P080 Target trial emulation to estimate the effectiveness of biologics using real-world data from the UK Juvenile Idiopathic Arthritis Biologics Register

Background/Aims There is uncertainty regarding the relative clinical effectiveness of biologic drugs for juvenile idiopathic arthritis (JIA) due to a lack of randomised trials comparing them head-to-head. Aim To estimate the relative effectiveness of etanercept and adalimumab for JIA using real-world data. Methods We applied the target trial emulation (TTE) method to data from the UK JIA Biologics Register on patients with non-systemic JIA, ≥2 years old, starting a first-line biologic following methotrexate from 2008 onwards. Patients with uveitis and/or zero active joints were excluded from the analysis. The closest outcome measurements to the time of biologic treatment start (baseline) and to 24 weeks were included. Clinical effectiveness was assessed using change in Juvenile Arthritis Disease Activity Score (JADAS-71) and odds of achieving an American College of Rheumatology Pediatric (ACR Pedi) 70% response. Amended ACR Pedi 70% response was also assessed to better reflect baseline disease activity in real-world patients. Inverse-probability-of-treatment weighting was used to mitigate confounding in the effect estimates due to non-random treatment allocation. Results 306 eligible patients were included; 215 starting etanercept, 91 starting adalimumab. At baseline, patients had similar demographic characteristics but those allocated etanercept had higher disease activity (Table 1). All patients beginning treatment before 2010 were allocated etanercept and most patients allocated adalimumab began treatment after 2015. It was unclear whether being allocated adalimumab vs. etanercept led to a greater decrease in disease activity, mean difference in change in JADAS -4.35 (95% confidence interval (CI): -11.46, 2.77). A similar proportion of patients achieved an ACR Pedi 70 response in both treatment groups, 56% (95% CI: 45%, 66%) for adalimumab and 55% (48%, 62%) for etanercept. However, after adjustment for non-random treatment allocation patients taking adalimumab vs. etanercept had greater odds of achieving a response, odds ratio (OR): 1.84 (1.15, 2.95). A similar result was seen for the amended ACR Pedi 70 response; OR: 1.76 (1.04, 3.01). Conclusion The TTE results suggest adalimumab may be more effective than etanercept in terms of treatment response in patients seen in UK clinical practice. Further research is required to clarify whether this depends on combination therapy with methotrexate. Disclosure J. Singh: None. L. Kearsley-Fleet: None. K.L. Hyrich: None. S. Bujkiewicz: None.