 5- to 20-year follow-up of 263 cases. 438. Turbat-Herrera EA, Hancock C, Cabello-
Acta Haematol 1997;97:174179. Inchausti B, Herrera GA. Plasma cell hyperplasia
428. Colls BM. Monoclonal gammopathy of and monoclonal paraproteinemia in human
undetermined signicance (MGUS) 31 year immunodeciency virus-infected patients. Arch
follow up of a community study. Aust N Z J Pathol Lab Med 1993;117:497501.
Med 1999;29:500504. 439. Radl J. Monoclonal gammapathies. An attempt
429. Larrain C. Transient monoclonal gammopathies at a new classication. Neth J Med 1985;28:
associated with infectious endocarditis. Rev Med 134137.
Chil 1986;114:771776. 440. van den Akker TW, Brondijk R, Radl J.
430. Arima T, Tsuboi S, Nagata K, Gyoten Y, Inuence of long-term antigenic stimulation
Tanigawa T. An extremely basic monoclonal IgG started in young C57BL mice on the
in an aged apoplectic patient with prolonged development of age-related monoclonal
References 213
gammapathies. Int Arch Allergy Appl Immunol Zorat F, Pozzato G. Effectiveness of leukocyte
1988;87:165170. interferon in patients affected by HCV-positive
441. Radl J, Liu M, Hoogeveen CM, van den Berg P, mixed cryoglobulinemia resistant to recombinant
Minkman-Brondijk RJ, Broerse JJ, Zurcher C, alpha-interferon. Clin Exp Rheumatol 2002;20:
van Zwieten MJ. Monoclonal gammapathies in 2734.
long-term surviving rhesus monkeys after lethal 452. Naarendorp M, Kallemuchikkal U, Nuovo GJ,
irradiation and bone marrow transplantation. Gorevic PD. Long-term efcacy of interferon-
Clin Immunol Immunopathol 1991;60:305309. alpha for extrahepatic disease associated with
442. Radl J, Valentijn RM, Haaijman JJ, Paul LC. hepatitis C virus infection. J Rheumatol
Monoclonal gammapathies in patients 2001;28:24662473.
undergoing immunosuppressive treatment after 453. Monti G, Saccardo F, Rinaldi G, Petrozzino MR,
renal transplantation. Clin Immunol Gomitoni A, Invernizzi F. Colchicine in the
Immunopathol 1985;37:98102. treatment of mixed cryoglobulinemia. Clin Exp
443. Jain A, Nalesnik M, Reyes J, et al. Posttransplant Rheumatol 1995;13 Suppl 13:S197199.
lymphoproliferative disorders in liver 454. Dispenzieri A. Symptomatic cryoglobulinemia.
transplantation: a 20-year experience. Ann Surg Curr Treat Options Oncol 2000;1:105118.
2002;236:429436. 455. Friedman G, Mehta S, Sherker AH. Fatal
444. Craig FE, Gulley ML, Banks PM. exacerbation of hepatitis C-related
Posttransplantation lymphoproliferative cryoglobulinemia with interferon-alpha therapy.
disorders. Am J Clin Pathol 1993;99:265276. Dig Dis Sci 1999;44:13641365.
445. Stevens SJ, Verschuuren EA, Pronk I, et al. 456. Yebra M, Barrios Y, Rincon J, Sanjuan I, Diaz-
Frequent monitoring of EpsteinBarr virus DNA Espada F. Severe cutaneous vasculitis following
load in unfractionated whole blood is essential intravenous infusion of gammaglobulin in a
for early detection of posttransplant patient with type II mixed cryoglobulinemia. Clin
lymphoproliferative disease in high-risk patients. Exp Rheumatol 2002;20:225227.
Blood 2001;97:11651171. 457. Brouet JC, Clauvel JP, Danon F, Klein M,
446. Dispenzieri A, Gorevic PD. Cryoglobulinemia. Seligmann M. Biologic and clinical signicance of
Hematol Oncol Clin North Am 1999;13: cryoglobulins. A report of 86 cases. Am J Med
13151349. 1974;57:775788.
447. Wintrobe MM, Buell MV. Hyperproteinemia 458. Liu F, Knight GB, Agnello V. Hepatitis C virus
associated with multiple myeloma. Bull Johns but not GB virus C/hepatitis G virus has a role in
Hopkins Hosp 1933;52:156165. type II cryoglobulinemia. Arthritis Rheum
448. Lerner AB, Watson CJ. Studies of cryoglobulins. 1999;42:18981901.
Unusual purpura associated with the presence of 459. Agnello V, Chung RT, Kaplan LM. A role for
a high concentration of cryoglobulin (cold hepatitis C virus infection in type II
precipitable serum globulins). Am J Med Sci cryoglobulinemia. N Engl J Med 1992;327:
1947;2:410. 14901495.
449. Durand JM, Cacoub P, Lunel-Fabiani F, et al. 460. Trendelenburg M, Schifferli JA. Cryoglobulins
Ribavirin in hepatitis C related cryoglobulinemia. are not essential. Ann Rheum Dis. 1998;57:35.
J Rheumatol 1998;25:11151117. 461. Trejo O, Ramos-Casals M, Garcia-Carrasco M,
450. Akriviadis EA, Xanthakis I, Navrozidou C, et al. Cryoglobulinemia: study of etiologic factors
Papadopoulos A. Prevalence of cryoglobulinemia and clinical and immunologic features in 443
in chronic hepatitis C virus infection and patients from a single center. Medicine
response to treatment with interferon-alpha. (Baltimore) 2001;80:252262.
J Clin Gastroenterol 1997;25:612618. 462. Feiner HD. Relationship of tissue deposits of
451. Mazzaro C, Colle R, Baracetti S, Nascimben F, cryoglobulin to clinical features of mixed
214 Conditions associated with monoclonal gammopathies
485. Howard TW, Iannini MJ, Burge JJ, Davis JS. paraprotein. Ann Clin BioChem 1991;28(Pt
Rheumatoid factor, cryoglobulinemia, anti-DNA, 3):229234.
and renal disease in patients with systemic lupus 497. Bakker AJ. Inuence of monoclonal
erythematosus. J Rheumatol 1991;18:826830. immunoglobulins in direct determinations of iron
486. Invernizzi F, Pietrogrande M, Sagramoso B. in serum. Clin Chem 1991;37:690694.
Classication of the cryoglobulinemic syndrome. 498. Bakker AJ, Bosma H, Christen PJ. Inuence of
Clin Exp Rheumatol 1995;13(Suppl 13): monoclonal immunoglobulins in three different
S123128. methods for inorganic phosphorus. Ann Clin
487. Keren DF, Di Sante AC, Mervak T, Bordine SL. BioChem 1990;27(Pt 3):227231.
Problems with transporting serum to the 499. Dalal BI, Collins SY, Burnie K, Barr RM. Positive
laboratory for cryoglobulin assay: a solution. direct antiglobulin tests in myeloma patients.
Clin Chem 1985;31:17661767. Occurrence, characterization, and signicance.
488. Gorevic PD, Galanakis D. Chapter 10. Am J Clin Pathol 1991;96:496499.
Cryoglobulins, cryobrinogenemia, and 500. Glaspy JA. Hemostatic abnormalities in multiple
pyroglobulins. In: Rose NR, Hamilton RG, myeloma and related disorders. Hematol Oncol
Detrick B., eds. Manual of clinical laboratory Clin N Am 1992;6:13011314.
immunology. Washington, DC: ASM Press, 501. Panzer S, Thaler E. An acquired
2002. cryoglobulinemia which inhibits brin
489. Grose MP. Clinical Immunology Tech Sample polymerization in a patient with IgG kappa
No. CI-2. ASCP Tech Sample 1988:1. myeloma. Haemostasis 1993;23:6976.
490. Okazaki T, Nagai T, Kanno T. Gel diffusion 502. Matsuzaki H, Hata H, Watanabe T, Takeya M,
procedure for the detection of cryoglobulins in Takatsuki K. Phagocytic multiple myeloma with
serum. Clin Chem 1998;44:15581559. disseminated intravascular coagulation. Acta
491. Gorevic PD, Kassab HJ, Levo Y, et al. Mixed Haematol 1989;82:9194.
cryoglobulinemia: clinical aspects and long-term 503. Richard C, Cuadrado MA, Prieto M, et al.
follow-up of 40 patients. Am J Med 1980;69: Acquired von Willebrand disease in multiple
287308. myeloma secondary to absorption of von
492. Meltzer M, Franklin EC. Cryoglobulinemia a Willebrand factor by plasma cells. Am J Hematol
study of twenty-nine patients. I. IgG and IgM 1990;35:114117.
cryoglobulins and factors affecting 504. Schwarzinger I, Stain-Kos M, Bettelheim P, et al.
cryoprecipitability. Am J Med 1966;40:828836. Recurrent, isolated factor X deciency in
493. Meltzer M, Franklin EC, Elias K, McCluskey myeloma: repeated normalization of factor X
RT, Cooper N. Cryoglobulinemia a clinical and levels after cytostatic chemotherapy followed by
laboratory study. II. Cryoglobulins with late treatment failure associated with the
rheumatoid factor activity. Am J Med 1966;40: development of systemic amyloidosis. Thromb
837856. Haemost 1992;68:648651.
494. Markel SF, Janich SL. Complexing of lactate 505. Redmon B, Pyzdrowski KL, Elson MK, Kay NE,
dehydrogenase isoenzymes with immunoglobulin Dalmasso AP, Nuttall FQ. Hypoglycemia due to
A of the kappa class. Am J Clin Pathol 1974;61: an insulin-binding monoclonal antibody in
328332. multiple myeloma. N Engl J Med 1992;326:
495. Backer ET, Harff GA, Beyer C. A patient with an 994998.
IgG paraprotein and complexes of lactate 506. Side L, Fahie-Wilson MN, Mills MJ.
dehydrogenase and IgG in the serum. Clin Chem Hypercalcaemia due to calcium binding IgM
1987;33:19371938. paraprotein in Waldenstroms
496. Pearce CJ, Hine TJ, Peek K. Hypercalcaemia due macroglobulinaemia. J Clin Pathol 1995;48:
to calcium binding by a polymeric IgA kappa- 961962.
216 Conditions associated with monoclonal gammopathies
507. Goodrick MJ, Boon RJ, Bishop RJ, Copplestone 510. Taft EG, Grossman J, Abraham GN, Leddy JP,
JA, Prentice AG. Inaccurate haemoglobin Lichtman MA. Pseudoleukocytosis due to
estimation in Waldenstroms cryoprotein crystals. Am J Clin Pathol 1973;60:
macroglobulinaemia: unusual reaction with 669671.
monomeric IgM paraprotein. J Clin Pathol 511. Haeney MR. Erroneous values for the total white
1993;46:11381139. cell count and ESR in patients with
508. Fohlen-Walter A, Jacob C, Lecompte T, Lesesve cryoglobulinaemia. J Clin Pathol 1976;29:
JF. Laboratory identication of cryoglobulinemia 894897.
from automated blood cell counts, fresh blood 512. Hambley H, Vetters JM. Artefacts associated
samples, and blood lms. Am J Clin Pathol with a cryoglobulin. Postgrad Med J 1989;65:
2002;117:606614. 241243.
509. Emori HW, Bluestone R, Goldberg LS. Pseudo- 513. Lesesve JF, Goasguen J. Cryoglobulin detection
leukocytosis associated with cryoglobulinemia. from a blood smear leading to the diagnosis of
Am J Clin Pathol 1973;60:202204. multiple myeloma. Eur J Haematol 2000;65:77.
7
Examination of urine for
proteinuria
Urine protein composition 217 Detection of MFLC in the urine and serum by
Measurement of total urine protein 219 electrophoresis and immunoxation 232
Concentration of urine samples 221 False negative MFLC in urine by electrophoresis 237
Proteinuria after minor injury 222 False positive MFLC in urine by
Overow proteinuria 222 immunoxation 238
Glomerular proteinuria 222 Nephelometry to measure total kappa and
Tubular proteinuria 225 total l light chains in urine 242
Factitious proteinuria 229 Techniques to measure free k and free l light
Monoclonal free light chains 229 chains in serum and/or urine 242
Renal damage caused by MFLC 230 References 246
Detection and measurement of MFLC 231 Appendix 255
URINE PROTEIN COMPOSITION albumin itself, its size and charge allow less than
0.1 per cent of its plasma concentration to pass
Evaluation of the protein content of urine
samples can provide useful information about the Albumin
Origin
location and degree of damage within the
nephron, as well as the presence of monoclonal Mild
proteins (Fig. 7.1). The protein composition of
urine depends upon both factors intrinsic to the
protein itself and on pathophysiological alter-
ations in the patients. Glomerular
into the glomerular ultraltrate.1 Yet, because of its g trace protein (12 kDa), and retinol-binding protein
large concentration in the serum, it is often the (20 kDa). More than 90 per cent of the low mole-
predominant protein detected in the urine even in cular weight proteins that pass into the glomerular
cases with only minor glomerular leakage (see ltrate under physiological conditions are
below). reabsorbed by binding to specic receptors on the
proximal convoluted tubules, where they are cata-
bolized. Absorption is also inuenced by the charge
Charge of the protein on the protein. Proteins that are relatively anionic
are more readily absorbed than cationic proteins.
Charge of the protein is another major inuence on Despite the reabsorption, a small amount of protein
glomerular permeability. The glomerulus has a net (less than 150 mg/24 h), mainly albumin, normally
negative charge because of the presence of negatively nds its way into the urine.
charged glycoproteins.2 Consequently, the poly-
anion albumin is repelled by this negatively charged
glomerular capillary surface. Neutral dextran of Protein in normal urine
similar molecular weight to albumin will pass much
more readily into the glomerular ltrate.3 In most urine samples, insufcient protein is present
for detection by standard electrophoretic tech-
niques. Consequently, normal urine must be con-
centrated to facilitate examination of the various
Hydrostatic pressure protein fractions. In concentrating the urine, one
must be concerned with the possible loss of low
Another factor that may alter protein composition
molecular weight proteins that are useful to deter-
of urine is the hydrostatic pressure within the sys-
mine the location of damage within the nephron.
temic circulation. As blood pressure increases,
Normally, adults excrete about 100150 mg/24 h
larger molecules pass through the glomerulus. This
and children excrete 60100 mg/24 h of protein
results in a glomerular ltrate with a dispropor-
(mainly albumin) in the urine. Often, a 24-h collec-
tionate number of molecules larger than 100 kDa.
tion is difcult to obtain. Because of this, the ratio
Consequently, among patients with primary hyper-
of total protein to creatinine on random samples of
tension and albuminuria, reduction of the blood
urine has been employed as an alternative method
pressure reduces the albuminuria.4
to study urine protein. For adults, the normal
protein/creatinine ratio is < 0.2, for children age
624 months it is < 0.5, and for children older than
Glomerular ltrate and tubular 24 months it is 0.20.25.6
absorption A faintly staining albumin band may be seen by
using electrophoresis on concentrated urine. The
Under normal circumstances, the glomerular ltrate albumin band in urine samples migrates closer to
contains about 10 mg protein/l.5 This is composed the anode than it does in the serum and there is
of numerous low molecular weight proteins and often anodal slurring that smears albumin toward
polypeptides, along with albumin. As mentioned the positive electrode. This migration of albumin is
above, despite its size and negative charge, its large attributed to the binding of anions to its surface.7 In
concentration in serum accounts for the presence of addition to albumin, a small amount of
albumin in the glomerular ltrate. The most com- TammHorsfall protein (80 kDa), the heavily gly-
mon small proteins in the glomerular ltrate include cosylated (about 30 per cent carbohydrate) secre-
a1-acid glycoprotein (orosomucoid, 40 kDa), a1- tory product from the cells in the thick ascending
microglobulin (27 kDa), b2-microglobulin (12 kDa), limb of the loop of Henle,8 makes up most of the
Measurement of total urine protein 219
other protein found in normal urine. Too little is when screening for MFLC; dipsticks are adequate
present for detection by electrophoresis. only as general screens for proteinuria. Further, in
Electrophoresis on the typical 50100 times con- assessing urine for the presence of MFLC it is
centrated sample of normal urine does not disclose important to discard the use of the classic heat
staining in the g-region. However, small amounts precipitation test of the urine. Most MFLC that are
of polyclonal free light chains are usually present in in the urine in large quantities will precipitate
concentrated urine from patients with mild when heated to 56C and upon further heating will
glomerular or tubular disease. It is important to dissolve; unfortunately, so will some polyclonal
distinguish these polyclonal FLC from the mono- light chains. Further, the heat test misses at least
clonal free light chains (MFLCs, formerly Bence 3050 per cent of true monoclonal free light chains
Jones proteins; discussed below). Heavy chains are because some MFLC do not precipitate and dis-
normally secreted by plasma cells only as part of solve under these conditions at any concentration
intact immunoglobulin molecules. Uncoupled and others are present in too small an amount to be
heavy chains remain in the endoplasmic reticulum detected by this insensitive method.12 Although sul-
and are degraded.9,10 Unlike normal heavy chains, fosalicylic acid (SSA) is more sensitive than
however, normal light chains may be secreted as Albusticks (Bayer AG, Leverkusen, Germany), I
part of the intact immunoglobulin molecule, or as have seen many cases of obvious MFLC protein-
free light chains.10,11 The excess free light chains uria detected by urine protein electrophoresis and
that occur as monomers (mainly k) with a molecu- immunoxation that gave negative SSA tests.
lar mass of about 25 kDa or as non-covalent Therefore, when screening urine for MFLC, I rec-
dimers (mainly l) with a mass of 50 kDa are small ommend performing urine protein electrophoresis
enough to pass through the glomerulus and are and immunoxation on concentrated urine (see
reabsorbed by the proximal convoluted tubules below). Examination of the urine sediment for the
where they are catabolized. presence of cells, casts, and crystals can be helpful
When excessive amounts of free light chains are in establishing etiology and prognosis. It is worth
present because of an increase in immunoglobulin repeating since a negative dipstick does not rule
production, or glomerular or tubular damage, they out MFLC.
are not completely reabsorbed by the tubules, result- There are several methods currently in use in clin-
ing in a light chain proteinuria. If the excreted light ical laboratories to measure urinary proteins. A
chains are polyclonal, they migrate broadly in the 24 h collection of urine provides a good sample on
b- and g-regions. Plasma cells in the bone marrow which to base the measurement of total protein.
of patients with polyclonal increases in the serum The classic biuret method is still used in a few lab-
gamma globulins often produce large amounts of oratories to measure total urine protein. It reacts
free polyclonal light chains. If one is unaware that with peptide bonds and provides an equal mea-
polyclonal free light chains can be found under surement of both MFLC and the other urine pro-
these conditions, an immunoelectrophoretic or teins.13 Other techniques do not detect MFLC to
immunoxation pattern could be mistaken for the same extent as they do other proteins.
MFLC because of the relatively restricted hetero- Although it does measure a variety of proteins with
geneity of free polyclonal light chains (Fig. 7.2). the same sensitivity, the biuret method is not
amenable to automation.14 Therefore, in recent
years, a wide variety of other methods have been
MEASUREMENT OF TOTAL URINE employed in clinical laboratories.
PROTEIN Because of their ease of automation, more
common assays now used to quantify urine protein
The most common test for proteinuria is the dip- are turbidimetry by trichloroacetic acid (TCA),15
stick. However a dipstick should never be used TCAPonceau S,16 and several dye-binding
220 Examination of urine for proteinuria
(a)
methods.17,18 Unfortunately, the dye-binding between the Coomassie Brilliant Blue and the pyro-
methods differ in their ability to react with proteins gallol redmolybdate dye-binding methods.
of different charge. Therefore, the reaction varies Lefevre et al.18 further noted that there was less
with the albumin to globulin ratio. One aid to con- than a 15 per cent difference between pyrogallol
sistency in performance of dye-binding assays is redmolybdate and pyrocatecholviolet (UPRO
standardization of the calibrator. Marshall and Vitros; Ortho-Clinical Diagnostics, Raritan, NJ,
Williams17 demonstrated that the use of a urinary USA) when measuring light chain proteinuria.
protein calibrator improved the agreement However, in samples with glomerular proteinuria,
Concentration of urine samples 221
they noted considerable differences in the total gel. Yet, Kaplan and Levinson13 caution against
protein obtained. using only 1020 times concentration, claiming it
No single particular automated method is consis- to be insufcient to detect some cases with MFLC.
tently superior to another. Whichever method is I have not found concentrating urine more than
chosen in the laboratory should be used for all 50100 times to be necessary to detect the pattern
samples on that patient in order to provide consis- of proteinuria, or, as discussed below, clinically
tent data when following a patient. If two different relevant amounts of MFLC by immunoxation.
methods are used on samples taken at two different The speed at which the stationary ltration devices
times, any apparent increase or decrease may be concentrate the urine is inversely proportional to the
caused by variability in the methods rather than a concentration of proteins in the sample. Therefore,
change in the patients condition. samples with large protein concentrations, such as
those from patients with a non-selective proteinuria,
will concentrate slowly. There is no reason to con-
CONCENTRATION OF URINE centrate such samples even 50 times. Our labora-
SAMPLES tory places the urine samples into the concentrators
and allows them to concentrate for up to 4 h. We
Urine samples need to be concentrated sufciently monitor them to prevent ones that concentrate
to allow visualization of the protein bands in most quickly from evaporating. Thus, one may concen-
techniques used by clinical laboratories for urine trate samples for a standard period of time as long
protein electrophoresis. Colloidal staining with as one records the nal concentration.
gold or silver does not require concentration of Concentration of urine is especially important to
urine samples.19 These staining methods are about detect small, but clinically relevant quantities of
200 times more sensitive than protein dye staining MFLC by immunoxation. Detection of even small
methods such as Ponceau S or Amido Black and quantities of MFLC may have importance in cases
others that are commonly used to stain urine of amyloid AL (amyloid associated with
protein electrophoresis.20,21 However, colloidal immunoglobulin light chain) or B-cell lymphopro-
methods are technically more laborious than the liferative disorders (see Chapter 6). To rule out amy-
dye methods, not readily automated and give loidosis, all adult patients with nephrotic syndrome
grainy results that may obscure subtle bands. should have an immunoxation of both serum and
Most laboratories use commercially available dif- urine.23 Unfortunately, with current techniques, not
ferential ltration techniques to concentrate urine. all cases of amyloid AL will have positive urine
Some are passive diffusion while others use cen- samples for the presence of MFLC by immunoxa-
trifugation to speed the process.21 The amount of tion,24,25 and biopsy is still the main method of diag-
concentrating required will vary depending on the nosing these cases. However, the advent of new,
technique used for electrophoresis and on the con- highly sensitive and specic nephelometric methods
centration of protein in a particular sample. For to quantify free light chains in both serum and urine
most agarose or cellulose acetate methods when (see below) has provided a method that is capable
stained by protein dyes, a 50100 times concentra- of detecting many cases of amyloid AL that formerly
tion is usually sufcient for urine specimens that were not detected by urine protein electrophoresis
have low concentrations of protein. While some or even by immunoxation.26
have suggested concentrating urine as much as The practical limit of detection of MFLC by
300600 times to detect tiny quantities of MFLC,22 immunoxation is related to the background of
such concentrating is technically laborious and can polyclonal free light chains that occur in the
result in highly viscous samples that are difcult to urine.27 The limited heterogeneity of these mole-
process.21 In samples with a large amount of pro- cules results in a banding pattern that makes reli-
teinuria, too much concentration may overload the able distinction of MFLC at concentrations similar
222 Examination of urine for proteinuria
to those of the bands found with polyclonal free Massive hemolysis and crush injuries will result in
light chains unreliable with currently available the release of large quantities of hemoglobin and
techniques. Because we readily see the ladder myoglobin, respectively, which will produce a
pattern on 50 concentrations of urine with even single large band in the urine (Fig. 7.3).44 Among
the small quantities of protein present in cases of other small molecules, b2-microglobulin,
mild tubular proteinuria, our laboratory does not eosinophil-derived neurotoxin and lysozyme have
concentrate samples more than that. been reported to cause a band in protein elec-
trophoresis that could be mistaken for MFLC (Fig.
7.4).4547 This is why immunoxation or measure-
PROTEINURIA AFTER MINOR ment of free light chains should be used to evaluate
INJURY unexplained bands seen on urine protein elec-
trophoresis to avoid an erroneous interpretation of
Relatively minor injury has been associated with MFLC.
transient proteinuria. Vigorous physical exercise
may cause signicant increases in the concentra-
tion of albumin in the urine.28 Congestive heart GLOMERULAR PROTEINURIA
failure, normal pregnancy, heavy alcohol con-
sumption and reactions to drugs are all associated Severe renal disease results in a profound protein-
with minor, usually transient proteinuria.2933 uria, the characteristics of which can be dened by
Patients with non-renal febrile illnesses also may urine protein electrophoresis in many cases. In
have proteinuria.3436 During such episodes, usually general, patients with glomerular disease excrete
only an albumin band is seen on electrophoresis. considerably more protein than do patients with
However, cases with a combined tubular and tubular disease. This reects the pathophysiology
glomerular proteinuria have been reported.37 The in which tubular dysfunction prevents absorption
etiology of the proteinuria in these conditions is
unclear. Within 2 days after such illness, most of
the proteinuria has resolved. Pre-eclampsia and
gestational hypertension are associated with rela-
tively large amounts of proteinuria that may
exceed 25 g/l. These patients may have glomerular,
glomerular with tubular leakage and even non-
selective glomerular patterns (see below).3841
OVERFLOW PROTEINURIA
When large amounts of relatively low molecular
weight proteins are present in the serum, they are Figure 7.3 The top urine sample has a tubular proteinuria
cleared through the glomeruli and overwhelm the pattern with a massive monoclonal free light chain (MFLC) in the
ability of the tubules to reabsorb them. Therefore, bg region. The middle urine sample contains a small amount of
this condition is termed overow proteinuria.1,42,43 albumin (A), but has been contaminated by the sample below. This
It is commonly seen in urine from patients with results in prominent staining in the albumin, a1- and b-regions only
in the lower portions of the second lane (indicated). The bottom
multiple myeloma, where large amounts of MFLC
sample is from a urine with considerable hemolysis and has a
are produced. However, other conditions can prominent hemoglobin band (arrow) that cannot be distinguished
result in the production of large amounts of from an MFLC without an immunoxation. (Paragon SPE2 system
protein that may mimic MFLC in the urine. stained with Paragon Violet.)
Glomerular proteinuria 223
1 1
2 2
3 3
4 4
5 5
6 6
7 7
SPE G BK BL FK FL
1 2 3 4 5 6
(a)
(b)
1 1
Figure 7.4 (a) Immunoxation of urine demonstrates a
2 2 prominent b-region band in the serum protein electrophoresis
3 3 (SPE) (acid-xed) lane. However, no immunoglobulins are
demonstrated by anti-IgG (G), anti-bound k (BK), anti-bound l
4 4 (BL), anti-free k (FK) or anti-free l (FL). (b) The densitometric
5 5 scan indicates the size of the b-region band. (c) Immunoxation
performed with antibodies against b2-microglobulin with the urine
6 6 at decreasing concentrations in lane 4 (note the prominent antigen
7 7 excess effect at this higher concentration of urine); lanes 5, and 6
demonstrate the identity of this band. Figure contributed by
SPE FK FL B2M B2M B2M
1 2 3 4 5 6 Beverly C. Handy,47 and used with permission from Archives of
(c) Pathology and Laboratory Medicine.
of the few small proteins that normally pass (Fig. 7.5).50 When sufcient quantities of protein are
through the glomerulus, whereas glomerular lost into the urine, the patient suffers from the
disease literally opens the oodgates to serum pro- nephrotic syndrome and displays the severe protein
teins of 50 kDa or more. loss pattern on serum protein electrophoresis (see
A wide variety of conditions result in sufcient Chapter 5).
damage of the glomerular capillary wall to permit All glomerular damage is not the same. The
large molecules to pass into the glomerular ltrate. amount of proteinuria and the size of the molecules
Diabetes mellitus and immunological renal dis- permitted to pass through the glomerular capillary
eases make up the vast majority of the cases with walls will depend on the degree of injury to the
glomerular proteinuria.2 When glomerular damage glomeruli. A useful measure of glomerular selectiv-
occurs, large molecules pass into the glomerular ity (the degree of damage) is the selectivity index.
ltrate along with smaller proteins. The smaller The concept of selectivity refers to the ability of
proteins and some of the larger molecular weight the glomerulus to permit only smaller proteins for
proteins can be reabsorbed by the renal tubules. passage through the glomerular basement mem-
However, the absorptive capacity of the tubules is brane. If very large molecules, such as IgG, IgM or
limited to approximately 10 g of protein over 24 h. a2-macroglobulin leak through the glomerulus in
Urine protein electrophoresis can distinguish similar proportions to smaller molecules such as
glomerular from tubular damage.48,49 In glomerular albumin or transferrin, then the proteinuria is
proteinuria patterns, large molecules such as albu- termed non-selective; that is, the glomerulus is not
min, a1-antitrypsin, and transferrin predominate able to discriminate between these two sizes. The
224 Examination of urine for proteinuria
Figure 7.5 Three examples of glomerular proteinuria. Note that the urine samples do not line up exactly because they were
electrophoresed on three separate runs. In all three, the main protein excreted is albumin with varying amounts of a1-antitrypsin,
transferrin and other globulins. The top and middle samples have moderate amounts of haptoglobin (H), indicating that the glomeruli have
lost some degree of selectivity. The discrete smaller bands which are seen in the a- and b-regions of patients with tubular proteinuria are
not seen. The middle sample has a small restriction around the origin that needs an immunoxation. (Top two urine samples concentrated
25-fold; bottom urine concentrated 100-fold. Panagel system stained with Coomassie Blue.)
greater the glomerular damage, the less selective is ulin they were able to distinguish cases with mini-
the proteinuria. The degree of glomerular damage mal change nephropathy and other transient
correlates with the level of the selectivity index conditions from those with crescentic glomeru-
(SI).51 lonephritis. The selectivity index based on IgG was
The SI is estimated from the ratio of the clearance not as useful because of its lower molecular weight.52
of one of these large molecules to that of albumin. The formula for the selectivity index using a specic
Several studies have examined the implications of high molecular weight (MW) analyte is:
the selectivity index as a marker for patients with a
variety of glomerular diseases.5254 Tencer et al.52,53 (high MW analyte ur/albumin ur)
Selectivity index =
reported that by using SI with IgM or a2-macroglob- (high MW analyte ser/albumin ser)
Tubular proteinuria 225
where: high MW ur = high molecular weight routine agarose or acetate electrophoresis and
analyte concentration in urine; albumin ur = albu- lacks a ready method for automation. A relatively
min concentration in urine; high MW ser = high simpler technique employing SDS-agarose gel elec-
molecular weight analyte concentration in serum; trophoresis to separate the urine proteins by mole-
and albumin ser = albumin concentration in serum. cular weight is now available for the same
The total amount of proteinuria increases with purpose.63 This technique has the advantage of dis-
decreasing selectivity of the glomerulus. Typically, pensing with the concentration step required for
mild glomerular damage (mainly albumin) has routine urine protein electrophoresis on agarose or
less than 1500 mg/24 h, moderate damage cellulose acetate. However, this technique is not as
(selective glomerular proteinuria) has from sensitive to detect MFLC as immunoxation on
15003000 mg/24 h and non-selective patterns concentrated urine.64
have greater than 3000 mg/24 h. When a random
urine sample has been obtained, the ratio of total
protein to creatinine for low-grade proteinuria TUBULAR PROTEINURIA
(mainly albumin) is 0.21.0, moderate proteinuria
(selective glomerular proteinuria) is from 1.0 to 5.0 When there is damage to the tubule reabsorptive
and values > 5.0 are seen in non-selective patterns function, small proteins in the glomerular ltrate
(nephrotic).6 In a case of minor glomerular damage (a1-acid glycoprotein, a1-microglobulin, b2-
where selectivity is maintained, albumin, a1-anti- microglobulin, g-trace protein, and retinol-binding
trypsin, occasionally a1-antichymotrypsin and protein) pass into the urine.65,66 Because of their low
transferrin are found on electrophoresis of concen- concentration in serum compared with other pro-
trated urine (Fig. 7.5). In the most severe cases of teins these small molecules are undetectable when
glomerular damage, selectivity is lost and the urine the serum is examined by electrophoretic tech-
protein electrophoresis resembles the normal serum niques used in clinical laboratories. However,
protein pattern. A non-selective pattern, then, will when they pass into the urine unencumbered by the
also display the large haptoglobin, a2-macroglobu- more plentiful larger molecules, they can be visual-
lin, and intact immunoglobulin molecules in addi- ized on concentrated specimens studied by elec-
tion to the molecules seen in selective glomerular trophoresis (Fig. 7.1).
proteinuria. Acute tubular necrosis, the most common cause of
Some laboratories use sodium dodecyl sulfate acute renal failure, can result from a wide variety of
polyacrylamide gel electrophoresis (SDS-PAGE) to injuries.6769 Some forms are reversible, such as
evaluate urine protein excretion.55 It delineates the those caused by transient ischemia, exposure to
molecular weight of proteins escaping from the heavy metals, toxins, or radiocontrast dyes.68,7074
damaged nephron, providing information about Chronic causes include congenital Fanconi syn-
the specic small proteins in the urine from drome, Dent disease (an X-linked condition caused
patients with tubular damage.5662 As such, it allows by inactivation of the renal chloride channel gene),
one to distinguish individuals that have combined and acquired Fanconi syndrome (the most common
selective glomerular and tubular proteinuria, from cause of which is myeloma kidney).75,76 Light chain
those that have non-selective proteinuria. Using Fanconi syndrome is often associated with intracy-
both SDS-PAGE and fractional excretion of a1- toplasmic crystals formed from MFLC in the prox-
microglobulin to divide glomerular selectivity into imal convoluted tubules,77 although some cases of
high (minimal damage), moderate or low selectiv- Fanconi syndrome caused by light chain deposition
ity, Bazzi et al.57 reported that the amount of do not form crystals, perhaps owing to the absence
glomerular selectivity correlates with nal outcome of side-chains in the CDR-L3 loop which are
and with response to therapy. However, SDS- needed for dimer formation.78
PAGE remains a more laborious technique than Patients with tubular disease have a normal
226 Examination of urine for proteinuria
glomerular selectivity because only small molecules amounts of proteinuria resulting from tubular
pass into the glomerular ltrate. These molecules bind disease (13 g/24 h), the urine must be concen-
to receptors on the proximal convoluted tubules. trated 50 for the protein to be visualized by elec-
After they are reabsorbed, the proteins are catabo- trophoresis. The albumin band is considerably
lized by the tubular epithelium into amino acids that fainter than that seen in glomerular proteinuria.
are reused by the body.79 When the tubules are dam- Typically, it will be similar in intensity, to slightly
aged, these small molecules pass into the urine. more intense, than the a1-microglobulin and b2-
The amount of proteinuria that results from microglobulin bands (Fig. 7.7).
tubular damage is only a fraction of that seen in When examining electrophoretic patterns from
glomerular disease. In tubular damage, traces of patients with tubular disease, one nds consider-
albumin accompany the smaller molecules into the able variation because the amount of damage
urine (Fig. 7.6). Because of the relatively small differs from one patient to another. However, a
Figure 7.6 Three examples of tubular proteinuria. Note again, that these three samples were performed on three different runs and
have slightly different migrations. With tubular proteinuria, the glomeruli are relatively intact and do not allow large amounts of albumin
and other large molecules to pass into the glomerular ltrate. Therefore, albumin is a relatively minor component, or slightly greater than
the many smaller molecules seen in the a- and b-regions. Note that the middle sample is from a patient with a k monoclonal free light
chain (indicated). This sample has two fused bands perhaps due to the formation of both monomer and dimer free monoclonal k chains.
