AB0373 A COMPARATIVE EFFECTIVENESS STUDY OF ABATACEPT AND TNFI AGENTS IN PATIENTS WITH RHEUMATOID ARTHRITIS USING REAL-WORLD DATA

Abstract

Background Introduction of multiple biologic DMARDs, targeted synthetic DMARDs and biosimilars has led to a paradigm shift in the management of patients (pts) with RA. There are, however, limited head-to-head studies comparing these agents to guide evidence-based treatment decisions. Abatacept versus adaliMumab comParison in bioLogic-naivE RA subjects with background MTX (AMPLE) was the first head-to-head randomised controlled trial to evaluate the efficacy of abatacept (ABA) and adalimumab. However, there are limited data replicating AMPLE findings in real-world settings. Objectives To compare effectiveness of ABA versus TNF inhibitors (TNFis) in pts with RA by line of therapy (LOT) and by LOT stratified by anti-citrullinated protein antibody (ACPA) status using data from real-world clinical practice. Methods Data from two independent RA registries comprising pts treated by physicians in routine care were used for the analysis. An RA disease-specific registry provided data on pts treated with TNFis; the other was a product-specific RA registry and provided data on pts treated with ABA.1,2 The disease-specific registry was a US-based single-centre registry; the product-specific registry was a multi-centre, 12-country registry excluding the US. Frequency matching on age (in years [yrs]; ±4 yrs’ window), RA disease duration (in yrs; ±4 yrs) and baseline CDAI category (remission [≤2.8], low activity [≤10], medium activity [≤22], high activity [>22]) between TNFi and ABA pts was done for biologic-naive and -experienced pts. Descriptive statistics were used to summarise baseline demographics, disease activity measures and serostatus for both treatment groups by LOT. Disease activity in the follow-up period was measured at the 12-month mark (±6 months). Change in disease activity after 12 months was evaluated for TNFi and ABA pts by LOT, and statistical comparison using Kruskal–Wallis test was performed. The analysis was repeated for pts stratified by ACPA status. Results Data for 105 first-line and 91 second- or further-line (subsequent-line) matched pairs of TNFi and ABA pts were included in the analysis (68 first-line and 60 subsequent-line matched pairs of TNFi and ABA pts, respectively, in ACPA+ cohort; 31 first-line and 21 subsequent-line matched pairs of TNFi and ABA pts, respectively, in ACPA– cohort; Table 1). In the biologic-experienced pts, ABA (vs TNFi) had a higher reduction in CDAI score (10.2 vs 5.2, p=0.035; Table 2). In the biologic-experienced ACPA+ pts, ABA (vs TNFi) had higher reduction in disease activity (CDAI: 13.3 vs 6.2, p=0.023; Table 2; SDAI: 13.9 vs 7.0, p=0.046). No difference in disease activity was observed between the two groups among the ACPA– pts. Conclusion Real-world RA registry data further confirm findings from the AMPLE study that the overall efficacy of ABA is similar to TNFi agents in biologic-naive pts with RA. Efficacy of ABA (vs TNFis) in biologic-experienced pts is greater, and greater reductions in joint disease activity in ACPA+ ABA (vs TNFis) pts were observed.