Analysis of fibrocalcific aortic valve stenosis: computational pre-and-post TAVR haemodynamics behaviours

Abstract

Fibro-calcific aortic valve (AV) diseases are characterized by calcium growth or accumulation of fibrosis in the AV tissues. Fibrocalcific aortic stenosis (FAS) rises specifically in females, like calcification-induced aortic stenosis (CAS), may eventually necessitate valve replacement. Fluid-structure-interaction (FSI) computational models for severe CAS and FAS patients were developed using lattice Boltzmann method and multi-scale finite elements (FE). Three parametric AV models were introduced: pathology-free of non-calcified tri-and-bicuspid AVs with healthy collagen fibre network (CFN), a FAS model incorporated a thickened CFN with embedded small calcification volumes, and a CAS model employs healthy CFN with embedded high calcification volumes. The results indicate that the interaction between calcium deposits, adjacent tissue and fibres crucially influences haemodynamics and structural reactions. A fourth model of transcatheter aortic valve replacement (TAVR) post-procedure outcomes was created to study both CAS and FAS. TAVR-CAS had a higher maximum contact pressure and lower anchoring area than TAVR-FAS, making it prone to aortic tissue damage and migration. Finally, although the TAVR-CAS offered a larger opening area, its paravalvular leakage was higher. This may be attributed to a similar thrombogenicity potential characterizing both models. The computational framework emphasizes the significance of mechanobiology in FAS and underscores the requirement for tissue modelling at multiple scales.

Keywords: calcific aortic valve; fibrosis; finite element; fluid–structure interaction; lattice Boltzmann method.

Figure 1.

Sex-specific clinical data comparison between males (in green) and females (in pink): (a) Calcium volume median and deviation: 447 ± 353 mm³ versus 1043 ± 500 mm³. (b) Stenotic aortic valve area median and deviation: 0.75 ± 0.15 cm² versus 0.85 ± 0.25 cm². (c) Calcium volume effect on aortic valve area normalized by body mass index.

Figure 2.

(a) Computational model of the parametric healthy tricuspid aortic valve with the collagen fibre network distribution. (b) Highly calcified and fibrocalcific aortic stenosis diseases computational models, including calcium deposits and fibres' architecture, reconstructed from their patients’ specific CT scans.

Figure 3.

(a) Calibrated aortic valve area of the different fibres' architecture types (I in yellow and II in green). (b) Aortic valve area variation due to fibres’ volume for the different architecture types. (c) Fluid–structure interaction boundary condition of time-dependent pressure gradient for healthy, highly calcified and fibrocalcific models.

Figure 4.

Fluid–structure interaction models using coupled LBM-FE: maximum velocity jet at systole peak and effective opening area for healthy tricuspid, non-calcified bicuspid, highly calcified and fibrocalcific aortic valves models.

Figure 5.

Fibres’, tissue and wall shear stress distributions at systole peak and mid-diastolic phase for (a) tricuspid, non-calcified bicuspid, (b) highly calcified and fibrocalcific models.

Figure 6.

Pre-TAVR computational models and their post-TAVR aortic valve area, anchoring contact area at each leaflet and maximum deployed contact pressure of highly calcified and fibrocalcific structural models.

Figure 7.

Velocity magnitude contours on different rotational views showing paravalvular leakage jets of highly calcified and fibrocalcific models at diastole peak time point. Paravalvular leakage quantification on sections AA’ and BB’ under the valve's annulus.

Figure 8.

Stress accumulation probability density function of TAVR implantation in logarithmic scale for highly calcified and fibrocalcific models in bulk flow (main plot) and tail region (inset). The vertical red line is Hellum's criterion threshold for particle activation.