Impact of Computational Modeling on Transcatheter Left Atrial Appendage Closure Efficiency and Outcomes

Abstract

BackgroundWhen performing transcatheter left atrial appendage (LAA) closure, peridevice leaks and device-related thrombus (DRT) have been associated with worse clinical outcomes—hence, their risk should be mitigated. ObjectivesThe authors sought to assess whether use of preprocedural computational modeling impacts procedural efficiency and outcomes of transcatheter LAA closure. MethodsThe PREDICT-LAA trial (NCT04180605) is a prospective, multicenter, randomized trial in which 200 patients were 1:1 randomized to standard planning vs cardiac computed tomography (CT) simulation–based planning of LAA closure with Amplatzer Amulet. The artificial intelligence–enabled CT-based anatomical analyses and computer simulations were provided by FEops (Belgium). ResultsAll patients had a preprocedural cardiac CT, 197 patients underwent LAA closure, and 181 of these patients had a postprocedural CT scan (standard, n = 91; CT + simulation, n = 90). The composite primary endpoint, defined as contrast leakage distal of the Amulet lobe and/or presence of DRT, was observed in 41.8% in the standard group vs 28.9% in the CT + simulation group (relative risk [RR]: 0.69; 95% CI: 0.46-1.04; P = 0.08). Complete LAA closure with no residual leak and no disc retraction into the LAA was observed in 44.0% vs 61.1%, respectively (RR: 1.44; 95% CI: 1.05-1.98; P = 0.03). In addition, use of computer simulations resulted in improved procedural efficiency with use of fewer Amulet devices (103 vs 118; P < 0.001) and fewer device repositionings (104 vs 195; P < 0.001) in the CT + simulation group. ConclusionsThe PREDICT-LAA trial demonstrates the possible added value of artificial intelligence–enabled, CT-based computational modeling when planning for transcatheter LAA closure, leading to improved procedural efficiency and a trend toward better procedural outcomes.