(Urine concentrated 100-fold. Panagel system stained with Coomassie Blue.)
Tubular proteinuria 227
Figure 7.7 Three urine samples with different patterns for comparison on the same gel. The top sample shows a striking tubular
proteinuria pattern with an albumin band (A) that stains no darker than the multiple bands in the a-, b- and g-regions. The middle sample
is a predominantly glomerular pattern of proteinuria, but the two bands in the a2-region (indicated) are consistent with a coexisting
tubular leakage, because they correspond to the type of staining seen with the a2 microglobulins rather than the broad haptoglobin
pattern shown in Fig. 7.5. The bottom sample has barely discernable albumin and a2-region bands (just below the indicated bands in the
middle sample). The strong band which lines up with the transferrin region (arrow) is an monoclonal free light chain. It would be
inconsistent to have such a large transferrin band with such a minor tubular proteinuria and almost no albumin. (Urine concentrated 100-
fold. Paragon SPE2 system stained with Paragon Violet.)
consistent nding is that albumin is much less make this distinction in samples with nonselective
dominant than in glomerular proteinuria. pattern.
Individuals with chronic renal disease will often Rarely, one may see analbuminuria in cases of
have a combined glomerular and tubular pattern tubular proteinuria. Sun et al.80 reported such a
indicative of widespread damage to the nephron case in a diabetic, hypertensive patient with
(Fig. 7.8). In these patients, the albumin content tubular proteinuria. They hypothesized that the
resembles that of glomerular proteinuria, but the patient may have had end-stage kidney disease
smaller bands in the a- and b-region can also be such that the glomerular ltration rate was so low
observed. The occurrence of combined glomerular as to allow practically no protein through the
and tubular proteinuria may be obscured in urine glomeruli. Yet, the tubules may have leaked the
from individuals that have a non-selective smaller proteins, possibly as a result of analgesic
glomerular leakage pattern. In those individuals, abuse in that case (Fig. 7.9).
either the use of SDS-PAGE (see above) or mea- Quantication of low molecular weight proteins
surement of specic small molecular weight pro- may be used as an indication of tubular proteinuria.
teins (see below) may be needed to detect the This may be especially important as an assessment
coexistence of tubular dysfunction. Since SDS- of renal function in patients with multiple
PAGE provides a separation of the molecules on myeloma.81 The urinary protein most often used to
the basis of molecular weight, it can be useful to evaluate this is b2-microglobulin, a molecule that
228 Examination of urine for proteinuria
Figure 7.8 Two samples from patients with combined glomerular and tubular proteinuria. Whereas albumin predominates in both
samples, many discrete bands are present in the a- and b-regions. (Urine concentrated 25-fold. Panagel system stained with Coomassie
Blue.)
has homology with immunoglobulins.8286 b2- dysfunction may result in a false negative.87,88
Microglobulin migrates between transferrin and C3 a1-Microglobulin is another plasma glycoprotein
on electrophoresis. While it is elevated in tubular that passes freely into the urine. Because it is rela-
proteinuria, it is unstable at pH < 6.5 and, there- tively stable at the acid pH of urine some have sug-
fore, reliance solely upon quantication of urine b2- gested that it may serve as a better indicator for renal
microglobulin to detect tubular damage or tubular dysfunction than b2-microglobulin.8991
Figure 7.9 Urine (top and bottom lanes) and serum (middle lane) are from a patient with diabetes, hypertension, and renal impairment.
Despite the presence of considerable tubular proteinuria, an albumin band is not present in this patients urine. Figure contributed by
Tsieh Sun.80
Monoclonal free light chains 229
However, Donaldson found that although b2- proteins. This reects confusion that I have
microglobulin is more susceptible to denaturation observed when some physicians equate the pres-
under acidic conditions, other marker proteins, ence of a monoclonal intact molecule, such as IgG,
including a1-microglobulin, also deteriorate below in the urine with a Bence Jones protein. It is not.
pH 6.0 and suggested that routine alkalinization of Further, it does not have the same signicance for
urine upon voiding be employed to enhance detec- the patients. Monoclonal free light chains are asso-
tion of these proteins.88 In addition to these two mol- ciated with a greater likelihood that the patient has
ecules, other suggested markers for detecting a dire condition such as multiple myeloma or
tubular proteinuria include retinol-binding protein, amyloid AL than does an intact monoclonal
lysozyme, and N-acetyl-b D-glucosaminidase.65,83,92,93 immunoglobulin. They are more likely to damage
renal tubules and deposit in glomeruli, tubules and
other locations.
FACTITIOUS PROTEINURIA Monoclonal free light chains have been referred
to as Bence Jones proteins to recognize the contri-
The protein bands identied can also give informa- bution made by Henry Bence Jones. His descrip-
tion about factitious proteinuria.9497 For example, tion of the rst reported case of multiple myeloma
in cases reported by Tojo et al.95 and Sutcliffe et recorded the characteristic of MFLC to precipitate
al.,96 the electrophoresis of urine demonstrated when weakly acidied urine that contained them
unusual protein bands. The albumin migrated was warmed to 4058C, only to redissolve upon
almost in the a1-antitrypsin region and the other heating to 100C.98 He called the protein a
major bands also did not line up. The problem was hydrated deutoxide of albumen in that paper. Dr
species: immunoxation conrmed that the protein Thomas Watson referred the patient, a 44-year-old
was egg albumin. Therefore, be suspicious when man who complained of chest pains after a fall, to
protein bands do not migrate correctly; ask for a Dr William MacIntyre. Doctor MacIntyre studied
new sample under controlled collection conditions. the urine, noting its peculiar thermal characteris-
Identication of the abnormal band is the key to tics and sent the specimen to Dr Jones who
correct diagnosis of factitious proteinuria. reported his studies in 1848. Two years later,
However, before suggesting this diagnosis, recall MacIntyre reported the urine characteristics in
that denaturation also may produce unusual detail.99,100 Doctor John Dalrymple examined the
bands. Denaturation of proteins in urine may slides of the bone marrow from the autopsy on this
occur because of deterioration of a sample with patient. He noted a large number of nucleated cells
time, bacterial overgrowth, the presence of pro- that varied in size and shape, also noting irregular
teases from leukocytes, or laboratory errors such cells with two or three nuclei (likely the malignant
as the accidental addition of a biuret reagent prior plasma cells).101,102
to electrophoresis. When abnormal bands are seen Doctor Jones was a prominent physician and
that are not monoclonal gammopathies, hemoglo- during his lifetime was described by Florence
bin, myoglobin, lysozyme, eosinophil-derived neu- Nightingale as being the best chemical doctor in
rotoxin, or b2-microglobulin, a repeat fresh urine London.103 Another odd fact concerns the hyphen
sample is often the best approach. that is occasionally placed between his middle and
last names. Although one often sees his name
hyphenated, he did not hyphenate his name in any
MONOCLONAL FREE LIGHT of the 40 papers and books that he wrote.102
CHAINS Further, in his contemporary reference books, he
was listed under Jones.102 The hyphen was added
In this book, I am encouraging the use of the term by some of his descendants.102 I presume they were
MFLC instead of the traditional term, Bence Jones members of the Bence clan. The nal point of note
230 Examination of urine for proteinuria
is that the test he and Dr MacIntyre used was quite macroglobulinemia, amyloid AL, B-cell non-
insensitive and should never be used today to Hodgkin lymphoma and B-cell leukemia. These
detect MFLC. conditions may result in production of excessive
In the 1960s, the recognition that Bence Jones amounts of MFLC from the clone of plasma cells
proteins in the urine were homogeneous popula- (or lymphoplasmacytic cells in the case of lym-
tions of immunoglobulin fragments played an phoma, leukemia and Waldenstrm macroglobu-
important role in the delineation of antibody struc- linemia). The free light chain products may exist as
ture.104 Although the normal product of plasma monomers, dimers, tetramers, or light chain frag-
cells is an intact immunoglobulin, these cells also ments. Occasionally, they may bind to other serum
produce excess free light chains.11 This differs from components.21,106 Patients suspected of having
heavy chains that normally are secreted only as amyloid AL should have urine immunoxation to
part of intact immunoglobulin molecules. When identify MFLC.23
light chains are not present, the normal heavy In myeloma, when large amounts of the MFLC
chains remain in the endoplasmic reticulum and are present, they overwhelm the capacity of the
are degraded.9,10 Unlike normal heavy chains, proximal convoluted tubules and can be detected
however, normal light chains may be secreted as in the urine (overow proteinuria). It is important
part of the intact immunoglobulin molecule, or as to separate polyclonal free light chains from MFLC
free light chains.10 Most of the k free light chains for diagnostic purposes. The presence of MFLC in
are secreted as monomers (25 kDa), although urine samples often portends a poorer prognosis,
dimerization certainly also occurs, whereas most l as this protein is much more frequently seen in
light chains exist as dimers (50 kDa).10,105 Both association with myeloma (5060 per cent) and
monomeric and dimeric light chains pass through amyloidosis (60 per cent) than it is with benign
the glomerulus and are reabsorbed by the proximal monoclonal gammopathy (14 per cent) that today
convoluted tubules where they are catabolized. would be termed MGUS.107109
Because of their different sizes, however, free k
chains have a clearance rate of approximately three
times that of l light chains.105 RENAL DAMAGE CAUSED BY MFLC
When there is a polyclonal increase in
immunoglobulin production, such as occurs in The pathology resulting from such proliferation of
chronic inammation and some autoimmune di- plasma cells relates to the lesions created by the
seases, the amount of polyclonal free light chains plasma cells themselves, interference with the pro-
produced also increases and, together with other duction of normal bone marrow elements, and
low molecular weight proteins, they may surpass damage caused by the monoclonal immunoglobu-
the ability of renal tubules to reabsorb them. This lin products of the clone on various organ systems.
results in the presence of overow proteinuria of Because the free light chains and their fragments
polyclonal free light chains in the urine, which may readily pass through the glomerulus and are reab-
produce a ladder pattern on examination of the sorbed by the renal tubules, a major portion of the
urine by immunoxation (see below).27 pathology resides in the kidney.110 Monoclonal free
There are many disorders of plasma cell and lym- light chains may damage renal tubules directly or
phoplasmacytic cells where proliferation of a single indirectly by facilitating release of intracellular
clone occurs, termed plasma cell dyscrasias. The lysozymes.26,111 This results in dysfunction of the
variety of conditions parallels the broad spectrum proximal convoluted tubules. In addition, light
of their clinical signicance. At the one end of the chain cast nephropathy or deposits of light chains
spectrum is multiple myeloma and at the other end in the glomeruli or around the tubules as light
is monoclonal gammopathy of undetermined sig- chain or as amyloid AL also result in renal dys-
nicance (MGUS). In between are Waldenstrm function.106,110,112 Almost all patients who have
Detection and measurement of MFLC 231
> 1 g/24 h of MFLC in the urine suffer from renal region (CDR3) of both k and l MFLC binds to
tubular dysfunction.26,113 TammHorsfall proteins resulting in the formation
When the proximal convoluted tubules are of the dense cast material present in the distal
damaged, the patients suffer Fanconi syndrome.114 tubules and collecting ducts in myeloma
This manifests as a loss of amino acids, glucose, kidney.8,112,121,122 The presence of these casts in the
phosphate and bicarbonate in the urine.114116 The distal tubules is associated with dilatation of the
MFLC-mediated damage to the proximal convo- tubular lumen, atrophy of the tubule epithelium,
luted tubules of patients with multiple myeloma is and signicantly worsens the survival compared
the most common cause of acquired Fanconi syn- with patients with pure light chain deposition
drome.76 In some of these cases, crystals can be disease.106,110 In addition to the light chainTamm
found in the renal tubules (proximal tubules, distal Horsfall protein complex, casts contain mono-
tubules and collecting ducts).117,118 These patients cytes, lymphocytes, multinucleated giant cells,
can have a paradoxically low serum calcium due to occasionally tubular epithelium, and neutrophils.106
loss into the urine.119 After therapy for myeloma Unfortunately, these consequences may recur in
and decline of MFLC, some cases of Fanconi syn- patients that receive renal transplantation. Short et
drome have shown improvement in tubular func- al.123 suggested that the histological pattern of
tion.120 damage in the patients kidneys helps to predict the
There have been several studies indicating the outcome of subsequent renal allograft. Individuals
potential toxic effects of MFLC on kidney tubules. that had light chain deposition disease with a pro-
Batuman et al.116 used MFLC puried from patients liferative glomerulonephritis in their native kidneys
with multiple myeloma to study their effect on had worse graft survival than those with cast
transport of phosphate and glucose by cultured rat nephropathy.123 The presence of light chain
proximal tubule cells. They reported that both k nephropathy is a serious complication that should
and l MFLC were able to inhibit the uptake of cause the clinician to consider the implementation
these analytes in a dose-dependent manner, of early dialysis.124
whereas albumin had no such effect. This group Deposits of the MFLC as either brillar amyloid
later demonstrated that NaK-ATPase activity of AL or non-brillar light chain deposition disease
primary cell cultures from rat proximal convoluted (LCDD) may occur in glomeruli or around tubules
tubule cells also was inhibited by monoclonal free in extracellular spaces.106,110 The same distribution
l light chain.114 NaK-ATPase is an important part is found in cases of heavy chain deposition disease
of the physiological mechanism for the sodium (HCDD). Amyloid AL deposits are irregular in
gradient in these cells. Further, Pote et al.112 used their distribution, Congo Red-positive and display
cultures of human proximal convoluted tubules to the characteristic apple-green birefringence under
demonstrate a direct toxic effect of MFLC on the conditions of polarized light (see Chapter 6). In
ability of these cells to proliferate as well as induc- contrast, LCDD deposits usually follow basement
ing apoptosis within 2 days of exposure to the light membranes, are not brillar and are Congo Red-
chains in vitro. These studies indicate that a direct negative.106
toxic effect by the MFLC on proximal convoluted
tubules may explain the occurrence of Fanconi syn-
drome in some patients with plasma cell DETECTION AND MEASUREMENT
dyscrasias. OF MFLC
Both the distal tubules and collecting ducts are
involved in myeloma cast nephropathy. They Guidelines for clinical and laboratory evaluation of
contain prominent renal casts and renal function patients with monoclonal gammopathy recommend
suffers from loss of concentrating ability and acid- that detection of MFLC is best achieved by
ication.112 The third complementarity-determining immunoxation of concentrated urine.125,126 The
232 Examination of urine for proteinuria
measurement of MFLC once demonstrated by been concentrated naturally and will provide excel-
immunoxation currently requires quantication of lent material for study. Note, however, that if a
the MFLC in a 24-h collection of urine.125,126 patient has a random sample taken at another
However, these guidelines were written before sen- time, a positive result is useful, but a negative result
sitive and specic automated assays for detecting free will not rule out MFLC. Indeed, even a negative
light chains (FLC) in serum were readily available. result on an early morning void may engender a
The availability of these assays (see below) makes it repeat analysis on a second sample from a clinician
possible to follow MFLC by convenient and consis- that has a high index of suspicion. Alternatively,
tent serum samples. Because most laboratories are the laboratory now may suggest the use of FLC
still using the urine electrophoresis and immunox- assays on serum in cases with a high index of sus-
ation tests at the time of this writing, I will review picion (see below). The new automated serum FLC
evaluation of urine in some detail. However, all lab- assays have been reported to detect abnormal k/l
oratories should be following the emerging literature ratios in a large number of patients that had been
on automated measurement of FLC in serum and thought to have non-secretory myeloma.129
consider incorporating them into their evaluation of After the initial detection of a MFLC on the early
patients with monoclonal gammopathies. morning urine, I recommend a 24-h sample to quan-
tify the amount of MFLC. This is especially impor-
tant in patients with light chain disease, as the
amount of urine MFLC is a key indicator of tumor
DETECTION OF MFLC IN THE burden and response to therapy. Once patients have
URINE AND SERUM BY been informed that a monoclonal protein is present
ELECTROPHORESIS AND in their urine, they should be sufciently motivated
IMMUNOFIXATION to provide a reliable 24-h collection. Once again,
the new automated serum FLC assays (discussed
There has been some controversy regarding the below) have been shown to provide good informa-
optimum specimen to use for detection of MFLC. tion to follow patients with MFLC and may be more
Although many authorities recommend a 24-h convenient and perhaps more accurate than the
urine collection for the initial detection of urine assays.26 Recently, Salomo et al.130 reported the
MFLC,125127 others accept random urine samples use of high-resolution Sebia Hydragel HR (Sebia,
for analysis.128 Brigden and coworkers128 reported Issy-les-Moulineaux, France) agarose electrophore-
that an early morning voided sample was as good sis on unconcentrated urine to measure the amount
as, and perhaps superior to, a 24-h sample. They of MFLC. They found that they could estimate the
noted, however, that random samples collected at concentration of MFLC in urine by comparing the
times other than the early morning void were densitometric scans of staining intensities of the
clearly inferior for the initial detection of MFLC MFLC bands relative to the staining intensities of
(Table 7.1). The early morning voided sample has albumin solutions.
Table 7.1 Detection of monoclonal free light chains in urine samples from patients with multiple myelomaa
24-h Sample + Early morning - Early morning + Random sample - Random sample
Positive = 17 17 0 14 3
Negative = 3 2 1 0 3
a
Data from Brigden et al.128
Detection of MFLC in the urine and serum by electrophoresis and immunoxation 233
For initial detection of the MFLC in urine, we is not necessary to use antisera against FLC (Fig.
rst review our les on that patient to determine if 7.11). In instances where the intact immunoglobu-
we have a previous serum protein electrophoresis lin and the light chain have both the same migra-
or immunoxation study demonstrating an M- tion and the same concentration, it is not possible
protein. If so, we use that information to determine to rule out MFLC by this method. In those cases,
the set-up for the immunoxation. For example, if laboratories may wish to perform immunoxation
the patient is known to have an IgGk M-protein, with antisera specic to FLC.21 We prefer the anti-
we perform immunoxation for IgG, k and l on sera to total light chains because we have found
the initial urine study. If there is no history, we them to be stronger than the anti-FLC antisera and
perform electrophoresis on concentrated urine and more sensitive in detecting small MFLC. Some of
immunoxation for k and l. I prefer to use antisera my evaluations of antisera against FLC have found
against total (bound and free) k and against total l their specicity to be disappointing.
for these studies rather than antisera that reacts If no suspicious bands are present, the sample is
solely with FLC. An MFLC will usually have dif- reported to be negative for MFLC. However, if one
ferent electrophoretic migration and/or concentra- or more M-protein bands are seen, the analysis is
tion from the intact immunoglobulin molecule. repeated with antisera against the most common
Therefore, as shown in Fig. 7.10, we report the heavy chains found in the urine (i.e. IgG and IgA).
MFLC when the light chain band has a distinctly If this is a 24-h urine sample, we perform a total
different electrophoretic migration than the heavy protein measurement (see above) and use a densito-
chain band. In cases where the intact immunoglob- metric scan of the known location of the MFLC
ulin and the light chain overlap in migration, they (correlating it with the immunoxation) to quantify
typically differ sufciently in concentration that it the 24-h MFLC content of the urine (Fig. 7.12). If
useful for the clinician. Hamilton et al.156 point out factitious proteinuria (see above) may produce pecu-
that myoglobin usually appears as a light brown color liar bands (usually in the albumin and a1-region).
whereas hemoglobin is usually red, especially in a A new source of unusual banding in the urine is
fresh sample. However, aged samples, unstable found in samples from patients with pancreas and
hemoglobin and methemoglobin will also appear a pancreasrenal transplants. Since 1987, the pre-
brownish color.156 The best solution to this issue is to ferred drainage of their exocrine pancreas secretions
perform an immunoassay to detect the myoglobin. is into the urinary bladder.161 It is well documented
As mentioned above, b2-microglobulin, eosinophil- that their urine contains discrete forms of the pan-
derived neurotoxin, and lysozyme, when present in creatic enzymes.162164 Indeed, evaluation of these
sufcient quantity, will produce a band in protein enzymes has been suggested as a means to follow
electrophoresis that could be mistaken for rejection in these patients.162164 Nonetheless, the
MFLC.45,46,47 Proteins added to the urine in cases of bands may be confusing to laboratories. I have seen
(a)
(b)
False negative MFLC in urine by electrophoresis 237
Figure 7.14 (a) A urine concentrated 100-fold (top) and a serum
diluted 1:3 from the same patient (bottom) are shown. The serum
shows a hypogammaglobulinemia and the urine has a suspicious
band in the fast g-region (arrow). (Paragon SPE2 system stained
with Paragon Violet.) (b) Immunoxation of the serum from (a)
shows a small IgA k monoclonal gammopathy (arrows) which had
been hidden in the b-region beneath the prominent b1-lipoprotein
and transferrin bands. In addition, there is a small, but suspicious
band in the k region (S). Dilutions of the serum used for
immunoxation of serum are shown below the immunoglobulin
antisera used. SPE, serum protein electrophoresis for comparison.
(Beckman Paragon system stained with Paragon Violet.) (c)
Immunoxation of 100-fold concentrated urine of the urine from
(a) shows a massive k monoclonal free light chain protein band, no
IgA band and a tiny l-band (indicated) thought to be part of the
ladder pattern. (Beckman Paragon system stained with Paragon
Violet; anode at the top.)
IgA K L
(c)
two such cases (Fig. 7.16), one of which has been against heavy chains to rule out the possibility of
recently documented as being due to pancreatic heavy chain disease. This should give a distinct
enzymes by Song et al.165 Although this is a some- band that matches the one in the serum. In per-
what unusual source of confusion it re-emphasizes forming immunoxation for heavy chains in urine,
the two key things that the laboratorian should do however, one must be aware that the a2-region of
when confronted by unusual bands that may be urine may demonstrate the presence of fragments
MFLC in urine: perform the immunoxation for k of free polyclonal g-chain that may be present in
and l, then if they are negative call the clinician to urine samples (Fig. 7.10).167 The presence of these
discuss the case. fragments is not uncommon in urine, although the
There is always the rare possibility of a false neg- exact cause is unclear. The fragments may result
ative immunoxation due to an antigen excess from breakdown of IgG by many factors such as
effect (see Chapter 3). When a large band seen on proteases or bacterial enzymes.
urine protein electrophoresis fails to yield a reac-
tion with k and l (yet is not consistent with the
sources mentioned above); one may wish to con- FALSE NEGATIVE MFLC IN URINE
sider a repeat immunoxation on 10- and 100-fold BY ELECTROPHORESIS
diluted urine to help rule out antigen excess prob-
lems.166 It is unusual that these dilutions yield a Roach et al.168 provided an excellent example of a
result that was missed with the original sample. potential source of false-negative urine for MFLC.
Antigen excess effects are usually easy to recognize. They presented a pattern in urine protein
One should also consider including antisera electrophoresis that resembled the ladder pattern
238 Examination of urine for proteinuria
Figure 7.17 Urine protein electrophoresis on a sample concentrated 50-fold. The g-region shows numerous closely spaced bands that
resemble a ladder pattern on immunoxation. Immunoelectrophoresis demonstrated that this was a case of l monoclonal free light chain.
When unusual patterns are seen on urine protein electrophoresis, immunological studies are needed. Case contributed by Drs Adrian O.
Vladutiu and Barbara M. Roach.
False positive MFLC in urine by immunoxation 239
Purify A
light B
A chains E Immunize
C
B
E D
C
D
Intact immunoglobulin
Serum
A
B Anti-A
E
C Anti-B
D Anti-C
Absorb with excess Anti-D
Anti-E
Anti-E Residual
anti-C
weak
anti-E
Theory Reality
Figure 7.23 Commercial polyclonal sera with reactivities to free light chain can be created because some light chain antigenic
determinants are hidden in intact molecules. In the example shown, determinant E in the intact molecule is not available to react with
antisera. When the light chains are separated from the heavy chains, this determinant is now expressed along with the many antigenic
determinants (A, B, C, D) which are also expressed in the intact molecule. When these light chains are used to immunize an animal,
antibodies against all of these determinants can result. By absorbing these antisera with intact molecules, in theory, only the antisera against
free light chain determinants will remain. However, these antisera are often very weak, and often crossreact with intact molecules. This is
why recent methods required further purication on Sepharose 4B columns coated with puried light chain.
Monoclonal antibodies have also been produced sents overwhelmingly intact immunoglobulins that
against the antigenic determinants of free light contain k or l light chains respectively) is 2:1.
chains that are hidden when bound to heavy However, the most recent studies of serum using
chains.173,174,178 This provides highly specic anti- highly specic polyclonal or monoclonal antibod-
sera; however, monoclonal proteins may not react ies that are able to distinguish bound from
with the entire spectrum of light chains produced. unbound light chains have demonstrated a free
Eventually, reagents composed of cocktails of k/free l ratio of approximately 1:2.105,174 Abe et
monoclonal antibodies may provide the breadth al.,174 who used monoclonal antibodies in an
and specicity of reaction required for optimal enzyme-linked immunosorbent assay hypothesized
results. However, at present, the highly puried that the disparity in free versus intact light chain
polyclonal products seem to have an edge. ratios may be due to a higher rate of production by
The serum ratio of total k to total l (which repre- l plasma cells than k plasma cells. They further
244 Examination of urine for proteinuria
Abe et al.174 16.6 6.1 33.8 4.8 2.96 1.84 1.07 0.69
105
Bradwell et al. 8.4 2.66 14.5 4.4 5.5 4.95 3.17 3.3
a
Table modied from Bradwell et al., Table 2, with permission.105
suggested that the quaternary structural differences reects the molecular sieve properties of the
between k and l free light chains also may play a glomerulus to favor passage of the smaller k mole-
role. Bradwell et al.,105 who used polyclonal anti- cules into the urine.
bodies in an automated immunoassay suggested Bradwell et al. reported a free light chain k/l
that since k chains exist predominately as ratio in serum to be 1:1.62 with a 95 per cent
monomers (25 kDa) they will be cleared more condence interval of 1:2.751:0.99.105 This free
quickly through the glomeruli than the mainly light chain k/l ratio in serum provided consistent
dimeric l free light chains (50 kDa). The normal discrimination between individuals with myeloma
serum and urine free k and l concentrations are and those with polyclonal increases in
noted in Table 7.2. Whereas the serum free light immunoglobulins.26,105 Sera from patients with
chain ratios are the opposite of serum intact light multiple myeloma or Waldenstrm macroglobu-
chain ratios, in urine, the free k/free l ratios in the linemia contain increased concentrations of the
two studies are 2.8:1 and 1.7:1; both similar to the free light chain type associated with the M-protein
total k/total l ratio usually quoted for serum. This in all 27 cases examined (Fig. 7.24).105 They also
10 000
Free light chain concentration in mg/l
1000
100
10
1
Kappa Lambda Kappa Lambda Kappa Lambda Kappa Lambda
Patients 24-h urine M-protein Serum k FLC Serum l FLC Serum k/l
g/24 h (normal 211.2 mg/l) (normal 6.825.2 mg/l) (normal 0.090.89)
k Patients
1 0.33 22.6 12.3 1.837
2 2.25 4370 10.1 432.673
3 1.77 699 14.3 48.881
4 1.36 1390 5.6 248.214
5 1.81 693 1.2 577.500
6 0.10 276 1.1 250.909
7 0.03 2820 8.1 348.148
8 0.30 1230 9.8 125.510
9 1.31 351 5.1 68.824
l Patients
10 0.29 10.4 235 0.044
11 1.53 1.0 1030 0.001
12 3.04 3.3 15 900 0.000
13 6.32 4.8 116 0.041
14 3.31 7.0 11 000 0.001
15 0.11 1.1 46.5 0.024
16 4.11 0.9 545 0.002
17 5.02 6.8 1390 0.005
18 1.23 7.2 1430 0.005
19 4.20 9.2 4690 0.002
20 1.46 3.1 71.8 0.043
21 7.14 1.2 65.4 0.018
22 10.81 6.9 2700 0.003
23 4.01 16.9 2360 0.007
24 6.48 10.5 1420 0.007
25 0.79 46.7 10 000 0.005
26 0.10 5.9 81.6 0.072
27 2.99 1.00 326 0.003
28 5.58 4.5 3570 0.001
a
Table modied from Abraham et al., Table 1, with permission.26
246 Examination of urine for proteinuria
investigated sera from 12 patients with systemic trophoresis. They detected increased concentra-
lupus erythematosus as an example of the effect of tions of one free light chain along with an abnor-
polyclonal proliferation of immunoglobulins on mal free k/l ratio in sera from 19 of the 28 patients
the k/l ratio. While the level of circulating free with the diagnosis of non-secretory multiple
light chain was increased for both k and l, the myeloma. Of course, these assays do not prove the
free k/free l ratios were in the reference range existence of a monoclonal protein. A polyclonal
they established.105 As would be expected in situa- increase in one or the other light chain type could
tions of decreased renal clearance, polyclonal free result in an abnormal free k/l ratio. However,
light chains also occur in patients receiving Drayson et al.s nding that the clinical changes
chronic hemodialysis.179 during follow-up of six patients correlated with the
In their studies, Bradwell et al.105 reported two changes in the free light chain concentration during
patients with a negative urine study for MFLC that time period suggest that this assay will be of
using radial immunodiffusion (RID) screening with use in many patients with non-secretory myeloma
a sensitivity of 40 mg/l, where the serum nephelo- (a name we may need to change pauci-secre-
metric immunoassay demonstrated increased free tory?).129
light chains of the monoclonal type. The technique of automated immunoassays for
Abraham et al.26 compared the urine and serum free light chain measurement in serum and urine is
levels for free k and free l light chains using the evolving rapidly. Currently, reagents are available
same nephelometric immunoassay with a normal from The Binding Site, Ltd, and assays have been
free k/l of 0.090.89.105 Their data is shown in performed on the Beckman IMMAGE and on the
Table 7.3. The selected population all had mono- Dade Behring BNII. The reader is encouraged to
clonal free light chains in the urine, but only three seek the most recent information about this tech-
of nine patients with k-secreting monoclonal nique.
gammopathies and 13 of 19 patients with l-secret-
ing monoclonal gammopathies had a monoclonal
peak on the serum protein electrophoresis. REFERENCES
Nonetheless, all of the patients had abnormal k/l
ratios (Table 7.3). As a control group, they evalu- 1. Waller KV, Ward KM, Mahan JD, Wismatt DK.
ated seven patients with lupus glomerulonephritis. Current concepts in proteinuria. Clin Chem
None of the seven patients had an abnormal k/l 1989;35:755765.
ratio, although one had an elevation of the free l 2. Kaysen GA, Myers BD, Couser WG, Rabkin R,
light chains in serum. This indicates that serum Felts JM. Mechanisms and consequences of
measurements of free k and free l light chains by proteinuria. Lab Invest 1986;54:479498.
automated immunoassays may provide a viable 3. Cooper EH. Proteinuria. Am Assn Clin Chem
alternative to the current 24-h urine collections to Specic Protein Analysis 1984;1:111.
follow the clinical response of patients with multi- 4. Ritz E, Nowicki M, Fliser D, Horner D, Klimm
ple myeloma.26 HP. Proteinuria and hypertension. Kidney Int
The serum free light chain assays may be of con- Suppl 1994;47:S7680.
siderable assistance with another difcult problem, 5. Guder WG, Hofmann W. Markers for the
the non-secretory cases of multiple myeloma. diagnosis and monitoring of renal tubular
Drayson et al.129 have reported the utility of auto- lesions. Clin Nephrol 1992;38:S37.
mated immunoassays for free light chains in serum 6. Wallach J. Interpretation of diagnostic tests.
from patients with the diagnosis of non-secretory Philadelphia: Lippincott, Williams & Wilkins,
multiple myeloma where no monoclonal protein in 2000.
either serum or urine could be demonstrated by 7. Wieme R. Agar gel electrophoresis. Amsterdam:
traditional methods of immunoxation and elec- Elsevier, 1965.
