Determination of sensitivities and specificities of cerebrospinal fluid free light chains to diagnose multiple sclerosis- a multicentric case-control study.

Background: CSF free light chains help diagnose multiple sclerosis, but no data is available on the Asian population. Our objective was to study the diagnostic utility of CSF free light chains for diagnosing multiple sclerosis in Indian patients.
Methods: Prospective multicentric case-control study. Cases included those who were tested for oligoclonal bands and fulfilled the modified McDonald criteria 2017 for multiple sclerosis and clinically isolated syndromes. Those tested for oligoclonal bands (OCB) but with other diagnoses- inflammatory and non-inflammatory were included as controls. Clinical details were collected from electronic medical records. CSF and serum kappa and lambda free light chains were measured, apart from oligoclonal bands, immunoglobulin, and albumin in paired serum and CSF samples.
Results: There were 70 patients (31 cases and 39 controls). The mean age was 43.41(SD 16.073) years, and 43(61.4%) were females. CSF kappa showed highest specificity 97.4%, at a cut off 2.06 mg/L (sensitivity 71%) and highest sensitivity 90.3%, at a cut off 0.47 mg/L (specificity 79.5%). Best balance of sensitivity and specificity for CSF kappa was seen at a cut-off of ≥ 0.63 mg/L {sensitivity 87·1 (CI - 70.17-96.37), and specificity 87·18 (CI -72.57-95.70)}. The ratio of Kappa/lambda showed highest specificity of 100%(similar to OCB) with a sensitivity of 71% at a cut off of 1.72. The ratio of sum of kappa and lambda light chains, and Qalb (∑CSF FLC/Qalb), showed the highest specificity (94.87%)among the blood brain barrier corrected ratios.
Conclusion: This study showed that the diagnostic utility of CSF kappa was comparable to OCB to diagnose multiple sclerosis in sensitivity, but not specificity, so can be a screening test before testing for OCB in our population.
Competing Interests: Declaration of Competing Interest None of the authors report any conflicts of interest.
(Copyright © 2023. Published by Elsevier B.V.)