One target, two evidence streams: Rituximab versus ocrelizumab in relapsing multiple sclerosis: A design-stratified systematic review, meta-analysis, and concordance assessment.
Yehia Nabil, Ahmed Shaheen, Anas Mansour et al.
✦ AI-curated · Sources linked
In 30 seconds
This design-stratified systematic review and meta-analysis compared rituximab with ocrelizumab in adults with relapsing multiple sclerosis, synthesizing randomized and longitudinal observational evidence separately. Thirteen studies were included: one randomized trial involving 216 participants and 12 observational studies comprising 14,753 treatment observations. In the trial, rituximab was noninferior to ocrelizumab for freedom from new or enlarging T2 lesions from months 6 to 24 (risk difference, -2.6 percentage points; 95% CI, -9.4 to 4.3; high certainty). Observational findings suggested more relapse activity with rituximab, but certainty was very low.
Key findings
- In the randomized trial, rituximab was noninferior for freedom from new or enlarging T2 lesions from months 6 to 24 (risk difference, -2.6 percentage points; 95% CI, -9.4 to 4.3; high certainty).
- Observational evidence suggested higher relapse activity with rituximab for time to first relapse (HR, 1.92) and annualized relapse rate (IRR, 1.80); certainty was very low.
- MRI and disability outcomes were similar, while rituximab was associated with earlier discontinuation.
- Serious-infection estimates were inconsistent.
Why it matters
The review separates randomized from observational evidence, showing that comparable MRI lesion suppression does not resolve uncertainty about comparative clinical efficacy. This distinction matters when interpreting real-world signals of greater relapse activity with rituximab, which were judged very uncertain. The findings also leave serious-infection risk unclear, relevant to weighing the evidence around these commonly compared therapies.
What to keep in mind
The randomized evidence came from one trial, whereas relapse comparisons relied on observational studies with very low certainty; serious-infection estimates were inconsistent.