Abstract / Summary
Background: Reference-arm annualized relapse rates (ARRs) in relapsing multiple sclerosis trials have declined. We examined whether baseline disease activity and the treatment environment during recruitment changed alongside this trend.
Methods: We analyzed 15 treated-reference arms from nine Phase 3 programs. ARR was related to recruitment era, prior-year relapse activity, MRI markers, and an Operational Selection Pressure Index (OSPI) combining recruitment geography with time since ocrelizumab approval. Associations were descriptive; disability outcomes were audited separately.
Results: Reference-arm ARR declined from 0.370 in TEMSO and 0.320 in TOWER to 0.109-0.140 in BTK inhibitor-era programs (ρ = -0.906; 95% CI, -0.985 to -0.279; n = 15). Prior-year relapse activity correlated with ARR (ρ = 0.739; 95% CI, 0.158 to 0.928; n = 13) and declined over time (ρ = -0.725; 95% CI, -0.965 to -0.157; n = 13). MRI findings were inconsistent. OSPI clustered at 0-7.3 in six programs and 65.2-80.5 in BTK inhibitor-era programs and correlated inversely with ARR (ρ = -0.892; 95% CI, -1.000 to -0.569; n = 9). The association persisted across sensitivities but was negligible after adjustment for recruitment midpoint (partial ρ = -0.020). Disability trends depended on reporting metric and effective follow-up.
Conclusions: Lower baseline relapse activity accompanied declining event yield. OSPI made a treatment-ecosystem transition measurable but could not be separated from calendar era or establish mechanism. Trials require conservative event assumptions, harmonized reporting, and validation using screening and country- or period-specific outcomes.