Quantifying Coronary Hemodynamics in Mice: Toward a Multiscale Computational Model of Cardiac Allograft Vasculopathy

Cardiac allograft vasculopathy (CAV) is the primary cause of chronic rejection in heart transplant (HTx) patients, affecting 50% of recipients within 10 years. It manifests as a progressive luminal narrowing of the graft coronary arteries, ultimately leading to heart failure. Currently, there are no effective treatments for CAV, and early diagnosis remains challenging, due to a lack of understanding of the mechanisms underlying the pathology’s development. To date, animal (in vivo) studies have been instrumental in deepening our comprehension of CAV risk factors. However, they are associated with ethical concerns, high resource demand, and can offer limited investigation time points. Computational (in silico) models can overcome these drawbacks, by providing a digital platform for investigating the multiscale processes occurring post-HTx. Additionally, they grant a high level of detail in studying the role of coronary morphology and hemodynamics, which are hardly quantifiable in vivo, yet widely recognized as one of the crucial factors in CAV initiation and progression. Therefore, in silico models can integrate and complement experimental research by testing relevant hypotheses to better elucidate CAV pathophysiology. In this context, a 2D agent-based model (ABM) has been recently developed to replicate CAV dynamics at the cellular level, under inflammatory and hemodynamic stimuli, in an idealized cross-section of a mouse coronary artery. We hypothesize that including in the ABM of CAV the biomechanical stimulus exerted by blood flow on the endothelium in terms of wall shear stress (WSS) derived from computational fluid dynamics (CFD) simulations performed in mouse coronary arteries will be instrumental in providing a valuable tool to better direct the preclinical research to improve the diagnosis and treatment of CAV. Being part of a large multi-node project, this thesis work focuses on CFD infrastructure development to replicate coronary hemodynamics in mice. Coronary arteries of n = 6 mice were reconstructed from heart micro-CT images using segmentation techniques, after being ex vivo perfused with a radiopaque polymer. The 3D geometry of the left coronary artery (LCA) was isolated and subjected to a series of pre-processing operations to make it suitable for CFD. Unsteady hemodynamic simulations were performed on each LCA using mouse-specific blood properties and boundary conditions. WSS- and helicity-based hemodynamic descriptors were extracted to characterize mouse coronary flow patterns under baseline conditions. The relationship between coronary artery morphology and hemodynamics in mice was investigated using both hemodynamic indices (e.g., time-averaged WSS, oscillatory shear index, transversal WSS) and geometric features (e.g., curvature, torsion, tortuosity), obtained through a centerline-based morphometric analysis. Due to their smaller size compared to humans, mice exhibit higher WSS values and lack significant helical flow. Moreover, the results indicate that WSS is not highly oscillatory, as observed in human coronary arteries. This work paves the way for future quasi-3D computational simulations of CAV, where the ABM describes molecular/cellular processes on the scale of hours/days and CFD replicates hemodynamics over seconds, offering a comprehensive description of CAV development key players. Integrating such a multiscale in silico model of CAV with preclinical experiments will aid in unveiling new CAV biomarkers to optimize anti-chronic rejection research.