IMMUNOPATHOLOGY
Original Article
Detection of Oligoclonal Bands in Cerebrospinal Fluid
by Immunofixation Electrophoresis
DOMINICK CAVUOTI, DO,1-2 LELAND BASKIN, MD,1-2 AND ISHWARLAL JIALAL, MD, PhD, FRCPath1"3
Multiple sclerosis is a severe demyelinating disease, the diagnosis of which is aided by biochemical tests, such as detection of
oligoclonal immunoglobulin bands in the cerebrospinal fluid
(CSF). Because interpretation of agarose gel electrophoresis
(AGE) of CSF for oligoclonal bands is often equivocal, we compared immunofixation electrophoresis (IFE) with AGE for 124
consecutive CSF specimens submitted to the Parkland Memorial
Hospital Clinical Chemistry Laboratory (Dallas, Tex) for detection of oligoclonal b a n d s . Both m e t h o d s u s e d the P a r a g o n
Electrophoresis Systems (Beckman Instruments, Brea, Calif).
A n t i - I g G a n t i s e r a w a s u s e d e x c l u s i v e l y on all s p e c i m e n s .
Oligoclonal bands were identified in 23 specimens (18.5%), while
the other 101 (81.5%) were interpreted as negative by both methods. Of the positive specimens, 17 (74%) were positive by both
methods, 5 (22%) by IFE alone, and 1 (4%) by AGE alone. Of the
23 patients with positive specimens represented, 17 (74%) had
been given a diagnosis of multiple sclerosis. The patient whose
specimen was positive by AGE alone had a diagnosis of HIV
infection with Guillain-Barre syndrome. The sensitivities (with
95% confidence intervals) of IFE and AGE were 73.9% (51.3-88.9)
and 56.5% (34.9-76.1), respectively. The specificities of both methods were identical at 95.0% (88.3-98.2). Subjective assessment of
the gels demonstrated that the IFE method is consistently easier
to interpret than AGE. The IFE method seems to be superior in
identifying oligoclonal bands and thus aiding in diagnosis of
d e m y e l i n a t i n g d i s o r d e r s . (Key w o r d s : M u l t i p l e s c l e r o s i s ;
Immunofixation electrophoresis; Cerebrospinal fluid) Am J Clin
Pathol 1998;109:585-588.
The detection of intrathecal immunoglobulin synthesis is the cornerstone of laboratory-based diagnosis of
multiple sclerosis (MS).1 While no method is completely specific for MS, the combination of history
and physical examination, magnetic resonance imaging, and electrophysiologic testing along with appropriate cerebrospinal fluid (CSF) studies can help clarify the diagnosis. 2
Examination of the CSF includes qualitative and
quantitative measures of intrathecal immunoglobulin
synthesis. A number of formulas have been devised
for the quantitative detection, including the IgG
index, IgG synthetic rate, IgG local synthesis, and the
albumin index. 3,4 Of these, the IgG index has been
found to be a sensitive indicator of i n t r a t h e c a l
immunoglobulin synthesis, while the albumin index
evaluates the integrity of the blood-brain barrier. It
has been shown that approximately 90% of patients
with MS have an elevated IgG index. 1 The classic
qualitative measure has been the detection of oligoclonal bands (OCBs) by CSF agarose gel electrophoresis (AGE). Occasionally, identification of these bands
can be equivocal with only a monoclonal band or a
diffuse heavy band detected. In an attempt to improve
the resolution, sensitivity, and ease of interpretation of
the gels, w e u s e d s t a n d a r d AGE followed by
immunofixation (IFE). Anti-IgG antibodies were used
because oligoclonal immunoglobulin is usually IgG.1
MATERIALS AND METHODS
Specimens
We analyzed 124 consecutive matched pairs of CSF
and serum submitted for OCB identification to the
Parkland Memorial Hospital (PMH) Clinical Chemistry
Laboratory (Dallas, Tex). The IgG index was measured
in the PMH Immunology Laboratory in 94 (75.2%) of
these specimens. Classification of patients by diagnosis
was based on chart review using the criteria for diagnosis of MS from Poser et al. 5 Clinically definite MS is
defined as (1) two attacks and clinical evidence of two
lesions or (2) two attacks and clinical evidence of one
From the ^Division of Clinical Chemistry, Parkland Memorial
Hospital and the Departments of2Pathology and 3Internal Medicine, The
University of Texas Southwestern Medical Center at Dallas.
