Concentrations of immunoglobulin free light chains in cerebrospinal fluid predict increased level of brain atrophy in multiple sclerosis

Recent studies showed that B cells play a major role in the pathogenesis of neurodegeneration in multiple sclerosis (MS). In this study, we aimed to determine the possible link between immunoglobulin free light chains (FLC) and brain atrophy in patients with MS. Ninety-two patients (32 males and 60 females) with MS were included. Kappa and lambda FLC concentrations in serum and cerebrospinal fluid (CSF) samples of MS patients were measured using ELISA assay. FLC quotients (Q-k and Q-λ, respectively) were calculated. In a cross-sectional group (n = 92), the MRI data were acquired within 6 months from the date of the lumbar puncture. Twenty patients from this cohort performed a follow-up MRI after 1 year of observation. Brain volumes were calculated with SIENAX and the brain atrophy (percentage brain volume change (PBVC)) was assessed with SIENA. Spearman’s test was performed to assess correlations. We have shown statistically significant correlation of Expanded Disability Status Scale (EDSS) level with normalized brain volume (NBV, r = − 0.2721, p = 0.0062), white matter volume (WMV, r = − 0.2425, p = 0.015), and gray matter volume (GMV, r = − 0.216, p = 0.0309). Multiple Sclerosis Severity Score (MSSS) score correlated with NBV (r = − 0.2521, p = 0.0352) and WMV (r = − 0.315, p = 0.0079). Neither EDSS, nor MSSS scores correlated with the age of patients and relapse rate during the first year and 5 years. In our study, we found statistically significant correlations of k-FLC in the CSF with NBV (r = − 0.311, p = 0.003) and with GMV (r = − 0.213, p = 0.0423). Q-k correlated only with NBV (r = − 0.340, p = 0.006) and Q-λ were negatively correlated with WMV (r = − 0.366, p = 0.003). We did not find correlations of k-FLC in CSF, λ-FLC in CSF, Q-k, and Q-λ with duration of MS course, EDSS, MSSS, number of relapses during the first year, and during the first 5 years of disease. Additionally, we subdivided the study population in accordance with level of k-FLC CSF, Q-k, and Q-λ on the 25th and 75th percentile subgroups (25-k-FLCCSF/75-k-FLCCSF; 25-λ-FLCCSF/75-λ-FLCCSF; 25-Q-k/75-Q-k; 25-Q-λ/75-Q-λ). We found statistically significant difference of NBV and GMV between 25-k-FLCCSF and 75-k-FLCCSF subgroups (p = 0.0047, p = 0.0297 respectively), NBV between 25-Q-k and 75-Q-k subgroups (p = 0.038), and NBV and WMV between 25-Q-λ and 75-Q-λ subgroups (p = 0.0446, p = 0.0026 respectively). PBVC in the prospective group showed negative correlation with kappa FLC in the CSF (r = − 0.4853, p = 0.0301) and Q-k (r = − 0.6132, p = 0.0224), but not with other clinical, epidemiological data. In this study, we showed a strong negative correlation of k-FLC, Q-k, and Q-λ with brain atrophy in MS patients. Additionally, patients with high concentration of FLC had lower brain volumes. We did not find correlations of FLC with the relapse rate, age of patients, and MS time course. In the prospective group, the rate of atrophy was correlated with k-FLC and Q-k. We suggest that level of intrathecal production of FLC can be a good prognostic biomarker for MS.