References 247
8. Ying WZ, Sanders PW. Mapping the binding 19. Cesati R, Dolci A. Simple gold overstaining
domain of immunoglobulin light chains for enhances sensitivity of automated electrophoresis
TammHorsfall protein. Am J Pathol 2001;158: of unconcentrated urine. Clin Chem 1996;42:
18591866. 12931294.
9. Bole DG, Hendershot LM, Kearney JF. 20. Matsuda K, Hiratsuka N, Koyama T, et al.
Posttranslational association of immunoglobulin Sensitive method for detection and
heavy chain binding protein with nascent heavy semiquantication of Bence Jones protein by
chains in nonsecreting and secreting hybridomas. cellulose acetate membrane electrophoresis using
J Cell Biol 1986;102:15581566. colloidal silver staining. Clin Chem 2001;47:
10. Dul JL, Aviel S, Melnick J, Argon Y. Ig light 763766.
chains are secreted predominantly as monomers. 21. Levinson SS, Keren DF. Free light chains of
J Immunol 1996;157:29692975. immunoglobulins: clinical laboratory analysis.
11. Shapiro AL, Scharff MD, Maizel JV, Uhr JW. Clin Chem 1994;40:18691878.
Synthesis of excess light chains of gamma 22. Pezzoli A, Pascali E. Urine collection for the
globulin by rabbit lymph node cells. Nature detection of Bence Jones proteinuria. Am J Clin
1966;211:243245. Pathol 1991;95:266268.
12. Kyle RA. Diagnosis and management of multiple 23. Gertz MA, Lacy MQ, Dispenzieri A.
myeloma and related disorders. Prog Hematol Immunoglobulin light chain amyloidosis and the
1986;14:257282. kidney. Kidney Int 2002;61:19.
13. Kaplan IV, Levinson SS. Misleading urinary 24. Hidaka H, Ikeda K, Oshima T, Ohtani H, Suzuki
protein pattern in a patient with H, Takasaka T. A case of extramedullary
hypogammaglobulinemia: effects of mechanical plasmacytoma arising from the nasal septum.
concentration of urine. Clin Chem 1999;45: J Laryngol Otol 2000;114:5355.
417419. 25. Lau CF, Fok KO, Hui PK, et al. Intestinal
14. Rice EW. Improved biuret procedure for obstruction and gastrointestinal bleeding due to
routine determination of urinary total proteins systemic amyloidosis in a woman with occult
in clinical proteinuria. Clin Chem 1975;21: plasma cell dyscrasia. Eur J Gastroenterol
398401. Hepatol 1999;11:681685.
15. Shahangian S, Brown PI, Ash KO. 26. Abraham RS, Clark RJ, Bryant SC, et al.
Turbidimetric measurement of total urinary Correlation of serum immunoglobulin free light
proteins: a revised method. Am J Clin Pathol chain quantication with urinary Bence Jones
1984;81:651654. protein in light chain myeloma. Clin Chem
16. Pesce MA, Strande CS. A new micromethod for 2002;48:655657.
determination of protein in cerebrospinal uid 27. Bailey EM, McDermott TJ, Bloch KJ. The
and urine. Clin Chem 1973;19:12651267. urinary light-chain ladder pattern. A product of
17. Marshall T, Williams KM. Total protein improved methodology that may complicate the
determination in urine: elimination of a recognition of Bence Jones proteinuria. Arch
differential response between the Coomassie blue Pathol Lab Med 1993;117:707710.
and pyrogallol red protein dye-binding assays. 28. Newman DJ, Pugia MJ, Lott JA, Wallace JF,
Clin Chem 2000;46:392398. Hiar AM. Urinary protein and albumin excretion
18. Lefevre G, Bloch S, Le Bricon T, Billier S, Arien corrected by creatinine and specic gravity. Clin
S, Capeau J. Inuence of protein composition on Chim Acta 2000;294:139155.
total urinary protein determined by 29. Ambuhl PA, Muller V, Binswanger U. Transient
pyrocatechol-violet (UPRO vitros) and pyrogallol proteinuria after infusion of a gelatin plasma
red dye binding methods. J Clin Lab Anal volume expander. Clin Nephrol 1999;52:
2001;15:4042. 399400.
248 Examination of urine for proteinuria
30. Ito S, Ueno M, Izumi T, Arakawa M. Induction 43. Voiculescu M. Advances in the study of
of transient proteinuria, hematuria, and proteinurias. I. The pathogenetic mechanisms.
glucosuria by ethanol consumption in Japanese Med Interne 1990;28:265277.
alcoholics. Nephron 1999;82:246253. 44. Lillehoj EP, Poulik MD. Normal and abnormal
31. Lipatov IS, Kupaev IA, Kozupitsa GS. Pregnancy aspects of proteinuria. Part I: Mechanisms,
outcomes in women with a pathological weight characteristics and analyses of urinary protein.
gain, vascular dysfunction, transitory edemas and Part II: clinical considerations. Exp Pathol
transient proteinuria. Akush Ginekol 1995;6: 1986;29:128.
1618. 45. Tardy F, Bulle C, Prin L, Cordier JF, Deviller P.
32. Nicot GS, Merle LJ, Charmes JP, et al. Transient High concentrations of eosinophil-derived
glomerular proteinuria, enzymuria, and neurotoxin in patients urine mimic lysozyme far-
nephrotoxic reaction induced by radiocontrast cathodic bands in agarose gel electrophoresis.
media. JAMA 1984;252:24322434. Clin Chem 1993;39:919920.
33. Reuben DB, Wachtel TJ, Brown PC, Driscoll JL. 46. Sexton C, Buss D, Powell B, OConnor M, Rainer
Transient proteinuria in emergency medical R, Woodruff R, Cruz J, Pettenati M, Rao PN,
admissions. N Engl J Med 1982;306: Case LD. Usefulness and limitations of serum and
10311033. urine lysozyme levels in the classication of acute
34. Alpert HC, Lohavichan C, Presser JI, Papper S. myeloid leukemia: an analysis of 208 cases. Leuk
Febrile proteinuria. South Med J 1974;67: Res 1996;20:467472.
552554. 47. Handy BC. Urinary beta2-microglobulin
35. Hemmingsen L, Skaarup P. Urinary excretion of masquerading as a Bence Jones protein. Arch
10 plasma proteins in patients with febrile Pathol Lab Med 2001;125:555557.
diseases. Acta Med Scand 1977;201:359364. 48. Bottini PV, Ribeiro Alves MA, Garlipp CR.
36. Mori M. Febrile proteinuria. Rinsho Byori Electrophoretic pattern of concentrated urine:
1979;(Suppl 36):8593. Comparison between 24-hour collection and
37. Jensen H, Henriksen K. Proteinuria in non-renal random samples. Am J Kidney Dis. 2002;39:E2.
infectious diseases. Acta Med Scand 1974;196: 49. Levinson SS. Urine protein electrophoresis and
7582. immunoxation electrophoresis supplement one
38. Derer K, Hauser C, Endler M, Nowotny C, another in characterizing proteinuria. Ann Clin
Lapin A, Balcke P. Proteinuria in normal Lab Sci 2000;30:7984.
pregnancy and in EPH gestosis. Acta Med 50. Killingsworth LM, Tyllia MM. Protein analysis
Austriaca. 1989;16:1318. nding clues to disease in urine. Diagn Med
39. Kaltenbach FJ, Wilhelm C. Selectivity of 1980;MayJune:6975.
glomerular proteinuria and liver function in 51. Myers B. In vivo evaluation of glomerular
gestosis. Zentralbl Gynakol 1994;116: permoselectivity in normal and nephrotic man.
340343. In: M A, ed. Proteinuria. New York: Plenum,
40. Sifuentes Alvarez A. Electrophoretic prole of 1985.
proteinuria in normal pregnancy and in 52. Tencer J, Bakoush O, Torffvit O. Diagnostic and
gestational hypertension. Ginecol Obstet Mex prognostic signicance of proteinuria selectivity
1995;63:147151. index in glomerular diseases. Clin Chim Acta
41. Dendorfer U, Anders HJ, Schlondorff D. Urine 2000;297:7383.
diagnosis: proteinuria. Dtsch Med Wochenschr 53. Tencer J, Torffvit O, Thysell H, Rippe B, Grubb
2001;126:13101313. A. Proteinuria selectivity index based upon alpha
42. Carroll MF, Temte JL. Proteinuria in adults: a 2-macroglobulin or IgM is superior to the IgG
diagnostic approach. Am Fam Physician based index in differentiating glomerular diseases.
2000;62:13331340. Technical note. Kidney Int 1998;54:20982105.
References 249
54. Kouri T, Harmoinen A, Laurila K, Ala-Houhala 64. Keren DF. Detection and characterization of
I, Koivula T, Pasternack A. Reference intervals monoclonal components in serum and urine. Clin
for the markers of proteinuria with a Chem 1998;44:11431145.
standardized bed-rest collection of urine. Clin 65. Bang LE, Holm J, Svendsen TL. Retinol-binding
Chem Lab Med 2001;39:418425. protein and transferrin in urine. New markers of
55. Marshall T, Williams KM. Electrophoretic renal function in essential hypertension and
analysis of Bence Jones proteinuria. white coat hypertension? Am J Hypertens
Electrophoresis 1999;20:13071324. 1996;9:10241028.
56. Bazzi C, Petrini C, Rizza V, et al. 66. Ikeda M, Moon CS, Zhang ZW, et al. Urinary
Characterization of proteinuria in primary alpha1-microglobulin, beta2-microglobulin, and
glomerulonephritis. SDS-PAGE patterns: clinical retinol-binding protein levels in general
signicance and prognostic value of low populations in Japan with references to cadmium
molecular weight (tubular) proteins. Am J in urine, blood, and 24-hour food duplicates.
Kidney Dis 1997;29:2735. Environ Res 1995;70:3546.
57. Bazzi C, Petrini C, Rizza V, Arrigo G, DAmico 67. Stark J. Acute renal failure. Focus on advances in
G. A modern approach to selectivity of acute tubular necrosis. Crit Care Nurs Clin
proteinuria and tubulointerstitial damage in North Am 1998;10:159170.
nephrotic syndrome. Kidney Int 2000;58: 68. Lameire N, Vanholder R. Pathophysiologic
17321741. features and prevention of human and
58. Bruning T, Thier R, Mann H, et al. Pathological experimental acute tubular necrosis. J Am Soc
excretion patterns of urinary proteins in miners Nephrol 2001;12(Suppl 17):S2032.
highly exposed to dinitrotoluene. J Occup 69. Weisberg LS, Allgren RL, Genter FC, Kurnik BR.
Environ Med 2001;43:610615. Cause of acute tubular necrosis affects its
59. Ikonomov V, Melzer H, Nenov V, Stoicheva A, prognosis. The Auriculin Anaritide Acute Renal
Stiller S, Mann H. Importance of sodium dodecyl Failure Study Group. Arch Intern Med 1997;157:
sulfate pore-graduated polyacrylamide gel 18331838.
electrophoresis in the differential diagnostic of 70. Spoto S, Galluzzo S, De Galasso L, Zobel B,
Balkan nephropathy. Artif Organs 1999;23: Navajas MF. Prevention of acute tubular necrosis
7580. caused by the administration of non-ionic
60. Koliakos G, Papachristou F, Papadopoulou M, radiologic contrast media. Clin Ter 2000;151:
Trachana V, Gaitatzi M, Sotiriou I. 323327.
Electrophoretic analysis of urinary proteins in 71. Dussol B, Reynaud-Gaubert M, Saingra Y,
diabetic adolescents. J Clin Lab Anal 2001;15: Daniel L, Berland Y. Acute tubular necrosis
178183. induced by high level of cyclosporine A in a lung
61. Schreiber S, Hamling J, Zehnter E, et al. Renal transplant. Transplantation 2000;70:12341236.
tubular dysfunction in patients with 72. Weisberg LS, Allgren RL, Kurnik BR. Acute
inammatory bowel disease treated with tubular necrosis in patients with diabetes
aminosalicylate. Gut 1997;40:761766. mellitus. Am J Kidney Dis. 1999;34:10101015.
62. Woo KT, Lau YK. Pattern of proteinuria in 73. Yanagisawa H, Nodera M, Wada O. Inducible
tubular injury and glomerular hyperltration. nitric oxide synthase expression in mercury
Ann Acad Med Singapore 1997;26:465470. chloride-induced acute tubular necrosis. Ind
63. Le Bricon T, Erlich D, Bengoufa D, Dussaucy M, Health 1998;36:324330.
Garnier JP, Bousquet B. Sodium dodecyl 74. Lo RS, Chan JC, Cockram CS, Lai FM. Acute
sulfateagarose gel electrophoresis of urinary tubular necrosis following endosulphan
proteins: application to multiple myeloma. Clin insecticide poisoning. J Toxicol Clin Toxicol
Chem 1998;44:11911197. 1995;33:6769.
250 Examination of urine for proteinuria
75. Igarashi T, Inatomi J, Ohara T, Kuwahara T, 86. Valles P, Peralta M, Carrizo L, et al. Follow-up
Shimadzu M, Thakker RV. Clinical and genetic of steroid-resistant nephrotic syndrome: tubular
studies of CLCN5 mutations in Japanese families proteinuria and enzymuria. Pediatr Nephrol
with Dents disease. Kidney Int 2000;58: 2000;15:252258.
520527. 87. Blumsohn A, Morris BW, Grifths H, Ramsey
76. Minemura K, Ichikawa K, Itoh N, et al. IgA-kappa CF. Stability of beta 2-microglobulin and retinol
type multiple myeloma affecting proximal and binding protein at different values of pH and
distal renal tubules. Intern Med 2001;40:931935. temperature in normal and pathological urine.
77. Markowitz GS, Flis RS, Kambham N, DAgati Clin Chim Acta 1991;195:133137.
VD. Fanconi syndrome with free kappa light 88. Donaldson MD, Chambers RE, Woolridge MW,
chains in the urine. Am J Kidney Dis 2000;35: Whicher JT. Stability of alpha 1-microglobulin,
777781. beta 2-microglobulin and retinol binding protein
78. Deret S, Denoroy L, Lamarine M, et al. Kappa in urine. Clin Chim Acta 1989;179:7377.
light chain-associated Fanconis syndrome: 89. Yu H, Yanagisawa Y, Forbes MA, Cooper EH,
molecular analysis of monoclonal Crockson RA, MacLennan IC. Alpha-1-
immunoglobulin light chains from patients with microglobulin: an indicator protein for renal
and without intracellular crystals. Protein Eng tubular function. J Clin Pathol 1983;36:
1999;12:363369. 253259.
79. Maker J. Tubular proteinuria: clinical 90. Bernard A, Vyskocyl A, Mahieu P, Lauwerys R.
implications. In: Avram M, ed. Proteinuria. New Effect of renal insufciency on the concentration
York: Plenum, 1985. of free retinol-binding protein in urine and
80. Sun T, Lien Y, Mailloux L. A case of proteinuria serum. Clin Chim Acta 1988;171:8593.
with analbuminuria. Clin Chem 1985;31: 91. Bernard AM, Lauwerys RR. Retinol binding
19051906. protein in urine: a more practical index than
81. Corso A, Serricchio G, Zappasodi P, et al. urinary beta 2-microglobulin for the routine
Assessment of renal function in patients with screening of renal tubular function. Clin Chem
multiple myeloma: the role of urinary proteins. 1981;27:17811782.
Ann Hematol 1999;78:371375. 92. Ginevri F, Piccotti E, Alinovi R, et al. Reversible
82. Saatci U, Ozdemir S, Ozen S, Bakkaloglu A. tubular proteinuria precedes microalbuminuria
Serum concentration and urinary excretion of and correlates with the metabolic status in
beta 2-microglobulin and microalbuminuria in diabetic children. Pediatr Nephrol 1993;7:2326.
familial Mediterranean fever. Arch Dis Child 93. Mengoli C, Lechi A, Arosio E, et al.
1994;70:2729. Contribution of four markers of tubular
83. Nomiyama K, Liu SJ, Nomiyama H. Critical proteinuria in detecting upper urinary tract
levels of blood and urinary cadmium, urinary infections. A multivariate analysis. Nephron
beta 2-microglobulin and retinol-binding protein 1982;32:234238.
for monitoring cadmium health effects. IARC Sci 94. Korneti P, Tasic V, Cakalaroski K, Korneti B.
Publ. 1992;118:325340. Factitious proteinuria in a diabetic patient. Am J
84. Boesken WH, Effenberger C, Krieger HP, Nephrol 2001;21:512513.
Jammers W, Stierle HE. Fractional clearance of 95. Tojo A, Nanba S, Kimura K, et al. Factitious
beta-2-microglobulin in the diagnostic and proteinuria in a young girl. Clin Nephrol
prognostic assessment of kidney diseases. Klin 1990;33:299302.
Wochenschr 1989;67:3136. 96. Sutcliffe H, Rawlinson PS, Thakker B, Neary R,
85. Tsukahara H, Fujii Y, Tsuchida S, et al. Renal Mallick N. Factitious proteinuria: diagnosis and
handling of albumin and beta-2-microglobulin in protein identication by use of isoelectric
neonates. Nephron 1994;68:212216. focusing. Clin Chem 1988;34:16531655.
References 251
97. Mitas JA 2nd. Exogenous protein as the cause of 110. Lin J, Markowitz GS, Valeri AM, et al. Renal
nephrotic-range proteinuria. Am J Med 1985;79: monoclonal immunoglobulin deposition disease:
115118. the disease spectrum. J Am Soc Nephrol
98. Jones H. On a new substance occurring in the 2001;12:14821492.
urine of a patient with mollities ossium. Phil 111. Winearls CG. Acute myeloma kidney. Kidney Int
Trans R Soc London 1848;138:5562. 1995;48:13471361.
99. Clamp JR. Some aspects of the rst recorded 112. Pote A, Zwizinski C, Simon EE, Meleg-Smith S,
case of multiple myeloma. Lancet 1967;ii: Batuman V. Cytotoxicity of myeloma light
13541356. chains in cultured human kidney proximal tubule
100. MacIntyre W. Case of mollities and fragilitas cells. Am J Kidney Dis 2000;36:735744.
ossium, accompanied with urine strongly charged 113. Cooper EH, Forbes MA, Crockson RA,
with animal matter. Med-Chir Trans 1850;33: MacLennan IC. Proximal renal tubular function
211232. in myelomatosis: observations in the fourth
101. Dalrymple J. On the microscopic character of Medical Research Council trial. J Clin Pathol
mollities ossium. Dublin Q J Med Sci 1846;2: 1984;37:852858.
8595. 114. Guan S, el-Dahr S, Dipp S, Batuman V.
102. Rosenfeld L. Henry Bence Jones (18131873): Inhibition of Na-K-ATPase activity and gene
the best chemical doctor in London. Clin Chem expression by a myeloma light chain in proximal
1987;33:16871692. tubule cells. J Invest Med 1999;47:496501.
103. Putnam FW. Henry Bence Jones: the best 115. Batuman V, Sastrasinh M, Sastrasinh S. Light
chemical doctor in London. Perspect Biol Med chain effects on alanine and glucose uptake by
1993;36:565579. renal brush border membranes. Kidney Int
104. Boffa GA, Zakin MM, Faure A, Fine JM. 1986;30:662665.
Contribution to the study of the relationships 116. Batuman V, Guan S, ODonovan R, Puschett
between the chains of gamma-G JB. Effect of myeloma light chains on phosphate
immunoglobulins and Bence-Jones proteins. and glucose transport in renal proximal tubule
Transfusion 1967;10:169181. cells. Ren Physiol Biochem 1994;17:
105. Bradwell AR, Carr-Smith HD, Mead GP, et al. 294300.
Highly sensitive, automated immunoassay for 117. Truong LD, Mawad J, Cagle P, Mattioli C.
immunoglobulin free light chains in serum and Cytoplasmic crystals in multiple myeloma-
urine. Clin Chem 2001;47:673680. associated Fanconis syndrome. A morphological
106. Picken MM, Shen S. Immunoglobulin light study including immunoelectron microscopy.
chains and the kidney: an overview. Ultrastruct Arch Pathol Lab Med 1989;113:781785.
Pathol 1994;18:105112. 118. Schillinger F, Hopfner C, Montagnac R, Milcent
107. Pick AI, Shoenfeld Y, Skvaril F, et al. T. IgG kappa myeloma with Fanconis syndrome
Asymptomatic (benign) monoclonal and crystalline inclusions. Immunohistochemical
gammopathy a study of 100 patients. Ann Clin and ultrastructural study. Presse Med 1993;22:
Lab Sci 1977;7:335343. 675679.
108. Kyle RA, Therneau TM, Rajkumar SV, et al. A 119. Horn ME, Knapp MS, Page FT, Walker WH.
long-term study of prognosis in monoclonal Adult Fanconi syndrome and multiple
gammopathy of undetermined signicance. myelomatosis. J Clin Pathol 1969;22:414416.
N Engl J Med 2002;346:564569. 120. Uchida S, Matsuda O, Yokota T, et al. Adult
109. Hobbs GA, Levinson SS. Hemoglobin Fanconi syndrome secondary to kappa-light
interference with urinary Bence Jones protein chain myeloma: improvement of tubular
analysis on electrophoresis. Ann Clin Lab Sci. functions after treatment for myeloma. Nephron
1996;26:7175. 1990;55:332335.
252 Examination of urine for proteinuria
121. Huang ZQ, Sanders PW. Localization of a single 133. Pascali E, Pezzoli A. Bence Jones proteins in the
binding site for immunoglobulin light chains on urine of patients with multiple sclerosis. Clin
human TammHorsfall glycoprotein. J Clin Chem 1989;35:15501551.
Invest 1997;99:732736. 134. Pezzoli A, Pascali E. Bence Jones proteinuria in
122. Sanders PW, Booker BB. Pathobiology of cast multiple sclerosis. Clin Chem 1987;33:
nephropathy from human Bence Jones proteins. 19231924.
J Clin Invest 1992;89:630639. 135. Mehta PD, Cook SD, Troiano RA, Coyle PK.
123. Short AK, ODonoghue DJ, Riad HN, Short CD, Increased free light chains in the urine from
Roberts IS. Recurrence of light chain nephropathy patients with multiple sclerosis. Neurology
in a renal allograft. A case report and review of 1991;41:540544.
the literature. Am J Nephrol 2001;21:237240. 136. Hobbs J. Bence Jones Proteins. Essays Med
124. Abbott KC, Agodoa LY. Multiple myeloma and Biochem 1975;1:105131.
light chain-associated nephropathy at end-stage 137. Laurell CB. Complexes formed in vivo between
renal disease in the United States: patient immunoglobulin light chain kappa, prealbumin
characteristics and survival. Clin Nephrol and/or alpha-1-antitrypsin in myeloma sera.
2001;56:207210. Immunochemistry 1970;7:461465.
125. Keren DF. Procedures for the evaluation of 138. Nabeshima Y, Ikenaka T. Primary structure of
monoclonal immunoglobulins. Arch Pathol Lab cryo Bence-Jones protein (Tog) from the urine of
Med 1999;123:126132. a patient with IgD myeloma. Mol Immunol
126. Keren DF, Alexanian R, Goeken JA, Gorevic PD, 1979;16:439444.
Kyle RA, Tomar RH. Guidelines for clinical and 139. Kanoh T, Niki T, Murata Y, Ohta M. Multiple
laboratory evaluation patients with monoclonal myeloma associated with cryo-Bence Jones
gammopathies. Arch Pathol Lab Med 1999;123: protein: report of a case and review of the
106107. literature. Nippon Naika Gakkai Zasshi
127. Kyle RA. Sequence of testing for monoclonal 1978;67:160165.
gammopathies. Arch Pathol Lab Med 1999;123: 140. Kojima M, Kobayashi Y, Murakawa E.
114118. Crystallizable, cryo-precipitable lambda Bence
128. Brigden ML, Neal ED, McNeely MD, Hoag GN. Jones protein in a case of IgD multiple myeloma.
The optimum urine collections for the detection Nippon Ketsueki Gakkai Zasshi 1978;41:
and monitoring of Bence Jones proteinuria. Am J 8189.
Clin Pathol 1990;93:689693. 141. Finazzi Agro A, Crifo C, Natali PG, Chersi A.
129. Drayson M, Tang LX, Drew R, Mead GP, Carr- Differential denaturation of a crystalline Bence-
Smith H, Bradwell AR. Serum free light-chain Jones type cryoprotein as monitored by
measurements for identifying and monitoring uorescence. Ital J Biochem 1978;27:3642.
patients with nonsecretory multiple myeloma. 142. Harris RI, Kohn J. A urinary cryo-Bence Jones
Blood 2001;97:29002902. protein gelling at room temperature. Clin Chim
130. Salomo M, Gimsing P, Nielsen LB. Simple Acta 1974;53:233237.
method for quantication of Bence Jones 143. Hirai H, Doi I, Kawai T. Autopsy case of
proteins. Clin Chem 2002;48:22022207. multiple myeloma with cryo-Bence Jones protein.
131. DelBuono L, Keren DF. Detection of Bence Jones Naika 1969;23:585591.
proteinuria by high-resolution electrophoresis 144. Kawai T, Tadano J. Cryo-Bence Jones protein.
and immunoxation. Am J Clin Pathol First case in Japan. Igaku To Seibutsugaku
1989;92:541(Abstr). 1967;74:251257.
132. Graziani MS, Righetti G. Immunoblotting for 145. Alper CA. Cryoglobulinuria: studies of a cryo-
detecting Bence Jones proteinuria. Clin Chem Bence Jones protein. Acta Med Scand Suppl
1987;33:10791080. 1966;445:200205.
References 253
146. Inoue N, Togawa A, Yawata Y. Tetrameric 157. Ross D, Prenant M, Bessis M. On the proper
Bence Jones protein case report and review of use of the Soret band for hemoglobin detection
the literature. Nippon Ketsueki Gakkai Zasshi. in erythrocytic cells. Blood Cells 1978;4:
1984;47:14561459. 361367.
147. Hom BL. Polymeric (presumed tetrameric) 158. Bush D, Keren DF. Over- and underestimation of
lambda Bence Jones proteinemia without monoclonal gammopathies by quantication of
proteinuria in a patient with multiple myeloma. kappa- and lambda-containing immunoglobulins
Am J Clin Pathol 1984;82:627629. in serum. Clin Chem 1992;38:315316.
148. Solling K, Solling J, Lanng Nielsen J. Polymeric 159. Su L, Keren DF, Warren JS. Failure of anti-
Bence Jones proteins in serum in myeloma lambda immunoxation reagent mimics alpha
patients with renal insufciency. Acta Med Scand heavy-chain disease. Clin Chem 1995;41:
1984;216:495502. 121123.
149. Inoue S, Nagata H, Yozawa H, Terai T, 160. Levinson SS. Studies of Bence Jones proteins by
Hasegawa H, Murao M. A case of IgG myeloma immunonephelometry. Ann Clin Lab Sci
with tetrameric Bence Jones proteinemia and 1992;22:100109.
abnormal brin polymerization. Rinsho Ketsueki 161. Prieto M, Sutherland DE, Goetz FC, Rosenberg
1980;21:200207. ME, Najarian JS. Pancreas transplant results
150. Kozuru M, Benoki H, Sugimoto H, Sakai K, according to the technique of duct management:
Ibayashi H. A case of lambda type tetramer bladder versus enteric drainage. Surgery.
Bence-Jones proteinemia. Acta Haematol 1987;102:680691.
1977;57:359365. 162. See WA, Smith JL. Urinary levels of activated
151. Togawa A, Imamura Y. Study on the Bence- trypsin in whole-organ pancreas transplant
Jones protein study on paraproteins detected in patients with duodenocystostomies.
myeloma patients with tetrameric Bence-Jones Transplantation. 1991;52:630633.
proteins (type kappa) (authors transl). Nippon 163. See WA, Smith JL. Urinary trypsin levels
Ketsueki Gakkai Zasshi. 1975;38:571581. observed in pancreas transplant patients with
152. Caggiano V, Dominguez C, Opfell RW, Kochwa duodenocystostomies promote in vitro
S, Wasserman LR. IgG myeloma with closed brinolysis and in vivo bacterial adherence to
tetrameric Bence Jones proteinemia. Am J Med urothelial surfaces. Urol Res 1992;20:
1969;47:978985. 409413.
153. Kosaka M, Iishi Y, Okagawa K, Saito S, 164. Zheng T, Lu Z, Merideth N, Lanza RP, Soon-
Sugihara J, Muto Y. Tetramer Bence Jones Shiong P. Early markers of pancreas transplant
protein in the immunoproliferative diseases. rejection. Am Surg 1992;58:630633.
Angioimmunoblastic lymphadenopathy, primary 165. Song L, Allison N, Lorah S, Seiple J. An
amyloidosis, and multiple myeloma. Am J Clin abnormal urine protein electrophoresis pattern
Pathol 1989;91:639646. associated with a patient who received
154. Carter PW, Cohen HJ, Crawford J. simultaneous pancreas kidney transplant. Clin
Hyperviscosity syndrome in association with Chem 2002;48:A29.
kappa light chain myeloma. Am J Med 1989;86: 166. Hess PP, Mastropaolo W, Thompson GD,
591595. Levinson SS. Interference of polyclonal free light
155. Khan P, Roth MS, Keren DF, Foon KA. Light chains with identication of Bence Jones
chain disease associated with the hyperviscosity proteins. Clin Chem 1993;39:17341738.
syndrome. Cancer 1987;60:22672268. 167. Charles EZ, Valdes AJ. Free fragments of gamma
156. Hamilton RW, Hopkins MB 3rd, Shihabi ZK. chain in the urine. A possible source of confusion
Myoglobinuria, hemoglobinuria, and acute renal with gamma heavy-chain disease. Am J Clin
failure. Clin Chem 1989;35:17131720. Pathol 1994;101:462464.
