Psychiatria Danubina, 2017; Vol. 29, Suppl.

2, pp 124-128 Original paper

© Medicinska naklada - Zagreb, Croatia

KAPPA FREE LIGHT CHAINS IN CEREBROSPINAL FLUID

OF PATIENTS WITH IDENTIFIED OLIGOCLONAL
IMMUNOGLOBULIN G
Marina Vasilj1,3, Miro Klarić2,3, Nada Vrkić4, Ivanka Mikulić1,3, Marijana Marković Boras1,3,
Nevenka Jelić-Knezović3 & Violeta Šoljić3
1
Department of Laboratory Medicine, University Clinical Hospital Mostar, Mostar, Bosnia and Herzegovina
2
Department of Psychiatry, University Clinical Hospital Mostar, Mostar, Bosnia and Herzegovina
3
Faculty of Health Studies, University of Mostar, Mostar, Bosnia and Herzegovina
4
Clinical Institution of Chemistry, Clinical Hospital Center „Sestre milosrdnice“, Zagreb, Croatia

received: 20.2.2017; revised: 15.3.2017; accepted: 27.3.2017

SUMMARY
Background: Production of kappa free light chains (KFLC) represents a part of humoral immune response, along with the
synthesis of intrathecal immunoglobulins. Increased concentrations of immunoglobulin G light chains, kappa and lambda chains,
were identified through research of numerous diseases of central nervous system. The qualitative method of isoelectric focusing
(IEF) followed by immunofixation currently represents the accepted standard in identifying oligoclonal bands (OCB), but
establishing a sensitive immunonephelometric method for quantification of kappa free light chains (KFLC) in cerebrospinal fluid
(CSF) has paved a way for new diagnostic possibilities. Andersson classified the pattern types of OCB, ranging from type 1 to type 5,
wherein types 2 and 3 indicate intrathecal synthesis. Our aim was to determine KFLC in CSF of patients with clinically isolated
syndrome (CIS) who had presented with type 2 and type 3 OCB, to determine if there is a difference in concentrations between those
two groups and to establish a borderline value of KFLC which would enable differential diagnostics.
Subjects and methods: 70 patients, who underwent lumbar punction for CSF analysis and had their blood sampled through the
cubital vein, participated in the study. Patients were classified according to Andersson as type 2 or type 3, which besides adulthood,
represented the inclusion criteria. The average age of patients classified as type 2 was 36 years, and those classified as type 3 was 39
years, where it is evident that there was not a statistically significant difference (p=0.0685). We used a qualitative electrophoretic
technique of IEF with agarose gel followed by immunofixation, and a quantitative immunonephelometric method. All results were
interpreted on a level of statistic significance of p<0.05.
Results: CSF KFLC concentrations in type 3 were statistically and significantly elevated with regard to type 2 (Mann-Whitney
test, p=0.0430). The median for KFLC in type 2 was 0.9 mg/L, while the median for KFLC in type 3 was 2.71 mg/L, and the detection
limit for both types was 0.18 mg/L. We used a statistical ROC curve to determine that KFLC concentration can be used for
differential diagnostics, meaning it can discriminate type 2 from type 3 with clinical sensitivity of 61% and clinical specificity of 71%
(AUC=0.641) (p=0.037).
Conclusion: Despite the obtained statistically significant differences in concentrations of KFLC between types of OCBs and
ROC analysis results, determination of KFLC by a nephelometric method, insufficiently strong clinical sensitivity and specificity does
not justify abandonment of IEF method followed by immunofixation.

