Keywords
serum immunofixation - serum free light chain - multiple myeloma - concordance - discordance
Introduction
The International Myeloma Working Group (IMWG; 2014) has elaborated criteria for evaluation
of multiple myeloma (MM). According to which, for any suspected case, the initial
investigation includes serum protein electrophoresis (SPEP) and serum immunofixation
electrophoresis (SIFE), 24-hour urine sample for urine protein electrophoresis and
urine immunofixation electrophoresis (UIFE), and estimation of serum free light chains
(SFLCs).[1] Of these, SIFE has been designated as the “gold standard” for confirming the presence
of monoclonal protein (M protein).[2]
While SPEP is utilized as a screening test for the presence of M protein, SIFE determines
both monoclonality and isotype. SFLC assay is utilized to estimate the levels of free
κ and λ light chains in the serum. A ratio of κ and λ further aids in the diagnosis
of monoclonal plasma cell disorders.[3]
The SFLC assays have been reported to be more sensitive than SIFE or SPEP in detecting
FLC M proteins. Interestingly, the disease is detected earlier with SFLC than IFE
and patients are followed-up with SFLC ratio.[4] Although it is speculated that negative SIFE plausibly have normal SLFC ratio as
well, however, we observed a significant discrepancy between SIFE and SFLC in MM patients
at our center. Recent studies have divulged sporadic reports with similar observations.[2]
[3]
[4]
Thus, the current study was undertaken to assess the concordance and discordance between
SIFE and SFLC at our center. In addition, the present study determines the sensitivity
and specificity of the turbidimetry-based SFLC assay as compared to electrophoretic
techniques.
Materials and Methods
This was a retrospective observational study for an 18-month period, from January
2021 till June 2022. All treatment-naive and follow-up cases of MM were based on SPEP
results, which were retrieved from the archival database of the Department of Immunopathology,
Postgraduate Institute of Medical Education and Research, Chandigarh, India. SIFE
was performed for confirmation and characterization of isotype of M proteins. SPEP
was performed using serum protein 6 band Agarose gel by Helena Biosciences platform
using SAS-1 SP-24 SB kit. SIFE was performed using Agarose gel-based electrophoresis
kit by Helena Biosciences (SAS-1 IFE-4 kit) that utilizes monospecific antisera for
immunoglobulin (Ig) G, IgM, and IgA heavy chains and κ and λ light chains. SFLC estimation
was carried out by turbidimetry using Optilite Freelite kappa free kit (LK016.OPT)
and Optilite Freelite lambda free kit (LK018.OPT) (The Binding Site Group Limited).
The measuring range for kappa and lambda free light chains are 0.6 to 127000 mg/L
and 1.3 to 139000 mg/L, respectively.
Monoclonal gammopathy was defined by the presence of a discrete monoclonal (M) band
in the gamma region or prominent bands in regions of other proteins in the serum.
Concordance between SIFE and SFLC was defined as similar results on both SIFE and
SFLC, that is, if SIFE was positive, SFLC ratio was also abnormal (< 0.26 or > 1.65)
and if SIFE was negative, SFLC ratio was also in normal range (0.26–1.65). Discordance
was defined as positive SIFE with normal SFLC ratio or negative SIFE with abnormal
SFLC ratio.
Statistical Analysis
Statistical analysis was done using IBM SPSS 20 (SPSS Inc, Chicago, United States).
The results were expressed as mean ± standard deviation (SD) for all continuous variables
and as percentages for categorical variables. To obtain the association of categorical
variables, chi-square test was applied. To find out the efficacy of two methods, McNemar
test was used. A p-value of < 0.05 was considered as statistically significant.
Results
The current study included a cohort of 377 patients, which contributed to 450 values
of both SIFE and SFLC. Their age ranged from 24 to 90 years with mean ± SD of 58.6 ± 9.1
years and median age of 59 years. The male-to-female ratio was 1.4:1. Majority values
(279/450, 62%) were patients on follow-up. The cohort included 39 patients with light
chain MM. Of these, 21 (53.8%) patients had lambda-associated MM and 18 (46.2%) patients
had kappa-associated MM.
Of the 450 values for SIFE and SFLC, there were 129 (28.7%) values which were discordant,
that is, positive SIFE with normal SFLC ratio or negative SIFE with abnormal SFLC
ratio. These cases with discordant values were further segregated into two groups:
SIFE positive-SFLC normal and SIFE negative-SFLC abnormal for further analysis ([Table 1]).
