Serum neurofilament as a predictor of 10-year gray matter atrophy and clinical disability in multiple sclerosis - a longitudinal study

Background The predictive value of serum neurofilament light chain (sNfL) on long-term prognosis in multiple sclerosis (MS) is still unclear. Objective Investigate the relation between sNfL levels over a 2-year period in patients with relapsing-remitting MS, and clinical disability and grey matter (GM) atrophy after 10 years. Methods 85 patients, originally enrolled in a multicentre, randomised trial of omega-3 fatty acids, participated in a 10-year follow-up visit. sNfL levels were measured by Simoa quarterly until month 12, and then at month 24. The appearance of new gadolinium-enhancing (Gd+) lesions was assessed monthly between baseline and month 9, and then at months 12 and 24. At the 10-year follow-up visit, brain atrophy measures were obtained using FreeSurfer. Results Higher mean sNfL levels during early periods of active inflammation (Gd+ lesions present or recently present) predicted lower total (beta=-0.399, p=0.040) and deep (beta=-0.556, p=0.010) GM volume, lower mean cortical thickness (beta=-0.581, p=0.010) and higher T2 lesion count (beta=0.498, p=0.018). Of the clinical outcomes, higher inflammatory sNfL levels were associated with higher disability measured by the dominant hand Nine-Hole Peg Test (beta=0.593, p=0.004). Mean sNfL levels during periods of remission (no Gd+ lesions present or recently present) did not predict GM atrophy or disability progression. Conclusion Higher sNfL levels during periods of active inflammation predicted more GM atrophy and specific aspects of clinical disability 10 years later. The findings suggest that subsequent long-term GM atrophy is mainly due to neuroaxonal degradation within new lesions.