Peridevice Leak Following Amplatzer Left Atrial Appendage Occlusion: Cardiac Computed Tomography Classification and Clinical Outcomes

• Published in: JACC. Cardiovascular interventions 2021-01, Vol.14 (1), p.83-93
• Main Authors: Korsholm, Kasper, Jensen, Jesper Møller, Nørgaard, Bjarne Linde, Samaras, Athanasios, Saw, Jacqueline, Berti, Sergio, Tzikas, Apostolos, Nielsen-Kudsk, Jens Erik
• Format: Article
• Language: English
• ISSN: 1876-7605
• DOI: 10.1016/j.jcin.2020.10.034

This study aimed to investigate cardiac computed tomography (CT) and transesophageal echocardiography (TEE) peridevice leak (PDL) assessments, and the clinical relevance of PDL. PDL assessment is integral during follow-up after left atrial appendage (LAA) occlusion. Comparative studies of TEE and cardiac CT are sparse, and the clinical relevance of PDL is uncertain. This was a single-center observational study of consecutive patients undergoing LAA occlusion with Amplatzer devices (Amplatzer Cardiac Plug/Amulet) between 2010 and 2018 (N = 415). Patients with both 8-week CT and TEE were included for analysis (n = 346). Images were analyzed by blinded investigators (K.K. and A.S.). PDL on cardiac CT was classified from grade 1 to 3, based on PDL at the device disc, device lobe, and LAA contrast patency. Primary clinical outcome was a composite of ischemic stroke, transient ischemic attack, systemic embolism, or all-cause death. PDL was present in 110 patients (32%) by TEE, with 29 (8%) >3 mm. By cardiac CT, 210 patients (61%) had PDL at the disc, with contrast patency in 204 patients (59%). A grade 3 PDL (gap at disc, lobe, and LAA contrast patency) was present in 63 patients (18%). Bland-Altman analysis showed poor agreement between CT and TEE for leak sizing. CT and TEE detected PDL was not significantly associated with worse outcome, hazard ratio: 1.82 (95 % confidence interval: 0.95 to 3.50); p = 0.07 and hazard ratio: 1.43 (95% confidence interval: 0.74 to 2.76); p = 0.28, respectively. PDL occurrence is substantially higher with CT compared with TEE, with a large discrepancy between modalities in leak quantification. A novel CT-based classification is proposed, yet PDL was not associated with worse clinical outcome.