Background: Peri-device leaks and device thrombi are assumed to lead to ischemic events after percutaneous left atrial appendage occlusion (LAAO). While these device related complications are well assessed until 12 months after LAAO, no data are available beyond this period. Consequently, the occurrence and impact on clinical outcome of such occluder related complications in the long-term phase after LAAO, where antithrombotic regime mainly consists of single antiplatelet therapy or even no antithrombotic medication, is completely unknown.

Objective: To assess peri-device leaks and device thrombi beyond 12 months after LAAO.

Methods: All patients who underwent successful percutaneous LAAO ≥ 12 months ago in the two participating centers were enrolled. After inclusion a structured transesophageal echocardiography (TEE) was performed to assess peri-device leaks, device thrombus and other device related complications.

Results: A total of 63 patients were enrolled into the present prospective two-center non-randomized single-arm trial. Median time from implantation until long-term TEE was 3.1 years (2.0 – 5.1) ranging from 1.0 to 7.5 years. The number of patients without detectable leak increased significantly at long-term TEE compared to the initial follow-up 6 weeks weeks following LAAO (p=0.04) (Figure 1A and B). Major peri-device leaks (> 5mm) were detected in 2 patients (3%). Peri-device leak size was 3mm in these 2 patients at the short-term 6-week TEE follow-up. Device thrombus was found in 8 patients (annual rate 4.2%, 4.1 per 100 patient-years).  Six of these patients had a device thrombus for the first time. The other 2 patients had recurrence of device thrombus after a first diagnosis at the 6-week TEE. Patients with device thrombus at long-term follow-up had larger peri-device leaks at 6-week follow-up compared to those without thrombus formation (3.0mm [2.7 – 3.5] vs. 1.4mm [0.0 – 2.6]; p=0.04).  During long-term follow-up stroke occurred in 5 patients, resulting in an annual stroke rate of 2.5%. Of those who suffered an ischemic stroke during long-term follow-up one patient showed a peri-device leak >5mm. None of the patients with detected device thrombus developed a stroke.

Conclusions: Peri-device leaks and device thrombi continue to occur beyond 12 months following LAAO. The clinical impact of these late occluder complications requires further evaluation in larger prospective trials. Furthermore, larger prospective trials should assess whether leaks with a certain size cut-off (3mm in the present analysis) at the short-term TEE control after LAAO may be a risk factor for late device complications and, therefore, should used as indicator to perform long-term TEE follow-up beyond 12 months.