Table 1
Results of SIFE and SFLC in study cohort (n = 450)
|
Test results
|
SIFE positive (%)
|
SIFE negative (%)
|
Total
|
|
κ/λ ratio abnormal
|
237 (52.7)
|
38 (8.4)
|
275 (61.1)
|
|
κ/λ ratio normal
|
91 (20.2)
|
84 (18.7)
|
175 (38.9)
|
|
Total
|
328 (72.9)
|
122 (27.1)
|
450
|
Abbreviations: SFLC, serum free light chain; SIFE, serum immunofixation electrophoresis.
SIFE positive-SFLC normal cases: This subgroup included 91/450, 20.2% values. Of these, majority (61, 67.1%) occurred
in follow-up cases. The dominant finding in 51/61, 83.6% of the follow-up cases was
M band on SPEP, while in 10/61 (16.4%) cases although SPEP was normal but monoclonal
bands were noted on SIFE. Similarly, in treatment-naive cases too, 28/30, 93.3% had
positive SPEP.
SIFE negative-SFLC abnormal cases: This subgroup accounted for 8.4% (38/450) values. On further analysis, 32 (84.2%)
values were observed in follow-up cases.
A higher proportion of discordance was observed in the subgroup SIFE positive-SFLC
normal cases as compared to SIFE negative-SFLC abnormal cases. Taking SIFE as a standard
for presence of M protein, the two methods were compared ([Table 2]).
Table 2
Comparison of SIFE and SFLC in new and follow-up cases of MM (n = 450)
|
Test
|
SIFE positive
|
SIFE negative
|
Sensitivity (%)
|
Specificity (%)
|
Accuracy (%)
|
p-Value
|
PPV (%)
|
NPV (%)
|
|
κ/λ ratio
abnormal
|
237
|
38
|
72.3
|
68.9
|
71.3
|
< 0.00001[a]
|
86.2
|
47.9
|
|
κ/λ ratio
normal
|
91
|
84
|
|
Abbreviations: MM, multiple myeloma; NPV, negative predictive value; PPV, positive
predictive value; SFLC, serum free light chain; SIFE, serum immunofixation electrophoresis.
a The chi-square statistic is 63.2324. The p-value is < 0.00001 (p-value < 0.05 is significant).
Discussion
IMWG defines complete response (CR) in MM as “negative SIFE and UIFE with absence
of soft tissue plasmacytomas and bone marrow plasma cells less than 5%” and stringent
CR (sCR) as “normal SFLC ratio along with absence of clonal bone marrow plasma cells,
demonstrable by immunohistochemistry or immunofluorescence.”[5]
[6] Thus, SIFE and SFLC have distinct roles in assigning response criteria in MM. To
simplify, SIFE negativity is required for CR and SFLC normalization is favored for
sCR in MM. On the other hand, the minimal residual disease (MRD) analysis requires
detection of very low levels of persistent or reemergent neoplastic plasma cells by
highly sensitive multicolor/next-generation flow cytometry and/or next-generation
sequencing on bone marrow aspirates in patients who have achieved CR.[7]
[8] However, there are no specified time intervals for MRD testing.[8]
[9]
SIFE is the most sensitive method for identification and characterization of M proteins.
It is a unique technique integrating the resolution offered by SPEP with specificity
of antigen-antibody reaction.[7] In initial “electrophoresis” phase, the serum gamma globulins are separated based
on their electrophoretic mobility under an electric field, followed by “fixation,”
whereby specific antisera are individually added to each migration lane to precipitate
out the heavy and light chains from gamma globulins in form of visible precipitin
band.[10]
[11] Turbidimetry-based measurement of SFLCs utilizes polyclonal and monospecific anti-κ
and anti-λ antibodies.[11]
[12] The relative sensitivities of SPEP, SIFE, and abnormal SFLC ratio have been estimated
as 77, 95, and 96%, respectively.[13] The turbidimetry-based FLC assays are reported to be 50 to 100 times more sensitive
than SIFE or SPEP in detecting M proteins, which enables detection of SFLC earlier
than SIFE.[4] Further, SFLC performed along with SPEP and SIFE is claimed to improve the overall
sensitivity for screening and prognostication of MM disease spectrum.
Previous studies attempted to compare these highly sensitive turbidimetry-based SFLC
assays with standard electrophoresis techniques ([Table 3]). The discordance rates between results of SIFE and SFLC in these studies are quite
variable, ranging from approximately 17% to as high as 50% ([Table 3]).[3]
[4]
[14]
[15]
[16]
[17] Our study revealed a discordance rate of 28.7%, which is comparable to the rates
reported by Wood at el and Singhal et al, respectively.[4] The normal range of SFLC ratio is 0.3 to 1.2; an abnormal ratio outside the acceptable
reference range is considered to favor monoclonal over polyclonal elevations of SFLC.[18]
[19] In MM, the reference range of diagnostic ratio is taken to be 0.26 to 1.65. However,
an exception to this reference ratio is made for patients with renal failure, where
the SFLCs are more retained and the ratio is revised to 0.37 to 3.17.[20]
Table 3
Isotype distribution of follow-up cases of MM where SFLC was within normal limits
but SIFE was positive (n = 58)
|
Isotype
|
No. of cases
|
|
IgG κ
|
21
|
|
IgA κ
|
6
|
|
IgG λ
|
22
|
|
IgA λ
|
6
|
|
κ light chain disease
|
1
|
|
λ light chain disease
|
2
|
|
Total
|
58
|
Abbreviations: Ig, immunoglobulin; SFLC, serum free light chain; SIFE, serum immunofixation
electrophoresis.