254 Examination of urine for proteinuria
168. Roach BM, Meinke JS, Sridhar N, Vladutiu AO. 174. Abe M, Goto T, Kosaka M, Wolfenbarger D,
Multiple narrow bands in urine protein Weiss DT, Solomon A. Differences in kappa to
electrophoresis. Clin Chem 1999;45:716718. lambda (kappa:lambda) ratios of serum and
169. Harrison HH. The ladder light chain or urinary free light chains. Clin Exp Immunol
pseudo-oligoclonal pattern in urinary 1998;111:457462.
immunoxation electrophoresis (IFE) studies: a 175. Brouwer J, Otting-van de Ruit M, Busking-van
distinctive IFE pattern and an explanatory der Lely H. Estimation of free light chains of
hypothesis relating it to free polyclonal light immunoglobulins by enzyme immunoassay. Clin
chains. Clin Chem 1991;37:15591564. Chim Acta 1985;150:267274.
170. Harrison HH. Fine structure of light-chain 176. Solling K. Free light chains of immunoglobulins
ladders in urinary immunoxation studies in normal serum and urine determined by
revealed by ISO-DALT two-dimensional radioimmunoassay. Scand J Clin Lab Invest
electrophoresis. Clin Chem 1990;36:15261527. 1975;35:407412.
171. MacNamara EM, Aguzzi F, Petrini C, et al. 177. Waldmann TA, Strober W, Mogielnicki RP. The
Restricted electrophoretic heterogeneity of renal handling of low molecular weight proteins.
immunoglobulin light chains in urine: a cause for II. Disorders of serum protein catabolism in
confusion with Bence Jones protein. Clin Chem patients with tubular proteinuria, the nephrotic
1991;37:15701574. syndrome, or uremia. J Clin Invest 1972;51:
172. Levinson SS. An algorithmic approach using 21622174.
kappa/lambda ratios to improve the diagnostic 178. Nelson M, Brown RD, Gibson J, Joshua DE.
accuracy of urine protein electrophoresis and to Measurement of free kappa and lambda chains in
reduce the volume required for serum and the signicance of their ratio in
immunoelectrophoresis. Clin Chim Acta patients with multiple myeloma. Br J Haematol
1997;262:121130. 1992;81:223230.
173. Axiak SM, Krishnamoorthy L, Guinan J, Raison 179. Wakasugi K, Sasaki M, Suzuki M, Azuma N,
RL. Quantitation of free kappa light chains in Nobuto T. Increased concentrations of free light
serum and urine using a monoclonal antibody chain lambda in sera from chronic hemodialysis
based inhibition enzyme-linked immunoassay. patients. Biomater Artif Cells Immobilization
J Immunol Methods 1987;99:141147. Biotechnol 1991;19:97109.
Appendix 255
0.22 0.22
13.200
0.18 0.18
0.14 0.14
13.371
Au
Au
0.10 0.10
12.558
12.079
0.06 0.06
14.008
14.867
10.554
12.258
13.433
13.525
0.02 0.02
7.329
0.02 0.02
0 5 10 15
Time (min)
A7.1 This is a capillary zone electropherogram of unconcentrated human serum showing predominantly albumin proteinuria. Total urine
protein is 0.17 g/l. The peak with the highest absorbance is at approximately 7.3 min. This is the urea/creatinine peak. Albumin is the
largest protein peak and occurs at 13.2 min. The smaller peaks before and after albumin are small amounts of other molecules that absorb
at 200 nm. The technique was performed on a Beckman MDQ CE instrument; assay buffer was 150 mM boric acid pH 9.7 containing
calcium lactate. Separation voltage was 18 kV and detection was at an absorbance of 200 nm. This gure was contributed by Margaret
Jenkins, Austin and Repatriation Medical Centre, Heidelberg, Australia.
0.40 0.40
0.34 0.34
0.28 0.28
10.454
0.22 0.22
7.412
Au
Au
11.087
0.16 0.16
12.692
13.067
0.10 0.10
11.567
12.504
13.196
13.633
13.942
14.125
0.04 0.04
10.667
10.917
0.02 0.02
0 5 10 15
Time (min)
A7.2 This is a capillary zone electropherogram of unconcentrated urine with a predominantly tubular pattern. Total urine protein is
1.40 g/l. Once again, the largest peak is due to urea/creatinine at approximately 7.3 min. The albumin peak at approximately 13.2 min is
difcult to see between the numerous tubular protein peaks. Conditions as in A7.1. This gure was contributed by Margaret Jenkins,
Austin and Repatriation Medical Centre, Heidelberg, Australia.
256 Examination of urine for proteinuria
0.22 0.22
0.18 0.18
0.14 0.14
13.054
8.300
Au
Au
0.10 0.10
7.996
8.188
13.792
14.033
0.06 0.06
14.308
13.392
14.683
13.221
9.721
0.02 0.02
7.325
0.02 0.02
0 5 10 15
Time (min)
A7.3 This is a capillary zone electropherogram of unconcentrated urine with monoclonal free light chain (MFLC) and albumin. Total urine
protein is 3.59 g/l. The urea/creatinine peak is at approximately 7.3 min. Immediately following this peak are two sharp peaks that merge at
their base. They are at 7.9 min and 8.3 min and represent a double free k MFLC. The albumin peak is prominent at approximately
13.1 min. Conditions as in A7.1. This gure was contributed by Margaret Jenkins, Austin and Repatriation Medical Centre, Heidelberg,
Australia.
12.846
0.20 0.20
0.16 0.16
7.592
14.154
0.12 12.413 0.12
Au
Au
0.08 0.08
13.346
10.275
13.196
11.575
9.479
11.742
0.04 0.04
13.000
13.675
0.00 0.00
0 5 10 15
Time (min)
A7.4 This is a capillary zone electropherogram of unconcentrated urine with combined glomerular and tubular proteinuria. Total protein
is 5.2 g/l. Again, the urea/creatinine is a good marker peak at approximately 7.6 min. The albumin peak is the second largest and occurs at
13.7 min. Conditions as in A7.1. This gure was contributed by Margaret Jenkins, Austin and Repatriation Medical Centre, Heidelberg,
Australia.
Appendix 257
A7.5 This is a capillary zone electropherogram of unconcentrated A7.6 This is a capillary zone electropherogram of unconcentrated
urine with a prominent monoclonal free light chain (MFLC). The urine with a combined glomerular and tubular pattern. The total
total protein is > 1.5 g/l. No urea/creatinine peak is present on the protein is > 1.5 g/l. No urea/creatinine peak is present on the
patterns performed on the Paragon CZE 2000; neither are the patterns performed on the Paragon CZE 2000. This pattern closely
precise elution times noted. The albumin peak (arrow) is the small resembles a normal serum protein electrophoresis pattern with a
peak toward the anode. A large peak due to the k MFLC is seen at few exceptions. The a2-region has the appearance of an irregular
the bg region interface. The technique was performed on a mesa and drops off sharply just before the transferrin band. There
Paragon CZE 2000, borate buffer, pH 10.0. Electrophoresis at is some bg bridging and irregularity in the g-region. The technique
10.5 kV, 24C for 4 min. This gure is courtesy of Cynthia was performed on a Paragon CZE 2000, borate buffer, pH 10.0.
Blessum, Beckman Coulter, Inc. Electrophoresis at 10.5 kV, 24C for 4 min. This gure is courtesy
of Cynthia Blessum, Beckman Coulter, Inc.
IgG
IgA
IgM
A7.8 This is a capillary zone electropherogram of immunosubtraction on an unconcentrated urine with l monoclonal free light chain.
Each square shows the pattern of migration after the urine was treated with beads coated with antibodies against the specic
immunoglobulin class noted in each square. The prominent g-region spike is present in all the sectors, except the one pretreated with
anti-l. The technique was performed on a Paragon CZE 2000, borate buffer, pH 10.0. Electrophoresis at 10.5 kV, 24C for 4 min. This
gure is courtesy of Cynthia Blessum, Beckman Coulter, Inc.
8
Approach to pattern interpretation in
cerebrospinal fluid
Early electrophoretic studies 259 Other conditions with CSF oligoclonal bands 272
Cerebrospinal uid protein composition 259 Detection of CSF leakage in nasal and aural uid
Electrophoretic methods to study CSF 264 following head trauma 273
Multiple sclerosis and oligoclonal bands 268 References 277
called CSF-specic slow transferrin) that exists in unique to CSF, the protein electrophoresis pattern
CSF along with the usual sialated form of transfer- seen with CSF differs considerably from the pat-
rin.6 As discussed below, the presence of this tern seen with serum (Fig. 8.1). Normally, the
desialated transferrin can be used as a marker of transthyretin (prealbumin) band in serum is barely
CSF leakage into nasal and aural uids as a result visible, whereas this band is increased relative to
of damage to the cranial vault.14 Although some of the other protein bands in concentrated CSF (Fig.
the desialated transferrin nds its way into the 8.2). This results from both a preferential trans-
blood, most of it is quickly taken up by receptors port of transthyretin because of its size and charge
on reticulo-endothelial cells that do not bind trans- characteristics as well as its local synthesis by the
ferrin containing the terminal sialic acid residues.6 epithelium of the choroid plexus.19,20 Because of
The presence of desialated transferrin (also termed this increase in its relative concentration,
carbohydrate-decient transferrin) in the blood has transthyretin had been used to detect CSF leakage
become a convenient marker for the presence of into nasal and aural uids.6 However, since the
alcoholism.1518 However, even in alcoholics, the advent of both immunoxation to detect
concentration of desialated transferrin is too low desialated transferrin and measurement of
to interfere with electrophoretic techniques that prostaglandin D synthase (formerly b-trace pro-
have been employed to detect CSF leakage. tein) (discussed below), I no longer recommend
studies for transthyretin to detect CSF leakage.21
Isoforms of albumin, transferrin and immuno-
Electrophoretic pattern of normal globulins comprise the vast majority of CSF
CSF proteins.5 Albumin is a major protein band in CSF,
as in serum, but it usually migrates more toward
Because of the molecular sieve action of the the anode in CSF than in serum. a1-Lipoprotein
choroid plexus and the presence of proteins tends to overlap albumin or migrates anodally to it
Haptoglobin
a2-Macroglobulin b1-Lipoprotein
S
E
R
U
M
C
S
F
Figure 8.1 Schematic comparison of serum versus concentrated cerebrospinal uid (CSF). The CSF transthyretin band is much stronger
than in the corresponding serum. In contrast the a2-region is considerably weaker in CSF because the high molecular weight haptoglobin
and a2-macroglobulin are restricted from passing across the bloodCSF barrier under normal circumstances. Increased protein in this
region is an indication of a disturbed bloodCSF barrier. In the b2-region of CSF an extra band is present, b2-transferrin that is not
normally present in serum. The g-region of CSF normally stains considerably lighter than the g-region in the corresponding serum.
Cerebrospinal uid protein composition 261
Tf
C3
Figure 8.3 The top sample is CSF concentrated 80-fold. To demonstrate the constituents of the b-region bands in the CSF,
immunoxation with anti-transferrin (Tf) and anti-C3 are shown immediately below. Note the two transferrin bands normally present in
the CSF. (Panagel system stained with Coomassie Blue.)
different individuals on the same gel, one should 1260 mg/dl (0.120.60 g/l).30 However, in chil-
consider the presence of some artifact either in that dren, there is a considerable age-dependent change
run or inherent to the system one is using. In the that must be taken into account when looking for
mid-g-region of some electrophoretic systems, one increased protein concentrations.31,32 Biou et al.33
may nd a slight sharpening of the g-band owing to reported that during the rst 6 months of life, there
restricted migration of CSF IgG, which could be con- is a dramatic decline in the total protein content of
fused with oligoclonal bands (Fig. 8.4).29 In some CSF. This difference reects the immature
electrophoretic methods, I have noticed two faint bloodCSF barrier of the newborn (especially of
bands in the fast g-region that have been mistaken premature infants) that permits larger amounts of
for oligoclonal bands (Fig. 8.5). They do not stain protein to transfer into the CSF than does the
with anti-immunoglobulin reagents. More recent bloodCSF barrier of older children (Table 8.1).
techniques use specic identication of oligoclonal Similarly, a gradual increase in CSF occurs,
bands by immunostaining on all samples. perhaps because of a less stable bloodCSF barrier,
Immunoxation of CSF has the advantage of requir- in individuals over the age of 45 years.10,11
ing much less CSF (since a concentration step can
be eliminated) and also allows one to rule out non-
specic bands caused by g-trace protein (Fig. 8.6). Damaged bloodCSF barrier
The tally of total CSF proteins reveals a content
about 1/350th of that in plasma. In adults up to Alterations of CSF protein patterns occur in a wide
about 50 years of age, the concentration is variety of conditions. However, there is little clini-
Cerebrospinal uid protein composition 263
Figure 8.4 Serum diluted 1:4 (top) and cerebrospinal uid (CSF) concentrated 80-fold (bottom) from the same patient are shown. Note
the single band (indicated) in the g-region of the CSF, which is not present in the corresponding serum. This is not an immunoglobulin by
immunoxation and should not be confused with oligoclonal bands seen in patients with multiple sclerosis. The presence of these bands
emphasizes the importance of using immunological identication of the bands. (Panagel system stained with Coomassie Blue.)
cal diagnostic signicance for alterations other elevated CSF albumin or CSF total protein can be
than those in the g-region. Meningitis results in an helpful in conrming the diagnosis of
elevated CSF total protein because of increased GuillainBarr syndrome in the face of a normal
bloodCSF barrier permeability. With increased CSF differential cell count.6 Thompson and Keir6
permeability the total protein content of the CSF also point out the useful nding of decreased CSF
increases, as does the proportion of larger proteins transthyretin concentration relative to the other
such as those in the a2-region. Yet, there are better CSF proteins as an indicator of obstructed CSF
laboratory methods to support the diagnosis of ow within the spinal cord.
meningitis, such as CSF differential cell counts and In addition to meningitis, a damaged bloodCSF
serum C-reactive protein levels to distinguish barrier can result from other sources of inamma-
between bacterial and aseptic meningitis.6,34 An tion (such as encephalitis from a variety of infec-
a
Data from Biou et al. expressed as mg/dl (g/l).33
264 Approach to pattern interpretation in cerebrospinal uid
ELECTROPHORETIC METHODS TO
STUDY CSF
A wide variety of methods have been used for the
electrophoretic evaluation of CSF. I currently rec-
ommend the use of methods that enhance the sen-
sitivity and specicity by using immunological
identication of bands, such as isoelectric focusing
Figure 8.5 Several sera diluted 1:3 and their corresponding CSF or immunoxation methods. The least sensitive are
concentrated 80-fold immediately below each serum is shown. methods where routine gel electrophoresis is used.41
Note the two faint bands that are indicated in each CSF sample. A recent study of a commercial agarose gel elec-
These are not immunoglobulins and should be ignored when trophoresis method to detect oligoclonal bands
examining samples for oligoclonal bands. Note that true oligoclonal
recorded a disappointing 53 per cent positive among
bands (seen in the slow g-regions of CSF specimens 2 and 3 stain
darker than these bands. Note also that the two bands indicated in individuals with clinically unambiguous multiple
the top three samples vary in staining intensity, roughly correlating sclerosis.42 Furthermore, in addition to relatively
with the amount of protein in the CSF. (Paragon SPE2 system poor sensitivity, even high-resolution agarose and
stained with Paragon Violet.) cellulose methods require that the sample be con-
centrated (typically 80-fold on a commercial ultra-
tious agents), cerebrovascular accidents, metastatic ltration device) before staining with a protein dye
or primary tumors of the central nervous system such as Coomassie Brilliant Blue. Unfortunately,
(CNS), hydrocephalus or herniated intervertebral this increases the volume of CSF required for analy-
discs.3538 The disturbed bloodCSF barrier permits sis. The use of silver stains on unconcentrated CSF
the larger a2 molecules to penetrate into the CSF has been advocated as a means to decrease the vol-
and enhances the staining in this region (Fig. 8.7). ume requirements while preserving the sensitivity of
Also, the total CSF protein is increased because of the assay.43,44 Recently, CSF has also been studied by
a proportionately larger amount of other proteins capillary zone electrophoresis (CZE).45 This method
passing into the CSF. On the electrophoretic strip is able to detect oligoclonal bands, has the advan-
itself, it is difcult to distinguish this pattern from tages of not requiring concentration or staining and
that of a traumatic tap, where some whole blood or has a shorter analysis time. However, none of the
plasma is mixed with the CSF. Often, with a trau- available CZE procedures is currently approved by
matic tap, the sample will have some hemoglobin, the US Food and Drug Administration (FDA) for
giving it a red or pink tinge. this type of analysis.
Electrophoretic methods to study CSF 265
1 2 3 4 5 6
N P MS P I SSPE MS MS N
Figure 8.8 Isoelectric focusing (anode at the top) followed by nitrocellulose blotting and immunoxation with antiserum against IgG is
one of the most sensitive methods to detect oligoclonal bands. This photograph compares oligoclonal banding of cerebrospinal uid (C)
and serum (S) pairs from the following situations: Normal (N), paraproteinemia (P), multiple sclerosis (MS), subacute sclerosing
panencephalitis (SSPE), and a peripheral inammatory response not within the central nervous system (I). Photograph provided by E. J.
Thompson.6
sensitivity and specicity of 83 per cent and 79 per serum must be run with CSF sample to improve
cent respectively for clinically denite multiple scle- the specicity of this information. Our laboratory
rosis using an immunoxation peroxidase method accepts a serum up to 2 weeks after the CSF sample
on unconcentrated CSF. With immunoxation on was run to assay as the control serum. If the serum
agarose gels, not as many bands are seen as on the had oligoclonal bands, they will still be present in a
isoelectric focusing methods and they are broader large enough concentration to be detected. This
in their migration. The number of bands and their cut-off is arbitrary and has not been rigorously
electrophoretic migration tend to remain constant investigated.
during active and inactive disease over a period of Although the bloodCSF barrier excludes most
years.52,53 However, the number of bands per se is immunoglobulins, some do cross the bloodCSF
not recommended to be used in clinical decision- barrier. For example, in Fig. 8.9, a patient with an
making.54 obvious monoclonal gammopathy in the serum has
I prefer batching several samples on the same gel had some of it transfer into the CSF. When CSF
to facilitate comparison of positive and negative samples contain one prominent band and we are
samples. Aside from the improvement in efciency not sent a corresponding serum, we recommend
that this offers, it makes artifacts due to specic gel that the clinician examine the serum and urine for
preparation relatively obvious. I recommend that the presence of a monoclonal gammopathy. In
serum from the patient accompany the CSF patients with prominent oligoclonal banding in the
sample. My laboratory will report a negative study serum, these immunoglobulins will also nd their
for CSF oligoclonal bands when serum is not pro- way into the CSF (Fig. 8.10). Typically, they stain
vided. However, when a CSF sample contains one more strongly in the serum than in the correspond-
or more bands in the CSF and no serum is pro- ing CSF, whereas, if the bands had originated in
vided, no nal interpretation can be made with the CSF (due to local synthesis in patients with
condence. The report is always appended with multiple sclerosis), they would not be detectable at
Electrophoretic methods to study CSF 267
Figure 8.10 A serum sample diluted 1:3 in the top lane shows
prominent a1- and a2-globulins along with three distinct oligoclonal
bands. The cerebrospinal uid (CSF) below is concentrated 80-fold
and has the same three oligoclonal bands barely visible. Because of
the density of staining of the bands in the serum and their faint
staining in the CSF, I report that since both CSF and serum have
the same oligoclonal bands that it is considered negative for CSF
oligoclonal bands. The patient had a systemic inammatory
Figure 8.9 The bottom cerebrospinal uid (CSF) is concentrated
condition, not multiple sclerosis. (Paragon SPE2 system stained
80-fold and has a small monoclonal band (arrow). This has likely
with Paragon Violet.)
diffused across the bloodCSF barrier from the serum, because
the corresponding serum (diluted 1:3) immediately above has a
much denser band in the same region. Whereas one cannot rule be seen as bands in the g-region. For example,
out involvement of the central nervous system by the monoclonal g-trace protein band or the slight mid-g restriction
proliferation of plasma cells, one would expect that the staining in
normally seen with g CSF proteins can be prob-
the CSF would be stronger or at least equal to that seen in the
serum (relative to the density of other bands). A normal serum lematic in cases with only a few bands. I also
and its corresponding CSF are in the top two lanes for excluded artifacts such as the two fast g restric-
comparison. (Paragon SPE2 system stained with Paragon Violet.) tions mentioned above seen on the Paragon SPE2
system (Figs 8.4 and 8.5). Once these artifacts are
known, these methods will demonstrate the more
all in the serum. Under conditions of damage to the prominent oligoclonal bands (Figs 8.11 and
bloodbrain barrier, one must be very suspicious 8.12).
of bands present in both the CSF and in the corre- The literature supports switching from high-
sponding serum. By using the lack of oligoclonal resolution agarose gel electrophoresis to isoelectric
bands in the corresponding serum as a criterion for focus or other techniques that use unconcentrated
reporting the presence of oligoclonal bands in the CSF and identify the bands as IgG. In the 2002
CSF, one improves the specicity of the assay with College of American Pathologists (CAP) Survey M-
only a slight decrease in sensitivity. In one study B, 218 laboratories (93 per cent) were listed as
using high-resolution agarose, the sensitivity of the using electrophoresis to perform analysis of oligo-
CSF oligoclonal band test was 84.9 per cent with a clonal bands, whereas only 17 (7 per cent) were
specicity of 78.9 per cent when the lack of these listed as performing isoelectric focusing.56 At that
bands in the corresponding serum sample was time our laboratory was performing the Sebia
required, however, when the requirement for no Hydragel CSF (Sebia, Issy-les-Moulineaux, France)
corresponding oligoclonal bands in the serum was by immunoxation, there was no category for this
withdrawn, the specicity dropped to 64.8 per type of testing and we listed the method as other.
cent.55 These gures indicate that while most investigators
It is necessary to dene what will be included as now recommend isoelectric focusing and/or
an oligoclonal band pattern. When using the less immunoxation methods to enhance the sensitivity
specic agarose gel-based methods, it is important and specicity of the examination of CSF for oligo-
to exclude other non-immunoglobulins that may clonal bands, the vast majority of clinical
268 Approach to pattern interpretation in cerebrospinal uid
Figure 8.11 The bottom lane contains cerebrospinal uid (CSF) concentrated 80-fold from a patient with multiple sclerosis and the
corresponding serum diluted 1:4 is immediately above. Several oligoclonal bands are evident in the CSF. (Panagel system stained with
Coomassie Blue.)
laboratories participating in this survey do not use provide sufcient evidence of its occurrence, or, if
these more sensitive and specic methods. objective information is needed, immunoassay for
Some authors note that IgM oligoclonal bands myelin basic protein in CSF is an excellent indica-
may be useful to document the onset of multiple tor of disease activity.61
sclerosis or to document acute relapse.4,24,5760
However, the clinical evidence of relapse itself may
MULTIPLE SCLEROSIS AND
OLIGOCLONAL BANDS
The most common reason for performing electro-
phoretic analysis of CSF is to help in the evaluation
of a patient suspected of having multiple sclerosis.4
Although detection of oligoclonal bands in the g-
region is not specic for multiple sclerosis,
examination of the CSF for the presence of oligo-
clonal bands is helpful because the clinical
diagnosis of multiple sclerosis can be difcult and
supportive laboratory data is useful to the clini-
cian.3
Multiple sclerosis is a disease predominately of
young adults (beginning 2040 years of age) and is
more frequent in women (2:1). It occurs in about
Figure 8.12 The top sample is a cerebrospinal uid (CSF) 100 per 100 000 individuals among Caucasian
concentrated 80-fold from a patient with multiple sclerosis and the populations.62 The incidence of multiple sclerosis
corresponding serum diluted 1:3 is immediately below. The g- is, however, much lower in Asian populations. In
region of the CSF from this patient has several densely staining Japan, the incidence is 0.73.8 cases per
oligoclonal bands (O) which are not in the corresponding serum.
100 000.63,64 Further, the Western type of multiple
The CSF is in the third lane and has a faint, slow g-band not seen in
its corresponding serum below. This single band is insufcient for
sclerosis tends to diffusely involve the CNS,
an interpretation of oligoclonal bands. (Paragon SPE2 system whereas the Asian type more selectively involves
stained with Paragon Violet). the optic nerves and spinal cord.64 The epidemio-
Multiple sclerosis and oligoclonal bands 269
logical differences are paralleled by a difference in immune response with predominately IgG1 sub-
the occurrence of oligoclonal bands. In Western class.73 Genetic linkage studies have indicated a
countries, over 90 per cent of individuals with variety of associations, but the strongest is with
diffuse multiple sclerosis have oligoclonal IgG HLA-DR2 (human leukocyte antigen) genes.62,74 A
bands in their CSF. Among Japanese patients with genetic predisposition would be consistent with
diffuse multiple sclerosis, only about half have ndings of distinctive idiotypes and a proclivity
oligoclonal bands and they are present in only toward development of autoantibody-secreting
about 10 per cent of individuals with the more cells in multiple sclerosis.75,76 Evidence that the IgG
selective form of multiple sclerosis.64,65 heavy chain repertoire in plaques from patients
In multiple sclerosis, localized destruction of with multiple sclerosis differs from that of their
myelin occurs in the CNS.66 The immune system peripheral blood lymphocytes implies a specic
has been implicated by the demonstration of CNS antigen-driven targeting that may have a
myelin-reactive T lymphocytes and the oligoclonal genetic basis.77
bands in CSF reecting local synthesis of Several tests can be performed by clinical labora-
immunoglobulins.62 Presenting clinical symptoms tories to aid in the diagnosis of multiple sclerosis:
and signs of multiple sclerosis are highly variable oligoclonal bands in CSF, CSF IgG synthesis (IgG
and include: weakness, diplopia, optic neuritis, Index), CSF myelin basic protein, and serum
paresthesias, numbness, poor vibration sensation antibodies against myeline oligodendrocyte glyco-
leading to difculty with coordinated movements, protein (MOG) and against myelin basic protein
absence of abdominal reexes, trigeminal neural- (MBP).78,79 Detection of oligoclonal bands is the
gia (in young adults), vertigo, and easy fatigabil- most sensitive for supporting the initial diagnosis.80
ity.67,68 The diagnosis of multiple sclerosis depends Seres et al.36 reported that, whereas 76 per cent of
on recurrent episodes of the above phenomena the 37 patients with clinically documented multiple
involving at least two anatomic sites. Magnetic sclerosis had an elevated IgG Index, 91 per cent
resonance imaging (MRI) is both a diagnostic had oligoclonal banding as demonstrated by
tool, and a marker to monitor the progress of the agarose gel electrophoresis. Anti-MOG and anti-
disease. It serves as a measurable baseline for MBP in serum are strong predictors for early
therapeutic trials.69,70 Because early clinical signs conversion of early to clinically denite multiple
and symptoms are non-specic, the diagnosis of sclerosis.78
multiple sclerosis at this stage can be quite dif- Patients with multiple sclerosis have an increased
cult. The clinical laboratory provides useful infor- local (CSF) synthesis of immunoglobulins. This is
mation to support the diagnosis in many of these demonstrated by the fact that oligoclonal bands
patients. are found in the CSF, but not in the corresponding
The etiology of multiple sclerosis is unknown, serum in over 90 per cent of these
but genetic factors have been implicated mainly by patients.3,36,41,55,79,81 Multiple sclerosis patients
family studies. The fact that 26 per cent of lacking oligoclonal bands in the CSF have fewer
monozygotic twins both develop multiple sclerosis plasma cells within the meninges and fewer
compared with only 2.3 per cent of dizygotic plaques at time of autopsy than patients whose
twins is strong evidence for a genetic basis for CSF contains oligoclonal bands. This suggests that
multiple sclerosis.71 Further support for this idea the oligoclonal bands are a reection of the local
comes from the observation of a higher risk for synthesis of immunoglobulin by plasma cells in the
developing multiple sclerosis in offspring of diseased tissue.82 Although there have been many
affected individuals than in spouses of affected studies on a wide variety of possible antigens
individuals.72 Alterations of the immune system (many viral), the specic antigen(s) against which
have been implicated in this process. For example, most these antibodies are being made has not been
patients with multiple sclerosis have a restricted identied. It is likely that many antigens are
270 Approach to pattern interpretation in cerebrospinal uid
autologous hematopoietic stem cell transplanta- has been referred to as the mirror pattern.4 An
tion the oligoclonal bands were found to persist occasional case of multiple sclerosis with oligo-
despite magnetic resonance imaging (MRI) evi- clonal bands in both locations does occur, but this
dence of reduction in some lesions.96 is not considered useful in conrming the presence
When I observe oligoclonal bands in both the CSF of multiple sclerosis.95
and serum, I consider this to be a result of diffusion The presence of oligoclonal bands in the CSF
of the serum oligoclonal bands into the CSF. This with no serum sample submitted is of question-
able signicance. When this occurs, I recommend
that a serum be sent within the next 2 weeks for
4 5 6 comparison with this CSF pattern. The 2 weeks is
an arbitrary period, which reects the fact that if
the oligoclonal bands came from a systemic
process, the IgG from those systemic clones
should have a half-life of about 23 weeks.
Therefore, at least half of the amount should be
present. Obviously, the ideal is to have the serum
accompany the CSF sample. But the patient can
avoid a needless repeat lumbar puncture if the
serum can be obtained relatively soon after the
rst sample.
When there is one small band in the CSF but
none in the serum, I note its presence but advise
the clinician that this is not sufciently strong evi-
dence to support the diagnosis of multiple sclero-
sis.
CSF S CSF S CSF S Finally, occasionally I have observed a relatively
large monoclonal band in the CSF; these cases also
Figure 8.13 Immunoxation performed on unconcentrated
have the band in the serum. When I observe this, I
cerebrospinal uid (CSF). Sample 4 has a negative CSF and negative
serum (S). Sample 5 has several oligoclonal bands in the CSF and recommend an evaluation of the patient for the
none in the corresponding serum. Sample 6 has a monoclonal band monoclonal gammopathy (which may be account-
in both the CSF and the serum. (Sebia CSF immunoxation gel.) ing for the neurological symptoms).
272 Approach to pattern interpretation in cerebrospinal uid
OTHER CONDITIONS WITH CSF Table 8.3 Conditions in which oligoclonal bands may be found in
cerebrospinal uid (CSF)
OLIGOCLONAL BANDS
One early sign of multiple sclerosis is optic neuritis. Multiple sclerosisa
Examination of the CSF for the presence of oligo- Subacute sclerosing panencephalitis
clonal bands and/or an elevated IgG index CreutzfeldtJakob disease
increases the risk that these individuals will go on Meningoencephalitis
to develop multiple sclerosis. However, a normal
Spinal cord compression
CSF study in this group cannot rule out that possi-
GuillainBarr syndrome
bility.97
A positive CSF oligoclonal band test is not patho- Syphilis
gnomonic for multiple sclerosis. Oligoclonal bands Peripheral neuropathy
can be found in a wide variety of neurological Optic neuritis
diseases, including inammation, neoplasia, cere- Hydrocephalus
brovascular accidents, structural CNS lesions,
Cerebrovascular accident
demyelinating diseases, and some peripheral
Immune complex vasculitis
neuropathies (Table 8.3).98107 Except for those
patients with subacute sclerosing panencephalitis, Systemic lupus erythematosus
the percentage of patients with these conditions Diabetes
who have oligoclonal bands is considerably less Whipples disease
than the approximately 90 per cent of multiple Neoplasms
sclerosis patients with oligoclonal bands. Acquired immune deciency syndrome (AIDS)
Fortunately, multiple sclerosis is not part of the dif-
Lyme disease
ferential diagnosis in most of these clinical
Fever of unknown origin
conditions.
