Biomarkers of response to ocrelizumab in relapsing-remitting multiple sclerosis

Objective: To ascertain the changes of serum neurofilament light chain (sNfL) and glial fibrillary acidic protein (sGFAP) values in relapsing-remitting multiple sclerosis (RRMS) patients treated with ocrelizumab and their association with treatment response.

Methods: Multicenter prospective study including 115 RRMS patients initiating ocrelizumab treatment between February 2020 and March 2022 followed during a year. Serum samples were collected at baseline and every 3 months to measure sNfL and sGFAP levels using single-molecule array (SIMOA) technology. Based on age and body mass index, sNfL values were standardized using z-score. NEDA (non-evidence of disease activity)-3 status was defined for patients free of disease activity after a year of follow-up. Inflammation (INFL) was considered when new relapses occurred during follow-up or new MRI lesions were found at 1-year exploration. PIRA (progression independent of relapse activity) was defined as disability progression occurring in the absence of relapses or new MRI activity.

Results: After a year on ocrelizumab, 85 patients (73.9%) achieved NEDA-3. Thirty patients did not achieve NEDA: 20 (17.4%) because of INFL and 10 (8.7%) because of PIRA. Of INFL patients, 6 (30.0%) had relapses, and 17 (85.0%) had at least one new MRI lesion at the 12-month examination. At baseline, INFL patients had higher sNfL (p = 0.0003) and sGFAP (p = 0.03) than the NEDA-3 group. PIRA patients mostly exhibited low sNfL and heterogeneous sGFAP levels. After a year, NEDA-3 and INFL patients showed similar decreases in sNfL (p < 0.0001) and sGFAP (p < 0.0001 for NEDA-3 and p = 0.001 for INFL ones). However, the decrease occurred earlier in NEDA-3 patients. Accordingly, sNfL > 1.5 z-score 3 months after ocrelizumab initiation indicated a higher risk of inflammation (OR = 13.6; p < 0.0001). Decrease in sGFAP values occurred later in both groups, with significant reductions observed at 12 months for INFL and 6 and 12 months for NEDA-3. No significant changes in sNfL or sGFAP were observed in PIRA patients.

Conclusion: Ocrelizumab induced normalization of sNfL and sGFAP in the majority of NEDA-3 and inflammatory patients but did not cause changes in the PIRA group. Our data suggest that normalization of sNfL and sGFAP is associated with the lack of inflammatory-associated disease progression but it may not affect non-inflammatory PIRA.