Manuscript received May 16,1997; revision accepted July 30,1997.
Address reprint requests to Dr Jialal: The University of Texas
Southwestern Medical Center at Dallas, 5323 Harry Hines Blvd,
Dallas, TX 75235-9072.
585
586
IMMUNOPATHOLOGY
Original Article
lesion and test evidence (such as magnetic resonance
imaging or evoked potentials) of a second lesion.
Classification as laboratory-supported definite MS
requires detection of OCBs or elevated IgG in the CSF
and one of three criteria: (1) two attacks and clinical or
test evidence of one lesion, (2) one attack and clinical evidence of two lesions, or (3) one attack and clinical evidence of one lesion and test evidence of a second lesion.5
Detection of OCBs
We performed AGE on the Paragon Electrophoresis
System (Beckman Instruments, Brea, Calif). The CSF specimens were concentrated at least 60x using a Minicon B15 concentrator (Amicon, Beverly, Mass). Serum samples
were diluted 1:2 in Paragon B2 barbital buffer. Three to 5
microliters of corresponding CSF and serum specimens
were applied via template to the surface of a 1% agarose
gel. Electrophoresis at 100 V was performed for 25 minutes. On completion, the gel was placed in an acetic acid
solution for 3 to 5 minutes. Following drying, it was
stained for 3 to 5 minutes with Paragon Blue. After staining, the gel was placed for 2 minutes in each of the following solutions: 5% acetic acid, acid alcohol (glacial acetic
acid:deionized water:ethanol in ratio of 1:3:6), and 5%
acetic acid again. The gel was then dried again.
We also performed IFE on the Paragon Electrophoresis
System using a procedure similar to that for AGE. The
electrophoresis time was 30 minutes. After electrophoresis, an antiserum template was aligned over the gel and
80 |rL of polyclonal goat antihuman IgG was added to
each trough. Following incubation at room temperature
for 10 minutes, the gel was washed in saline buffer and
blotted dry. The gel was dried for 5 minutes, stained with
Paragon Blue for 3 minutes, soaked in acetic acid, and
dried in an oven. For both methods, normal CSF specimens were included as negative controls. Positivity for
OCBs was defined as 2 or more discrete bands in the
gamma region of AGE or IFE that were not present in the
concurrent serum sample.
Serum albumin, serum IgG, CSF IgG, and CSF
albumin were measured via rate nephelometry using
the Array 360 system (Beckman Instruments). We
used the following equations:
IgG Index = [CSF IgG/Serum IgG]/
[CSF Albumin/Serum Albumin]
(reference range, 0.34 to 0.7)
Albumin Index = 100 x CSF Albumin/
Serum Albumin
(reference range, 0 to 9)
In these equations, units are chosen appropriately
to produce dimensionless indices. 3
Statistical
Analysis
Confidence intervals for the sensitivities and specificities were calculated using the binomial distribution. The sensitivities and specificities were compared
using McNemar's %2 test with Edwards correction for
continuity. Overall agreement between methods was
assessed by Cohen's K.6
RESULTS
Of the 124 specimens, 101 (81.5%) were negative by
both AGE and IFE. Six of the patients represented had
been given a diagnosis of MS. Oligoclonal bands were
detected in a total of 23 specimens (18.5%). Seventeen
(74%) of these were positive by both methods, 5 (22%)
by IFE alone, and 1 (4%) by AGE alone (Table). Of the
17 patients whose specimens were positive by both
methods, 13 (76%) had a diagnosis of MS. The remaining 4 patients had the following diagnoses: multifactorial dementia, 2; AIDS neuropathy, 1; and central
nervous system (CNS) toxoplasmosis, 1. The patient
whose specimen was positive by AGE alone had a
diagnosis of HIV infection and Guillain-Barre syndrome. Of the 5 patients whose specimens were positive by IFE alone, 4 had been given a diagnosis of MS
and 1 had amyotrophic lateral sclerosis.
