Endocardial left atrial appendage occlusion in atrial fibrillation: computational fluid dynamics simulations to assess stroke risk

Abstract

Abstract Funding Acknowledgements Type of funding sources: None. Background Percutaneous endocardial left atrial appendage (LAA) occlusion (LAAO) in non-valvular atrial fibrillation (AF) seems comparable to anticoagulation therapy (OAC) as regards thromboembolic risk reduction with a possible additional decrease in major bleeding. LAAO is currently limited to patients with contraindications to OAC, due to its high costs and procedural risks, but better pre-procedural planning and operative techniques might improve the outcome widening practical indications. Computational fluid dynamics (CFD) represents a valuable non-invasive approach to estimate physiologically significant hemodynamic parameters in a complex fluid dynamics system. It might provide a helpful in silico simulation of blood flow patterns within the LA and LAA by using 3D patient-specific models, allowing LAAO planning and effects prediction. Purpose This study’s aim was to simulate the fluid dynamics effects of LAAO in AF patients to predict patient-specific hemodynamic changes caused by applying the two most popular devices. Methods LAAO was simulated on the 3D LA anatomical models obtained from CT data in 5 AF patients, considering the device specific shape. CFD simulations in AF condition were performed both on the whole LA model and on the models with the LAAO performed with the two devices. Significant fluid dynamics indices (blood velocity, vortex structures, LAA ostium velocity, LA blood stasis) were computed to evaluate the changes in the flow patterns after LAAO in relation to the thrombogenic risk. Results Overall we found a more effective washout within the LA after LAAO, in terms of a different spatial distribution of velocities (see figure for a qualitative evaluation of LA blood flow velocity in one patient: (A) model with LA and LAA; models after LAAO applying the Amulet (B) and the Watchman (C) device) and vortex structures (after LAAO, they were decreased in number and increased in size). Moreover, a higher velocity at the mitral valve and at the LAA ostium (peak velocity: 12-17 cm/s in the models with LAA, 40-60 cm/s in LAAO_A and 35-65 in LAAO_W) was detected together with  a slightly improved washout effect in terms of blood stasis with the Watchman device (stasis: 3.1-5.7% in the models with LAA, 1.9-4.1% in LAAO_A, 1.7-3.7% in LAAO_W). Conclusions A workflow for simulating the fluid dynamics effects of endocardial LAAO in AF was developed and tested. CFD provides a valuable tool to quantify hemodynamic changes after LAAO and assess thrombogenic risk in patient-specific LA and LAA. Our preliminary results suggest that endocardial LAAO favourably affects blood fluid dynamics in the LA. Abstract figure