Key words: kappa free light chains - immunoglobulin G - intrathecal synthesis - isoelectric focusing - oligoclonal bands

* * * * *

INTRODUCTION diseases of CNS and through analysis of CSF, which

also changed their CSF/serum ratio (Rudick et al.
The synthesis of intrathecal immunoglobulins is 1989, Krakauer et al. 1998, Goffette et al. 2004,
usually present during inflammation of central nervous Kaplan et al. 2010). Kappa quotient (Q KFLC) repre-
system (CNS) (DeCarli et al. 1987). Considering that sents the KFLC CSF/KFLC serum ratio. In evaluation
the blood-brain barrier (BBB) greatly prevents immuno- of patients with documented first episode that was
globulins from entering the bloodstream, they start defined as clinically isolated syndrome (CIS), and with
depositing in the CNS (Compston & Coles 2008). the suspected diagnosis of multiple sclerosis (MS), the
Plasma cells, which are present in the intrathecal space, detection of IgG intrathecal synthesis, which can be
mainly produce immunoglobulin G (IgG) (Montalban et proven by a laboratory analysis of CSF, is of great
al. 2010, Freedman et al. 2005). Production of immuno- importance (Link 1991).
globulins is a part of humoral immune response, along The test which represents the accepted standard and
with the production of FLC. a diagnostic criteria for MS is identification of two or
FLCs show as kappa isotypes (KFLC) and lambda more oligoclonal IgG zones with the qualitative method
isotypes (LFLC). Elevated levels of light chains, espe- of isoelectric focusing followed by immunofixation.
cially kappa chains, were found through researching Oligoclonal zones are present in approximately 85%

S124
 Marina Vasilj, Miro Klarić, Nada Vrkić, Ivanka Mikulić, Marijana Marković Boras, Nevenka Jelić-Knezović & Violeta Šoljić:
KAPPA FREE LIGHT CHAINS IN CEREBROSPINAL FLUID OF PATIENTS WITH IDENTIFIED OLIGOCLONAL IMMUNOGLOBULIN G
Psychiatria Danubina, 2017; Vol. 29, Suppl. 2, pp 124-128