Because the oligoclonal band test is non-specic, a
In multiple sclerosis about 90% of patients will have
it should be used in dened situations, such as in oligoclonal bands in the CSF. In most of the other conditions
the case of a patient that has had few clinical listed, such bands are uncommon but may be seen.
episodes suggestive of multiple sclerosis, but the
diagnosis is not yet secure. The presence of oligo-
clonal bands in those patients is useful supportive
information. A negative test will cause the clini- Central nervous system systemic
cian to review the clinical features, as 8090 per lupus erythematosus (CNS lupus)
cent of patients with multiple sclerosis should
have oligoclonal bands in the CSF. The labora- Some patients with systemic lupus erythematosus
tory test should never be used as the sole evidence (SLE) develop central nervous system (CNS)
of multiple sclerosis. Using a sensitive polyacry- involvement, that manifests a variety of symptoms:
lamide gel electrophoresis technique, Coret et psychosis, cranial nerve palsy, seizures, cerebrovas-
al.108 reported the diagnoses associated with CSF cular accidents, and transverse myelopathy.109,110
oligoclonal bands in a study of 488 patients suf- The incidence of CNS manifestations in patients
fering from neurological disease (Table 8.4). with SLE varies widely from 25 per cent, reported
While multiple sclerosis represented the vast in a large clinical series, to as high as 75 per cent in
majority of cases, almost half of their patients a retrospective postmortem series.111 Central
with infectious diseases had oligoclonal bands, as nervous system involvement can be the cause of
did 11 per cent of patients with vascular malig- death in as many as 13 per cent of these patients.112
nancies. Unfortunately, symptoms of CNS lupus can be
Detection of CSF leakage in nasal and aural uid following head trauma 273
Table 8.4 Occurrence of cerebrospinal uid (CSF) oligoclonal oligoclonal band test only as a conrmatory test
bands in patients with neurological disease
for patients suspected of having multiple sclerosis.
The increased IgG levels in CSF of some patients
Diagnosis Per cent of with CNS lupus result primarily from an impaired
patients with bloodCSF barrier.111
CSF O-bands Serum antibody against ribosomal P has been
suggested as a marker for patients with lupus psy-
Denite multiple sclerosis 84 chosis.117120 These antibodies give a pattern on the
Probable multiple sclerosis 46 uorescent antinuclear antibody (ANA) test that
Inammatory infectious diseases 43 shows both cytoplasmic and nuclear staining
Possible multiple sclerosis 7
similar to that seen with mitochondrial antibody.
Previously, conrmation of this antibody required
Vascular malignancies 11
either Western blot or a specic immunoassay for
Other neurological diseases 4 ribosomal P.121 However, a recently reported
Control CSF 0 enzyme-linked immunosorbent assay (ELISA) that
a
used a 22 amino acid peptide that corresponds to a
Data from Coret et al.108
common epitope on ribosomal P0, P1, and P2 had
an 83 per cent concordance with Western blot.122
When this ELISA was used to test sera from 178
mimicked by steroid psychosis, and the clinician is consecutive patients with SLE and 28 others with
occasionally faced with a patient with known SLE CNS lupus, the presence of anti-ribosomal P was
who is receiving steroid therapy and demonstrating associated with clinically active disease, high levels
psychotic symptoms. Should the steroids be of anti-dsDNA and decreased C4 levels.122
increased (for CNS lupus) or should they be Unfortunately, only 11 of the 28 patients with
tapered (for steroid psychosis)? Over the years, CNS lupus were positive for anti-ribosomal P.
there have been many attempts at establishing lab- While this is signicantly greater than in unselected
oratory tests that would help with this differential SLE patients, it leaves more than 60 per cent of
diagnosis. Largely, they have failed. Levels of C3, individuals with CNS lupus as false negatives.
C4, anti-DNA, oligoclonal bands and, more Therefore, in its present form, a positive suggests
recently, interleukin-1 and interleukin-6 levels in more clinically active disease and is supportive, but
CSF are, at best, only partly helpful.113 One recent not diagnostic evidence for CNS lupus. Also, the
improvement in the use of these assays is the adop- absence of the antibody does not rule out CNS
tion of the ratios (Q) of CSF C3 to serum C3, and lupus.123
CSF C4 to serum C4 rather than the absolute
values. The CSFQ3 and Q4 are increased in some
patients with CNS lupus.114 Nonetheless, the
clinical picture remains the gold standard for DETECTION OF CSF LEAKAGE IN
whether the patient has CNS lupus.109,110,115 NASAL AND AURAL FLUID
The presence of oligoclonal bands in the CSF also FOLLOWING HEAD TRAUMA
has been suggested as an aid in the diagnosis of
CNS lupus. Unfortunately, this is a poor marker Leakage of CSF into nasal or aural cavities is most
because only a minority of patients with CNS often caused by trauma, but also results from
lupus have such bands.116 Further, a negative oligo- intracranial surgical procedures, infection, hydro-
clonal band test in a patient with clinically cephalus, congenital malformations, and neo-
suspected CNS lupus will not cause the clinician to plasms.124 In order to prevent the development of
withhold therapy. I recommend use of the CSF meningitis, the CSF leakage must be differentiated
274 Approach to pattern interpretation in cerebrospinal uid
from allergic rhinitis or infectious rhinosinusitis as As mentioned earlier in this chapter, the presence
soon as possible.125,126 There are two highly specic of neuraminidase in the CNS causes desialation of
methods available to detect CSF leakage: b2-trans- some transferrin molecules in the CNS. The loss of
ferrin demonstration by immunoxation or these negatively charged sialic acid groups from
immunoblotting and measurement of b-trace pro- some transferrin molecules results in two transfer-
tein (prostaglandin D synthase). At present, the rin bands by electrophoresis: the b1 fraction (the
detection of b2-transferrin is more commonly same as that found in serum), and a more cathodal
employed. However, it requires non-standard b2-transferrin band (t fraction, also called CSF-
immunoxation or immunoblotting techniques, as specic transferrin).124 b2-Transferrin is not
described below. In contrast, prostaglandin D syn- normally present in serum, tears, saliva, sputum,
thase may be measured by either nephelometry or nasal or aural uid, perilymph or endolymph,
immunoxation and will likely become the pre- although it is present in aqueous and vitreous
ferred method for this situation.125,127131 humor.132 However, one must be cautious to
control for genetic variants of transferrin. gels are washed twice in isotonic saline, dried,
Immunological identication of b2-transferrin pro- stained with Paragon Violet for 5 min, and
vides a sensitive and specic tool to distinguish destained in two washes of 10 per cent glacial
CSF leakage from serous nasal or aural uid. acetic acid.14
Immunoxation, or more sensitive immuno- Western blotting may also be performed to detect
blotting procedures on these uids have proven some of this leakage as well as genetic variants of
useful to make this distinction.14,124,133 transferrin.134136 This involves performing serum
My laboratory performs immunoxation on 10 protein electrophoresis and then blotting the pro-
concentrated samples of the nasal or aural uid. teins onto nitrocellulose paper. The paper is then
After concentration, 35 ml of unknown uid are incubated with anti-transferrin (Fig. 8.15). By
applied onto the gel in two lanes with the patients using immunoenzyme conjugates, this technique
serum diluted 1:3 in an adjacent lane to control for may provide greater sensitivity and lower back-
genetic variants of transferrin. A CSF control is ground than immunoxation. Normansell et al.137
applied to two other lanes as a positive control for report a sensitivity of 1 mg/ml and their procedure
b2-transferrin (Fig. 8.14). After a 5-min diffusion has the further advantage of not requiring a con-
time, the gels are gently blotted and the samples are centration step. By using a combination of
electrophoresed at 100 V for 30 min. The gels are isoelectric focusing on polyacrylamide gel, direct
then overlaid with 80 l of antiserum against immunoxation and silver staining, Roelandse et
human transferrin. Following a 35 min incubation al.138 have further improved the sensitivity of this
at room temperature in a moisture chamber, the procedure.
React with
anti-transferrin,
wash, labeled
second antibody
Normal
Normal transferrin desialated transferrin
detect nasal and aural leakage of CSF. More recent Skoog I, Wikkelso C, Svennerholm L. Protein
nephelometric studies substantiate the claimed sen- analysis in cerebrospinal uid. III. Relation to
sitivity of greater than 90 per cent with a bloodcerebrospinal uid barrier function for
concentration of prostaglandin D synthase (b-trace formulas for quantitative determination of
protein) of 6 mg/l or higher and a specicity of 100 intrathecal IgG production. Eur Neurol
per cent. Detection of this enzyme may soon 1993;33:134142.
become the method of choice to identify CSF 11. Blennow K, Fredman P, Wallin A, et al. Protein
leakage.21,125,128131,142145 analysis in cerebrospinal uid. II. Reference
values derived from healthy individuals 1888
years of age. Eur Neurol 1993;33:129133.
REFERENCES 12. Moos T. Immunohistochemical localization of
intraneuronal transferrin receptor
1. Kabat EA, Landow H, Moore DH. immunoreactivity in the adult mouse central
Electrophoretic patterns of concentrated nervous system. J Comp Neurol 1996;375:
cerebrospinal uid. Proc Soc Ex Biol Med 675692.
1942;49:260263. 13. Kissel K, Hamm S, Schulz M, Vecchi A,
2. Kabat EA, Moore DH, Landow H. An Garlanda C, Engelhardt B. Immunohistochemical
electrophoretic study of the protein components localization of the murine transferrin receptor
in cerebrospinal uid and their relationship to (TfR) on blood-tissue barriers using a novel anti-
the serum proteins. J Clin Invest 1942;21: TfR monoclonal antibody. Histochem Cell Biol
571577. 1998;110:6372.
3. Falip M, Tintore M, Jardi R, Duran I, Link H, 14. Zaret DL, Morrison N, Gulbranson R, Keren
Montalban X. Clinical usefulness of oligoclonal DF. Immunoxation to quantify beta 2-
bands. Rev Neurol 2001;32:11201124. transferrin in cerebrospinal uid to detect
4. Sindic CJ. CSF analysis in multiple sclerosis. Acta leakage of cerebrospinal uid from skull injury.
Neurol Belg 1994;94:103111. Clin Chem 1992;38:19081912.
5. Sickmann A, Dormeyer W, Wortelkamp S, 15. Rukstalis MR, Lynch KG, Oslin DW, et al.
Woitalla D, Kuhn W, Meyer HE. Towards a high Carbohydrate-decient transferrin levels reect
resolution separation of human cerebrospinal heavy drinking in alcohol-dependent women
uid. J Chromatogr B Analyt Technol Biomed seeking treatment. Alcohol Clin Exp Res
Life Sci 2002;771:167196. 2002;26:15391544.
6. Thompson EJ, Keir G. Laboratory investigation 16. Arndt T, Kropf J. Alcohol abuse and
of cerebrospinal uid proteins. Ann Clin carbohydrate-decient transferrin analysis: are
Biochem 1990;27(Pt 5):425435. screening and conrmatory analysis required?
7. Fishman RA. Cerebrospinal uid in diseases of Clin Chem 2002;48:20722074.
the nervous system. Philadelphia: WB Saunders, 17. Suzuki Y, Saito H, Suzuki M, Hosoki Y, Sakurai
1992. S, Fujimoto Y, Kohgo Y. Up-regulation of
8. Jeppson JO, Laurell CB, Franzen B. Agarose transferrin receptor expression in hepatocytes by
gel electrophoresis. Clin Chem 1979;25: habitual alcohol drinking is implicated in hepatic
629638. iron overload in alcoholic liver disease. Alcohol
9. Grifn DE, Giffels J. Study of protein Clin Exp Res 2002;26:26S-31S.
characteristics that inuence entry into the 18. Hermansson U, Helander A, Brandt L, Huss A,
cerebrospinal uid of normal mice and mice Ronnberg S. The Alcohol Use Disorders
with encephalitis. J Clin Invest 1982;70: Identication Test and carbohydrate-decient
289295. transferrin in alcohol-related sickness absence.
10. Blennow K, Fredman P, Wallin A, Gottfries CG, Alcohol Clin Exp Res 2002;26:2835.
278 Approach to pattern interpretation in cerebrospinal uid
19. Duan W, Cole T, Schreiber G. Cloning and of serum and cerebrospinal uid by
nucleotide sequencing of transthyretin immunoxation. Acta Neurol Scand
(prealbumin) cDNA from rat choroid plexus and 1978;58:141147.
liver. Nucl Acids Res 1989;17:3979. 29. Laurell CB. Composition and variation of the gel
20. Harms PJ, Tu GF, Richardson SJ, Aldred AR, electrophoretic fractions of plasma, cerebrospinal
Jaworowski A, Schreiber G. Transthyretin uid and urine. Scand J Clin Lab Invest Suppl
(prealbumin) gene expression in choroid plexus is 1972;124:7182.
strongly conserved during evolution of 30. Lott JA, Warren P. Estimation of reference
vertebrates. Comp Biochem Physiol B intervals for total protein in cerebrospinal uid.
1991;99:239249. Clin Chem 1989;35:17661770.
21. Kleine TO, Damm T, Althaus H. Quantication 31. Tibbling G, Link H, Ohman S. Principles of
of beta-trace protein and detection of transferrin albumin and IgG analyses in neurological
isoforms in mixtures of cerebrospinal uid and disorders. I. Establishment of reference
blood serum as models of rhinorrhea and values. Scand J Clin Lab Invest
otorrhea diagnosis. Fresenius J Anal Chem 1977;37:385390.
2000;366:382386. 32. Link H, Tibbling G. Principles of albumin and
22. Whicher JT. The interpretation of electrophoresis. IgG analyses in neurological disorders. II.
Br J Hosp Med 1980;24:348356. Relation of the concentration of the proteins in
23. Blennow K, Fredman P. Detection of serum and cerebrospinal uid. Scand J Clin Lab
cerebrospinal uid leakage by isoelectric focusing Invest 1977;37:391396.
on polyacrylamide gels with silver staining using 33. Biou D, Benoist JF, Nguyen-Thi C, Huong X,
the PhastSystem. Acta Neurochir (Wien) Morel P, Marchand M. Cerebrospinal uid
1995;136:135139. protein concentrations in children: age-related
24. Rijcken CA, Thompson EJ, Teelken AW. An values in patients without disorders of the
improved, ultrasensitive method for the central nervous system. Clin Chem
detection of IgM oligoclonal bands in 2000;46:399403.
cerebrospinal uid. J Immunol Methods 34. Tatara R, Imai H. Serum C-reactive protein in
1997;203:167169. the differential diagnosis of childhood meningitis.
25. Hiraoka A, Arato T, Tominaga I, Eguchi N, Oda Pediatr Int 2000;42:541546.
H, Urade Y. Analysis of low-molecular-mass 35. Chamoun V, Zeman A, Blennow K, Fredman P,
proteins in cerebrospinal uid by sodium dodecyl Wallin A, Keir G, Giovannoni G, Thompson EJ.
sulfate capillary gel electrophoresis. Haptoglobins as markers of blood CSF barrier
J Chromatogr B Biomed Sci Appl 1997;697: dysfunction: the ndings in normal CSF.
141147. J Neurol Sci 2001;182:117121.
26. Link H. Isolation and partial characterization of 36. Seres E, Bencsik K, Rajda C, Vecsei L. Diagnostic
trace proteins and immunoglobulin G from studies of cerebrospinal uid in patients with
cerebrospinal uid. J Neurol Neurosurg multiple sclerosis. Orv Hetil
Psychiatry 1965;28:552559. 1998;139:19051908.
27. Stibler H. The normal cerebrospinal uid 37. Qin D, Ma J, Xiao J, Tang Z. Effect of brain
proteins identied by means of thin-layer irradiation on bloodCSF barrier permeability of
isoelectric focusing and crossed chemotherapeutic agents. Am J Clin Oncol
immunoelectrofocusing. J Neurol Sci 1997;20:263265.
1978;36:273288. 38. Hallgren R, Terent A, Wide L, Bergstrom K,
28. Laurenzi MA, Link H. Localization of the Birgegard G. Cerebrospinal uid ferritin in
immunoglobulins G, A and M, beta-trace protein patients with cerebral infarction or bleeding.
and gamma-trace protein on isoelectric focusing Acta Neurol Scand 1980;61:384392.
References 279
39. Pakulski C, Drobnik L, Millo B. Age and sex as 49. Cawley LP, Minard BJ, Tourtellotte WW, Ma BI,
factors modifying the function of the blood- Chelle C. Immunoxation electrophoretic
cerebrospinal uid barrier. Med Sci Monit techniques applied to identication of proteins in
2000;6:314318. serum and cerebrospinal uid. Clin Chem
40. Haussermann P, Kuhn W, Przuntek H, Muller T. 1976;22:12621268.
Integrity of the bloodcerebrospinal uid barrier 50. Cavuoti D, Baskin L, Jialal I. Detection of
in early Parkinsons disease. Neurosci Lett 2001; oligoclonal bands in cerebrospinal uid by
300:182184. immunoxation electrophoresis. Am J Clin
41. Lunding J, Midgard R, Vedeler CA. Oligoclonal Pathol 1998;109:585588.
bands in cerebrospinal uid: a comparative study 51. Richard S, Miossec V, Moreau JF, Taupin JL.
of isoelectric focusing, agarose gel electrophoresis Detection of oligoclonal immunoglobulins in
and IgG index. Acta Neurol Scand 2000;102: cerebrospinal uid by an
322325. immunoxationperoxidase method. Clin Chem
42. Niederwieser G, Bonelli RM, Brunner D, et al. 2002;48:167173.
Diagnostic accuracy of an agarose gel 52. Nilsson K, Olsson JE. Analysis for cerebrospinal
electrophoretic method in multiple sclerosis. Clin uid proteins by isoelectric focusing on
Chem 2001;47:144. polyacrylamide gel: methodological aspects and
43. Lane JR, Bowles KJ, Normansell DE. The normal values, with special reference to the
detection of oligoclonal IgG bands in alkaline region. Clin Chem 1978;24:
unconcentrated cerebrospinal uid by isoelectric 11341139.
focusing on thin-layer agarose gels and silver 53. Olsson JE, Link H, Muller R. Immunoglobulin
staining. Arch Pathol Lab Med 1986;110: abnormalities in multiple sclerosis. Relation to
2629. clinical parameters: disability, duration and age
44. Wybo I, van Blerk M, Malfait R, Goubert P, of onset. J Neurol Sci 1976;27:233245.
Gorus F. Oligoclonal bands in cerebrospinal uid 54. Avasarala JR, Cross AH, Trotter JL. Oligoclonal
detected by PhastSystem isoelectric focusing. Clin band number as a marker for prognosis in
Chem 1990;36:123125. multiple sclerosis. Arch Neurol 2001;58:
45. Ivanova M, Tzvetanova E, Jetcheva V, Kilar F. 20442045.
Abnormal protein patterns in blood serum and 55. Davenport RD, Keren DF. Oligoclonal bands in
cerebrospinal uid detected by capillary cerebrospinal uids: signicance of
electrophoresis. J Biochem Biophys Methods corresponding bands in serum for diagnosis of
2002;53:141150. multiple sclerosis. Clin Chem 1988;34:
46. Verbeek MM, de Reus HP, Weykamp CW. 764765.
Comparison of methods for the detection of 56. College of American Pathologists (CAP).
oligoclonal IgG bands in cerebrospinal uid and Cerebrospinal Fluid Survey M-B. Participant
serum: results of the Dutch Quality Control Summary Rep 2002:6.
survey. Clin Chem 2002;48:15781580. 57. Villar LM, Gonzalez-Porque P, Masjuan J,
47. Hansson LO, Link H, Sandlund L, Einarsson R. Alvarez-Cermeno JC, Bootello A, Keir G. A
Oligoclonal IgG in cerebrospinal uid detected sensitive and reproducible method for the
by isoelectric focusing using PhastSystem. Scand detection of oligoclonal IgM bands. J Immunol
J Clin Lab Invest 1993;53:487492. Methods 2001;258:151155.
48. Caudie C, Allausen O, Bancel J. Detection of 58. Nelson D, Fredman P, Borjeson J. A high-
oligoclonal IgG bands in cerebrospinal uid by sensitivity immuno-chemiluminescence technique
immunoxation after electrophoretic migration for detection of oligoclonal IgG and IgM in
in the automated Hydrasys sebia system. Ann unconcentrated cerebrospinal uid. Scand J Clin
Biol Clin (Paris) 2000;58:376379. Lab Invest 1994;54:5154.
280 Approach to pattern interpretation in cerebrospinal uid
59. Sharief MK, Thompson EJ. Intrathecal 72. Ebers GC, Yee IM, Sadovnick AD, Duquette P.
immunoglobulin M synthesis in multiple Conjugal multiple sclerosis: population-based
sclerosis. Relationship with clinical and prevalence and recurrence risks in offspring.
cerebrospinal uid parameters. Brain Canadian Collaborative Study Group. Ann
1991;114(Pt 1A):181195. Neurol 2000;48:927931.
60. Sindic CJ, Monteyne P, Laterre EC. Occurrence 73. Vandvik B, Natvig JB, Wiger D. IgG1 subclass
of oligoclonal IgM bands in the cerebrospinal restriction of oligoclonal IgG from cerebrospinal
uid of neurological patients: an immunoafnity- uids and brain extracts in patients with multiple
mediated capillary blot study. J Neurol Sci sclerosis and subacute encephalitides. Scand J
1994;124:215219. Immunol 1976;5:427436.
61. Ohta M, Ohta K, Ma J, et al. Clinical and 74. Fogdell A, Olerup O, Fredrikson S, Vrethem M,
analytical evaluation of an enzyme immunoassay Hillert J. Linkage analysis of HLA class II genes
for myelin basic protein in cerebrospinal uid. in Swedish multiplex families with multiple
Clin Chem 2000;46:13261330. sclerosis. Neurology 1997;48:758762.
62. Hellings N, Raus J, Stinissen P. Insights into the 75. LaGanke CC, Freeman DW, Whitaker JN.
immunopathogenesis of multiple sclerosis. Cross-reactive idiotypy in cerebrospinal uid
Immunol Res 2002;25:2751. immunoglobulins in multiple sclerosis. Ann
63. Kuroiwa Y, Igata A, Itahara K, Koshijima S, Neurol 2000;47:8792.
Tsubaki T. Nationwide survey of multiple 76. Sellebjerg F, Jensen CV, Christiansen M.
sclerosis in Japan. Clinical analysis of 1084 cases. Intrathecal IgG synthesis and autoantibody-
Neurology 1975;25:845851. secreting cells in multiple sclerosis.
64. Nakashima I, Fujihara K, Itoyama Y. Oligoclonal J Neuroimmunol 2000;108:207215.
IgG bands in Japanese multiple sclerosis patients. 77. Owens GP, Burgoon MP, Anthony J,
J Neuroimmunol 1999;101:205206. Kleinschmidt-DeMasters BK, Gilden DH. The
65. Fukazawa T, Kikuchi S, Sasaki H, et al. The immunoglobulin G heavy chain repertoire in
signicance of oligoclonal bands in multiple multiple sclerosis plaques is distinct from the
sclerosis in Japan: relevance of immunogenetic heavy chain repertoire in peripheral blood
backgrounds. J Neurol Sci 1998;158: lymphocytes. Clin Immunol 2001;98:
209214. 258263.
66. Zipp F. Apoptosis in multiple sclerosis. Cell 78. Berger T, Rubner P, Schautzer F et al. Antimyelin
Tissue Res 2000;301:163171. antibodies as a predictor of clinically denite
67. Lublin FD. The diagnosis of multiple sclerosis. multiple sclerosis after a rst demyelinating
Curr Opin Neurol 2002;15:253256. event. N Engl J Med 2003;349:139145.
68. Bernard CC, Kerlero de Rosbo N. Multiple 79. Gerson B, Cohen SR, Gerson IM, Guest GH.
sclerosis: an autoimmune disease of Myelin basic protein, oligoclonal bands, and IgG
multifactorial etiology. Curr Opin Immunol in cerebrospinal uid as indicators of multiple
1992;4:760765. sclerosis. Clin Chem 1981;27:19741977.
69. Arnold DL, Matthews PM. MRI in the diagnosis 80. Bloomer LC, Bray PF. Relative value of three
and management of multiple sclerosis. Neurology laboratory methods in the diagnosis of multiple
2002;58:S2331. sclerosis. Clin Chem 1981;27:20112013.
70. McFarland HF. The emerging role of MRI in 81. Correale J, de Los Milagros Bassani Molinas M.
multiple sclerosis and the new diagnostic criteria. Oligoclonal bands and antibody responses
Mult Scler 2002;8:7172. in multiple sclerosis. J Neurol 2002;249:
71. Ebers GC, Bulman DE, Sadovnick AD, et al. A 375389.
population-based study of multiple sclerosis in 82. Farrell MA, Kaufmann JC, Gilbert JJ,
twins. N Engl J Med 1986;315:16381642. Noseworthy JH, Armstrong HA, Ebers GC.
References 281
Wisniewski HM. Effect of treatment on 112. Dubois EL, Wierzchowiecki M, Cox MB, Weiner
oligoclonal IgG bands and intrathecal IgG JM. Duration and death in systemic lupus
synthesis in sequential cerebrospinal uid and erythematosus. An analysis of 249 cases. JAMA
serum from patients with subacute sclerosing 1974;227:13991402.
panencephalitis. J Neurol Sci 1992;109:6468. 113. Alcocer-Varela J, Aleman-Hoey D, Alarcon-
103. Izquierdo G, Aguilar J, Angulo S, Giron JM. Segovia D. Interleukin-1 and interleukin-6
Oligoclonal bands in serum and cerebrospinal activities are increased in the cerebrospinal uid
uid in patients with suspected neuroborreliosis. of patients with CNS lupus erythematosus and
Med Clin (Barc) 1992;98:516517. correlate with local late T-cell activation
104. Dalakas M. Oligoclonal bands in the markers. Lupus 1992;1:111117.
cerebrospinal uid of post-poliomyelitis muscular 114. Jongen PJ, Doesburg WH, Ibrahim-Stappers JL,
atrophy. Ann Neurol 1990;28:196197. Lemmens WA, Hommes OR, Lamers KJ.
105. Skotzek B, Sander T, Zimmermann J, Kolmel Cerebrospinal uid C3 and C4 indexes in
HW. Oligoclonal bands in serum and immunological disorders of the central nervous
cerebrospinal uid of patients with HIV system. Acta Neurol Scand 2000;101:
infection. J Neuroimmunol 1988;20: 116121.
151152. 115. van Dam AP. Diagnosis and pathogenesis of
106. Gessain A, Caudie C, Gout O, Vernant JC, CNS lupus. Rheumatol Int 1991;11:111.
Maurs L, Giordano C, Malone G, Tournier- 116. Wineld JB, Shaw M, Silverman LM, Eisenberg
Lasserve E, Essex M, de-The G. Intrathecal RA, Wilson HA 3rd, Kofer D. Intrathecal IgG
synthesis of antibodies to human T lymphotropic synthesis and bloodbrain barrier impairment in
virus type I and the presence of IgG oligoclonal patients with systemic lupus erythematosus and
bands in the cerebrospinal uid of patients with central nervous system dysfunction. Am J Med
endemic tropical spastic paraparesis. J Infect Dis 1983;74:837844.
1988;157:12261234. 117. Weiner SM, Otte A, Schumacher M, et al.
107. Jamieson DG, Mehta PD, Lavi E. Oligoclonal Diagnosis and monitoring of central nervous
immunoglobulin bands in cerebrospinal uid of a system involvement in systemic lupus
patient with lymphocytic choriomeningitis. Ann erythematosus: value of F-18 uorodeoxyglucose
Neurol 1986;19:386388. PET. Ann Rheum Dis 2000;59:377385.
108. Coret F, Vilchez JJ, Enguidanos MJ, Lopez- 118. Isshi K, Hirohata S. Differential roles of the anti-
Arlandis J, Fernandez-Izquierdo S. The presence ribosomal P antibody and antineuronal antibody
of oligoclonal bands in the cerebrospinal uid in in the pathogenesis of central nervous system
various neurologic diseases. Rev Clin Esp involvement in systemic lupus erythematosus.
1989;185:231234. Arthritis Rheum 1998;41:18191827.
109. Small P, Mass MF, Kohler PF, Harbeck RJ. 119. Georgescu L, Mevorach D, Arnett FC, Reveille
Central nervous system involvement in SLE. JD, Elkon KB. Anti-P antibodies and
Diagnostic prole and clinical features. Arthritis neuropsychiatric lupus erythematosus. Ann N Y
Rheum 1977;20:869878. Acad Sci 1997;823:263269.
110. Jennekens FG, Kater L. The central nervous 120. Bonfa E, Elkon KB. Clinical and serologic
system in systemic lupus erythematosus. Part 2. associations of the antiribosomal P protein
Pathogenetic mechanisms of clinical syndromes: antibody. Arthritis Rheum 1986;29:981985.
a literature investigation. Rheumatology 121. Teh LS, Bedwell AE, Isenberg DA, Gordon C,
(Oxford) 2002;41:619630. Emery P, Charles PJ, Harper M, Amos N,
111. Johnson RT, Richardson EP. The neurological Williams BD. Antibodies to protein P in systemic
manifestations of systemic lupus erythematosus. lupus erythematosus. Ann Rheum Dis 1992;
Medicine (Baltimore) 1968;47:337369. 51:489494.
References 283
122. Tzioufas AG, Tzortzakis NG, Panou-Pomonis E, aqueous humor and is not unique to
et al. The clinical relevance of antibodies to cerebrospinal uid. Exp Eye Res 1990;50:
ribosomal-P common epitope in two targeted 541547.
systemic lupus erythematosus populations: a 133. Rouah E, Rogers BB, Buffone GJ. Transferrin
large cohort of consecutive patients and patients analysis by immunoxation as an aid in the
with active central nervous system disease. Ann diagnosis of cerebrospinal uid otorhea.
Rheum Dis 2000;59:99104. Arch Pathol Lab Med 1987;111:756757.
123. Hay EM, Isenberg DA. Autoantibodies in central 134. Sloman AJ, Kelly RH. Transferrin allelic variants
nervous system lupus. Br J Rheumatol may cause false positives in the detection of
1993;32:329332. cerebrospinal uid stulae. Clin Chem
124. Meurman OH, Irjala K, Suonpaa J, Laurent B. A 1993;39:14441445.
new method for the identication of 135. Porter MJ, Brookes GB, Zeman AZ, Keir G. Use
cerebrospinal uid leakage. Acta Otolaryngol of protein electrophoresis in the diagnosis of
1979;87:366369. cerebrospinal uid rhinorrhoea. J Laryngol Otol
125. Arrer E, Meco C, Oberascher G, Piotrowski W, 1992;106:504506.
Albegger K, Patsch W. beta-Trace protein as a 136. Keir G, Zeman A, Brookes G, Porter M,
marker for cerebrospinal uid rhinorrhea. Clin Thompson EJ. Immunoblotting of transferrin in
Chem 2002;48:939941. the identication of cerebrospinal uid otorrhoea
126. Oberascher G. Cerebrospinal uid otorrhea and rhinorrhoea. Ann Clin Biochem 1992;29(Pt
new trends in diagnosis. Am J Otol 1988;9: 2):210213.
102108. 137. Normansell DE, Stacy EK, Booker CF, Butler
127. Arrer E, Gibitz HJ. Detection of beta 2-transferrin TZ. Detection of beta-2 transferrin in otorrhea
with agarose gel electrophoresis, immunoxation and rhinorrhea in a routine clinical laboratory
and silver staining in cerebrospinal uid, setting. Clin Diagn Lab Immunol
secretions and other body uids. J Clin Chem 1994;1:6870.
Clin Biochem 1987;25:113116. 138. Roelandse FW, van der Zwart N, Didden JH,
128. Bachmann G, Achtelik R, Nekic M, Michel O. van Loon J, Souverijn JH. Detection of CSF
Beta-trace protein in diagnosis of cerebrospinal leakage by isoelectric focusing on polyacrylamide
uid stula. HNO 2000;48:496500. gel, direct immunoxation of transferrins, and
129. Bachmann G, Nekic M, Michel O. Clinical silver staining. Clin Chem 1998;44:351353.
experience with beta-trace protein as a marker 139. Verheecke P. On the tau-protein in cerebrospinal
for cerebrospinal uid. Ann Otol Rhinol uid. J Neurol Sci 1975;26:277281.
