Abstract
BACKGROUND: Optimising periprocedural management of direct oral anticoagulation in patients with atrial fibrillation on chronic treatment undergoing major surgeries is an important aspect of balancing the risk of surgery-related bleeding with the risk of thromboembolic events, which may vary by surgery type. METHODS: This subanalysis of the prospective EMIT-AF/VTE programme assessed periprocedural-edoxaban management, according to physicians' decisions, and bleeding and thromboembolic event rates in patients who underwent major vs. nonmajor surgeries. Edoxaban interruption and clinical outcomes were compared between major vs. nonmajor surgeries and between renal function subgroups (creatinine clearance [CrCL] ≤ 50 mL/min vs. > 50 mL/min). RESULTS: We included 276 major and 512 nonmajor surgeries. The median pre- and postprocedural duration of edoxaban interruption in major vs. nonmajor surgeries was 4 vs. 1 days, whereas median duration of interruption for those with preprocedural-only and postprocedural-only interruption was 2 vs. 1 days and 2 vs. 0 days, respectively (P < 0.0001). Rates of all bleeding and clinically relevant nonmajor bleeding were numerically higher in major vs. nonmajor surgeries. Event rates (number of events per 100 surgeries) were low overall (< 6 events per 100 surgeries), independent of renal function subgroups. CONCLUSION: In this subanalysis of the EMIT-AF/VTE programme, periprocedural-edoxaban interruption was significantly longer in patients undergoing major vs. nonmajor surgery. This clinician-driven approach was associated with low rates of bleeding and thromboembolic events following both major and nonmajor surgeries. TRIAL REGISTRATION: NCT02950168, registered October 31, 2016; NCT02951039, registered November 1, 2016.