CIS patients and they are an independent risk factor for The included patients were separated into two
MS conversion. Considering that this method is quali- groups (type 2 and type 3). Out of total 70 patients, 31
tative and technically very complicated, sometimes even (11 males, 20 females) were classified as type 2, while
an experienced analyst has troubles identifying OCB the other 39 patients were classified as type 3 (10 males,
because of their low resolution and weak visualization. 29 females). The average age of patients classified as
A need for a quantitative specific immunochemistry type 2 was 36 years (18-54), and those classified as type
method in controlled environment, such as nephelo- 3 was 39 years (22-76).
metry, has arisen because of the absence of standardi- The study was approved by the Ethics Committee of
zation and quality control (Freedman et al. 2010, te Clinical Hospital Center „Sestre milosrdnice“ and all
Velthius et al. 2010). Although the qualitative detection patients gave their written consent for using the surplus
of intrathecal IgG synthesis through IEF followed by of CSF and serum. The study was conducted in
immunofixation is generally regarded as a highly accordance with the Helsinki Declaration.
reproducible method, this reproducibility still has not
been thoroughly researched on a higher number of Methods
patients (Olsson et al. 1984, Keir et al. 1990).
According to Andersson, there are five types of OCB detection
OCB IgG which are classified as: type 1 – normal CSF; Immunoglobulins were separated by electropho-
type 2 – OCB present in CSF only, which represents retic technique of IEF on agarose gel in pH gradient of
intrathecal synthesis; type 3 – many OCBs present in 6-10, fixated (immunoprecipitated) with an antibody
CSF, some of which can be found in serum, which also for IgG and visualised with chromogen. Patient's
represents intrathecal synthesis; type 4 – identical OCBs serum and CSF were applied to agarose gels (gel,
in CSF and serum, which do not demonstrate intrathecal chromogen and Hydrasis instrument for IEF, Sebia,
synthesis; type 5 – strongly expressed monoclonal France) in order to compare and detect OCBs, which,
bands in CSF and serum, which also do not demonstrate if there are at least two in CSF, indicate intrathecal
intrathecal synthesis (Andersson et al. 1994). synthesis.
The aim of this study was to compare the qualitative
Determination of kappa free light chains
and quantitative methods and the possibility of quanti-
tative immunonephelometric method to prove intrathe- Concentrations of KFLC in serum and CSF were
cal synthesis. Furthermore, the aim is to explore the determined with an immunonephelometric method on
possibility of differentiation between type 2 and type 3 an automated analyzer BN II (Siemens, Germany). The
according to Andersson, which are already discri- reagent contains specific monoclonal antibodies (N
minated by the number of OCBs through IEF followed Latex FLC assay, Siemens, Marburg, Germany).
by immunofixation.
Statistical analysis
SUBJECTS AND METHODS Statistical analysis was conducted using MedCalc
software (MedCalc Statistical Software version 14.8.1,
Patients Mariakerke, Belgium, http://www.medcalc.org; 2014).
For normality testing, the Kolmogorov-Smirnov test
The study was conducted on 70 samples of CSF and
was used. Non-parametric Mann-Whitney test was
serum at the Clinical Institute for Medical Chemistry of
used to compare two independent groups. We tried to
the Clinical Hospital Center „Sestre milosrdnice“. Pa-
determine the borderline value for KFLC which
tients with the diagnosis of CIS underwent lumbar
provides the best sensitivity and specificity in
punction in the Neurology Department and had their
comparing OCBs in type 2 and type 3. All results were
blood sampled simultaneously in the morning. Patients'
interpreted at the level of statistical significance
evaluation was ordinated by a specialist neurologist,
p<0.05.
which included demonstration of oligoclonal immuno-
globulins as one of important differential-diagnostic
procedures which can indicate a localized immuno- RESULTS
logical reaction in the central nervous system (CNS).
The remaining CSF and serum samples of patients who Seventy patients classified as type 2 (n=31) and
were classified as type 2 or type 3 were used for a type 3 (n=39) participated in this study. Concentrations
quantitative immunonephelometric determination of of KFLC in CSF were significantly higher in type 3
KFLC in CSF and serum. than those in type 2, which is shown in Table 1. Con-
The inclusion criteria were patients' adulthood and centrations of KFLC in serum in type 2 and type 3,
classification into type 2 or type 3 of oligoclonal IgGs. along with Q KFLC, are also presented in Table 1.
Patients whose CSF was hemorrhagic and who were Values in Table 1 are expressed as medians with
classified as type 1, type 4 or type 5, were excluded interquartile ranges. We used the 25-75 percentile as
from the study. interquartile range.

S125
 Marina Vasilj, Miro Klarić, Nada Vrkić, Ivanka Mikulić, Marijana Marković Boras, Nevenka Jelić-Knezović & Violeta Šoljić:
KAPPA FREE LIGHT CHAINS IN CEREBROSPINAL FLUID OF PATIENTS WITH IDENTIFIED OLIGOCLONAL IMMUNOGLOBULIN G
Psychiatria Danubina, 2017; Vol. 29, Suppl. 2, pp 124-128

Table 1. Demographic and biochemical characteristics of patients

Type of IgG accor- Number Age CSF FLC Serum FLC
Q FLC***
ding to Andersson (M/F) (years) (mg/L) (mg/L)
Type 2 31 (11/20) 36 (18-54) 0.90* (0.617-2.532)** 10.10* (9.090-13.925)** 0.1178* (0.0504-0.2449)**
Tip 3 39 (10/29) 39 (22-76) 2.71* (0.810-6.123)** 12.10* (10.200-16.325)** 0.2009* (0.0689-0.4849)**
KFLC – kappa free light chains; Q FLCt – FLCCSF/FLCserum ratio; *Values expressed as median; distribution does not
follow Gaussian distribution; **Values expressed as interquartile range; ***kappa light chain quotient;
N – number (male/female); CSF FLC – cerebrospinal fluid free light chains; Serum FLC – serum free light chains;
Q FLC – free light chains quotient