Laryngol 2000;109:10991102. 140. Hamill RL, Woods JC, Cook BA. Congenital
130. Bachmann G, Petereit H, Djenabi U, Michel O. atransferrinemia. A case report and review of
Predictive values of beta-trace protein the literature. Am J Clin Pathol 1991;96:
(prostaglandin D synthase) by use of laser- 215218.
nephelometry assay for the identication of 141. Kanaoka Y, Ago H, Inagaki E, et al. Cloning and
cerebrospinal uid. Neurosurgery 2002;50: crystal structure of hematopoietic prostaglandin
571577. D synthase. Cell 1997;90:10851095.
131. Petereit HF, Bachmann G, Nekic M, Althaus H, 142. Sri Kantha S. Prostaglandin D synthase (beta-
Pukrop R. A new nephelometric assay for beta- trace protein): a molecular clock to trace the
trace protein (prostaglandin D synthase) as an origin of REM sleep? Med Hypotheses
indicator of liquorrhoea. J Neurol Neurosurg 1997;48:411412.
Psychiatry 2001;71:347351. 143. Felgenhauer K, Schadlich HJ, Nekic M. Beta
132. Tripathi RC, Millard CB, Tripathi BJ, Noronha trace-protein as marker for cerebrospinal uid
A. Tau fraction of transferrin is present in human stula. Klin Wochenschr 1987;65:764768.
284 Approach to pattern interpretation in cerebrospinal uid
144. Tumani H, Nau R, Felgenhauer K. Beta-trace 145. Bachmann G, Petereit H, Djenabi U, Michel O.
protein in cerebrospinal uid: a bloodCSF Measuring beta-trace protein for detection of
barrier-related evaluation in neurological perilymph stulas. HNO 2002;50:129133.
diseases. Ann Neurol 1998;44:882889.
9
Laboratory strategies for diagnosing
monoclonal gammopathies
Guidelines for clinical and laboratory evaluation Clues to detecting monoclonal gammopathies 296
of monoclonal gammopathies 285 Maintain an active le of all monoclonal
Initial screen by serum protein electrophoresis proteins 297
and Penta immunoxation 290 Screening and follow-up of MFLC 297
k/l Total (not free) quantication and the Monoclonals which may be difcult to diagnose 297
diagnosis of monoclonal gammopathies 290 Final words 299
Following monoclonal gammopathies 296 References 300
In the past decade there have been several changes manual method. These decisions need to be based
to the testing available for the clinical evaluation of on the technical demands, available skills and
patients suspected of harboring a monoclonal economic realities of the individual laboratory.
gammopathy. As detailed in Chapters 2 and 3, When reviewing the alternatives presented in this
improved resolution on gels and capillary zone chapter, consider how each strategy would work in
electrophoresis, together with automated and semi- your specic laboratory situation.
automated systems of electrophoresis and
immunoxation, have provided more efcient and
sensitive methods for these studies. Recent GUIDELINES FOR CLINICAL AND
immunochemical methods to measure free light LABORATORY EVALUATION OF
chains in serum and urine promise to enhance MONOCLONAL GAMMOPATHIES
further our ability to detect and follow monoclonal
proteins in these patients. Because of the sensitivity The major reason for incorporating serum protein
and efciency of the new methodologies, our labo- electrophoresis and immunoxation into the
ratory has changed our method of evaluating clinical laboratory is to improve the detection of
monoclonal gammopathies in the past few years. monoclonal gammopathies. In 1998, the College
These methods allow for the efcient detection and of American Pathologists Conference XXXII con-
immunochemical characterization of most mono- vened a panel of experts to provide recommenda-
clonal gammopathies in 1 day. tions for the clinical and laboratory evaluation of
There is no perfect strategy to detect monoclonal patients suspected of having a monoclonal gammo-
gammopathies. A laboratory with a relatively large pathy. As a result of that conference and several
volume of testing may prefer to use automated subsequent teleconferences, guidelines were agreed
screening methods, such as capillary zone electro- upon and reported in the Archives of Pathology
phoresis, a laboratory with a more modest volume and Laboratory Medicine.1 The ndings of the
may choose a semi-automated gel-based method expert panel were reported as nine guidelines and
and one with a smaller volume may prefer to use a published with detailed articles to provide a basis
286 Laboratory strategies for diagnosing monoclonal gammopathies
for selection of the best available strategy to detect, included in these guidelines, it is important to
characterize and follow patients with monoclonal recognize that individuals with immunodeciency
gammopathies. Since the publication of these diseases (congenital, acquired or iatrogenic) also
guidelines, I have received suggestions from several may present with monoclonal gammopathies
sources to improve the information presented. I and/or prominent oligoclonal banding that may
will include some of the suggestions, together with require sensitive techniques such as immunoblot-
a review of the guidelines in the following discus- ting for detection (see Chapter 6).711
sion. The reader is encouraged to review the guide- Guideline 2 recommends the use of immunoxa-
lines document and the articles that accompanied tion to dene the abnormal protein type. It also
them to esh out the details involved in selecting noted that in cases where serum protein electro-
patients to be tested, optimal methods for follow- phoresis screen is negative, but there is a high
ing those patients, and evaluation of unusual cir- clinical suspicion that the patient may harbor a
cumstances, such as cryoglobulins.16 plasma-cell dyscrasia, immunoxation with anti-
Guideline 1 recommends that electrophoretic sera against k and l may be useful to detect more
techniques capable of high-resolution be used to subtle M-proteins. Since publication of these
evaluate serum and urine on samples suspected of guidelines, the semi-automated immunoxation
containing a monoclonal gammopathy. Because technique with pentavalent antisera (Penta: one
there are many techniques available for electro- reagent antiserum that detects IgG, IgA, IgM, k
phoresis at various levels of resolution (see Chapter and l) has become available and provides the
2), the practical denition offered in one of the advantage of being able to detect heavy chain
papers that accompanied the guidelines was that diseases as well as monoclonal gammopathies in
the method provide crisp separation of the trans- one lane while providing a control serum protein
ferrin (b1) and C3 (b2) bands.5 Direct examination electrophoresis in the adjacent lane (Fig. 9.1).
of the gel itself was encouraged. However, at the Guideline 2 also recommended that immunoxa-
time of the conference, capillary zone electro- tion may be useful to investigate subtle bands that
phoresis (CZE) with high-resolution electrophero- cause asymmetry in the g-region or distortions of
grams, such as those presented in this book, were the b-region bands. Finally, Guideline 2 discour-
not widely used in clinical laboratories. Although aged the use of immunoelectrophoresis because it
capillary zone electrophoresis provides virtual gel is less sensitive than immunoxation, often more
images, the basic information is the electrophero- difcult to interpret and slower. The advantages
gram. This is the opposite of the situation with and disadvantages of those techniques are detailed
gels, where the gels are the basic information and in Chapter 3. Immunoblotting was not mentioned
the densitometric scans provide useful adjunctive in the Guidelines because it is not commonly used
information about the quantity of protein in in clinical laboratories of the individuals at the
specic regions of the gel. Use of methods provid- conference. However, it is a sensitive method to
ing low resolution was discouraged. Guideline 1 detect subtle M-proteins and oligoclonal bands. It
recommended use of densitometry (I currently is especially useful for evaluating patients with
include electropherograms) to quantify the M- monoclonal proteins as a result of immunode-
protein peak thereby providing an estimate of ciency problems, detecting monoclonal free light
tumor burden and a reproducible way to follow chains (MFLC) in urine and the more subtle bands
the patients course. that occur in patients with lymphoma and
Guideline 1 also noted that these recommenda- leukemia.1218
tions apply most commonly to clinical disorders Guideline 3 recommends that after an M-protein
that suggest the presence of malignant B- is identied, it should be followed by measurement
cell/plasma-cell lymphoproliferative disorders, as of the spike on densitometry (electropherogram
detailed in Chapter 6 of this book. Although not measurements are the equivalent) in preference to
Guidelines for clinical and laboratory evaluation of monoclonal gammopathies 287
immunonephelometric, immunoturbidimetric or free light chains (FLCs) in the serum and the urine,
radial immunodiffusion techniques that are stan- it may soon be possible to follow the MFLC by
dardized against polyclonal rather than mono- merely studying the FLC in a serum sample.2023
clonal immunoglobulins (see later). The only However, at present, there is relatively little infor-
exception is when a small M-protein is obscured by mation in the literature to acquaint clinicians with
a serum protein band, such as C3, in which case the possibilities of these techniques. One of our
measurement of the immunoglobulin type may be tasks in the laboratory is to encourage the use of
more accurate. This guideline also noted that there efcient sensitive new techniques. This seems to t
is no reason to repeat immunoxation on a pre- that description. By the end of 2003 an article will
viously characterized M-protein unless there has be published in our laboratory publication (The
been a change in the electrophoretic migration, Warde Report) that will be sent to clinicians who
development of an additional M-protein, or for use our laboratory. Please feel free to review mate-
conrmation of remission after treatment. Other- rial from the Warde Report (www.Wardelab.com).
wise, repeating immunoxation on a previously Guideline 6 recommends intervals that are useful
characterized M-protein is wasteful. When I to follow patients with previously identied mono-
receive requests for a redundant immunoxation I clonal gammopathies in serum or urine. The
append a note to the report that repeat immunox- Guidelines recognized that the follow-up time
ation on previously characterized M-proteins is not would vary depending on the clinical circum-
indicated. stances. For individuals that are actively being
Guideline 4 recommends that clinicians order treated for a lymphoplasmacytic neoplasm, exami-
measurements of serum immunoglobulins at the nation of both serum and urine are recommended
time of detection of the M-protein to determine the at 1- to 2-month intervals. At the other end of the
level of the uninvolved immunoglobulins. This spectrum are patients that have been diagnosed
Guideline also cautions that immunoglobulin mea- after careful clinical and laboratory evaluation to
surement should not be the sole primary screening have monoclonal gammopathy of undetermined
technique for M-proteins. Nephelometry and tur- signicance (MGUS). These individuals only
bidimetry were preferred to radial immunodiffu- require annual evaluation of serum and urine along
sion as methods to measure the total with their physical examination, unless there has
immunoglobulins (IgG, IgA, and IgM). been a change in their clinical condition.
Guideline 5 recommends that all patients sus- Guideline 7 addresses the issue of hyperviscosity
pected of having plasma-cell dyscrasias have a syndrome. On occasion, hyperviscosity requires
24-h urine sample studied as well as the serum. As intervention by emergency plasma exchange. By
discussed in Chapter 7, some studies indicate that performing serum viscosity and serum protein
an early morning void may be adequate for the electrophoresis prior to the rst plasma exchange,
initial screening if a 24-h urine sample is not one may be able to correlate the level of the M-
obtained.19 However, once a MFLC has been protein with symptoms. Then by following the
detected, a 24-h collection is currently the standard M-protein as described in Guidelines 3 and 6, the
to use as a baseline to follow these patients. clinician may anticipate the need for plasma
Guideline 5 also noted that the urine should not be exchange by a rise of the M-protein toward the
tested with dipsticks, sulfosalicylic acid or the level where the patient experiences viscosity
ancient acidied heat precipitation tests as screens. problems.
Urine should have the total protein measured by Guideline 8 recognizes that cryoglobulins are
one of the techniques discussed in Chapter 7 and unique problems in evaluating patients with mono-
then have both immunoxation and urine protein clonal gammopathies, as well as some autoimmune
electrophoresis performed on concentrated urine. and infectious conditions. Not all patients with M-
With the recent availability of immunoassays for proteins require study for cryoglobulins. This
288 Laboratory strategies for diagnosing monoclonal gammopathies
Guideline discouraged screening for cryoglobulins precipitate that forms in the 4C tube, but not in
in individuals with vague, non-specic symptoms. the 37C tube should be measured and character-
It is recommended that evaluation for cryoglobu- ized as described in Chapter 6.
lins be performed on individuals who have specic Guideline 9 reafrms the importance of the tech-
clinical features reecting cold sensitivity (detailed niques used in evaluating M-proteins. It recognized
in Chapter 6). For the evaluation, the initial collec- that either gel- or capillary-based electrophoretic
tion and transport of the specimen are known to be techniques of high-resolution were preferred.
critical.2,24,25 A 10 ml sample of blood should be Guideline 9 also recommended immunoxation and
collected in a prewarmed tube (37C) and trans- immunoselection (for documenting cases of heavy
ported to the laboratory at that temperature chain disease). Immunoselection, however, is a
(unless the venepuncture may be performed in the highly specialized technique and should only be per-
laboratory). After separating the serum from the formed by laboratories that have experience with its
clot, the sample should be split into two tubes, one use (see Chapter 3). The guidelines also recognize
kept at 37C the other at 4C or up to 7 days. A the potential value for immunosubtraction.
(a)
Guidelines for clinical and laboratory evaluation of monoclonal gammopathies 289
1 2 3 4
5 6 7 8
9 10 11 12
(b)
Figure 9.1 (a) Twelve serum samples are assayed on this Penta (pentavalent) screening immunoxation gel. Each sample is placed in two
lanes. The rst lane of each sample provides an acid xation of the serum protein electrophoresis for orientation, whereas the second
lane is the immunoxation with the Penta reagent (Sebia, Penta immunoxation). (b) Electropherograms by capillary zone electrophoresis
for the same twelve samples shown in (a). Note in specimen 1 a normal electropherogram is present. This corresponds to the normal
pattern of sample 1 in (a) and the Penta immunoxation in 1 shows a diffuse pattern. In contrast, in specimen 2 of the electropherogram,
there is a small, but obvious M-protein in the mid-g-region. In (a) this corresponds to the restriction seen in 2 and conrmation in 2 that
it is an immunoglobulin. A more difcult sample is number 9. On the electropherogram is shown only a bridging between b1- and b2-
globulins. However, in (a), the arrow in 9 demonstrates that this is a small M-protein that requires further characterization. Similarly, in
the electropherogram for specimen 11 a tiny restriction is seen in the fast g-region where we commonly see a brinogen band in samples
with inadequate clotting. However, the Penta of sample 11 in (a) shows that the restriction is caused by an immunoglobulin (arrow) and
deserves further evaluation. (Paragon CZE 2000.)
These guidelines form the basis for developing the relevant bands and distinguish them from bands
strategy that a laboratory could adopt in evaluat- that are not clinically relevant? Unfortunately,
ing serum and urine for monoclonal gammo- there is no such test. Small monoclonal gam-
pathies. The use of serum protein electrophoresis mopathies are occasionally part of infectious con-
with immunoxation or even more sensitive tech- ditions, often part of MGUS, which may be
niques such as immunoblotting increases the detec- transient.2631 All monoclonal gammopathies must
tion of small monoclonal gammopathies. Is there be followed for the remainder of the patients life
a screening technique that will detect clinically because some are a reection of a malignant,
290 Laboratory strategies for diagnosing monoclonal gammopathies
potentially malignant process or a dysregulation of immunoglobulin bands when using CZE is the
the immune system. presence of radiocontrast dyes.33 The proteins that
cause non-immunoglobulin restrictions are dis-
cussed in Chapter 6.
When using automated immunosubtraction, all
INITIAL SCREEN BY SERUM evaluations are for IgG, IgA, IgM, k and l.
PROTEIN ELECTROPHORESIS AND However, with immunoxation, we can tailor the
PENTA IMMUNOFIXATION reaction depending on the electrophoretic migra-
tion of the M-protein. A slow g-migrating restric-
Regardless of which other technique is used, the tion results in immunoxation with antisera
initial screen should be performed by electrophore- against IgG, k and l. If the monoclonal protein is
sis on both serum and urine. As mentioned in the not identied by this, immunoxation is repeated
guidelines, methods that provide crisp resolution with IgA, and IgM antisera; IgD and IgE are only
of the b-region bands are recommended. Currently, evaluated when the other antisera fail to identify
when serum is sent to our laboratory to evaluate the monoclonal protein.
for a possible monoclonal gammopathy by A fast g- or b-restriction is evaluated by
immunoxation, we rst perform serum protein immunoxation with antisera against IgG, IgA,
electrophoresis using a high-resolution capillary IgM, k and l. Monoclonal restrictions of k or l in
zone method. If no suspicious band is seen, we the serum with no corresponding restriction of
perform a Penta immunoxation that will detect IgG, IgA or IgM by immunoxation are usually
IgG, IgA, IgM, k and l (Fig. 9.1). The combination caused by light chain disease. However, these cases
of a normal serum protein electrophoresis and a require evaluation for possible IgD or IgE mono-
normal Penta immunoxation should detect virtu- clonal proteins. One may perform immunoxation
ally all monoclonal proteins in serum. When a for IgD and IgE or one may quantify the levels of
band suggesting an M-protein is seen on the initial these isotypes. A urine study for MFLC by
serum protein electrophoresis, we go straight to immunoxation is always recommended when a
immunosubtraction or immunoxation to identify monoclonal protein is suspected, regardless of the
the heavy and light chain type, foregoing the Penta serum immunoxation ndings. The early morning
screen. If the band is quite small, immunoxation void has been shown to be as sensitive as a 24-h
is chosen. Immunosubstraction by the capillary sample (although the latter is needed to quantify
zone method is highly reliable, requires very little the MFLC once it has been demonstrated).19
technologist time and can be performed in under
an hour.32 However, with small M-proteins it is
sometimes difcult to detect the effect of the
immunosubtraction, and for those I prefer k/l TOTAL (NOT FREE)
immunoxation. It is unusual that a case requires QUANTIFICATION AND THE
measurement of immunoglobulins to identify the DIAGNOSIS OF MONOCLONAL
monoclonal protein. GAMMOPATHIES
The availability of semi-automated screening
techniques such as the Penta is helpful in dealing Before the availability of automated high-resolution
with the occasional subtle restrictions or distortion screening, CZE-based automated immuno-
of b-region bands. I use the Penta immunoxation subtraction and the automated Penta immuno-
to demonstrate that the cause of the restriction is or xation, our laboratory used a combination of
is not an immunoglobulin. If it is an immunoglob- serum protein electrophoresis (using a high-
ulin, an immunoxation or immunosubtraction resolution method) and quantication of IgG, IgA,
will identify it. The most common cause for non- IgM with measurement of total k and l in serum to
k/l Total (not free) quantication and the diagnosis of monoclonal gammopathies 291
identify many of the M-proteins encountered. I no electropherogram, the heavy chain isotype of the
longer nd this to be a useful method to identify the monoclonal protein and the light chain type of the
M-protein. The method took advantage of the nor- monoclonal protein would be very close in con-
mal ratio k/l in the serum in our population, centration (unless a considerable amount of
1.22.6 ( 2 SD).34 When patients have a chronic MFLC is present). This is often not the case. For
infectious disease with a marked elevation of example, in Fig. 9.2, compare IgG, IgA, IgM, k,
immunoglobulins, there is a polyclonal expansion and l concentrations and the total protein of the
of B-cell clones. In this situation, although the total spike (by densitometry) in selected patients with
amounts of k and l are elevated, the k/l ratio usu- prominent monoclonal gammopathies. Note that
ally remains in the normal range. In contrast, when in some individuals, there is a very poor correla-
a monoclonal gammopathy is present owing to a tion between the concentration of the light chain
malignant process such as multiple myeloma, there and the heavy chain. In patient RK, the concentra-
usually is a marked alteration in the k/l ratio.34 This tion of the spike is similar to that of the k light
results from the combined effect of the increase in chain, but the IgM concentration is twice as high
the monoclonal protein of the single light chain type as the light chain concentration. In patient FC, the
and the suppression of normal polyclonal IgA concentration is twice as high as k and the
immunoglobulin-secreting clones that occur in spike is intermediate in amount. The concentra-
myeloma. However, a k/l ratio cannot be used to tion of l in patient SP is half again greater than
predict whether one is dealing with MGUS or mul- that of the spike, whereas the IgA heavy chain has
tiple myeloma. an intermediate concentration (this could be
Unfortunately, in some circumstances, serum explained by a large MFLC, but this was not the
protein electrophoresis, immunoglobulin quanti- case in patient SP). Yet, in contrast to these exam-
cation and the k/l ratio are inadequate to charac- ples, patients NG, MS, and EK have reasonably
terize a monoclonal gammopathy. For example, in good correlation between the densitometric scan
a double (biclonal) gammopathy, where one neo-
plastic clone secretes a k-containing monoclonal
IgG
protein and the other clone secretes a l-containing IgA
monoclonal protein, the k/l ratio may give mis- 9000
IgM
kappa
leading information. If one does not have a serum
Concentration (mg/dl)
lambda
protein electrophoresis pattern to demonstrate an spike mg/dl
6000
obvious double gammopathy, one could seriously
underestimate the process. Because of the presence
of normal polyclonal immunoglobulins, the k/l 3000
ratio is too insensitive to detect relatively small
monoclonal gammopathies. The k/l ratio has great
0
difculty detecting monoclonal gammopathies that RK FC SP NG MS EK
are smaller than 400 mg/dl.35 Immunoglobulin
Figure 9.2 Immunoglobulin and monoclonal spike (from
measurements should not be used without a serum densitometry) values on selected patients with monoclonal
protein electrophoresis to include or rule out the gammopathies. Note that in the last serum protein electrophoresis
presence of a monoclonal gammopathy. patients (NG, MS, and EK) there is reasonably good correlation of
When there are massive increases in the amount the spike and the monoclonal proteins as measured by
of a monoclonal protein, nephelometric determi- nephelometry. However, in the rst serum protein electrophoresis
patients (RK, FC, and SP), there is disparity between these values.
nations of a particular immunoglobulin compo-
Therefore, in following those patients, if one were to use the
nent can grossly overestimate or underestimate the densitometric information one time and the light chain or heavy
amount present.36, 37 Ideally, the amount of the chain number another time, one would not have consistent
serum M-protein measured by densitometry or results.
292 Laboratory strategies for diagnosing monoclonal gammopathies
these reagents, the values were less likely to agree y = 0.35612 + 0.90759x R^2 = 0.939
RK H/L 20
FC H/L
3 SP H/L
NG H/L
MS H/L
Heavy chain/light chain
EK H/L 10
2
0
1 0 10 20 30
Polyclonal samples kappa (Beckman) g/l
Sample (g/l)
Method A B C D E
a
Data from Bush and Keren.37
b
Gamma optical density (OD) = densitometric scans of the g-region of serum protein electrophoresis gels.
AA
clonal immunoglobulins. However, in extreme E A
D
E
C
D
B
concentrations of monoclonal immunoglobulins B
D C
C
A
and suppression of the normal polyclonal popula- D
tion, nephelometric techniques may give, at best, Polyconal solution Large precipitate
rough estimates of the amount of monoclonal
protein present.
The differences between results obtained on the
same monoclonal proteins with different reagent C E
B
B B B
antisera may result from the lack of standardiza- Anti-A
B D
tion of polyclonal reagents for nephelometry. Take D
+ Anti-C D D
A B D B
the example of determinants A, B, C, D, and E on Anti-E
the polyclonal immunoglobulins shown in the C D D
E D D
B
solution in Fig. 9.6. The reagent antibody shown
C C
D
has strong reactivity against determinants A, C and
E. It produces an excellent precipitate with poly- Monoclonal solution Small precipitate
Step 1. Analyse serum protein electrophoresis and IgG, IgA, IgM, k, and l
Step 2. Decide whether classication can be attempted:
if (a) single band and k/l > 0.75a and < 2.30a
or (b) multiple bands are present
or (c) no bands are present and k/l > 0.75a and < 2.30a
then report immunoxation is necessary
Step 3. Assign light chain class:
if k/l gt; 2.30a, then light chain class = k
if k/l < 0.75a, then light chain class = l
Step 4. Assign heavy chain class:
heavy chain = immunoglobulin in highest concentration (expressed as multiples
of the standard deviation of the reference range to correct for the normally higher
presence of IgG than IgA or IgM)
Step 5. Calculate monoclonal light chain concentration (LC:MC):
if LC = k, then
[k:MC] = [k:total] - {[1:total] 2.20a}
(accounting for the polyclonal k which is assumed to be present)
if LC = l, then [l:MC] = [l:total] - [k:total]/0.95a (accounting for
the polyclonal l, which is assumed to be present)
Step 6. Detect free light chains:
if [light chain:total] > [heavy chain:total] 0.8,a
then free light chains are present
Step 7. Check conrmatory criteria:
if [heavy chain] > [light chain:monoclonal component] 0.75a
and [heavy chain]SDb > - 0.35a
then report heavy chain:light chain (e.g. IgG k) and free light chain (e.g. k), if present
Step 8. Check for possible IgD or free light chain disease:
if [light chain:total] > [heavy chain:total] 0.80a
and [IgG]SD < - 0.80a
and [IgA]SD < - 0.80a
and [IgM]SD < - 0.80a
then report free light chain disease or IgD is possible and immunoxation is
necessary for classication
Step 9. Unable to classify:
report immunoxation is necessary for classication
a
Number derived from iterative procedure described in Jones et al.40, 41
b
[Heavy chain]SD = concentration expressed as multiple of the standard deviation of the reference range.
296 Laboratory strategies for diagnosing monoclonal gammopathies
when viewing a serum protein electrophoresis Tetrameric light chain disease must be followed by
pattern. Similarly, routine hematological screening serum samples, as the molecules are too large to
tests that demonstrate plasma cells on the differen- pass into the urine.45 As with the serum, any change
tial, rouleaux formation on the blood smear, or a in the electrophoretic migration or the develop-
bone marrow with > 10 per cent plasma cells are ment of other suspicious bands should trigger a
highly suspicious. These relevant laboratory nd- reinvestigation, complete with immunoxation to
ings should result in evaluation of serum and urine determine if the patient is developing a double
for a monoclonal gammopathy. gammopathy, or if the course of the condition has
altered.
When the clinician sends urine to be evaluated for
MFLC, we perform serum protein electrophoresis
MAINTAINING AN ACTIVE FILE OF on 50-fold concentrated urine. When the urine has
ALL MONOCLONAL PROTEINS considerable protein, it will not readily concentrate
to this level. As the guidelines recommend, the acid-
To conserve time and important laboratory
ied heat test, sulfosalicylic acid test and urine dip-
resources, one should maintain a le on all patients
sticks are inadequate screens for MFLC.46 Urine
with monoclonal proteins. When a sample is
immunoxation has replaced immunoelec-
received with a request to evaluate for monoclonal
trophoresis on all samples in our laboratory. With
protein, the old le is checked for previous nd-
immunoxation, a false negative may occur because
ings.43 Serum protein electrophoresis is performed,
the dilution used may be inappropriate; therefore,
and the monoclonal band (if in the g-region) is
occasionally, an additional dilution may be useful
quantied by densitometry.44 If there is no change
to be certain of the nal result when an antigen
in pattern, there is no reason to perform immuno-
excess effect is seen. In practice, we have found that
xation because the monoclonal protein was
such dilutions are rarely needed. Most commercial
characterized on the original sample. However, any
antibodies are sufciently strong that even when an
change in the pattern of migration, or the appear-
antigen excess situation occurs, it is usually obvi-
ance of a second band should result in an
ous. We end up repeating less than 1 per cent of our
immunoxation. By comparing the present results
urine immunoxation with additional dilutions.
with previous ones, the laboratory helps clinicians
When using immunoxation on urine, one must be
follow the patient by determining the change (if
aware of the ladder patterns that occur mainly with
any) of the monoclonal protein quantity. The same
k and to a lesser extent with l (Chapter 7). As
quantication method should be used to follow a
always, clinicians should be encouraged to send a
particular patient. For example, if the densitometric
second sample (preferably the early morning void
scan of the g-region was used to estimate the
or a 24-h urine) if the rst is negative and myeloma
concentration of a g-migrating monoclonal gammo-
or amyloidosis is still part of their differential diag-
pathy, continue to use this on subsequent samples,
nosis. I also recommend serum assays for FLC in
unless there is a change in the pattern (such as the
these situations because of its proven sensitivity in
development of a second monoclonal band).
detecting light chain myeloma and some cases of
non-secretory myeloma.42,47
Figure 9.7 Several samples show a brinogen band (indicated) because the samples were not allowed to clot completely before
separation of the serum. (Panagel system stained with Coomassie Blue; anode at the left.)
Final words 299
chain disease can be missed because the serum Polyclonal free light chains in the urine can occa-
spike is often in the b-region, where it may be sionally obscure the presence of a monoclonal
confused with other proteins, and because of gammopathy. When the immunoxation pattern
molecular size the light chains in this situation do of urine gives dense staining, a repeat immunoxa-
not appear in the urine as MFLC. Fibrinogen in tion should be performed in order to rule out a
an incompletely clotted specimen may be mis- monoclonal process.48 Uncommonly, it may not be
taken for a monoclonal protein (Fig. 9.7). This possible to distinguish between polyclonal and
error will be avoided by checking the serum for a monoclonal free light chains migrating in a ladder-
clot (if a small one is present, it indicates that banding pattern (see Chapter 7). In those instances,
some brinogen was left), and repeating the I advise the clinician to follow both the urine and
sample (always the best choice when one is not serum after 23 months to see if the process
certain of the diagnosis). If still uncertain, evolves.
immunoxation with Penta, or anti-light chain
antisera or even antisera against brinogen will
reveal the true nature of the protein. A possible FINAL WORDS
trap, however, lies in the non-specic reactivity of
some commercial antisera. We have found that The strategies reviewed in this chapter provide
commercial antisera against immunoglobulins efcient processing of specimens, beneting the
occasionally will react with brinogen, C3, C4 patient, the clinician, and the laboratory. They
and transferrin. For example, in the case shown help prevent inappropriate ordering and over-uti-
in Fig. 9.8, a brinogen band reacted with anti- lization of the laboratory. With the incorporation
sera against IgM. When this antisera was reacted of newer methods such as semi-automated gel-
against normal plasma, it gave the same line. based and CZE, immunosubtraction and Penta
Therefore, when evaluating new lots of reagents immunoxation screens, more rapid turnaround
used for immunoxation in the laboratory, it is a times are possible, which have been appreciated
good idea to test it against a sample of plasma to by clinicians. In general, when I discuss unusual
be sure that this will not be a problem, for cases with clinicians, they appear to like being
example, in specimens from patients on anticoag- involved in the evaluation of challenging speci-
ulants. mens. An occasional clinician has objected to the
report of a small monoclonal protein or hypogam-
maglobulinemia in patients that were clinically
well. However, a few of these prove to be associ-
ated with lymphoproliferative processes. We are
not able to determine the meaning of each abnor-
mality detected by these sensitive methods, but we
do know (as has been discussed) the important
conditions that need to be ruled in or ruled out.
When we have a very small monoclonal gam-
aM mopathy, about which we are uncertain after talk-
ing to the clinician and studying the urine, we
recommend repeating the evaluation in a few
months. If it represents an early neoplastic mono-
clonal process, it will still be there or may have
Figure 9.8 Immunoxation of a sample with a prominent
brinogen band gave a false positive reaction (arrow) with antisera progressed to the point where it will be readily
against IgM. (Panagel system stained with Coomassie Blue; anode detectable. Myeloma is not treated in the early
at the left). clinical stages, therefore, there has been no
300 Laboratory strategies for diagnosing monoclonal gammopathies
problem with delay in diagnosis. If it was merely a result, repeat the procedure, speak to the clini-
an oligoclonal expansion due to some infection or cian, perform further studies, occasionally send for
other process, it will likely have resolved by this a fresh sample, or send the sample off for reference
time. work.
It is important to view our diagnostic process in
context of the causes for monoclonal gammo-
pathies. For example, when classifying the etiology REFERENCES
of monoclonal gammopathies as caused by B-cell
malignancies and B-cell benign processes, Radl and
colleagues14 estimate that the newer, more sensitive 1. Keren DF, Alexanian R, Goeken JA, Gorevic PD,
techniques detect one true malignancy for 100 Kyle RA, Tomar RH. Guidelines for clinical and
benign processes. Typically, the monoclonal laboratory evaluation patients with monoclonal
product is much larger in malignant than in benign gammopathies. Arch Pathol Lab Med 1999;123;
processes. However, size of the gammopathy alone 106107.
is inadequate to distinguish MGUS from malignant 2. Kallemuchikkal U, Gorevic PD. Evaluation of
or potentially malignant processes. When dealing cryoglobulins. Arch Pathol Lab Med 1999;123;
with small monoclonal bands in a serum sample, I 119125.
recommend that the clinicians follow both the 3. Alexanian R, Weber D, Liu F. Differential
serum and the urine. diagnosis of monoclonal gammopathies. Arch
Occasionally, we receive requests for immunox- Pathol Lab Med 1999;123;108113.
ation on serum samples from children. These are 4. Goeken JA, Keren DF. Introduction to the report
relevant in the evaluation of children with a wide of the consensus conference on monoclonal
variety of immunodeciency diseases. Myeloma in gammopathies. Arch Pathol Lab Med 1999;123;
children is vanishingly rare. The normal k/l ratio 104105.
in children increases with age from about 1.0 at 5. Keren DF. Procedures for the evaluation of
4 months to the adult values of about 2.0 by mid- monoclonal immunoglobulins. Arch Pathol Lab
adolescence (15 years).49,50 When I receive a Med 1999;123;126132.
request for an immunoxation on a child, I con- 6. Kyle RA. Sequence of testing for monoclonal
tact the clinician to nd out what they are consid- gammopathies. Arch Pathol Lab Med 1999;123;
ering in the differential diagnosis of the case. 114118.
