Absence of association between abatacept exposure and initial infection in patients with juvenile idiopathic arthritis

Objective. To assess the relationship between infection risk and abatacept (ABA) exposure levels in patients with polyarticular-course juvenile idiopathic arthritis (pJIA) following treatment with subcutaneous (SC) and intravenous (IV) ABA. Methods. Data from 2 published studies (ClinicalTrials.gov: NCT01844518, NCT00095173) of ABA treatment in pediatric patients were analyzed. One study treated patients aged 2.17 years with SC ABA and the other treated patients aged 6.17 years with IV ABA. Association between serum ABA exposure measures and infection was evaluated using Kaplan-Meier plots of probability of first infection vs time on treatment by ABA exposure quartiles and log-rank tests. Number of infections by ABA exposure quartiles was investigated. Results. Overall, 343 patients were included in this analysis: 219 patients received SC ABA and 124 patients received IV ABA. Overall, 237/343 (69.1%) patients had ≥ 1 infection over 24 months. No significant difference in time to first infection across 4 quartiles of ABA exposure levels was observed in the pooled (P = 0.45), SC (2.5 yrs: P = 0.93; 6.17 yrs: P = 0.48), or IV (P = 0.50) analyses. Concomitant use of methotrexate and glucocorticoids (at baseline and throughout) with ABA did not increase infection risk across the ABA exposure quartiles. There was no evidence of association between number of infections and ABA exposure quartiles. No opportunistic infections related to ABA were reported. Conclusion. In patients aged 2.17 years with pJIA, no evidence of association between higher levels of exposure to IV ABA or SC ABA and incidence of infection was observed.