The IFE gels were consistently easier to interpret
t h a n the AGE. The b a n d i n g p a t t e r n of IFE w a s
sharper, with multiple bands often present compared
with the occasionally indistinct banding seen with
AGE (Figure). The sensitivity (with 95% confidence
interval) of IFE was 73.9% (51.3-88.9), while that of
AGE was 56.5% (34.9-76.1). Although no statistically
significant difference was detected between sensitivities (P = .13), a definite trend toward increased sensitivity by IFE was present. The specificities of the
two methods were identical at 95.0% (88.3-98.2).
Overall agreement between the two methods was
good (K = 0.82).
We included 94 specimens in the IgG index evaluation; some were excluded owing to blood-brain barrier leaks or because the index was not performed. Of
the 18 specimens positive by IFE, 16 (89%) had an IgG
index greater than 0.7, with the most common diagnosis being MS (n = 13). The remaining diagnoses were
as follows: multifactorial dementia, 1; AIDS neuropathy, 1; and CNS toxoplasmosis, 1. One patient with an
IgG index less than 0.7 had MS, while the other had
AJCP • May 1998
587
CAVUOTIET AL
Detection of Oligoclonal Bands in CSF by Immunofixation Electrophoresis
amyotrophic lateral sclerosis. Only two patients with
MS had neither an elevated IgG index nor OCBs.
In the group in whom OCB and IgG indices were
performed, the sensitivities of the IFE and IgG indices
were 70.0% (45.7-87.2) and 85.0% (61.1-96.0), respectively. The specificities of the IFE and IgG indices were
94.6% (86.0-98.3) and 85.1% (74.5-92.0), respectively.
Although the IgG index seems to be more sensitive,
this difference is not statistically significant (P = .37).
However, the specificity of IFE w a s significantly
greater than that of the IgG index (P = .046). The overall agreement between these two methods (K = 0.60)
was less than that for IFE and AGE (K = 0.82). This lack
of agreement suggests that different characteristics are
being measured. Thus, a combination of IFE and IgG
index may be more effective than either test alone. The
combination of a positive result for either IFE or IgG
index increased the sensitivity to 90.0% (66.9-98.2)
while maintaining a specificity of 83.8% (73.0-91.0).
MS may represent up to 20% of cases.7'8 The absence of
OCBs should prompt a review of all diagnostic criteria
with possible repeated testing. Autopsy studies of OCBnegative MS have demonstrated low numbers of plasma
cells and histologic evidence of plaque inactivity.8 The
lack of OCBs in patients with confirmed MS suggests a
better prognosis related to the less pronounced humoral
immune response. 9
Isoelectric focusing has been shown to be the most
sensitive technique for detecting OCBs.10 Many laboratories have incorporated isoelectric focusing for detecting OCBs, and as a result, most reported improved sensitivities when using this technology.1 However, based
on our preliminary experience, a skilled laboratorian
A
B
C
D
DISCUSSION
In this study, IFE and AGE were compared for the
detection of OCBs in paired CSF and serum specimens.
IFE was clearly easier to interpret. The bands present on
IFE were typically more numerous and more sharply
defined than the bands seen with AGE. Also, IFE displayed greater, although not statistically significant, sensitivity than AGE. It is possible that the failure to obtain
a statistically significant increased sensitivity was related
to the relative small sample showing positivity for
OCBs. The majority of patients with detectable OCBs
had a diagnosis of MS as did those with an elevated IgG
index. Patients without a diagnosis of MS had other
CNS inflammatory diseases, which is consistent with
other reported findings. Our sensitivity was lower than
that of other reported studies in which sensitivity
approaches 90% or more. 2 Extensive chart review and
use of the criteria from Poser et al 5 confirmed the diagnoses of MS in six patients whose specimens were negative for OCBs. Although these criteria are helpful, they
are not 100% sensitive or specific. True OCB-negative
IFE
I I I:
AGE
COMPARISON OF IFE AND AGE FOR DETECTION OF
OLIGOCLONAL BANDS
Agarose Gel Electrophoresis
Immunofixation
Positive
Negative
electrophoresis
Positive
Negative
17
1
5
101
Representative samples of oligoclonal bands on immunofixation
electrophoresis (IFE) and agarose gel electrophoresis (AGE) gels.