Figure 1. shows the comparison of CSF KFLC bet- serum concentration of KFLC and for Q KFLC showed
ween type 2 and type 3 where we established that worse results (AUC=0.627 and 0.628; sensitivity =66%
concentration of CSF KFLC was significantly higher in and 46%; specificity =67% and 83%) with p=0.0671
type 3 (p=0.0430). Median with its belonging interquar- and 0.0619.
tile ranges for KFLC in type 2 was 0.90 mg/L (0.61-
2.53), whereas median for KFLC in type 3 was 2.71
mg/L (0.810-6.123). Lower detection limit for both
types was 0.18 mg/L. Nine patients had values lower
than the lower detection limit. Out of these 9 patients, 3
of them (9.6%) were classified as type 3, whereas other
6 patients (15.3%) were classified as type 2.
We also determined Q KFLC values from a mathe-
matical relation between CSF KFLC and serum KFLC.
Median for Q KFLC was 0.1178 (0.0504-0.2449) for
type 2, and 0.2009 (0.0689-0.4849) for type 3. Q KFLC
between type 2 and type 3 reached borderline signi-
ficance (p=0.06).

P=0.0430

Figure 2. ROC analysis of kappa free light chains in

CSF for possibility of differentiation between oligoclo-
nal IgG type 2 and type 3

DISCUSSION
OCB immunoglobulins are synthesized in plasma
blasts and plasma cells in CSF or CNS, and their
Figure 1. Comparison of CSF KFLC concentrations determination represents an international reccommen-
between oligoclonal IgG type 2 and type 3 dation for detection of intrathecal inflammation and
synthesis of oligoclonal immunoglobulins which origi-
Serum KFLC concentrations of type 3 are higher nate from two or more B lymphocite clones (Awad et al.
than type 2, but without statistical significance 2010, Passerini et al. 2016). The activity of intrathecal
(p=0.069). The same results were found for Q KFLC IgG synthesis from plasma cells can be qualitatively
comparison – higher values of type 3 did not show shown through IEF followed by immunofixation, which
statistical significance (p=0.068). is a very demanding and very often equivocal method,
Figure 2. shows the ROC (Receiver Operating Cha- especially if the OCBs are weakly expressed or presen-
racteristics) analysis. By using ROC analysis, we tried ted in CSF only (Franciotta & Lolli 2007, Andlovic et
to determine the borderline value for KFLC which gives al. 2012, Link & Huang 2006). IgG FLCs, which are
the best sensitivity and specificity in discriminating type synthesized through B lymphocites, consist of two
2 and type 3. With the borderline value of 1.89 mg/L, heavy chains and two light chains, which exist as kappa
the diagnostic sensitivity was 61.5%, whereas the and lambda isotypes (Kaplan et al. 2011, van der
diagnostic specificity was 71.0%. The ROC analysis for Heijden et al. 2006). KFLCs have already been reported

S126
 Marina Vasilj, Miro Klarić, Nada Vrkić, Ivanka Mikulić, Marijana Marković Boras, Nevenka Jelić-Knezović & Violeta Šoljić:
KAPPA FREE LIGHT CHAINS IN CEREBROSPINAL FLUID OF PATIENTS WITH IDENTIFIED OLIGOCLONAL IMMUNOGLOBULIN G
Psychiatria Danubina, 2017; Vol. 29, Suppl. 2, pp 124-128

