Drug-specific risk, and associated factors, of vasculitis-like events in patients exposed to tumour necrosis factor alpha inhibitor therapy: results from the British Society for Rheumatology Biologics Register for Rheumatoid Arthritis (BSRBR-RA)

Abstract

BackgroundThe association between TNF inhibitors and vasculitis-like events, possibly secondary to induction of autoantibodies, has been well reported. However, the incidence, drug-specific differences, and associated factors have been poorly characterised. The aim of this study was to assess the drug-specific risk, and associated factors, of vasculitis-like events in patients with rheumatoid arthritis treated with TNF inhibitors and non-biologic disease modifying antirheumatic drugs (nbDMARDs). MethodsThis analysis included two cohorts from the British Society for Rheumatology Biologics Register for Rheumatoid Arthritis (BSRBR-RA) recruited from Oct 1, 2001, to May 31, 2015: patients starting first TNF inhibitors (adalimumab, etanercept, infliximab, certolizumab) and a comparison cohort receiving nbDMARDs. After identification of vasculitis-like events, additional information from hospital consultants was sought for verification. The risk of an event was compared between the two cohorts with Cox proportional hazard models, adjusted using propensity scores (inverse probability of treatment weighting). FindingsThere were 95 incident vasculitis-like events (14 in 3640 patients on nbDMARD, 81 in 12 745 patients starting first TNF inhibitors) with 20 365 and 52 428 patient-years of follow-up generating crude incidence rates of 7 per 10 000 and 16 per 10 000 person-years, respectively. After adjustment, the hazard ratio (HR) of vasculitis-like events in patients on TNF inhibitors versus nbDMARD was 1·3 (95% CI 0·4–4·0). Drug specific HRs were highest in patients on infliximab (HR 2·7, 95% CI 1·4–5·2) and etanercept (2·6, 1·4–4·8), but these associations were not significant after adjustment. Factors associated with vasculitis-like events included baseline disease activity (HR 1·4, 95% CI 1·2–1·7), rheumatoid factor positive status (1·8, 1·2–2·8), health assessment questionnaire score (1·7, 1·2–2·3), and sulfasalazine use (0·6, 0·3–0·9). InterpretationIn one of the largest registers of patients receiving biological drugs, the absolute risk of vasculitis-like events was low, irrespective of whether patients received TNF inhibitors or nbDMARDs. There were no significant differences in risk between TNF inhibitors after adjustment. Higher baseline disease activity and seropositive status were associated with higher rates of vasculitis events, whereas baseline sulfasalazine use was associated with lower rates. To our knowledge, this is the first study to date to assess the risk of this outcome in a prospective, observational cohort. FundingMedical Research Council.