Invariably, in my experience, the clinicians are 7. Radl J. Monoclonal gammapathies. An attempt
concerned about the possibility of an immunode- at a new classication. Neth J Med 1985;28;
ciency disease such as WiskottAldrich syndrome, 134137.
Brutons X-linked agammaglobulinemia, an 8. Radl J. Differences among the three major
immunoglobulin subclass deciency, or a comple- categories of paraproteinaemias in aging man
ment deciency because the child has had recur- and the mouse. A minireview. Mech Ageing Dev
rent pyogenic infections. This is helpful 1984;28;167170.
information because individuals with various con- 9. Radl J, Liu M, Hoogeveen CM, et al.
genital and acquired immune disorders can have Monoclonal gammopathies in long-term
monoclonal or oligoclonal gammopathies with surviving rhesus monkeys after lethal
abnormal k/l ratios.51,52 irradiation and bone marrow transplantation.
Finally, a cooperative relationship between the Clin Immunol Immunopathol 1991;60;
laboratory and clinician is one of the critical points 305309.
to success in dealing with difcult cases. 10. Radl J, Valentijn RM, Haaijman JJ, Paul LC.
Contacting the ordering physician for more clinical Monoclonal gammapathies in patients
information can be helpful. When uncertain about undergoing immunosuppressive treatment after
References 301
renal transplantation. Clin Immunol immunoglobulin free light chains in serum and
Immunopathol 1985;37;98102. urine. Clin Chem 2001;47;673680.
11. Radl J. Light chain typing of immunoglobulins in 21. Abraham RS, Clark RJ, Bryant SC, et al.
small samples of biological material. Correlation of serum immunoglobulin free light
Immunology 1970;19;137149. chain quantication with urinary Bence Jones
12. Norden AG, Fulcher LM, Heys AD. Rapid protein in light chain myeloma. Clin Chem
typing of serum paraproteins by immunoblotting 2002;48;655657.
without antigen-excess artifacts. Clin Chem 22. Abraham RS, Charlesworth MC, Owen BA,
1987;33;14331436. et al. Trimolecular complexes of lambda light
13. Nooij FJ, van der Sluijs-Gelling AJ, Jol-van der chain dimers in serum of a patient with
Zijde CM, van Tol MJ, Haas H, Radl J. multiple myeloma. Clin Chem 2002;48;
Immunoblotting techniques for the detection of 18051811.
low level homogeneous immunoglobulin 23. Katzmann JA, Clark RJ, Abraham RS, et al.
components in serum. J Immunol Methods Serum reference intervals and diagnostic ranges
1990;134;273281. for free kappa and free lambda immunoglobulin
14. Radl J, Wels J, Hoogeveen CM. Immunoblotting light chains: relative sensitivity for detection of
with (sub)class-specic antibodies reveals a high monoclonal light chains. Clin Chem 2002;48;
frequency of monoclonal gammopathies in 14371444.
persons thought to be immunodecient. Clin 24. Gorevic PD, Galanakis D. Chapter 10.
Chem 1988;34;18391842. Cryoglobulins, cryobrinogenemia, and
15. Beaume A, Brizard A, Dreyfus B, Preudhomme pyroglobulins. In: Rose NR, Hamilton RG,
JL. High incidence of serum monoclonal Igs Detrick B, eds. Manual of clinical laboratory
detected by a sensitive immunoblotting technique immunology. Washington, DC: ASM Press,
in B-cell chronic lymphocytic leukemia. Blood 2002.
1994;84;12161219. 25. Keren DF, Di Sante AC, Mervak T, Bordine SL.
16. Withold W, Rick W. An immunoblotting Problems with transporting serum to the
procedure following agarose gel electrophoresis laboratory for cryoglobulin assay: a solution.
for subclass typing of IgG paraproteins in human Clin Chem 1985;31;17661767.
sera. Eur J Clin Chem Clin Biochem 1993;31; 26. Godeau P, Herson S, De Treglode D, Herreman
1721. G. Benign monoclonal immunoglobulins during
17. Withold W, Reinauer H. An immunoblotting subacute infectious endocarditis. Coeur Med
procedure following agarose gel electrophoresis Interne 1979;18;312.
for detection of Bence Jones proteinuria 27. Herreman G, Godeau P, Cabane J, Digeon M,
compared with immunoxation and quantitative Laver M, Bach JF. Immunologic study of
light chain determination. Eur J Clin Chem Clin subacute infectious endocarditis through the
Biochem 1995;33;135138. search for circulating immune complexes.
18. Koide N, Suehira S. Clinical application of Preliminary results apropos of 13 cases. Nouv
immunoblotting to the detection of monoclonal Presse Med 1975;4;23112314.
immunoglobulins and Bence Jones protein. 28. Keren DF, Morrison N, Gulbranson R. Evolution
Rinsho Byori 1987;35;638643. of a monoclonal gammopathy (MG) documented
19. Brigden ML, Neal ED, McNeely MD, Hoag GN. by high-resolution electrophoresis (HRE) and
The optimum urine collections for the detection immunoxation (IFE). Lab Med 1994;25:
and monitoring of Bence Jones proteinuria. Am J 313317.
Clin Pathol 1990;93;689693. 29. Kanoh T. Fluctuating M-component level in
20. Bradwell AR, Carr-Smith HD, Mead GP, et al. relation to infection. Eur J Haematol 1989;42;
Highly sensitive, automated immunoassay for 503504.
302 Laboratory strategies for diagnosing monoclonal gammopathies
51. Haraldsson A, Weemaes CM, Kock-Jansen MJ, 52. Haraldsson A, Jaminon M, Bakkeren JA,
et al. Immunoglobulin G, A and M light chain Stoelinga GB, Weemaes CM. Immunoglobulin G,
ratio in children. Ann Clin Biochem 1992;29(Pt A, and M light chain ratios in some humoral
3):271274. immunological disorders. Scand J Immunol
1992;36;5761.
10
Case studies for interpretation
This chapter allows one to interpret some cases cases processed with a variety of techniques to give
from my les. In most of these cases, I only know a broad overview of the possible patterns. In the
the age and sex of the individual at the time of the past, some readers have sent me correspondence by
initial interpretation. I hope that you will review mail. I welcome these, but if you want a quicker
the electrophoretic information, make your inter- response (perhaps) you could e-mail me at
pretation and, where appropriate, any suggestions kerend@wardelab.com. I will get back to you as
for the clinicians. The following discussion reviews soon as I can.
my interpretations on each case. I have included
Case studies for interpretation 305
IgG SPE
IgA K
(a) (b)
Figure 10.1 (a) Capillary zone electropherogram performed on serum from a 42-year-old man. (b) Immunosubtraction of the serum
from (a). (Paragon CZE 2000.)
INTERPRETATION FOR FIG. 10.1 The serum protein electrophoresis (SPE) recapit-
ulates the ndings in the capillary zone electro-
The absolute value of albumin is decreased. The pherogram. When serum preincubated with beads
a1-, a2-, and b-globulins are all decreased along coated with anti-IgG was evaluated, the beads
with albumin in their relative percentage of the subtracted out the massive spike, indicating that it
serum proteins. This results from the massive was overwhelmingly due to IgG. Neither preincu-
increase in g-globulin that has a total concentra- bation with beads coated with anti-IgA nor with
tion of 7.42 g/dl. Such a massive increase in beads coated with anti-IgM decreased the peak.
g-globulin causes one to consider multiple However, what about the light chains? The beads
myeloma. The peak of this massive g-globulin coated with anti-k removed most of the peak, but
region is relatively sharp, yet the base of the not all. How much was removed? Were not given
g-globulin extends throughout the entire g-globulin an exact amount but at least three-quarters of the
region. Nonetheless, such a massive increase peak seems to be gone. If this was a monoclonal
deserves an immunosubtraction or immunoxa- gammopathy, I would expect the peak to be
tion to rule out monoclonality. completely removed by the beads. Yet, could one
306 Case studies for interpretation
see an antigen excess situation where the amount exact quantities. However, if this were a massive
of antibody on the beads was insufcient to IgG k monoclonal gammopathy, one would
remove all of the antibody? Now consider the expect virtually all of it would be k-containing
results in the L (anti-l) electropherogram. At rst, and almost none l. Therefore, the anti- beads
it may appear that the anti-l beads have not should have had no noticeable effect. Since it did,
affected the massive g-globulin region. It is not this is a massive polyclonal increase in IgG. Such
readily apparent what has happened unless a increases are uncommon. The differential includes
frame of reference is established. I use albumin. viral infections such as human immunodeciency
Note that in the SPE electropherogram the - virus, hepatitis C virus, a combination of the two,
globulin peak is considerably higher than the and EpsteinBarr virus. T-cell immunoblastic lym-
albumin peak. But, in the anti-l electropherogram phoma (angioimmunoblastic lymphadenopathy)
the beads have clearly reduced the peak to well may also produce a picture of decreased albumin
below the height of albumin. This indicates that it and polyclonal increase in g-globulin with bg
has subtracted perhaps one-quarter of the peak. In bridging. Fortunately, these conditions can be
a polyclonal response the ideal is if the k-contain- tested for serologically and with molecular stud-
ing IgG accounts for two-thirds of the response ies. I have also seen an idiopathic case where no
and the l-containing IgG accounts for one-third. underlying cause has been found for a period of
Here, the ratio seems to be three-quarters to one- over 5 years.
quarter. One can quibble because we do not have
(a)
Figure 10.2 (a) Densitometric scan of serum protein electrophoresis from a 3-year-old girl with a history of a skin rash and anemia. IgA
level was elevated, IgM was low and IgG was in the upper limit of normal.
Case studies for interpretation 307
1 1
2 2
3 3
4 4
5 5
6 6
7 7
(b)
Figure 10.2 (contd) (b) Immunoxation of this patients serum (Paragon immunoxation). Case contributed by Joseph M. Lombardo.
K L
(a) (b)
Figure 10.3 (a) Four urine samples are shown with the anode indicated to the left. The bottom sample came from a 51-year-old man.
(Sebia b1,2 gel). (b) Immunoxation for anti-k (K) and anti-l (L) on the bottom sample from (a).
INTERPRETATION OF FIG. 10.3 that normally pass through the glomerulus and are
reabsorbed by the tubules. However, in the bg
The bottom urine sample (arrowed) has a tubular region there is a prominent band. It is suspicious
proteinuria pattern. The albumin band is small for a monoclonal free light chain (MFLC, or Bence
compared with the two samples above it that have Jones protein). However, it does not produce a
glomerular and tubular proteinuria. The numerous band with k or l immunoxation. We tested the
small bands and diffuse staining in the a- and b- urine for myoglobin and found that it had a value
regions are consistent with the many small proteins of 9642 ng/ml (normal is < 90 ng/ml).
Case studies for interpretation 309
Reference Ranges
Rel % g/dl
ALBUMIN 49.7 64.4 3.60 5.00
ALPHA 1 4.8 10.1 0.30 0.70
ALPHA 2 8.5 15.1 0.60 1.10
BETA 7.8 13.1 0.60 1.00
GAMMA 10.5 19.5 0.70 1.40
(a)
IgG SPE
IgA K
(b) IgM L
Figure 10.4 (a) Serum from a 73-year-old woman. (Capillary zone electropherogram, Paragon CZE 2000).
(b) Immunosubtraction of the serum from (a).
310 Case studies for interpretation
ELP G D E K L
2 4 2 2 4 4
(c) (d)
Figure 10.4 (contd) (c) Immunoxation to rule out IgD and IgE. (d) Urine: the small albumin band is to the left and the massive MFLC is
obvious.
Reference Ranges
Rel % g/dl
ALBUMIN 49.7 64.4 3.55 5.04
ALPHA 1 4.8 10.1 0.25 0.74
ALPHA 2 8.5 15.1 0.55 1.14
BETA 7.8 13.1 0.55 1.04
GAMMA 10.5 19.5 0.65 1.44
(a)
(b)
Figure 10.5 (a) Capillary zone electropherogram for an 80-year-old man (Paragon CZE 2000). (b) Penta (pentavalent) immunoxation of
serum from (a) (Sebia Penta Immunoxation).
312 Case studies for interpretation
IgG SPE
IgA K
IgM L
(c)
D E M K L
2 2 4 8 8
(d) (e)
Figure 10.5 (contd) (d) Immunoxation for IgD, IgE, IgM, k, and l. (e) Densitometric scan of urine.
INTERPRETATION FOR FIG. 10.5 though small band. Also, the small band between a2
and b is still present. The IgA immunosubtraction has
The electropherogram has a slight increase in the a2- no effect on the band between a2 and b, but the k
region, but otherwise the values are normal. immunosubtraction makes the small band in the fast
However, there is a small band at the interface end of the g-region more obvious (having removed the
between the a2- and b-regions. Sometimes a small k-containing polyclonal IgG, the non-k monoclonal
hemopexin band may be seen here, but this seems too gammopathy is more evident). The l immunosub-
large. In the g-region, there is an asymmetry to the traction discloses that both the band between a2 and
usually almost normal distribution of the immuno- b as well as the band at the anodal end of the g-region
globulins here. This asymmetry may represent a poly- has been removed. They are both l monoclonal
clonal increase in a subclass. However, because of gammopathies. But do they have a heavy chain
these two features, I wanted to see a pentavalent attached? We do not know yet. Thus, an immuno-
(Penta) immunoxation of this sample. The Penta xation was performed for IgM, k, l, IgD and IgE
immunoxation in Fig. 10.5b demonstrates two (Fig. 10.5d). Dilutions of the patients serum are
bands in the a2-region and a more subtle band in the shown in each lane. Here the band at the anodal end
g-region. This nding indicates that a monoclonal of the g-region is found to be an IgD l monoclonal
gammopathy is likely present. Immunosubtraction gammopathy and the one between a2 and b is found
was performed in Fig. 10.5c, but there was a prob- to be two bands (the cathodal one quite faint I hope
lem. No record was obtained from the capillary that that you see it), both l MFLC (possibly a monomer
contained the results of the IgM immunosubtraction, and a dimer). Lastly, the urine densitometric scan in
so we just see a at line. No conclusions may be drawn Fig. 10.5e demonstrates the urine contains a large
about IgM from this study. The SPE lane provides the amount of l MFLC (by immunoxation, the band
frame of reference for the other areas. In the IgG had no IgD heavy chain attached to it). The weak
immunosubtraction, much of the g-region is removed, staining of the where it is bound to IgD is not
but its anodal end (where we saw the asymmetry on unusual and occasionally leads to an erroneous
the initial electropherogram) now has a well-outlined, impression of heavy chain disease.
314 Case studies for interpretation
Fractions % Ref. %
ALBUMIN 44.7 < 45.3 67.7
ALPHA 1 6.3 2.9 6.8
ALPHA 2 16.9 > 6.2 14.9
BETA 17.4 8.1 18.0
GAMMA 14.7 8.8 24.5
(a) (b)
Figure 10.6 (a) Electropherogram for a 72-year-old man (Sebia Capillarys). (b) Penta (pentavalent) immunoxation screen of serum from
(a) (Sebia Penta immunoxation).
Case studies for interpretation 315
ELP G A M K L G A K L
(c) (d)
Figure 10.6 (contd) (c) Immunoxation of serum from same case. (d) Immunoxation of urine from same case.
INTERPRETATION OF FIG. 10.6 than the rest. In Fig. 10.6b, the Penta immunoxa-
tion screen demonstrates two bands in the a2- and
This sample was sent for an immunoxation of the b-regions. The g-region of the Penta has a few tiny
serum. In our laboratory, a request for immunox- oligoclonal bands consistent with the electrophero-
ation requires both a capillary zone electrophoresis gram. Because of the a2- and b-region bands, the
(CZE) and the serum Penta immunoxation. The immunoxation in Fig. 10.6c was performed. It
serum electropherogram has a slightly decreased demonstrates that the two k are MFLC. Another
percentage of albumin with a modest increase in the immunoxation (not shown) ruled out IgD and IgE.
a2-globulin. There is a small restriction present at The urine immunoxation in Fig. 10.6d has the
the anodal end of the b1-region. The g-region has a same nding as the serum, although the two k-
few irregularities consistent with oligoclonal bands. bands are more prominent. It is likely that the two
The most cathodal band is a little more prominent k MFLC represent a monomer and a dimer.
316 Case studies for interpretation
Reference Ranges
Rel % g/dl
ALBUMIN 49.7 64.4 3.60 5.00
ALPHA 1 4.8 10.1 0.30 0.70
ALPHA 2 8.5 15.1 0.60 1.10
BETA 7.8 13.1 0.60 1.00
GAMMA 10.5 19.5 0.70 1.40
IgG SPE
IgA K
(b) IgM L
Figure 10.7 (a) An 84-year-old woman who presented with a viral syndrome aches, pains, and dehydration (capillary zone
electropherogram, Paragon CZE 2000). (b) Immunosubtraction from same sample.
Case studies for interpretation 317
G K L
(c)
Reference Ranges
Rel % g/dl
ALBUMIN 49.7 64.4 3.55 5.04
ALPHA 1 4.8 10.1 0.25 0.74
ALPHA 2 8.5 15.1 0.55 1.14
BETA 7.8 13.1 0.55 1.04
GAMMA 10.5 19.5 0.65 1.44
(a)
Figure 10.8 (a) Serum from a 64-year-old woman. (Capillary zone electropherogram, Paragon CZE 2000.)
INTERPRETATION OF FIG. 10.8 removed almost the entire peak, including the
shoulder. The residual amount left in the g-region
This serum has a decrease in albumin and a after subtracting the IgA shows us the patients
markedly elevated g-globulin. The elevation is normal IgG. It is very low. Similarly, subtracting
unusual because it has a relatively broad base and the l removes the same amount of the peak leav-
a shoulder on the left side. These features are most ing a little normal k-containing IgG as a residual.
likely a monoclonal gammopathy. The cathodal It would not be a bad idea here to go back and
shoulder could represent the normal IgG or per- look at the immunosubtraction from Fig. 10.1.
haps a second M-protein. This question is quickly There, the massive IgG increase was polyclonal
answered by the immunosubtraction pattern in and immunosubtraction of k-containing
Fig. 10.8b. Using the SPE as the frame of refer- immunoglobulins removed about three-quarters of
ence, it is clear that subtracting out the IgA it while immunosubtraction of l-containing
Case studies for interpretation 319
IgG SPE
IgA K
IgM L
(b)
immunoglobulins removed about one-quarter of Fig. 10.1 had the same basic shape. The breadth of
it. Contrast that with the removal of well over the present IgA l as well as the presence of the
9095% of the spike by the relevant immunosub- shoulder likely reects two features of IgA mono-
traction in the present case. Further, the residual g- clonal gammopathies. They tend to be heavily gly-
region protein left after the k-immunosubtraction cosylated and thus have a broader migration. They
in the present case is normal polyclonal IgG, also tend to form multimers that could explain the
whereas both the k- and l-immunosubtractions in shoulder.
Reference Ranges
Rel % g/dl
ALBUMIN 49.7 64.4 3.55 5.04
ALPHA 1 4.8 10.1 0.25 0.74
ALPHA 2 8.5 15.1 0.55 1.14
BETA 7.8 13.1 0.55 1.04
GAMMA 10.5 19.5 0.65 1.44
(a)
Figure 10.9 (a) Serum from a 59-year-old man. (Capillary zone electropherogram, Paragon CZE 2000).
Case studies for interpretation 321
Figure 10.9 (contd) (b) Penta (pentavalent) immunoxation for this case and for Fig. 10.10 (Sebia Penta immunoxation).
INTERPRETATION OF FIG. 10.9 cases, it is clear that the Penta immunoxation can
detect even quite small M-proteins. So what is this
This sample has a decrease in albumin with a band? I called the clinician and learned that the
modest increase in g-globulin and possibly a slight patient had received a radiocontrast dye during the
bg bridge. There is a prominent band in the performance of a stent procedure. This is a
cathodal end of the a2-region. In Fig. 10.9b, the problem with the CZE technique because these
Penta immunoxation for this case (number 8 and dyes absorb at the same wavelength that peptide
8') demonstrates that the a2-region restriction is bonds do. As shown on the Penta gel, however,
not an immunoglobulin. On the gel-based Penta they will not stain with the protein dye. A table of
system, a protein dye is used and no such band radiocontrast dyes that are known to produce this
even appears in the acid-xed lane (8). In previous artifact is present in Table 2.3 (Chapter 2).
322 Case studies for interpretation
Reference Ranges
Rel % g/dl
ALBUMIN 49.7 64.4 3.55 5.04
ALPHA 1 4.8 10.1 0.25 0.74
ALPHA 2 8.5 15.1 0.55 1.14
BETA 7.8 13.1 0.55 1.04
GAMMA 10.5 19.5 0.65 1.44
Reference Ranges
Rel % g/dl
ALBUMIN 49.7 64.4 3.55 5.04
ALPHA 1 4.8 10.1 0.25 0.74
ALPHA 2 8.5 15.1 0.55 1.14
BETA 7.8 13.1 0.55 1.04
GAMMA 10.5 19.5 0.65 1.44
(b)
Figure 10.10 (a) Serum from a 54-year-old woman (capillary zone electropherogram, Paragon CZE 2000). (b) Serum from a 74-year-old
man (capillary zone electropherogram, Paragon CZE 2000).
Case studies for interpretation 323
Reference Ranges
Rel % g/dl
ALBUMIN 49.7 64.4 3.55 5.04
ALPHA 1 4.8 10.1 0.25 0.74
ALPHA 2 8.5 15.1 0.55 1.14
BETA 7.8 13.1 0.55 1.04
GAMMA 10.5 19.5 0.65 1.44
(c)
Figure 10.10 (contd) (c) Serum from a 67-year-old man (capillary zone electropherogram, Paragon CZE 2000).
INTERPRETATION FOR FIG. 10.10 However, in this case it does not. Look at the same
area just anodal to the transferrin band in Figures
Serum 10.10a is has a decrease in albumin with a 10.10b, and 10.10c. Both of these also contain
relative increase in the percentage of a1- and a2- some type of shoulder or suggestion of a double
globulins. The g-globulin is considerably decreased transferrin band. All three samples were run in the
which, together with the low albumin, suggests a same capillary as the rst sample. By processing
protein loss pattern. However, it could also be con- these specimens in other capillaries, the band dis-
sistent with immunosuppression due to light chain appears. It is an artifact. Both the Paragon CZE
multiple myeloma. Also, MFLC will cause renal 2000 and the Sebia Capillarys have several capil-
damage possibly resulting in a pattern like this. laries that serum passes through. Occasionally, if a
When I looked at the b-region expecting to see a highly lipemic sample is processed, the capillary
decrease in transferrin to conrm an impression of may give spurious results for the next few assays.
an acute-phase pattern, I noticed a discrete band Furthermore, as the capillary ages similar artifacts
just anodal to transferrin (b1). The same sample is can occur. Therefore, when I observe an unusual
processed in lanes 5 and 5 on the Penta gel from band in a sample, I check other samples that passed
Fig. 10.9b. There is no evidence of a monoclonal through that particular capillary on the same run
band. This could represent a transferrin variant. prior to rendering my interpretation.
324 Case studies for interpretation
(a) (b)
Figure 10.11 (a) Electrophoresis of ve urine samples concentrated up to 50-fold is shown for interpretation (Sebia b1,2 gel). (b)
Electrophoresis of ve more urine samples concentrated up to 50-fold is shown for interpretation (Sebia b1,2 gel)
Reference Ranges
Rel % g/dl
ALBUMIN 49.7 64.4 3.60 5.00
ALPHA 1 4.8 10.1 0.30 0.70
ALPHA 2 8.5 15.1 0.60 1.10
BETA 7.8 13.1 0.60 1.00
GAMMA 10.5 19.5 0.70 1.40
(a)
G A K L
(b) (c)
Figure 10.12 (a) Capillary zone electropherogram for a 41-year-old woman (Paragon CZE 2000). (b) Penta (pentavalent) immunoxation
on the sample from (a). (c) Immunoxation on the same sample.
INTERPRETATION OF FIG. 10.12 C3 bands both are slender and symmetrical. Here,
the transferrin band is broad and slightly asymmet-
The capillary zone electropherogram has two rical showing a subtle cathodal shoulder. The
unusual features. First the transferrin band appears second area of concern was the bg bridge. A bg
to be broader than usual. Typically, transferrin and bridge usually reects a polyclonal increase in IgA.
326 Case studies for interpretation
It can be seen in a wide variety of circumstances, the location of transferrin). There is a second
but one of the more common is cirrhosis and some- broader area in the bg-region that likely reects
times hepatitis. This pattern looks like neither of the bridging described in the electropherogram. In
those. Further, considering the normal amount of Fig. 10.12c, the immunoxation demonstrates that
total g-globulin, the bridge has the unusual appear- the restriction in the transferrin-region was due to
ance of going up into the b-region rather than just the small IgG l monoclonal gammopathy. The
being a at bridge across. Because of these con- bridge, however, is a polyclonal increase in IgA.
cerns we performed the Penta immunoxation This can be determined by looking at the k and l
shown in Fig. 10.12b. It demonstrates an lanes. There is a broad increase in both light chain
immunoglobulin band (likely an M-protein) at the types in the same area of migration that has an
anodal end of the immunoxation (approximately increase in the IgA.
Reference Ranges
Rel % g/dl
ALBUMIN 49.7 64.4 3.60 5.00
ALPHA 1 4.8 10.1 0.30 0.70
ALPHA 2 8.5 15.1 0.60 1.10
BETA 7.8 13.1 0.60 1.00
GAMMA 10.5 19.5 0.70 1.40
(a)
Figure 10.13 (a) Serum protein electrophoresis from 51-year-old woman (capillary zone electropherogram, Paragon CZE 2000).
Case studies for interpretation 327
IgG SPE
IgA K
IgM L
(b)
SPE anti-
fibrinogen
G A K L
(c) (d)
Figure 10.13 (contd) (c) Immunoxation of same serum (Sebia immunoxation). (d) Immunoxation for brinogen.
INTERPRETATION OF FIG. 10.13 been subtracted out; however, that is where poly-
clonal IgA normally migrates and one would
This serum has a decrease in albumin, a relative assume that this would decrease the overall height
increase in the percentage of a1-globulin, a modest of the b2-region because of that. The shoulder to
increase in a2-globulin, and a considerable increase C3 remains, however, even in the IgA immunosub-
in b-globulin. The transferrin band is decreased traction. Because of this, we performed the
and is consistent, together with the above ndings, immunoxation shown in Fig. 10.13c. Since we
with an acute-phase response. However, forming a already knew that the IgM had no effect on the
cathodal shoulder to the C3 band is a broad, but restriction, we did not include this in the
denite restriction. The g-region has a slight immunoxation. There is a subtle restriction in the
restriction in its anodal end. But the concentration middle of the IgG lane that corresponds to a
is in the normal range. Because of the presumption similar subtle restriction in the middle of the k
that the slow b-region band was a monoclonal lane. This may represent a tiny IgG k monoclonal
gammopathy, I skipped the Penta step and per- gammopathy that is most likely part of the acute-
formed the immunosubtraction shown in Fig. phase response, but what is responsible for the
10.13b. The SPE electropherogram provides the shoulder? Finally, I performed an immunoxation
frame of reference for the immunosubtraction for brinogen (Fig. 10.13d). The mystery protein is
study. It demonstrates the unusual cathodal cling- brinogen. With the earlier versions of the Paragon
ing to the C3 band. Unfortunately, it is not CZE 2000 brinogen did not show up as a band,
removed by any of the immunosubtractions. There but with the new buffer systems it does.
is a slight decrease in the b2-region after IgA has
Case studies for interpretation 329
Reference Ranges
Rel % g/dl
ALBUMIN 49.7 64.4 3.60 5.00
ALPHA 1 4.8 10.1 0.30 0.70
ALPHA 2 8.5 15.1 0.60 1.10
BETA 7.8 13.1 0.60 1.00
GAMMA 10.5 19.5 0.70 1.40
(a)
(b)
Figure 10.14 (a) Capillary zone electropherogram of a 56-year-old man (Paragon CZE 2000). (b) Measurement of suspicious band in (a).
330 Case studies for interpretation
IgG SPE
IgA K
IgM L
(c)
(a)
Figure 10.15 (a) Capillary zone electropherogram on a 60-year-old woman ( Sebia Capillarys).
Case studies for interpretation 333
IgG SPE
IgA K
IgM L
(b)
Figure 10.15 (contd) (b) Immunosubtraction of serum from (a) (Paragon CZE 2000).
334 Case studies for interpretation
G K L
(c)
(a)
Figure 10.16 (a) Serum protein electrophoresis on four samples (Paragon SPE2 system stained with Paragon Violet).
336 Case studies for interpretation
(b) (c)
(d) (e)
Figure 10.16 (contd) (b) Densitometric scan of sample in the top lane of (a), which was from a 61-year-old woman. (c) Densitometric
scan of sample in the second lane of (a), which was from an 81-year-old man with dizziness. (d) Densitometric scan of sample in the third
lane of (a), which was from a 29-year-old man. (e) Densitometric scan of sample in the bottom lane of (a), which was from a 78-year-old
man.
Case studies for interpretation 337
(f)
Figure 10.16 (contd) (f) A somewhat underexposed second photograph of the gel in (a).
INTERPRETATION FOR FIG. 10.16 mation helps here by letting us know they were
quantitatively in the normal range.
Top lane
Second lane
The top lane in Fig. 10.16a is a normal serum. The
gel stained relatively lightly; in this case a1-regions The albumin band is moderately decreased as is the
do not show to advantage, especially in the top and b2-region. In the g-region, there is a relatively large
bottom lanes. However, the densitometric infor- spike. By immunoxation, the spike is character-
338 Case studies for interpretation
Bottom lane
Third lane
All of the fractions are normal except g-globulin.
The albumin band is decreased and there is a There is a tiny restriction in the slow g-region.
marked increase in the g-globulin. Because of the This can be seen directly on the gel and by not-
extreme density of the staining of the g-globulin ing the sharp drop-off at the cathodal end of the
region, a second photograph of this gel is given densitometric scan (Fig. 10.16e). Compare this to
(Fig. 10.16f). This photograph is somewhat under- the smoother decrease in the cathodal end of the
exposed to emphasize the presence of oligoclonal normal sample in Fig. 10.16b. This very tiny
bands in the g-region. These bands can be appreci- restriction is present in an otherwise normal g-
ated by holding the gel up to a strong light. On the globulin region. I noted that a tiny g-globulin
densitometric scan, the irregularities in the g-region restriction was present and that the signicance
are the counterparts to these oligoclonal bands. of such tiny restrictions is unclear. I recom-
The presence of a massive polyclonal increase in g mended that this serum and a urine have an
with oligoclonal bands usually indicates profound immunoxation. I also recommended that the
infectious disease. Rarely, patterns like this one serum be re-evaluated in 36 months to see if the
are seen in patients with angioimmunoblastic process resolves.
(a)
Figure 10.17 (a) Serum protein electrophoresis on two samples is shown (Paragon SPE2 system stained with Paragon Violet).
Case studies for interpretation 339
(b) (c)
(d)
Figure 10.17 (contd) (b) Densitometric scan of sample in top lane of (a), which was serum from a 59-year-old woman. (c) Densitometric
scan of the sample in bottom lane of (a), which was serum from a 28-year-old woman. (d) Immunoxation of serum from sample from
bottom lane of (a) (Paragon system stained with Paragon Violet; anode at the top).
340 Case studies for interpretation
IgM K L IgM K L
Untreated Treated
(e)
Figure 10.17 (contd) (e) Immunoxation of same sample as in (d) before and after removal of IgG fraction by a commercial column
(Paragon system stained with Paragon Violet; anode at the top).
