Diagnostic Accuracy of Serum Free Light Chains in Patients with Suspected Central Nervous System Disorders: A Retrospective Observational Study

Abstract

Background: Free light chains (FLCs), particularly kappa (?)- and lambda (?)-FLCs, are immunoglobulin components that may reflect systemic or intrathecal immunoglobulin activity. FLC-based biomarkers have been investigated extensively in inflammatory and demyelinating neurological disorders, particularly multiple sclerosis. However, most established evidence concerns cerebrospinal fluid (CSF) ?-FLC or the CSF/serum ?-FLC index, whereas the diagnostic utility of serum FLC measurements alone in patients undergoing evaluation for suspected central nervous system (CNS) disease remains less clearly defined. Objective: To evaluate the diagnostic accuracy of serum ?-FLC, ?-FLC, and the ?/? FLC ratio for distinguishing inflammatory/demyelinating CNS disorders from non-inflammatory neurological disorders and to assess their discriminatory performance using receiver operating characteristic (ROC) curve analysis. Materials and Methods: This retrospective observational diagnostic-accuracy study was conducted in the Department of Neurology and department of biochemistry, Mahavir Institute of Medical Sciences, Vikarabad, from June 2025 to March 2026. A total of 120 patients aged ?18 years who underwent evaluation for suspected CNS disease and had available serum FLC measurements were included. Based on the final documented clinical diagnosis, 42 patients had inflammatory/demyelinating CNS disorders and 78 had non-inflammatory neurological disorders. Serum ?-FLC and ?-FLC were measured using the Freelite® assay (The Binding Site Ltd., Birmingham, UK) on an Optilite® turbidimeter. The laboratory reference intervals were 3.3–19.4 mg/L for ?-FLC, 5.7–26.3 mg/L for ?-FLC, and 0.26–1.65 for the ?/? ratio. For the primary diagnostic analysis, serum FLC positivity was defined as ?-FLC >19.4 mg/L. The final clinical diagnosis served as the reference standard. ROC analysis was performed using continuous serum ?-FLC concentration as the test variable and inflammatory/demyelinating CNS disease as the binary reference outcome. Results: Among 120 patients, 42 (35.0%) had inflammatory/demyelinating CNS disorders and 78 (65.0%) had non-inflammatory neurological disorders. Mean serum ?-FLC concentration was higher in the inflammatory/demyelinating group than in the non-inflammatory group (23.8 ± 9.1 vs 15.8 ± 8.3 mg/L; p<0.001). Using a predefined ?-FLC cutoff of >19.4 mg/L, 38 of 42 patients with inflammatory/demyelinating CNS disorders were classified as positive compared with 13 of 78 patients in the non-inflammatory group. Sensitivity was 90.5% (95% CI, 77.9–96.2%), specificity was 83.3% (95% CI, 73.5–90.0%), positive predictive value was 74.5% (95% CI, 61.1–84.5%), negative predictive value was 94.2% (95% CI, 86.0–97.7%), and overall accuracy was 85.8% (95% CI, 78.5–91.0%). ROC analysis of continuous serum ?-FLC concentration demonstrated an AUC of 0.91 (95% CI, 0.85–0.96; p<0.001). The optimal ROC threshold was 19.4 mg/L, corresponding to a sensitivity of 90.5% and specificity of 83.3%. Conclusion: Serum ?-FLC demonstrated good discriminatory performance for differentiating inflammatory/demyelinating CNS disorders from non-inflammatory neurological disorders in this retrospective cohort. The relatively high negative predictive value suggests potential utility as an adjunctive screening biomarker. However, serum FLC measurement reflects circulating immunoglobulin activity and should not be interpreted as a direct measure of intrathecal immunoglobulin synthesis. Clinical findings, neuroimaging, CSF analysis, oligoclonal band testing, and other relevant investigations remain essential for definitive diagnosis. Prospective multicentre studies using standardized assays and laboratory-specific thresholds are required to validate these findings. Keywords: free light chains; kappa free light chain; lambda free light chain; ?/? ratio; central nervous system disorders; inflammatory neurological disease