as a surrogate marker of intrathecal immunoglobulin

synthesis, but in a differential diagnostics of multiple Acknowledgements:
sclerosis they need to be complemented with clinical This study was supported by the Federal Ministry of
observations, anamnestic data, imaging, biochemical, Education and Science of Bosnia and Herzegovina.
cytological and microbiological analyses. Besides these,
the MRZ reaction (antigen-specific immunoglobulin G Conflict of interest: None to declare.
to Measles, Rubella and Varicella Zoster Virus)
improves the specificity of oligoclonal IgG findings. Contribution of individual authors:
The quantitative determination of immunoglobulins is Marina Vasilj: design study, analyses of data results,
still not being used in these diagnostics (Passerini et al. statistical analyses, literature searches;
2016). In contrast to IEF followed by immunofixation, Miro Klarić: design study, analyses of data results,
determination of KFLC is a simple, quantitative method statistical analyses, literature searches;
which can be performed by using ELISA or Nada Vrkić: design study, analyses of data results,
nephelometry (Senel et al. 2014). statistical analyses, literature searches;
Ivanka Mikulić: analyses of data results, literature
In this study, we report KFLC measuring as a
searches;
possible method for detection of intrathecal synthesis Marijana Marković Boras: analyses of data results,
which allows incorporation of KFLC analysis in routine literature searches;
laboratory diagnostic algorithms (Presslauer et al. Nevenka Jelić-Knezović: analyses of data results,
2014). Comparing CSF KFLC values in two evaluated literature searches;
groups, somewhat higher KFLC concentrations were Violeta Šoljić: analyses of data results, literature
found in type 3 than in type 2. We have to emphasize searches.
that during this study we used laboratory data only and
excluded all radiological data. For the complete
overview of significance and clinical value of KFLC References
quantification, it is necessary to correlate these with the 1. Andersson M, Alvarez-Cermeno J, Bernardi G, Cogato I,
findings of MRI and clinical presentation in the future. Fredman P et al.: Cerebrospinal fluid in the diagnosis of
It is well known that other authors found signifi- multiple sclerosis: a consensus report. J Neurol
cantly elevated Q KFLC values with present IgG OCBs, Neurosurg Psychiatry 1994; 57:897-902.
with Q KFLC being elevated in patients classified as 2. Andlovic A, Babic M, Accetto S, Rot U: Comparison of two
type 2 and type 3 which represent intrathecal IgG methods for the detection of oligoclonal bands in a large
synthesis in regard to type 1 which excludes MS (Senel number of clinically isolated syndrome and multiple scle-
et al. 2014, Rudick et al. 1989). However, in this study rosis patients. Clin Neurol Neurosurg 2012; 114:659–62.
3. Awad A, Hemmer B, Hartung HP, Kieseier B, Bennett JL,
we tried to distinguish type 2 from type 3 based on the
Stuve O: Analyses of cerebrospinal fluid in the diagnosis
laboratory data of KFLC and its quotient. Determination and monitoring of multiple sclerosis. J Neuroimmunol
of CSF KFLC was statistically significant (p<0.05), 2010; 219:1–7.
whereas comparing Q KFLC between type 2 and type 3 4. Compston A, Coles A: Multiple sclerosis. Lancet 2008;
reached borderline significance only (p=0.06). ROC 372:1502-17.
analysis of KFLC in type 2 and type 3 indicated a diag- 5. DeCarli C, Menegus MA, Rudick RA: Free light chains in
nostic sensitivity of 61.5% and diagnostic specificity of multiple sclerosis and infections of the CNS. Neurol 1987,
71.0%. The area under the curve was 0.641. 37:1334–8.
6. Franciotta D, Lolli F: Interlaboratory reproducibility of
isoelectric focusing in oligoclonal band detection. Clin
CONCLUSION Chem 2007; 53:1557–8.
7. Freedman MS, Thompson EJ, Deisenhammer F,
Although we obtained a statistically significant Giovanonni G, Grimsley G et al.: Recommended standard
difference in KFLC determination in this study, which of cerebrospinal fluid analysis in the diagnosis of multiple
indicates that a nephelometric method could be used to sclerosis: a consensus statement. Arch Neurol 2005;
distinguish two types in Andersson's classification, its 62:865-70.
low sensitivity and specificity does not confirm the 8. Goffette S, Schluep M, Henry H, Duprez T, Sindic CJ:
commonality of findings, so further using of technically Detection of oligoclonal free kappa chains in the absence
more demanding method of IEF followed by immuno- of oligoclonal IgG in the CSF of patients with suspected
multiple sclerosis. J Neurol Neurosurg Psychiatry 2004;
fixation is still recommended. Additional determination
75:308–10.
of KFLC with a nephelometric method would be useful
9. Van der Heijden M, Kraneveld A, Redegeld F: Free
for patients with unclear results of IEF followed by immunoglobulin light chains as target in the treatment of
immunofixation and with low resolution of KFLC, chronic inflammatory diseases. Eur J Pharmacol 2006;
therefore this method should further be assessed in 533:319–26.
multicentered prospective studies on a larger number of 10. Kaplan B, Aizenbud BM, Golderman S, Yaskariev R, Sela
patients classified according to Andersson as type 2 or BA: Free light chain monomers in the diagnosis of
type 3. multiple sclerosis. J Neuroimmunol 2010; 229:263–71.

