Computational fluid dynamics simulations of mitral paravalvular leaks in the human heart

Abstract Funding Acknowledgements Type of funding sources: Public Institution(s). Main funding source(s): Warsaw University of Technology - BIOTECHMED-3 project under the program Excellence Initiative: Research University. Background Presence of paravalvular leaks (PVL) can lead to hemolysis and high shear stress that affects erythrocytes passing through PVL channels may be the main reason behind this pathology. In vitro, shear stress forces above 300 Pa cause erythrocyte destruction. Distribution of shear stress can currently be precisely determined with computational fluid dynamics (CFD). CFD is a numerical modelling method based on solving equations of differential balance of mass, momentum and energy for a small section of fluid known as a computational cell. Purpose To investigate the magnitude of shear stress forces that affect erythrocytes in PVL channels and define properties of PVL channels that increase the likelihood of hemolysis. Methods Models of left-sided heart chambers with artificial mitral valves and PVLs were created based on cardiac computed tomography data from subjects with PVLs. Computational fluid dynamics principles were applied to create patterns of blood flow through PVL channels. 4 channels with cross-sectional areas (CSA) of 26.2 mm2 (model I), 15.7 mm2 (model 2), 20.4 mm2 (model 3) and 32 mm2 (model 4) were examined. Length of PVL channels was 6.5, 9.5, 7.3 and 3.25 mm respectively. In models I,III and IV the PVL channel was located anteriorly around the mitral annulus, whereas in model II a posterior location was described. Blood flow was simulated using a commercial CFD software (ANSYS Fluent 2020R1). Simulations were performed for heart rate of 60 beats per minute. The following variables were examined regarding blood flow through PVL channels: wall shear stress, shear stress in fluid, volume of PVL channels in which shear stress exceeded 300 Pa and duration of exposure of red blood cells to shear stress above 300 Pa. Obtained values were compared with lactate dehydrogenase (LDH) activities in subjects whose cardiac CTs were used to create PVL models. One of the models of the left ventricle is presented in picture 1. Studied PVL channels are presented in picture 2. Results In all models shear stress value of 300 Pa was exceeded. Maximal wall shear stress and maximal shear stress in fluid values were not related to CSA of the PVL channel. Lowest values of all examined variables were obtained for model 2, which had the longest PVL channel and a posterior location. The lowest LDH activity was noted in model with the lowest volume of the PVL channel with shear stress above 300 Pa (model 2) and the highest LDH activity in model with both the highest volume of the PVL channel with shear stress above 300 Pa and longest exposure to shear stress exceeding 300 Pa (model III). Conclusions Presence of PVLs is always accompanied by some degree of hemolysis. Shear stress values in PVL channels are not related to CSA of these channels. Volume of PVL channels in which shear stress exceeds 300 Pa and duration of exposure of erythrocytes to shear stress above 300 Pa may influence the severity of hemolysis.