Endocardial (Watchman) vs epicardial (Lariat) left atrial appendage exclusion devices: Understanding the differences in the location and type of leaks and their clinical implications.

Background: Watchman and Lariat left atrial appendage (LAA) occlusion devices are associated with LAA leaks postdeployment.

Objective: The purpose of this study was to compare the incidence, characteristics, and clinical significance of these leaks.

Methods: We performed a multicenter prospective observational study of all patients who underwent LAA closure. Baseline, procedural, and imaging variables along with LAA occlusion rates at 30-90 days and 1-year postprocedure were compared.

Results: A total of 478 patients (219 with the Watchman device and 259 with the Lariat device) with successful implants were included. Patients in the Lariat group had a higher CHADS2 (congestive heart failure, hypertension, age >74 years, diabetes, stroke) score and a larger left atrium and LAA. A total of 79 patients (17%) had a detectable leak at 1 year. More patients in the Watchman group had a leak compared with those in the Lariat group (46 [21%] vs 33 [14%]; P = .019). All the leaks were eccentric (edge effect) in the Watchman group and concentric (gunny sack effect) in the Lariat group. The size of the leak was larger in the Watchman group than in the Lariat group (3.10 ± 1.5 mm vs 2.15 ± 1.3 mm; P = .001). The Watchman group had 1 device embolization requiring surgery and 2 pericardial effusions requiring pericardiocentesis. In the Lariat group, 4 patients had cardiac tamponade requiring urgent surgical repair. Three patients in each group had a cerebrovascular accident and were not associated with device leaks.

Conclusion: The Lariat device is associated with a lower rate of leaks at 1 year as compared with the Watchman device, with no difference in rates of cerebrovascular accident. There was no correlation between the presence of residual leak and the occurrence of cerebrovascular accident.