S127
 Marina Vasilj, Miro Klarić, Nada Vrkić, Ivanka Mikulić, Marijana Marković Boras, Nevenka Jelić-Knezović & Violeta Šoljić:
KAPPA FREE LIGHT CHAINS IN CEREBROSPINAL FLUID OF PATIENTS WITH IDENTIFIED OLIGOCLONAL IMMUNOGLOBULIN G
Psychiatria Danubina, 2017; Vol. 29, Suppl. 2, pp 124-128

11. Kaplan B, Livneh A, Sela BA: Immunoglobulin free light and avidin-biotin amplification. Clin Chem 1984,
chain dimers in human diseases. Sci World J 2011; 30:1246-9.
11:726–35. 18. Passerini G, Dalla Costa G, Sangalli F, Moiola L,
12. Keir G, Luxton RW, Thompson EJ: Isoelectric focusing of Colombo B, Locatelli M et al.: Free light chains and intra-
cerebrospinal fluid immunoglobulin G: an annotated thecal B cells activity in multiple sclerosis: A prospective
update. Ann Clin Biochem 1990; 27:436-43. study and meta-analysis. Mult Scler Int 2016; 9.
13. Krakauer M, Schaldemose Nielsen H, Jensen J, Sellebjerg 19. Presslauer S, Milosavljevic D, Huebl W, Parigger S,
F: Intrathecal synthesis of free immunoglobulin light Schneider-Koch G, Bruecke T: Kappa Free Light Chains:
chains in multiple sclerosis. Acta Neurol Scand 1998; Diagnostic and Prognostic Relevance in MS and CIS.
98:161–5. PloS ONE 2014; 9:2.
14. Link H: The cerebrospinal fluid in multiple sclerosis. Clin 20. Rudick RA, French CA, Breton D, Williams GW: Relative
Neurol 1991; 1128-39. diagnostic value of cerebrospinal fluid kappa chains in
15. Link H, Huang YM: Oligoclonal bands in multiple MS: comparison with other immunoglobulin tests. Neurol
sclerosis cerebrospinal fluid: An update on methodology 1989; 39:964–8.
and clinical usefulness. J Neuroimmunol 2006; 180:17– 21. Senel M, Tumani H, Lauda F, Presslauer S, Mojib-
28. Yezdani R, Otto M et al.: Cerebrospinal fluid
16. Montalban X, Tintoré M, Swanton J, Barkhof F, Fazekas immunoglobulin Kappa light chain in clinically isolated
F, Filippi M et al.: MRI criteria for MS in patients with syndrome and multiple sclerosis. PloS ONE 2014; 9:4.
clinically isolated syndromes. Neurol 2010; 74:427-34. 22. Te Velthius H, Knop I, Stam P, van den Broek, Klaasse
17. Olsson T, Kostulas V, Link H: Improved detection of Bos H, Hol S et al.: N Latex FLC – new monoclonal high-
oligoclonal IgG in cerebrospinal fluid by isoelectric- performance assays for determination of free light chain
focusing in agarose, double-antibodyperoxidase labeling, kappa and lambda. Clin Chem Lab Med 2011; 49:1323-32.

Correspondence:
Marina Vasilj, Master of Chemistry and Biology
Department of Laboratory Medicine, University Clinical Hospital Mostar
Bijeli Brijeg bb, 88000 Mostar, Bosnia and Herzegovina
E-mail: [email protected]

S128