Objective: To examine the effects of tumor necrosis factor inhibitors on the risk for serious infections and other influencing factors in a registry.
Methods: Patients exposed for the first time to etanercept, adalimumab, or methotrexate and serious infections were identified in the German Biologic Registry for Pediatric Rheumatology (BIKER) registry. Serious infection rates per 1,000 observation-years and relative risks were calculated. Cox regression identified risk factors and provided hazard ratios (HRs) for occurrence of infections.
Results: A total of 3,350 patients with 5,919 observation-years fulfilled the inclusion criteria for the study. The first biologic agents were etanercept (1,720 cases) and adalimumab (177 cases). A total of 1,453 patients were treated with methotrexate and no biologic agent. In total, 28 serious infections were reported in 26 patients (4.7 per 1,000 patient-years), 5 with methotrexate (1.6 per 1,000 patient-years), 21 with etanercept (8.1 per 1,000 patient-years), and 2 with adalimumab (9.7 per 1,000 patient-years). Significant univariate risk factors for infection were therapy with biologic agents, disease duration before therapy start, corticosteroid medication, nonbiologic premedications, higher clinical Juvenile Arthritis Disease Activity Score including maximal 10 joints (cJADAS10) at therapy start, and higher mean cJADAS10 during therapy. In multivariate Cox regression, only biologic therapy and cJADAS10 at therapy start remained significant. Risk for infection was increased by etanercept (univariate HR 6.0 [95% confidence interval (95% CI) 2.0-17.5]) or adalimumab (HR 7.3 [95% CI 1.3-40.0]) compared to methotrexate as well as by an elevated cJADAS10 at therapy start (HR 1.1 [95% CI 1.0-1.2] per unit increase).
Conclusion: The total rate of serious infections reported in the BIKER registry seems low. Treatment with etanercept or adalimumab increases the risk for serious infection slightly, compared to methotrexate. Disease activity expressed by cJADAS10 appears to be an independent risk factor.
© 2016, American College of Rheumatology.