Drug survival of the first and second course of anti-tumour necrosis factor agents in juvenile idiopathic arthritis

Objectives: To evaluate drug survival (continuation rates on drug) of anti-tumour necrosis factor (TNF) agents in juvenile idiopathic arthritis (JIA) and predictors for treatment discontinuation. Methods: A retrospective observational study on JIA patients taking etanercept (n  =  105) or infliximab (n  =  104) with at least one year follow-up. Kaplan–Meier curves and log-rank statistics were used to compare treatments and a proportional hazards model to assess risk factors for discontinuation. Results: Etanercept versus infliximab treatment survival at 12 months was 83% versus 80%, at 24 months 68% versus 68%, at 36 months 64% versus 53%, at 48 months 61% versus 48% (p = 0.194), respectively. Reasons for discontinuing the first biological treatment were inefficacy (etanercept 28% vs infliximab 20%, p = 0.445), adverse events (7% vs 22%, p = 0.002) or inactive disease (10% vs 16%, p = 0.068). Women (hazard ratio (HR) 2.8, 95% CI 1.3 to 5.8), patients with systemic JIA (HR 7.8, 95% CI 1.7 to 34.9) or those taking infliximab (HR 2.0, 95% CI 1.2 to 3.3) were at higher risk of treatment discontinuation. One-third of the patients were switched to the second anti-TNF therapy, which was discontinued less frequently than the first. At 12 months treatment survival of etanercept was 60%, infliximab 58% and adalimumab 66% as the second-line anti-TNF therapy. Conclusions: Although infliximab was discontinued more often than etanercept because of adverse events, during a 48-month follow-up the overall treatment survival of etanercept and infliximab as the first biological agent in JIA was comparable. A switch from one anti-TNF agent to another appears a reasonable therapeutic option.