on behalf of the PRINTO and PRCSG working groups. Background: Infections are the most common expected AEs linked to biologic (b) DMARDs in paediatric patients (pts) with juvenile idiopathic arthritis (JIA). Blood concentrations achieved with bDMARDs vary greatly between individual pts. It is not known if higher abatacept (ABA) exposure is linked to higher infection risk in paediatric populations. Objectives: To assess the relationship between the incidence of infection and SC (50-125 mg weekly)

more »

... and IV (10 mg/kg monthly) ABA exposure in pts with polyarticular-course JIA (pJIA). Methods: Data from the 4 month open-label periods of a Phase III SC ABA study (NCT01844518; weight-tiered ABA: 10-<25 kg [50 mg], 25-<50 kg [87.5 mg],!50 kg [125 mg]; 219 pts aged 2-17 years) and an IV ABA study (NCT00095173; ABA 10 mg/kg monthly; 184 pts aged 6-17 years) in pts with pJIA were analysed. The association between serum ABA exposure measures (steady-state trough [C minss ], maximum [C maxss ] and time-averaged [C avgss ] concentrations) estimated by population pharmacokinetic analysis and time to first infection (regardless of seriousness) was assessed. Kaplan-Meier (KM) plots of infection probability versus time to first infection by ABA exposure quartiles were created and log-rank test was performed to test the differences in distribution of time to first infection across exposure quartiles. Box plots of ABA exposure measures over time to Month 4 were generated, stratified by first infection occurrence (yes/no). Data for SC and IV ABA were assessed separately and pooled. Results: Baseline demographic and clinical characteristics were comparable in the SC and IV studies. 1 2 Overall, 135/403 pts (33.5%) had !1 infection over 4 months: 77/219 (35.2%) with SC ABA and 58/184 (31.5%) with IV ABA. KM plots for pooled SC and IV ABA showed no statistically significant difference in infection probability across four quartiles of ABA C minss (Fig A; p=0.2317; log-rank test), C maxss (p=0.5501) or C avgss (p=0.3808). Consistent results were seen for individually studied SC and IV ABA C minss (Fig B, C) , C maxss and C avgss (not shown). In addition, there was no difference in median ABA exposure measures by infection occurrence (yes/no) in pooled and separate SC and IV analyses. Abstract THU0549 - Figure 1 . Kaplan-Meier