Cardiac allograft vasculopathy (CAV) is one of the leading causes of morbidity and morality in orthotopic heart transplant (HTx) patients. While disturbed flow patterns have been linked to the spatial localization of atherosclerosis, the role of hemodynamics in CAV development has not been examined. HTx patients (n = 5) requiring percutaneous coronary intervention (PCI) for a focal, epicardial lesion were studied. Angiographic images were retrospectively obtained from baseline (i.e., in the presence of no observed disease) and follow-up catheterizations (i.e., at the time of PCI; 12.4 ± 2.6 years post-HTx). Patient-specific computational models were created from baseline images. Computational fluid dynamic techniques were employed to quantify the hemodynamic environment, which was expressed as normalized time-averaged WSS (TAWSSnorm; measure of temporal WSS magnitude) and normalized WSS angle deviation (WSSADnorm; measure of instantaneous WSS vector oscillation) values. Baseline hemodynamic and follow-up angiographic data were co-registered to investigate the association between WSS and subsequent occlusive CAV lesion location. Results indicate a high degree of co-localization between baseline low WSS data and follow-up occlusive CAV lesion. Local minima in TAWSSnorm were located 2.5 ± 0.6 mm from the site of PCI. Furthermore, local maxima in WSSADnorm were located 3.9 ± 0.7 mm from the site of PCI. In 3 patients, the occlusive lesion formed in a region that was subjected to both low and oscillatory WSS at baseline. There was discernable spatial co-localization between baseline disturbed flow patterns and follow-up CAV lesions requiring PCI. These results suggest a role of fluid mechanics in the development of focal, flow-limiting CAV lesions.
Keywords: Cardiac allograft vasculopathy; Cardiac transplant; Computation fluid dynamics; Hemodynamics; Wall shear stress.