Abstract
OBJECTIVES: End-to-side anastomosis is common in coronary artery bypass grafting, although restrictive suturing can narrow the anastomosis. We evaluated ex vivo end-to-side models by numerically simulating fluid dynamics to compare various degrees of stenotic anastomoses to predict haemodynamic effects. METHODS: A carotid artery was grafted via an end-to-side anastomosis onto the left anterior descending artery of a porcine heart, with liquid silicone injected into the vessels. The end-to-side image was acquired via multidetector computed tomography for reference, and models of longitudinal shortening and bilateral narrowing were created with 25%, 50%, 75%, along with 90%, and 100% stenosis in the native coronary artery. Haemodynamics were analysed using computational fluid dynamics simulations to calculate streamlines, wall shear stress and oscillatory shear index. RESULTS: In the reference model, the graft inflow impinged on the floor of the native artery, creating a recirculating vortex and a high oscillatory shear index region near the heel. As the graft flow angle increased with longitudinal stenosis, bilateral stenosis generated helical flow near the lateral wall of the native artery, worsening with increased stenosis. At 75% stenosis, both longitudinal shortening and bilateral narrowing caused abnormal flow separation, with low wall shear stress and high oscillatory regions forming distal to the toe of the anastomosis. CONCLUSIONS: Computational fluid dynamics modelling predicts that end-to-side anastomoses with 75% longitudinal or bilateral stenosis are at a risk of intimal hyperplasia causing graft failure, while anastomotic stenosis <50% indicates acceptable haemodynamics. Future studies should explore long-term clinical outcomes with suboptimal surgical anastomotic construction. CLINICAL REGISTRATION NUMBER: Not applicable.