The saphenous vein remains one of the most important conduit sources for coronary artery bypass grafting (CABG) because of its availability, length, and technical versatility1. However, saphenous vein grafts exhibit a failure rate of up to 61% at 10 years after CABG2. This increases the need for repeat revascularization and is associated with a higher risk of in-hospital mortality. In recent years, accumulating evidence has suggested that hemodynamic factors, particularly competitive flow and abnormal wall shear stress, are major contributors to late graft failure3,4,5. However, the lack of suitable animal models has created a bottleneck for related research. At present, most preclinical studies of vein graft failure (VGF) still rely primarily on small animals, especially rats, mice, and rabbits, because these models are inexpensive and technically accessible, whereas large-animal graft models remain relatively limited6. The most commonly used method for establishing small animal vein graft models involves transecting the common carotid artery and connecting the two stumps with an autologous inferior vena cava7,8,9. However, due to the substantial differences in heart rate and blood pressure between small animals and humans10, findings derived from small-animal models may not fully reflect the pathophysiological mechanisms in humans.
Pigs are ideal models for cardiovascular surgery because their hearts closely resemble human hearts in cardiac size, coronary anatomy, and surgical handling11,12. In addition, porcine and human coronary arteries are highly similar in both anatomic and hemodynamic characteristics, and pigs have been shown to recapitulate the pathological phenotype of vein graft disease13,14. Therefore, performing CABG with vein grafts in a porcine model may provide a superior animal model for investigating VGF.
In this study, three on-pump double-vessel CABG models in Bama miniature pig models were established using bilateral autologous internal mammary veins (IMVs). One IMV graft was constructed from the aorta to the left anterior descending coronary artery (LAD), and the other from the aorta to the left circumflex coronary artery (LCX). In one representative model, intraoperative ultrasonographic assessment confirmed the patency of both vein grafts, and postoperative computed tomography angiography (CTA) demonstrated the patency of the aorta-to-LAD IMV graft. Notably, the target native coronary arteries were not ligated in this model. Therefore, this model is most suitable for investigating the effects of competitive flow in the target coronary arteries on graft hemodynamics and graft failure. However, competitive flow in the target coronary arteries was not directly measured in the present study. Thus, this study mainly demonstrates the technical feasibility of model establishment.