Dimethyl Fumarate, But Not Rituximab, Reduces Serum <scp>GFAP</scp> Levels and <scp>PIRMA</scp> in Relapsing–Remitting <scp>MS</scp>

OBJECTIVE: Serum neurofilament light chain (sNfL) and glial fibrillary acidic protein (sGFAP) levels are believed to reflect mainly acute and chronic disease processes in multiple sclerosis (MS), respectively. In this study, we investigated whether dimethyl fumarate (DMF) and rituximab (RTX) differentially affect these biomarkers. METHODS: RIFUND-MS was a 2-year, rater-blinded, 1:1 randomized controlled multicenter trial comparing DMF and RTX in relapsing-remitting multiple sclerosis (RRMS). Serum samples for analysis of sNFL and sGFAP were collected at baseline and 0, 6, 12 and 24. Log-transformed biomarker data were analyzed with linear mixed models, based on intention to treat (ITT), per protocol (PP) and accounting for therapy switches. Cox proportional hazards models were performed to evaluate progression outcomes. RESULTS: Of 200 participants, 197 were analyzed. Based on ITT, sNfL decreased significantly in both arms from baseline to month 24; by 50.7% (CI 43.7%-56.8%; p < 0.001) with RTX, and by 46.4% (CI 38.6%-53.2%; p < 0.001) with DMF, no differences between treatments (global p-value: ITT = 0.06; PP = 0.08; switch group = 0.15). In contrast, sGFAP remained stable in RTX (3.6% decrease; CI -7.8%-13.8%, p = 0.81) but decreased with DMF (18.4%; CI 8.5%-27.2%; p < 0.001). Global analyses favored DMF (ITT = 0.02; PP = 0.004; switch group = 0.74). The risk of progression independent of relapse and MRI activity (PIRMA) was higher with RTX (HR 3.3, CI 1.1-10, p = 0.04). INTERPRETATION: Both RTX and DMF reduced sNfL levels, consistent with suppression of acute inflammatory disease activity. However, only DMF was associated with a sustained reduction in sGFAP and a lower risk of non-inflammatory disability progression. These findings suggest that DMF may exert additional effects on astrocyte-related or compartmentalized CNS pathology beyond peripheral immune modulation.