Note: Total λ-associated cases: 30; Total κ-associated cases: 28.
Several plausible causes have been put forth in the literature to explain the discordance
between the two testing modalities. According to Böer and Deufel, not just MM but
non-neoplastic conditions such as infections, autoimmune disorders, chronic liver
disease, and neurological disorders as well as certain malignancies produce hypergammaglobulinemia
or elevated serum levels of κ and λ.[21] Bhole et al and Heaton et al attributed issues such as excess antigen, nonlinear
antigen-antibody reaction, and polymerization of FLC molecules, which lead to falsely
elevated or diminished values of SFLC.[22]
[23] Associated renal dysfunction and aggregator property of SFLC can also yield erroneous
results which do not correspond to SIFE as mentioned by Singh.[2] Further, Udd et al in their study concluded that the FLCs need to be sufficiently
elevated to be able to be detected via FLC assays.[24] Since discrepancy was more in follow-up cases, it may also be reflective of the
effect of stage of disease, individual heterogeneity in disease progression, and disease
biology on the estimation of SFLC as demonstrated by Habib et al.[25] It has been observed that approximately 36% of patients may demonstrate abnormal
SFLC ratio even without monoclonal gammopathy, which is mainly κ-associated.[26] Normal SFLC ratio despite positive SIFE is ascribed to shorter half-lives of κ and
λ light chains.[4] Also, false negative SFLC ratios are encountered more frequently with λ-light chain
disorders owing to underproduction and/or underdetection of excess λ FLC.[2] In the current study, we noted an increased number of λ-associated MM cases on follow-up
when SFLC was within normal limits, however, SIFE reported positive ([Table 4]).[2]
[3]
[4]
[15]
[16]
[17] Interestingly, κ-associated SFLC ratios may be seen in λ chain–associated MM postautologous
stem cell transplantation contributing to unexpected results.[20] Thus, it can be concurred, that SFLC ratio does not always score above SIFE for
diagnosis and monitoring in a subset of MM patients.[2]
[3]
[4]
[12]
[15]
[16]
[17]
Table 4
Studies on discordance between SIFE and SFLC
|
Serial no.
|
Authors
|
Country
|
Cohort
|
Methodology
|
Results
|
Conclusion
|
|
1
|
Singhal et al (2009)
|
USA
|
No. of patients = 122 (values available = 2,648)
|
SFLC: Freelite, The Binding Site on Dade-Behring nephelometer
|
Discordance between SIFE and SFLC-33.9%
No of SIFE positive-SFLC normal patients was much greater than SIFE negative-SFLC abnormal patients
|
Normal SFLC cannot rule out residual disease
Normal SFLC more likely to be in presence of positive SIFE than negative SIFE in presence
of abnormal SFLC
|
|
2
|
Wood et al (2010)
|
USA
|
No. of samples = 501
|
IFE: Paragon IFE kit (Beckman Coulter, Brea, CA); SFLC: Freelite,
(The Binding Site Ltd, Birmingham) on Dade
Behring BNII automated nephelometer
|
Discordance between SIFE and SFLC-24.6%
82% light chain disease patients had abnormal SFLC
|
Proportion of patients with normal SFLC similar in IgG and IgA subgroups
|
|
3
|
Kim et al (2014)
|
Korea
|
No. of samples = 157
|
SFLC: N Latex FLC assays (Siemens Healthcare
Diagnostics GmbH, Marburg, Germany)
and Freelite assays (The Binding Site Ltd., Birmingham) performed on Behring Nephelometer
II (Siemens Healthcare Diagnostics GmbH, Germany)
IFE: Agarose gel (Helena Laboratories, Beaumont, USA)
|
Discordance between SIFE and SFLC: 17.8% (N latex assay), 17.2% (Freelite assay)
|
Essential to perform SIFE with SFLC to detect M-protein in cases with very low M-protein
levels, CKD, polyclonal gammopathy, biclonal gammopathy, IgM-type monoclonal gammopathy
|
|
4
|
Li et al (2015)
|
China
|
No. of patients = 16
|
SPEP: Sebia Hydragel 15 protein kit (Sebia, Lisses, France); SIFE: Sebia Hydragel
4 Immunofixation PE kit on Hydrasys
system (Sebia, Lisses, France);SFLC:Immunonephelometry (Freelite, Binding Site Ltd,
Birmingham)Tests were performed on Beckman Coulter Immage 800 (Beckman Coulter, Brea,
CA)
|
50% patients with positive SIFE but normal SFLC
