SAT0386 Adjudication of infections from the pharmacovigilance in juvenile idiopathic arthritis patients (PHARMACHILD) treated with biologic agents and/or methotrexate: update on results with a focus on opportunistic infections

Background Pharmachild is a pharmacovigilance registry on children with JIA treated mainly with biologics±methotrexate (MTX). Little evidence exists in literature about the role of JIA or its immunosuppressive therapy in determining infections, especially caused by opportunistic pathogens. Objectives To provide an update on opportunistic infections (OI) revised by an independent Safety Adjudication Committee (SAC) (3 paediatric rheumatologists and 2 paediatric infectious disease specialists). Methods The participating centres were asked to report all infections encountered by their JIA patients. PRINTO and the medical monitor (MM) classified events based on MedDRA dictionary. Moderate/serious/severe/very severe infections were then revised blindly by the SAC, who was asked to answer 6 questions. The events with consensus of at least 3/5 experts on the first 3 questions (‘Is this an infection?’, ‘Is it common?’, ’Is it opportunistic?’) were retained for the analysis. With referral to the recommendations by Withrop et al.1, for the first time a list of OIs in children with JIA on immunosuppressive therapy was elaborated and approved by consensus, through three Delphi steps. Finally, we compared the OI list defined by the SAC to the list of OI approved by the specialists. Results A total of 772 safety events related to 634 patients were submitted to the SAC. 689 (89.2%) events received consensus among the experts on the 3 questions and, of these, 682 (99.0%) were considered as infections, corresponding to 153 different Preferred Terms (PT), according to MedDRA dictionary. Among the 682 infections, 603 (88.4%) were defined by the experts as common and 119 (17.4%) as opportunistic. For 92 (60%) of the 153 PT, the MM and SAC used the same PT, while the remaining 40% was adjudicated by a third examiner, who analysed again the case reports and assigned the PT which was the most appropriate taking into account the experts’ opinion. Among the final number of infections emerged that herpes viral, respiratory and EBV infections were the most frequent (Table). Analysing the infections by PT (n=149), the experts adjudicated: 22 as OI, 119 as not OI, 8 discordant. Comparing the experts’ adjudication with the approved list of OI by PT, there was full agreement for the 22 PT classified as OI, while 19/117 (16.2%) PT resulted in the list, but were not classified as OI by the experts. Conclusions Our analysis showed a significant number of OI in JIA patients on immunosuppressants and provided an approved document stating the most frequent OI in children with JIA. The most frequent opportunistic pathogens resulted herpes virus, excluding varicella, and mycobacteria, but the list of definite and probable OIs needs to be validated/revised with the analysis of future datasets. Reference [1] Winthrop, et al. Ann Rheum Dis2015;74:2107–16. Disclosure of Interest None declared