(P183) DIRECT ORAL ANTICOAGULATION VERSUS WARFARIN IN PATIENTS WITH ATRIAL FIBRILLATION AND BIOPROSTHETIC HEART VALVES: A RETROSPECTIVE, REAL WORLD COHORT STUDY

Background: Data are limited on the efficacy and safety of direct oral anticoagulants (DOACs) compared to warfarin in atrial fibrillation (AF) patients with bioprosthetic heart valves, particularly in the real world. In this retrospective cohort study of consecutive patients with AF and surgical or transcatheter implantation of a bioprosthetic valve, we compared the efficacy and safety of DOACs and warfarin. The primary efficacy outcome was the composite of death and thrombosis (stroke, transient ischemic attack, systemic embolism, and myocardial infarction). The principal safety outcome was the composite of death and bleeding requiring hospitalization.

METHODS AND RESULTS: Using linked health administrative databases housed at the Institute for Clinical Evaluative Sciences, consecutive patients in Ontario (Canada), 65 years of age or older with AF who underwent bioprosthetic valve replacement between April 1, 2012, and March 31, 2017, were identified. A time-varying Cox model was used to examine the relationship between the type of anticoagulant and time to thrombosis or bleeding events after adjustment for baseline thrombosis and bleeding risks using the CHA2DS2-VASc and HASBLED scores, respectively, with prespecified subgroup analyses according to whether valve implantation was surgical or transcatheter. We identified 2245 eligible patients. Median age was 79 years, 41% were female, and 39% had transcatheter aortic valve replacement. Median follow-up was 795 (IQR 470-1196) days. Most patients (83.1%) had preoperative AF, and 63% were taking anticoagulation preoperatively. Mean CHA2DS2-VASc and HASBLED scores were 4.0 (SD 1.3) and 2.5 (SD 0.9), respectively. There was no difference in risk of death or thrombosis between direct oral anticoagulants and warfarin after adjustment for CHA2DS2-VASc score (hazard ratio [HR] 1.02, 95% confidence interval [CI], 0.83-1.25). There was no difference in risk of death or bleeding between direct oral anticoagulants and warfarin after adjustment for HAS-BLED score (HR 0.89, 95% CI 0.75-1.07). Subgroup analyses of surgical or transcatheter valves were consistent with overall results.

Conclusion: In a real-world population of patients with AF and bioprosthetic valve replacement, DOACs may be associated with an increased risk of thrombosis and bleeding after adjustment for the baseline risk of thrombosis and bleeding. Comparison of adjusted and unadjusted analyses suggests that baseline risk of bleeding and thrombosis are key determinants of outcomes. These findings suggest that patient selection is critical when choosing between an anticoagulant in such patients, and highlight the need for a randomized trial to determine whether DOACs are effective and safe for AF patients with bioprosthetic valves.