Background Serious infections are a major concern in JIA patients treated with immunosuppressants and biologics. An increased risk has been described also for JIA patients without such treatments. An additional effect of TNF inhibitors on the risk for serious infections and factors of influence are studied here.
Methods Pts exposed to Etanercept (ETA), Adalimumab (ADA) and Methotrexate (MTX) but no biologics and serious infections were identified in the BIKER registry. Descriptive statistics and infection rates per 1000 patient years with confidence intervals (CI) were calculated. Cox-regression was used to identify risk factors and calculate Hazard ratios (HR) for occurrence of infections. Potential analysed risk factors were therapy (MTX, ETA or ADA), JIA category, gender, age at register entry, disease duration, ANA, HLA B27, JADAS10 at therapy start, concomitant or pre-medication with corticosteroids and daily dosage, pre-medication with DMARDS or MTX, mean JADAS10 level during the observation period. Observation time for survival analyses was time to 1. infection and censored at a maximum of 5 years.
Results A total of 3350 pts. with 5929 exposure years were identified in the German BIKER registry data base. First biologic was ETA in 1720 and ADA in 177 cases. 1353 patients were not exposed to biologics. 28 serious infections have been reported (4.7/1000 pt-years). MTX patients had 5 events, Etanercept patients 21 and patients with Adalimumab therapy 2 events.The serious infections were of bacterial origin in 16, viral in 10 and unknown in 2. Total infection incidence per 1000 person years was 4.72 (CI 3.26 – 6.84). The highest rate was found under ADA (9.73, CI 2.43-38.91) followed by ETA (8.08, CI 5.27 – 12.40) and the lowest rate of 1.6 (CI 0.67 – 3.84) with MTX.
Univariate Cox regression revealed a number of significant risks for infection, beside therapy (p=0.004) also JADAS10 at start of therapy, mean JADAS level during therapy, corticosteroids as pre- or concomitant medication as well as MTX and DMARDS as pre-medication were relevant. In multivariate Cox regression only therapy and mean JADAS10 during observation time remained significant. Both ETA (HR=4.88) and ADA (HR=10.06) showed an increased risk compared to MTX, whereas ADA and ETA differed not significally. Risk for infection was significantly increased by an elevated mean JADAS10 level (HR=1.12). Gender, JIA category, age at start of disease, disease duration, ANA status were not significant. Kaplan Meier analysis confirmed the influence of disease activity as demonstrated by the JADAS10 (figure).
Conclusions Risk factors for infections were identified. Treatment with ETA and ADA as well as higher disease activity (JADAS) contribute to the risk for serious infections. The overall risk for serious infections is low.
Disclosure of Interest G. Horneff Grant/research support: Abbvie, Pfizer, Roche/Chugai, I. Becker: None declared