Rate of discontinuation and drug survival of biologic therapies in rheumatoid arthritis: a systematic review and meta-analysis of drug registries and health care databases

Objectives. To assess the proportion of RA patients who discontinued biologics in world registries and health care databases and to identify causes and predictors of discontinuation. Methods. Medline, Embase, Cochrane Library and Web of Science electronic databases and ACR and EULAR meeting abstracts were used. The selection of studies from world registries and health care databases including RA patients treated with biologics was independently performed. Data extracted from articles and abstracts were combined using a random effects model. Meta-analyses of percentages and hazard ratios were used to assess discontinuation. Results. Ninety-eight studies with >200 000 patients frdiscontinuation rates of TNF inhibitors at 0.5, 1, 2, 3 and 4 years were 21, om 11 242 articles and 119 abstracts met the inclusion criteria. Overall 27, 37, 44 and 52%, respectively. Discontinuation of etanercept was significantly lower at 3 and 4 years (35% and 41 %, respectively) than infliximab and adalimumab (46% and 52%, respectively). Predictors of time to discontinuation were etanercept [hazard ratios (HRs) 0.58 and 0.77 versus infliximab and adalimumab, respectively), concomitant use of DMARDs (HR 0.77), disease duration (HR 1.01) and female sex (HR 1.18). Studies from registries conducted after 2005 and from countries with lower biologics access showed higher percentages of discontinuation. Relevant data on abatacept and tocilizumab were missing. Conclusion. In RA, treatment with etanercept has a lower percentage of discontinuation than infliximab and adalimumab. Concomitant use of DMARDs, disease duration before treatment with a biologic and female sex predict time to discontinuation.