Residual peri-device leaks of any size after left atrial appendage occlusion (LAAO) were associated with modestly higher risk of thromboembolic and bleeding events, according to new registry data. Large leaks (>5 mm) were not associated with adverse events, but higher percentages of these patients had to keep taking anticoagulants, Mohamad Alkhouli, MD, of Mayo Clinic, added during his presentation Sunday at the American College of Cardiology’s 71st Annual Scientific Session in Washington, D.C. These findings were simultaneously published online in JACC: Clinical Electrophysiology. Asking whether residual leaks after LAAO matter, Alkhouli pointed out that both PROTECT-AF trial and the ACP Registry showed no association of leaks with thrombotic events. However, these trials were underpowered to detect an association because they each had a low number of events, he said. He added that the Vanderbilt Registry’s combined endpoint of failure to stop oral anticoagulants, transient ischemic attack (TIA), stroke, device-related thrombus or need for leak closure was higher in patients with leaks >3 mm (69% vs. 34%; p=0.002). Finally, he pointed out a study presented at the American Heart Association Scientific Sessions 2021, which showed no association at 45 days, but an association at 1 year, between residual leaks and higher rates of ischemic stroke and systemic embolism. Alkhouli and colleagues queried the National Cardiovascular Data Registry LAAO Registry to identify patients undergoing LAAO from 2016 to 2019. Patients were classified according to leak size on echocardiography at 45 ± 14 days – no leak (0 mm), small leak (>0-5 mm) or large leak (>5 mm). The analysis included 51,333 patients. Of these, 37,696 (73.4%) had no leak, 13,258 (25.8%) had small leaks, and 379 (0.7%) had large leaks. Patients in the different leak groups had “modest but statistically significant differences” in their baseline profiles, including percentages of non-paroxysmal atrial fibrillation (no leak 43.3%, small leak 48.2%, large leak 53.8%; p<0.001) and cardiomyopathy (no leak 19.8%, small leak 22.1%, large leak 24.0%; p<0.001). The groups also had different LAA orifice diameters (no leak 21.1 ± 4.2 mm, small leak 22.3 ± 4.3 mm, large leak 23.7 ± 4.4 mm; p<0.001). The large-leak cohort had a higher proportion of patients on warfarin at 45 days (44.9% than the small- (34.4%) and no-leak (32.4%, p<0.001). At 6 and 12 months, anticoagulants were used less, but they were still more frequently used in patients with large leaks. Thromboembolic events were uncommon in all three groups. However, patients with small leaks had a slightly higher odds of stroke/transient ischemic attack/systemic embolization than those with no leak (adjusted hazard ratio [aHR], 1.152; 95% confidence interval [CI], 1.025-1.294). The same was true for major bleeding (HR, 1.11; 95% CI, 1.029-1.120) and any major adverse events (HR, 1.102; 95% CI, 1.048-1.160). There were no significant differences in adverse events between patients with large leaks and patients with small or no leaks. Study limitations included that this was an observational registry, that imaging for leaks after 45 days is uncommon in practice, there was variation in peri-device leak size measurement, only the first-generation Watchman LAAO device (Boston Scientific) was included, data on interventional leak management was not available, and follow-up was limited to 1 year. “Our study shows that any leak matters, and we should find ways to optimize the procedure to minimize the proportion of patients who end up with leaks,” Alkhouli said in a news release announcing the results. “Because this is a preventative procedure, it is important to find ways to master this and prevent leaks from happening. Fortunately, there are indications that leaking might be less common with newer devices and improved techniques.” The study was funded by a grant from Boston Scientific.