INTERPRETATION FOR FIG. 10.17 unusual appearance of the b-regions of the densito-
metric scans. The b1-lipoprotein band creates a
third band that is confusing when looking at the
Top lane densitometric scans in the absence of the gel. This
is one of the reasons I always show the gel for com-
The top lane in Fig. 10.17a is a normal serum. parison. With CZE this is not usually a problem.
Note that the densitometric scan in Fig. 10.17b As demonstrated in Chapter 4, b1-lipoprotein usu-
(and the other) barely separates the a1-region from ally migrates in the slow a2-region and the b-region
albumin, yet by looking at the gel, one can clearly bands are transferrin and C3. Unlike the capillary
see the a1-region band. Contrast this with the zone electropherograms where I demonstrated the
clarity of earlier CZE gures depiction of the a1- ability to detect subtle monoclonal proteins by
region. With gel-based techniques, it was usually identifying distortions of the usually sharp and
impossible to see the a1-acid glycoprotein band, slender transferrin and C3 bands, the muddle of
but this is commonly seen with CZE, especially in the densitometer b-region patterns makes this very
cases with acute-phase reactions (see Chapter 5). hard. However, the gels themselves provide a rea-
Another important feature of these samples is the sonably good look at this region.
Case studies for interpretation 341
(a)
Figure 10.18 (a) Serum protein electrophoresis on four samples is shown (Paragon SPE2 system stained with Paragon Violet).
Case studies for interpretation 343
(b) (c)
(d) (e)
Figure 10.18 (contd) (b) Densitometric scan of sample in the top lane in (a) (54-year-old woman). (c) Densitometric scan of sample in
the second lane in (a) (55-year-old man). (d) Densitometric scan of sample in the third lane in (a) (51-year-old man). (e) Densitometric
scan of sample in the bottom lane in (a) (76-year-old woman).
344 Case studies for interpretation
INTERPRETATION FOR FIG. 10.18 anodal slurring (often seen when the bilirubin is
elevated), and the a- and b-globulin fractions are
normal.
Top lane
Although all of the densitometric scan informa- Third lane
tion is normal, note how much better the informa-
tion is by looking directly at the gel. The gel in the When comparing the major bands on this gel, it
top lane of Fig. 10.18a shows a dense b1-lipopro- becomes apparent that the anodal edge of albumin
tein band in the middle of the b-region with the for the case in lane three has a slightly faster anodal
normal transferrin and smaller normal C3 band migration than the other albumin bands on this
also visible. However, the densitometric scan of gel. In addition, there is a decrease in the g-globu-
this sample in Fig. 10.18b indicates the spike due lin region. The other bands are unremarkable. The
to the b1-lipoprotein, but it is difcult to perceive interpretation of this case notes the presence of the
the transferrin band and the C3 band does not anodal slurring and its possible relation to drug
show up well at all. Look back to the earlier cases binding (commonly antibiotics or heparin) and
that used electropherograms to contrast the con- emphasizes the isolated hypogammaglobulinemia.
sistent crisp b2-region bands that these electro- I recommend a urine be evaluated for the presence
pherograms display. For the present case, direct of MFLC in any case of isolated hypogammaglob-
inspection of the g-globulin region of the gel dis- ulinemia.
closes the presence of at least two small restric-
tions. These bands are most likely related to an
inammatory process, but there is no correspond- Bottom lane
ing acute-phase reaction. I recommended a follow-
up sample in 36 months. Despite the presence of an obvious mid-g mono-
clonal gammopathy in case D, there is no increase
in the g -globulin region overall. This patient had
Second lane been being followed for a known IgG k mono-
clonal gammopathy (characterized by a previous
There is polyclonal increase in g-globulins with bg immunoxation) for many years. We always
bridging. The presence of the bg bridge usually examine our records of patients with monoclonal
corresponds to a polyclonal increase in IgA. gammopathies when a new sample is sent to us.
Although the bg bridge is traditionally associated There had been no signicant change in the
with cirrhosis, this pattern is also seen in other amount or migration of this monoclonal protein
individuals with chronic inammatory processes since the examination 12 months previously. This
that share an increase in IgA. Autoimmune condi- was noted on the report. A repeat was not neces-
tions and infections may have this pattern. This sary because there was no change in the migration
patient has normal albumin with no evidence of of the monoclonal protein.
Case studies for interpretation 345
(a)
(b) (c)
Figure 10.19 (a) Serum protein electrophoresis on three samples (Paragon SPE2 system stained with Paragon Violet). (b) Densitometric
scan of sample in the top lane in (a) (76-year-old man). (c) Densitometric scan of sample in the second lane in (a) (61-year-old man).
346 Case studies for interpretation
Middle lane
There is a slight increase in the absolute amount of
a1-globulin and a relative increase in both a1- and
a2-globulins. This is consistent with a mild acute-
phase reaction pattern. The transferrin band
appears normal, however, and no C-reactive
protein band is seen.
Bottom lane
There is a moderate increase in the a1-globulin,
which has not separated as well as I like to see from
the albumin band. Once again, looking at the pic-
ture of the gel in Fig. 10.19a provides a better view
of this than the densitometric scan which does,
however, provide quantitative information. This
(d)
may be due to an increase in the a1-acid glycopro-
Figure 10.19 (contd) (d) Densitometric scan of sample in the tein (orosomucoid) which migrates just anodally to
third lane in (a) (41-year-old woman). a1-antitrypsin. But unlike the capillary zone elec-
tropherograms, even in cases with a marked
increase in this band, it is difcult to see on most
gel-based systems. Both proteins increase as part of
the acute-phase response. Alternatively, there may
be an increase in a1-lipoprotein which may obscure
the region between albumin and a1-antitrypsin.
The a2- and transferrin-bands are both at relatively
high normal levels. Therefore, the combination of
elevated a1-lipoprotein, a1-antitrypsin and trans-
INTERPRETATION FOR FIG. 10.19 ferrin may reect a hyperestrogen effect rather than
an acute-phase response. The most important nd-
ing in the case is that of a slow-migrating g band. A
Top lane band in this position is almost always due to a
monoclonal gammopathy. In urine from a patient
There is anodal slurring of the albumin band in with myelogenous leukemia, this band could be
Fig. 10.19a. Although it is difcult to discern by due to lysozyme. The previous records from this
looking at the photograph of this gel, the patient indicated that she had an identical slow g
densitometric information in Fig. 10.19d demon- band 6 months previous to this sample.
strates a decrease in albumin. The other bands Immunoxation at that time revealed an IgG k
are normal, although there is a relative increase monoclonal gammopathy. There was no change in
in the g-globulin region. In addition, there is a migration or amount of the monoclonal protein on
slight bg bridge in this sample. When I spoke to the present sample, therefore, immunoxation was
the clinician about this patient, I learned that the not repeated. Urine did not contain a MFLC.
patient had hyperbilirubinemia and cirrhosis. Annual follow-up was recommended.
Case studies for interpretation 347
(a)
Figure 10.20 (a) Serum protein electrophoresis on four samples is shown. (Paragon SPE2 system stained with Paragon Violet).
348 Case studies for interpretation
(b) (c)
(d)
Figure 10.20 (contd) (b) Densitometric scan of sample in the second lane in (a) (35-year-old man). The tube was marked as grossly
hemolyzed. (c) Densitometric scan of sample in the third lane in (a) (45-year-old man). (d) Densitometric scan of sample in the bottom
lane in (a) (75-year-old man).
Case studies for interpretation 349
(e)
Figure 10.20 (contd) (e) Immunoxation of sample in (d) (Paragon SPE2 system stained with Paragon Violet).
INTERPRETATION FOR FIG. 10.20 specimen from a patient with a polyclonal increase
in g-globulin. The serum was bright red. The
markedly increased b1-region band is too large for
Top lane transferrin in an iron-decient patient. It could rep-
resent a monoclonal gammopathy. That possibil-
This is a normal electrophoretic pattern (no ity, however, is unlikely in the face of the gross
densitometry shown). hemolysis and the polyclonal increase in g. If one is
uncertain of the nature of such a band, an
immunoxation will rule out a monoclonal
Second lane gammopathy.
albumin and a1-antitrypsin with the serum above large for C3 under any circumstances. On these
and below. There is also a modest polyclonal gels, brinogen migrates between the b- and the g-
increase in g-globulin with no bg bridging. This regions. Therefore, the large band in the C3 region
pattern is most consistent with a chronic inam- is almost certainly another monoclonal band. In
matory process. addition to the large g-region spike mentioned
above, there is a decrease in the staining density of
the remaining g-globulin. The immunoxation of
Bottom lane this patients sample (Fig. 10.20e; Paragon SPE2
system stained with Paragon Violet) demonstrates
The serum has a decrease in albumin and an that the b-band is due to l MFLC and the mid-g
increase in both b- and g-globulins by the densito- band is due to an IgG l monoclonal protein. Upon
metric scan information in Fig. 10.20d. Although discussing this case with the clinician, I learned
ones attention is drawn to the massive g-globulin that this patient has multiple myeloma. A 24-h
spike, one must not ignore the other signicant urine specimen was recommended to quantify the
ndings on this gel. The b2-region band is far too amount of MFLC.
(a)
Figure 10.21 (a) Serum protein electrophoresis on three samples (Paragon SPE2 system stained with Paragon Violet).
Case studies for interpretation 351
(b) (c)
(d)
Figure 10.21 (contd) (b) Densitometric scan of sample in the top lane in (a) (61-year-old man with lymphocytosis). (c) Densitometric
scan of sample in the middle lane in (a) (69-year-old man). (d) Densitometric scan of sample in the bottom lane in (a) (83-year-old
woman).
352 Case studies for interpretation
(a)
Figure 10.22 (a) Serum protein electrophoresis on four samples (Paragon SPE2 system stained with Paragon Violet).
354 Case studies for interpretation
(b) (c)
(d) (e)
Figure 10.22 (contd) (b) Densitometric scan of sample in the top lane in (a) (68-year-old man). (c) Densitometric scan of sample in the
second lane in (a) (67-year-old man). (d) Densitometric scan of sample in the third lane in (a) (55-year-old man). (e) Densitometric scan of
sample in the bottom lane in (a) (43-year-old man).
Case studies for interpretation 355
(a)
Figure 10.23 (a) Serum protein electrophoresis on three samples (Paragon SPE2 system stained with Paragon Violet).
Case studies for interpretation 357
Top lane
Typical for lyophilized controls, the C3 (b2-region)
band stains very weakly. No densitometric scan is
shown.
Middle lane
This serum has a markedly decreased serum
albumin and a moderate decrease in g-globulin. In
addition, there is a marked increase in the a2-glob-
ulin region and a prominent b1-lipoprotein band.
These features are typical of nephrotic syndrome.
(a)
(b)
Figure 10.24 (a) Serum protein electrophoresis on three samples is shown. Case X is from a 46-year-old man with liver disease.
(Panagel system stained with Amido Black). (b) Densitometric scan of case X.
Case studies for interpretation 359
INTERPRETATION FOR FIG. 10.24 human; I know of no disease that could cause
such a pattern.
The top and bottom serum sample in this gure This is an artifact due to partial denaturation of
are normal. The electrophoretic pattern (X) is the serum proteins. On questioning the techno-
markedly abnormal. Two dense bands are seen in logist, it became clear that at the same time he was
the albumin region. The densitometric scan in monitoring this serum protein electrophoresis gel,
Fig. 10.24b demonstrates that the two albumin he used adjacent tubes for the total protein deter-
bands are unequal in height. Further, the cathodal mination (biuret technique). We suspected that a
albumin band has a broad shoulder that covers drop of biuret reagent may have fallen into the
up the a1-antitrypsin band. There is a diffuse haze patients sample. We were able to reproduce this
between the second albumin region band and artifact by placing a drop of biuret reagent into the
transferrin. No distinct a2 band is seen. C3 is serum and performing electrophoresis. When
absent although a diffuse haze extends from the unusual samples like this one are found and the
transferrin band to about the origin. No staining interpretation is unclear, a repeat analysis should
is found in the g-globulin region. One interpreter be done. In this case, the repeat revealed a normal
thought that this was a bisalbuminemia in a electrophoretic pattern. If the unusual pattern
patient with severe liver disease. When I rst saw repeats, one should call the clinician and ask for a
this pattern, I doubted that it came from a new sample.
360 Case studies for interpretation
(a)
(b)
Figure 10.25 (a) Serum protein electrophoresis on two samples. Case X is from a 72-year-old woman (Paragon SPE2 stained with
Paragon Violet). (b) Densitometric scan of serum from case X.
Case studies for interpretation 361
Figure 10.25 (contd) (c) Immunoxation of serum from case X (Paragon system stained with Paragon Violet; anode at the top).
INTERPRETATION FOR FIG. 10.25 ciency. The immunoxation in Fig. 10.25c reveals
an IgG k monoclonal gammopathy. This is an
The electrophoretic pattern for both samples unusual location for an IgG monoclonal gammo-
demonstrates irregularities (best seen at the catho- pathy. I had expected to nd an IgA monoclonal
dal end of the albumin bands) that reect inade- protein in this location. Note that at the dilution
quate blotting of the gel prior to application of the used, the broad polyclonal nature of the IgA pro-
sample. The top sample is normal. The bottom tein is obvious. Only faint staining is seen in the l
sample (X) has an enormous band in the transfer- reaction. This is too dilute for optimal conditions
rin region. The densitometric scan information in with the reagent we were using at the time. It should
Fig. 10.25b conrms the visual impression of the have been diluted to about half the concentration of
large b1-region band. A transferrin band would k for a better reaction (1:4 or 1:5). I also recom-
never be this large, even in the presence of iron de- mended that urine be evaluated for MFLC.
362 Case studies for interpretation
(a)
(c)
Figure 10.26 (contd) (c) Immunoxation of serum from case X (Paragon system stained with Paragon Violet; anode at the top).
(a)
IgA K L IgA K L
1:12 1:8 1:4
Serum Urine
(b)
Figure 10.27 (a) Serum (top lane) and urine (bottom lane) protein electrophoresis from a 76-year-old man. (Paragon SPE2 stained with
Paragon Violet). (b) Immunoxation of serum and urine from serum and urine of case shown in (a) (Paragon system stained with Paragon
Violet; anode at the top).
Case studies for interpretation 365
INTERPRETATION OF FIG. 10.27 and urine patterns described above are the result of
the presence of biclonal gammopathy and one with
The a1-region band in the serum (top lane) stains the further complication of monoclonal free light
weakly. The a2-region is darker and broader than chains. IgA k and IgA l monoclonal proteins are in
normal. There is a diffuse haze between the a2- the serum while k and l MFLC are in the urine.
region band and the dense sharp band (presumed Note how much more densely the Bence Jones pro-
to be b1-lipoprotein). This serum was not hemol- teins stain in the immunoxation than in the urine
ysed. The b1-region band (transferrin) is beneath protein electrophoresis gel. This illustrates why
the dark b1-lipoprotein band. The b2-region has merely performing serum protein electrophoresis
two bands. The rst is C3, the second is not identi- on concentrated urine samples is an inadequate
ed. In the mid-g-region of the serum there is a screen for MFLC. The recent availability of
weakly staining band that is barely visible. The g- immunoassays for free light chains in serum and
region stains very weakly indicating that hypo- urine are already available to aid in the diagnostic
gammaglobulinemia is present. The corresponding process (see Chapter 7). Also, note that tiny
urine has an artifactual restriction at the origin. amounts of the intact IgA monoclonal proteins are
Two small bands are present in the g-region of the also present (albeit barely visible) in the urine, but
urine. their light chain counterparts are not seen at the
Immunoxation of the serum and urine shown in dilution of urine used for this study (concentrated
Fig. 10.27b demonstrates that the complex serum 100 times).
366 Case studies for interpretation
(a)
Figure 10.28 (a) Serum protein electrophoresis on four samples. Case X is from an 82-year-old man (his name is not Lirpa Loof!)
(Paragon SPE2 stained with Paragon Violet). (b) Densitometric scan of serum from case X. (c) Immunoxation of serum from case X
(Paragon system stained with Paragon Violet; anode at the top).
Case studies for interpretation 367
(b)
INTERPRETATION OF FIG. 10.28 densitometric scan in Fig. 10.28b shows these three
irregularities. In Fig. 10.28c, the immunoxation
This is a true story (though perhaps embellished demonstrates that the large band in the brinogen
with time). On April 1, not quite 20 years ago, I region is an IgA l monoclonal gammopathy. The
was presented with a sample demonstrating a tri- breadth of the IgA and associated l light chain
clonal gammopathy to interpret on a patient sup- band probably reects the glycosylation of many
posedly named Lirpa Loof. My residents were able IgA molecules. Interestingly, migrating at the same
to contain themselves for only a few minutes while location is a smaller IgM k monoclonal gammo-
I waxed poetic about the theoretical possibility of pathy, a band I was not able to appreciate on the
this happening, and its even greater likelihood in serum protein electrophoresis or on the densito-
this era of acquired immunodeciency syndrome metric scan. This same combination (IgM k) is
(AIDS) (which was then a new disease). After my responsible for the mid-g-region band seen on the
monologue, they gleefully pointed out that Lirpa electrophoresis in Fig. 10.28a and on the densito-
Loof was April Fool backward, and showed me the metric scan. It is possible that this is a multimer of
serum gels from the three separate monoclonal the IgM k band seen in the b-region. However, it
patients they had mixed to produce the artifact. could be a product of an entirely separate clone.
The serum X from Fig. 10.28a is not a mixture of The slowest band is due to an IgG l monoclonal
sera, and this 82-year-old man does not have gammopathy. Note that the light chain does not
AIDS. The serum shows three distinct monoclonal stain as well as the IgG on this immunoxation
bands. The darkest band is in the brinogen pattern. The patient did not have a lymphoprolif-
region. In the mid-g-region a small, but distinct erative process.
band is present. Just cathodal to the mid-g-band If my former residents are reading this, all I can
is a slightly fainter, but also distinct band. The say is Lirpa Loof lives!
Case studies for interpretation 369
(a)
Figure 10.29 (a) Serum protein electrophoresis on two samples is shown. Case X is a College of American Pathologists (CAP) survey
sample EC-08 (Panagel system stained with Paragon Violet). (b) Immunoxation of Case X. (Paragon system stained with Paragon Violet;
anode at the top).
370 Case studies for interpretation
INTERPRETATION FOR FIG. 10.29 other methods missed this small, but obvious
monoclonal gammopathy. Earlier examples
The electrophoretic pattern shows a small fast g- demonstrated that small serum monoclonal gam-
restriction. Note that the normal sample below has mopathies may be associated with excretions of
a tiny origin artifact which should not be mistaken large amounts of MFLC found most readily in the
for a monoclonal gammopathy. The immunoxa- urine by electrophoresis and immunoxation.
tion of the College of American Pathologists (CAP) Further, as discussed in Chapter 6, small IgM
sample identies this band as an IgM k monoclonal monoclonal gammopathies are associated with
gammopathy. As discussed in Chapter 2, as many peripheral neuropathies.
as two-thirds of the laboratories that used the
Case studies for interpretation 371
(a)
Figure 10.30 (a) Serum protein electrophoresis on four samples. Case X is from a 46-year-old woman with anemia, elevated calcium and
lytic bone lesions (Paragon SPE2 stained with Paragon Violet).
372 Case studies for interpretation
2100 5.0
600 600 1800 1200
1800 4.0
500 500 1500 1000
1500 3.0
400 400 1200 800
1200 2.0
300 300 900 600
900 1.0
200 200 600 400
600 0.50
100 100 300 200
300 0.25
0 0 0 0
(b)
Figure 10.30 (contd) (b) Immunoglobulin measurements on the serum from case X on the Beckman Nephelometer.
Case studies for interpretation 373
2100 5.0
600 600 1800 1200
1800 4.0
500 500 1500 1000
1500 3.0
400 400 1200 800
1200 2.0
300 300 900 600
900 1.0
200 200 600 400
600 0.50
100 100 300 200
300 0.25
0 0 0 0
(c)
Figure 10.30 (contd) (c) Immunoglobulin measurements same data for IgG, IgM, and k, but IgA and l remeasured on the serum diluted
1:10 prior to analysis.
374 Case studies for interpretation
(d)
Figure 10.30 (contd) (d) Immunoxation of serum from case X (Paragon system stained with Paragon Violet; anode at the top).
Case studies for interpretation 375
INTERPRETATION FOR FIG. 10.30 not t. a heavy chain disease occurs among indi-
viduals in their second and third decades in the
A broad, massive band is present in the b-region. Middle East and Mediterranean regions. It pre-
The g-region is markedly decreased compared with sents as a gastrointestinal disease and does not
the other samples on this gel. This is certainly a cause lytic bone lesions or elevated calcium values.
massive monoclonal gammopathy. The immuno- This is a woman in the latter half of her fth
globulin measurements in Fig. 10.30b disclose an decade. So what is wrong? Even the k/l ratio is
extraordinary increase in the amount of IgA only barely abnormal. This is a demonstration of
present that correlates well with the amount of the problems of using nephelometry alone to detect
protein seen on the serum protein electrophoresis monoclonal proteins. As discussed in Chapter 9,
gel. Please note that the light chain measurements some monoclonal proteins do not react well with
are presented as the total concentration of the antisera standardized against polyclonal immuno-
immunoglobulin they are attached to (i.e. calcu- globulins. The immunoxation in Fig. 10.30d
lated as the molecular weight of the molecule to clearly demonstrates an IgA l monoclonal gammo-
which they are attached rather than as the weight pathy. When the immunoglobulin quantications
of the free light chain). At the time this study was were repeated prediluting the sample 1:10 as
performed, that was the standard. More recently, shown in Fig. 10.30c, more of the monoclonal l
only the molecular weight of the light chain is con- light chains were detected. However, further dilu-
sidered in calculating the quantity of light chain tions did not allow better approximation of the
present. These numbers will likely differ from amount of IgA present. When a patient appears to
current numbers you may see in your patients. have a heavy chain disease because of the lack of
Therefore, with the study numbers, in theory k and reactivity of a light chain either by nephelometry or
l should add up to the total of IgG + IgA + IgM. In by immunoxation, be careful. Use another tech-
this case they do not. Indeed, there seems to be no nique such as immunosubtraction, or send it to a
corresponding light chain for the massive IgA. laboratory that performs immunoselection to make
Could this be a case of a heavy chain disease? No, a denitive diagnosis. This patient has classic mul-
but situations like this are often mistaken for heavy tiple myeloma.
chain disease despite the fact that the history does
376 Case studies for interpretation
(a)
(b)
Case studies for interpretation 377
(c)
Figure 10.31 (a) Serum protein electrophoresis on three samples. Case X is from a 67-year-old man with recurrent pneumonia (Paragon
SPE2 stained with Paragon Violet). (b) Densitometric scan of serum from case X. (c) Immunoxation of serum from case X (Paragon
system stained with Paragon Violet; anode at the top).
INTERPRETATION OF FIG. 10.31 and a couple of faintly stained bands are barely
discernable. Immunoxation was ordered on this
Although the albumin band on the photograph of sample. It demonstrates that the slightly broader
case X looks similar to the other two samples on C3 band was really due to a tiny IgG l band that
the gel, the densitometric scan in Fig. 10.31b doc- migrates in the C3 region. There are also two
uments a decrease in albumin concentration. Both slower-moving IgG k bands and a third slow-
the a1- and a2-regions are increased, consistent migrating IgG l band. This oligoclonal expansion
with an acute inammatory response. The trans- in the context of a borderline hypogammaglobu-
ferrin band stains fainter on the gel specimen linemia can be seen in patients with B-cell
(Fig. 10.31a) than the transferrin bands in the lymphoproliferative disorders (discussed in
two samples below it. The C3 band is slightly Chapter 6). A call to the clinician revealed that
broader than the two samples below it (possibly this patient has chronic lymphocytic leukemia
indicating a subacute inammation). The g-region with a leukocyte count of 70 000 (virtually all
stains more weakly than the two samples below mature lymphocytes).
378 Case studies for interpretation
(a)
Figure 10.32 (a) Serum protein electrophoresis on a 79-year-old man with endocarditis. All sera on this gel are from this patient. Top to
bottom: January 15, 18, 27, 29, and February 3 (Paragon SPE2 stained with Paragon Violet). (b) Immunoxation of serum from January 27
sample (Paragon system stained with Paragon Violet; anode at the top). (c) Immunoxation of urine from January 27 sample (Paragon
system stained with Paragon Violet; anode at the top).
Case studies for interpretation 379
(b)
(c)
380 Case studies for interpretation
(d)
2100 5.0
600 600 1800 1200
1800 4.0
500 500 1500 1000
1500 3.0
400 400 1200 800
1200 2.0
300 300 900 600
900 1.0
200 200 600 400
600 0.50
100 100 300 200
300 0.25
0 0 0 0
(e)
Figure 10.32 (contd) (d) Patient serum from March 5 on top and control serum on bottom (Paragon SPE2 stained with Paragon Violet).
(e) Immunoglobulin measurements from January 27 sample.
Case studies for interpretation 381
INTERPRETATION OF FIG. 10.32 the serum and a ladder pattern with both kappa
and lambda. No MFLC (Bence Jones protein) is
The earliest electrophoretic sample on the gel from seen.
Fig. 10.32a shows a barely discernable mid-g band. With time, this pattern evolved into an oligo-
The second sample shows an increase in a1- and a2- clonal pattern (Fig. 10.32d), demonstrating that
globulin and the presence of a faint band in the the original monoclonal band was likely an early
mid-g-region. The third through the bottom sera response of a prominent B-cell clone to the infec-
all have anodal slurring of albumin (presumably tious agent causing the endocarditis. Transient
due to the antibiotic therapy), an acute-phase reac- monoclonal gammopathies have been reported
tion and a prominent mid-g-region band. The with endocarditis and other infectious diseases and
immunoxation performed on the serum from are discussed in Chapter 6. Typically, they do not
January 27 (Fig. 10.32b) shows an obvious IgG k have accompanying MFLC. Also, in cases with
monoclonal gammopathy. The urine immunoxa- reactive clonal bands, the k/l ratio is often in the
tion from the same date (Fig. 10.32c) shows a very normal range, as it was in this case (Fig. 10.32e).
faint IgG band which corresponds to the band in
382 Case studies for interpretation
Figure 10.33 (a) Immunoxation on serum from an 82-year-old man with dizziness, paresthesias, and skin ulcers surrounded by
erythematosus plaques on his legs (Paragon system stained with Paragon Violet; anode at the top). (b) Repeat of the immunoxation on
the same sample, this time, instead of overlaying the sample with anti-IgG, saline was placed in that lane (Paragon system stained with
Paragon Violet; anode at the top).
384 Case studies for interpretation
Figure 10.34 (a) Immunoxation on serum from a 49-year-old man with shoulder pain. (Paragon system stained with Paragon Violet;
anode at the top).
Case studies for interpretation 385
2100 5.0
600 600 1800 1200
1800 4.0
500 500 1500 1000
1500 3.0
400 400 1200 800
1200 2.0
300 300 900 600
900 1.0
200 200 600 400
600 0.50
100 100 300 200
300 0.25
0 0 0 0
(b)
Figure 10.34 (contd) (b) Immunoglobulin measurements from the same serum.
INTERPRETATION FOR FIG. 10.34 major problem in this case. There has been grossly
inadequate migration. (Someone turned off the
The immunoxation in Fig. 10.34a shows a dense power too soon.) The poor separation of the
band in each of the immunoglobulin classes. The protein led to this confusing picture. The
pattern resembles, somewhat, the origin artifacts immunoglobulin quantications are consistent
that are often seen in patients with cryoglobuline- with a normal sample, but the immunoxation
mia. Examination of the SPE lane discloses the needs to be repeated.
386 Case studies for interpretation
(a)
Figure 10.35 (a) Serum protein electrophoresis on four samples is shown. The sample in the top lane is from a patient at the oncology
clinic (Paragon SPE2 stained with Paragon Violet). (b) Immunoxation of serum from case in top lane in (a) (Paragon system stained with
Paragon Violet; anode at the top). (c) Immunoxation of serum from same case using antisera against IgG subclasses.
Case studies for interpretation 387
INTERPRETATION OF FIG. 10.35 k is diluted 1:30, whereas the l is diluted only 1:8.
This broad band is not a cryoglobulinemia because
The top lane of the electrophoretic pattern in Fig. no such bands occur in the IgM lane (diluted 1:2),
10.35a shows a broad, but distinct band in the although a tiny origin artifact is seen at this con-
slow b- to fast g-region. Although one might centration. A logical supposition for this pattern is
suspect that this increase would be associated with that there is a polyclonal increase in one subclass of
an elevated serum IgA, it was in the mid-normal IgG. The IgG4 subclass was elevated but the other
range (233 mg/dl). However, the IgG was elevated subclasses were not. The immunoxation, in Fig.
(3070 mg/dl, about twice the normal upper limit). 10.35c, performed using subclass antisera demon-
The immunoxation shown in Fig. 10.35b demon- strates that the IgG4 subclass corresponds to the
strates that both k and l lanes show a broad band same migration. Follow-up on the patient revealed
similar to that seen in the serum protein electro- he had an epithelial malignancy and chronic respi-
phoresis lane. Although the k lane stains more ratory disease, not a monoclonal gammopathy.
weakly than the l lane in this region, note that the This case contributed by Dr A. C. Parekh.
Case studies for interpretation 389
Figure 10.36 Case A: cerebrospinal uid (CSF) concentrated 80-fold and serum (immediately below) diluted 1:3 from a 28-year-old
woman. Case B: CSF concentrated 80-fold and serum (immediately below) from a 42-year-old man. Case C: CSF only is shown from a 12-
year-old girl (no serum was sent) (Paragon SPE2 system stained with Paragon Violet).
INTERPRETATION FOR FIG. 10.36 techniques are preferred for sensitivity and for
identication of the bands as immunoglobulins.
I no longer recommend using serum protein However, many clinical laboratories are still using
electrophoresis to detect oligoclonal bands. Either these techniques. Therefore, Ive included this
isoelectric focusing or immunoxation-based example.
390 Case studies for interpretation
Case A: the cerebrospinal uid (CSF) is negative tion in a negative CSF. However, if there had been
for oligoclonal bands. Case B: the CSF is positive oligoclonal bands in this CSF, we would have
for oligoclonal bands; the corresponding serum is requested that a serum be sent to rule out systemic
negative for these bands. Case C: the CSF is oligoclonal bands that had diffused across the
negative for oligoclonal bands. The lack of a corre- bloodbrain barrier.
sponding serum does not interfere with interpreta-
Case studies for interpretation 391
Figure 10.38 Urine (concentrated 100-fold) from a 56-year-old man (Paragon system stained with Paragon Violet; anode at the top).
INTERPRETATION FOR FIG. 10.38 interpretation in this case is Negative for mono-
clonal free light chains (Bence Jones protein).
The urine protein electrophoresis lane shows Ladder patterns merely reect the limited hetero-
mainly albumin with a couple of smaller bands in geneity of polyclonal free light chain migration.
the a2-region. A diffuse IgG band is seen. No IgA Terms such as oligoclonal or minimonoclonal are
or IgM is detectable. There is a classic ladder confusing and ambiguous.
pattern in k and a fainter ladder pattern in l. The
Case studies for interpretation 393
Figure 10.39 Urine (concentrated 100-fold) from a 68-year-old man with a known serum IgG k monoclonal gammopathy from our les
(Paragon system stained with Paragon Violet; anode at the top).
INTERPRETATION FOR FIG. 10.39 lanes. The slow g-region band is the counterpart of
the IgG k monoclonal protein seen previously in
The urine protein electrophoresis lane has a single this patients serum. However, in addition, there is
dense band in the ab-region, no albumin band is a free k MFLC that migrates in the fast g-region.
visible. There is also a tiny band in the slow g- The interpretation noted the MFLC and recom-
region. The ab-region band is hemoglobin and mended a 24-h protein quantication of this com-
the urine was red. This band causes a faint artifac- ponent. I also reported the presence of the IgG k
tual staining in all of the immunoglobulin antisera monoclonal protein.
394 Case studies for interpretation
Figure 10.40 Urine (concentrated 100-fold) from a 53-year-old woman (rule out monoclonal free light chain) (Paragon system stained
with Paragon Violet; anode at the top).
Note Page numbers in bold type refer to gures and italic type indicates tables.