RESUMO
BACKGROUND AND OBJECTIVE: Observational studies have shown that a significant proportion of patients interchanging between tumor necrosis factor-α inhibitor biosimilars withdraws from the new treatment because of adverse effects. We aim to analyze adverse events related to interchanging from tumor necrosis factor-α (TNF-α) inhibitor reference products to biosimilars and between biosimilars reported in the World Health Organization pharmacovigilance database. METHODS: We extracted all cases reporting the Medical Dictionary for Regulatory Activities term "Product substitution issue (PT)" for TNF-α inhibitors. Then, we analyzed and categorized all adverse events reported in more than 1% of cases. We compared the adverse events reported according to reporter qualification, type of switch, and type of TNF-α inhibitor using Chi2 tests. We conducted a network analysis coupled with a clustering approach to identify syndromes of co-reported adverse events. RESULTS: In the World Health Organization pharmacovigilance database, 2543 cases and 6807 adverse events related to TNF-α inhibitor interchangeability have been reported up to October 2022. Injection-site reactions were the most reported adverse events with 940 cases (37.0%), followed by modifications in drug effect in 607 cases (23.9%). Musculoskeletal, cutaneous, and gastrointestinal disorders linked to the underlying disease were reported in 505 (20.0%), 145 (5.7%), and 207 (8.1%) cases, respectively. Adverse events non-related to the underlying disease were nonspecific (n = 458, 18.0%), neurologic (n = 224, 8.8%), respiratory (n = 132, 5.2%), and psychological disorders (n = 64, 2.5%). Injection-site reactions and infection-related symptoms (e.g., nasopharyngitis, urinary tract infection, lower respiratory tract infection) were more reported by non-healthcare professionals while adverse events related to reduced clinical efficacy (e.g., drug ineffective, arthralgia, psoriasis) were more reported by healthcare professionals. The proportions of injection-site reactions were higher when switching between biosimilars of the same reference product, but the proportions of adverse events related to reduced clinical efficacy (e.g., psoriasis, arthritis, psoriatic arthropathy) were more reported when switching from a reference product. The main differences in the proportions of reported cases between adalimumab, infliximab, and etanercept were driven by symptoms related to the underlying targeted diseases, except for a higher reporting rate of injection-site pain with adalimumab. Adverse events evocative of hypersensitivity reactions were reported in 192 (7.6%) cases. Most of the network clusters concerned non-specific adverse events or were related to reduced clinical efficacy. CONCLUSIONS: This analysis highlights the burden of patient-reported adverse events when interchanging between TNF-α inhibitor biosimilars, notably injection-site reactions, non-specific adverse events, and symptoms related to reduced clinical efficacy. Our study also highlights differences in reporting patterns between patients and healthcare professionals and depending on the type of switch. The results are limited by missing data, the lack of precision of the coded Medical Dictionary for Regulatory Activities terms, and by the variability of reporting rate of adverse events. Thus, incidence rates of adverse events cannot be inferred from these results.