Ofatumumab Versus Intravenous Ocrelizumab after Sphingosine-1-Phosphate Receptor Modulators in Patients with Relapsing–Remitting Multiple Sclerosis: A Real-World Inverse Probability of Treatment Weighting Multicenter Study

Background and Aims: Evidence directly comparing ocrelizumab (OCR) and ofatumumab (OFA) after sphingosine-1-phosphate receptor modulators (S1PRMs) withdrawal is limited, even though this transition period carries an increased risk of disease reactivation and the two anti-CD20 agents differ in molecular properties that may influence post-S1PRM outcomes. We compared effectiveness, safety, and tolerability of OFA versus OCR in relapsing–remitting multiple sclerosis (RRMS) after S1PRM therapy. Methods: We retrospectively analyzed adult (> 18 years) patients with RRMS from 23 Italian centers who switched from S1PRMs to OCR or OFA (October 2016 to July 2024). Clinical outcomes included annualized relapse rate (ARR), cumulative relapse incidence, confirmed disability progression (CDP), progression independent of relapse activity (PIRA), confirmed disability improvement (CDI), and variation in the expanded disability status scale score (ΔEDSS) from baseline to last follow-up. Magnetic resonance imaging (MRI) outcomes included time to overall MRI activity, new/enlarging T2 lesions, and new T1 Gd-enhancing lesions. Persistence on anti-CD20 and safety were also assessed. We applied inverse probability of treatment weighting (IPTW) to balance baseline covariates, and used Cox, Andersen–Gill, and regression models, further adjusted for variables with residual imbalance, to compare outcomes. Results: We included 225 subjects (mean age 43.3 ± 10.6 years; OCR = 131, OFA = 94), with a median follow-up of 33 months. After IPTW, groups were well balanced. Compared with OCR, OFA was associated with lower ARR (adjusted relative risk [aRR] 0.22, 95% CI 0.06–0.86, p = 0.030) and relapse hazard (adjusted hazard ratio [aHR] 0.21, 95% CI 0.05–0.83, p = 0.026). When restricting washout to ≤ 10 weeks and applying spline-based Andersen–Gill models, results remained consistent, with OFA showing a lower relapse risk than OCR across short-to-intermediate washout intervals. CDP hazard was lower (aHR 0.38, 95% CI 0.15–0.94, p = 0.038), while PIRA and CDI did not differ. ΔEDSS favored OFA (adjusted mean difference [aMD] − 0.26, 95% CI − 0.51 to − 0.02, p = 0.046). OFA showed a lower hazard of overall MRI activity (aHR 0.41, 95% CI 0.19–0.92, p = 0.030), with similar risk for new T1 Gd+ lesions. A sensitivity analysis restricted to patients who switched to OFA or OCR during the same availability window yielded results consistent with the primary analysis. Treatment persistence was shorter on OFA, but absolute discontinuation rates were low in both groups. Severe AEs were rare and did not differ. Conclusions: In S1PRM-exposed RRMS, OFA was associated with reduced relapse risk, disability accumulation, EDSS worsening, and MRI activity compared with OCR, with overall good tolerability and safety.