IntroductionObservational studies have shown low bleeding rates in patients with atrial fibrillation (AF) treated by left atrial appendage closure (LAAC); however, data from randomized studies are lacking. This study compared bleeding events among patients with AF treated by LAAC and nonvitamin K anticoagulants (NOAC). MethodsThe Prague-17 trial was a prospective, multicenter, randomized trial that compared LAAC to NOAC in high-risk AF patients. The primary endpoint was a composite of a cardioembolic event, cardiovascular death, and major and clinically relevant nonmajor bleeding (CRNMB) defined according to the International Society on Thrombosis and Hemostasis (ISTH). ResultsThe trial enrolled 402 patients (201 per arm), and the median follow-up was 3.5 (IQR 2.6-4.2) years. Bleeding occurred in 24 patients (29 events) and 32 patients (40 events) in the LAAC and NOAC groups, respectively. Six of the LAAC bleeding events were procedure/device-related. In the primary intention-to-treat analysis, LAAC was associated with similar rates of ISTH major or CRNMB (sHR 0.75, 95% CI 0.44-1.27, p = 0.28), but with a reduction in nonprocedural major or CRNMB (sHR 0.55, 95% CI 0.31-0.97, p = 0.039). This reduction for nonprocedural bleeding with LAAC was mainly driven by a reduced rate of CRNMB (sHR for major bleeding 0.69, 95% CI 0.34-1.39, p = .30; sHR for CRNMB 0.43, 95% CI 0.18-1.03, p = 0.059). History of bleeding was a predictor of bleeding during follow-up. Gastrointestinal bleeding was the most common bleeding site in both groups. ConclusionDuring the 4-year follow-up, LAAC was associated with less nonprocedural bleeding. The reduction is mainly driven by a decrease in CRNMB.