Comparative Safety and Efficacy of Apixaban and Rivaroxaban During Extended Anticoagulation for Venous Thromboembolism: A Non-randomised Post Hoc Comparison Within the RENOVE Trial

Background and aims Some evidence reported that rivaroxaban use is associated with a higher risk of clinically relevant bleeding (CRB) than apixaban during the first months of treatment of venous thromboembolism (VTE). We aimed to determine whether the risk of CRB is higher with rivaroxaban than apixaban during extended treatment, either at full or reduced dose. Methods The RENOVE open-label randomised controlled trial compared full-dose versus reduced-dose of apixaban or rivaroxaban in patients who had completed at least 6 months of full-dose treatment for VTE and required extended anticoagulation. Drug choice was not randomised. In this post-analysis, the primary outcome was centrally-adjudicated CRB events. Treatment effects were estimated using crude hazard ratio (HR), adjusted HR, and an inverse probability of treatment weighting (IPTW). Results Among 2768 enrolled patients, 1513 received rivaroxaban (755 full-dose, 758 reduced-dose) and 1255 apixaban (630 full-dose, 625 reduced-dose). Median age was 62.6 versus 66.5 years and proportion of female was 32.1% versus 38.7%, respectively. Median follow-up was 37 months. The 5-year cumulative incidence of CRB was 11.7% with rivaroxaban and 13.7% with apixaban (adjusted-HR 0.90 [95%CI 0.70–1.17], HR IPTW 0.92 [0.70–1.21]). The incidence of major bleeding, recurrent VTE, composite outcome, arterial thromboembolic events, and all-cause death were similar. Serious adverse event not related to VTE and CRB was 7.7% and 10.8%, respectively (adjusted-HR 0.74 [95% CI 0.57–0.97], HR IPTW 0.76 [0.55–1.04]). Conclusions During extended anticoagulation, no difference in clinically relevant bleeding or recurrent VTE was observed between rivaroxaban and apixaban, although that non-randomised treatment allocation possibly influenced outcomes.