Neurofilament light chain (NfL) serum level is associated with disease activity in multiple sclerosis (MS). Its practical value for detecting subclinical disease activity as currently assessed by brain MRI has not yet been evaluated.

To investigate the ability of serum NfL levels to identify brain MRI activity.

163 patients with relapsing-remitting MS patients (as per 2017 McDonald criteria) from the SET study, treated with intramuscular interferon beta1a were investigated. Serum NfL levels were evaluated at 12, 24 and 36 months from the first MS event using a Single Molecule Array assay. Age-adjusted NfL percentiles from healthy controls were defined as per Disanto; Ann Neurol 2017. Brain MRI scans were performed within 4 months from MS onset and at 12, 24 and 36 months, using single 1.5T scanner and uniform scanning protocol. The accuracy of different NfL cut-offs in identifying active lesions (new/enlarging or enhancing brain lesions) on annual scans was assessed.

NfL levels ≥90th percentile were present in 15.1% of samples. 89% of patients with NfL ≥90th percentile showed activity on their brain MRI during the preceding year. NfL levels <10th percentile were observed in 180 (44.4%) of samples and were reflective of negligible MRI activity: only 2% developed ≥3 active lesions and 12% developed ≥2 active lesions during the preceding year. 39% of these patients showed at least one active lesion (≥1 active lesion). Results were confirmed in sub-analyses of clinically stable patients and in a pooled analysis with multivariable generalized linear mixed models.

Low serum NfL levels help identify MS patients with very low recent radiological disease activity. Subject to external validation, this result suggests that NfL assessment may supplement the need for annual brain MRI monitoring in clinically stable patients with very low NfL levels.