|
IFE is more sensitive than SPEP and SFLC assay in detection of relapse
|
|
5
|
Singh (2017)
|
USA
|
No. of patients = 468 (values available = 2,409)
|
SPEP: Agarose gel electrophoresis using Helena SPIFE 3000 system
SFLC: Siemens ADVIA 2400 instrument, using Freelite kits and reagents from the Binding
Site
|
Concordance rate of
electrophoretic method greater than for SFLC ratio (p << 0.00001)
Discordant rate of SPEP/SIFE:0.58%
Discordance rate for SFLC ratio: 3.3%
|
Electrophoretic tests are superior to SFLC ratio in diagnosis and monitoring of MG
Nonconcordant observations in
SFLC had detectable monoclonal proteins by electrophoretic method
Nonconcordance rate was nearly twice as good for electrophoretic
methods compared to SFLC ratio
|
|
6
|
Kuriakose et al (2019)
|
India
|
No. of patients = 46
|
SPEP: Agarose gel zone electrophoresis on Interlab Genios fully automated
Machine
SIFE: Interlab Genios fully automated machine
SFLC: Immunoturbidimetry
on Beckman Coulter AU 2700 analyzer
|
19% discordance between SIFE and SFLC
|
Combination of SPEP, SIFE, and SFLC has more diagnostic potential in the identification
of MG than in isolation
Sensitivity, specificity, positive, and NPV of SIFE with respect to SFLC were 81.3,
78.6, 89.7, and 64.7%, respectively
Accuracy of SIFE came to be 80% compared to SFLC
|
|
7
|
Current study (2022)
|
India
|
No. of patients = 377 (values available = 450)
|
SPEP: serum protein 6 band gel by Helena biosciences (SAS-1 SP-24 SB kit)
SIFE: gel-based electrophoresis kit by Helena biosciences (SAS-1 IFE-4 kit)
SFLC: Turbidimetry using Optilite Freelite Kappa and Lambda Free Kit by The Binding
Site group
|
28.7% values had discordance between SIFE and SFLC ratios
More follow-up cases with normal SFLC ratio but positive SIFE than abnormal SFLC ratio
and negative SIFE
|
Normal SFLC ratio alone cannot exclude residual disease as defined conventionally
by positive SIFE
|
Abbreviations: CKD, chronic kidney disease; Ig, immunoglobulin; SFLC, serum free light
chain; SIFE, serum immunofixation electrophoresis; SPEP, serum protein electrophoresis.
For any suspected case, an abnormal SFLC ratio must be confirmed by electrophoretic
studies. An electrophoretic evidence of M protein is a reliable diagnostic marker
of MM. However, a substantial rate of false negative SFLC ratio, in patients with
detectible M protein argues against using it as the only modality for guiding evaluation
of MM, especially in follow-up phase. Furthermore, serial measurements of SFLC ratio
are only appropriate and reliable when read in parallel with SIFE or bone marrow findings.[2] Several studies have upheld the use of combination or sequential tests for accurate
diagnosis owing to the discrepant results of SFLC vis-a-vis electrophoresis.[3]
[17]
[21]
[22]
[23]
[25]
[26] In the current study, we encountered 28.9% discordance between SIFE and SFLC results.
Since normal SFLC ratio heralds an impending negative SIFE, we propose a more practical
laboratory-based approach to follow-up MM patients. All old cases of MM may be followed
up with SFLC measurement. After attaining normal SFLC ratio, further follow-up should
be done with SIFE and that may be considered as the actual, deepest response to therapy.
This algorithm will help in optimizing the available resources in the laboratory in
the best possible manner. Furthermore, few authors have reported SIFE to be more sensitive
and useful than SFLC for detection of residual disease.[16] By adopting this algorithm, we can optimize the utilization of the tests and resources
as well.[23]
However, in this laboratory-based study, long-term follow-up was not available due
to limited study time period. Also, the cohort comprising of light chain multiple
myeloma patients was small.
Conclusion
The current study denotes that normal SFLC ratio cannot exclude residual disease as
defined conventionally by positive SIFE. Larger studies are needed to explore the
temporal relation between normalization of SFLC ratio and serum/urine M protein clearance.
