Prognostic value of neurofilament light chain in chronic inflammatory demyelinating polyneuropathy

Chronic inflammatory demyelinating polyneuropathy is a neuroinflammatory disorder with considerable variation in clinical phenotype, disease progression and therapy response among patients. Recently, paranodal antibodies associated with poor response to intravenous immunoglobulin therapy and more aggressive disease course have been described in small subsets of patients, but reliable serum-based prognostic biomarkers are not yet available for the general population. In current retrospective longitudinal study, we utilized logistic regression models to investigate the associations of serum neurofilament light chain levels with 1-year disease progression and therapy response during follow-up in chronic inflammatory demyelinating polyneuropathy. One-year disease progression was defined as a decrease of four or more points (the minimal clinically important difference) on an 80-point Medical Research Council sum-score scale 1 year after sampling. Patients who, compared to treatment received at time of sampling, required therapy switch during follow-up due to insufficient effect were classified as non-responders. Serum neurofilament light chain was measured by electrochemiluminescence assay in clinical residual serum samples of 76 patients diagnosed with probable (13 patients) or definite (63 patients) chronic inflammatory demyelinating polyneuropathy according to European Federation of Neurological Societies/Peripheral Nerve Society diagnostic criteria. Eleven (15%) patients were female, and the mean (standard deviation) cohort age was 61.5 (11.7) years. In both univariate and multivariable (including demographics) models, elevated serum neurofilament light chain harboured increased odds for 1-year disease progression (respectively odds ratio, 1.049; 95% confidence interval, 1.022-1.084 and odds ratio, 1.097; 95% confidence interval, 1.045-1.169; both P = 0.001). Patients with levels above the median cohort neurofilament light chain level (28.3 pg/ml) had largely increased odds of 1-year disease progression (univariate: odds ratio, 5.597; 95% confidence interval, 1.590-26.457; P = 0.01; multivariable: odds ratio, 6.572; 95% confidence interval, 1.495-39.702; P = 0.02) and of insufficient treatment response (univariate: odds ratio, 4.800; 95% confidence interval, 1.622-16.442; P = 0.007; multivariable: odds ratio, 6.441; 95% confidence interval, 1.749-29.357; P = 0.009). In a combined approach analysis, patients with levels above median cohort serum neurofilament light chain level reported strongly increased odds of demonstrating 1-year disease progression and/or therapy non-response during follow-up (univariate: odds ratio, 6.337; 95% confidence interval, 2.276-19.469; P < 0.001; multivariable: odds ratio, 10.138; 95% confidence interval, 2.801-46.404; P = 0.001). These results show that in various logistic regression models, serum neurofilament light chain was associated with both 1-year disease progression and therapy response during follow-up in chronic inflammatory demyelinating polyneuropathy. Hence, our findings warrant further prospective research regarding the value of neurofilament light chain as potential prognostic biomarker in chronic inflammatory demyelinating polyneuropathy.