POS0519 RELATIONSHIP BETWEEN DISEASE ACTIVITY AND MAJOR ADVERSE EVENTS IN PATIENTS WITH RHEUMATOID ARTHRITIS ON TOFACITINIB OR TNF INHIBITORS: A POST HOC ANALYSIS OF ORAL SURVEILLANCE

Abstract

Background: Uncontrolled rheumatoid arthritis (RA) activity and acute disease fares are associated with higher risk of adverse outcomes such as cardiovascular (CV) disease, venous thromboembolism (VTE), malignancy and infection.1‐4 Objectives: To evaluate associations of acute and cumulative Clinical Disease Activity Index (CDAI) measurements with major CV, malignancy, or infectious adverse events (AEs) of special interest in ORAL Surveillance. Methods: ORAL Surveillance (NCT02092467) was a post‐authorisation safety study of tofacitinib vs TNF inhibitors (TNFi) in patients (pts) aged ≥50 yrs with active RA despite methotrexate (MTX), and ≥1 additional CV risk factor. Pts were randomised 1:1:1 to tofacitinib 5 or 10 mg twice daily (BID) or subcutaneous TNFI. Two post hoc analyses were performed: (1) a time‐varying multivariate Cox model examined risks of major AEs when pts were in CDAI‐defned low (>2.8‐≤10; LDA), moderate (>10‐≤22; MDA) or high (>22; HDA) disease activity vs remission (≤2.8). The Cox model also included pt demographics, medical history, RA characteristics, prior treatments, baseline (BL) medications and treatment arm, pre‐selected using backward selection; (2) area under the curve (AUC) per yr for CDAI prior to event or to study end (pts without event) was calculated and compared using an analysis of variance model with treatment arm, event status and interaction (supportive). Nominal p values <0.10 were considered evidence of associations. Results: 4362 pts were included. Mean RA duration at BL was approximately 10 yrs. All pts were on MTX at BL, and 28% had previously been on one other synthetic disease‐modifying antirheumatic drug (DMARD). Overall, 10% of pts had been on one biologic DMARD. Hazard ratios suggested that when pts had LDA, MDA or HDA vs remission, they were potentially at higher risk of developing major adverse CV events (MACE), VTE and non‐serious infections (NSIs) excluding herpes zoster (HZ), but not malignancies, serious infections or HZ (Figure 1). Similarly, mean CDAI AUC trended higher for MACE, VTE and NSIs (Table 1). Conclusion: In ORAL Surveillance, the risk of MACE, VTE and NSIs excluding HZ appeared higher when pts had active disease than when in remission. Greater cumulative RA disease activity was seen in pts who suffered these AEs vs those who did not. Our fndings support treat‐to‐target recommendations for RA.