The specimens in lanes A, B, and C contain IgG oligoclonal bands
that are clearly visible by IFE, whereas only the specimen in lane C
contains unequivocal oligoclonal bands by AGE. Lane D contains a
negative cerebrospinal fluid control.
Vol. 109 • No. 5
IMMUNOPATHOLOGY
588
Article
must perform the test. This may account for some of
the discrepancies in sensitivity of OCB detection in MS
in different studies. The majority of OCBs are IgG, but
IgM and IgA have also been implicated. Sindic et al11
demonstrated IgM OCBs in the CSF of patients with
MS, which was associated with acute relapses and first
manifestations of the disease. However, in their series,
more than 98% of patients also had IgG OCBs. Thus,
because only IgG antisera was used, OCBs that are
exclusively IgM and IgA could have gone undetected,
but this seems unlikely according to the high concordance reported by Sindic et al.11
Agarose gel electrophoresis followed by immunofixation with anti-IgG represents a reasonable alternative for
OCB testing, using the current electrophoresis system, in
the smaller laboratory without the workload to justify
isoelectric focusing instrumentation. In addition, the IgG
index performed on the sample will clearly increase the
sensitivity in the appropriate clinical setting.
REFERENCES
1. Bentz J. Laboratory investigation of multiple sclerosis. Lab Med.
1995;26:393-399.
2. Rolak LA. The diagnosis of multiple sclerosis. Neurol Clin.
1996;14:27-43.
3. Peter JB, Bowman RL. Intra-blood-brain barrier synthesis of
IgG: comparison of IgG synthesis formulas in a computer
model and in 1,629 consecutive specimens. Neurology.
1992;42:510-515.
4. Peter JB, McKeown KL, Agopian MS. Assessment of different
methods to detect increased autochthonous production of
immunoglobulin G and oligoclonal immunoglobulins in
multiple sclerosis. Am } Clin Pathol. 1992;97:858-860.
5. Poser CM, Paty DW, Scheinberg L, et al. New diagnostic criteria
for multiple sclerosis: guidelines for research protocols. Ann
Neurol. 1983;13:227-231.
6. Fleiss JL. Statistical Methods for Rates and Proportions. 2nd ed.
New York, NY: John Wiley & Sons. 1981;13-15, 113-119,
212-225.
7. Zemon AZJ, Kidd D, McLean BN, et al. A study of oligoclonal
band negative multiple sclerosis. / Neurol Neurosurg
Psychiatry. 1996;60:27-30.
8. Farrell MA, Kaufmann JCE, Gilbert JJ, et al. Oligoclonal band
negative multiple sclerosis: clinical pathologic correlation.
Neurology. 1985;35:212-218.
9. Stendahl-Brodin L, Link H. Relation between benign course of
multiple sclerosis and low grade humoral immune response
in cerebrospinal fluid. / Neurol Neurosurg Psychiatry.
1980;43:102-105.
10. Andersson M, Alvarez-Cermeno J, Bernardi G, et al.
Cerebrospinal fluid in the diagnosis of multiple sclerosis: a
consensus report. / Neurol Neurosurg Psychiatry.
1994;57:897-902.
11. Sindic CJ, Monteyne P, Laterre EC. Occurrence of oligoclonal
IgM bands in the cerebrospinal fluid of neurological patients:
an affinity-mediated capillary blot study. / Neurol Sci.
1994;124:215-219.
AJCP • May 1998