Drug Retention of First-Line Biologic Therapies in Rheumatoid Arthritis: Real-World Evidence From a Moroccan Cohort

Drug retention is a meaningful real-world composite endpoint reflecting both long-term effectiveness and tolerability of biologic therapies in rheumatoid arthritis (RA). Data from North African cohorts remain scarce, and sociocultural determinants of treatment persistence are poorly explored. We conducted a retrospective observational study including RA patients who initiated first-line biologic therapy at a Moroccan tertiary center. Drug retention was defined as the time from initiation to permanent discontinuation for any reason. To compare drug retention across biologic agents, the Kaplan-Meier method was applied alongside the log-rank test. Cox proportional hazards regression was subsequently used to determine baseline factors independently associated with treatment discontinuation. One hundred and thirty patients met the inclusion criteria, with a clear female predominance (n = 112, 86.2%) and a mean age of 58.7 ± 12.9 years. Rituximab was the most frequently prescribed first-line biologic (n = 72, 56.1%), followed by Etanercept and Tocilizumab (n = 21, 16.2% each). Drug retention differed significantly across biologics (log-rank p = 0.018). Taking Adalimumab as the reference, three biologics showed significantly better retention: Etanercept (hazard ratio (HR) = 0.14, p = 0.001), Rituximab (HR = 0.26, p = 0.006), and Tocilizumab (HR = 0.07, p < 0.001). Additional independent predictors of better retention included concomitant conventional synthetic disease-modifying antirheumatic drug (csDMARD) therapy (HR = 0.39, p = 0.014), lower baseline Simplified Disease Activity Index (SDAI) (HR = 0.96, p = 0.020), and primary education level (HR = 0.32, p = 0.033). In this real-world Moroccan cohort, non-TNF (tumor necrosis factor) biologics demonstrated favorable retention compared to Adalimumab. Beyond treatment type, both clinical and sociocultural factors were associated with persistence. Given the retrospective observational design and the inherent risk of channeling bias, these results should be interpreted with appropriate caution.