Competitive Flow Identification and Hemodynamic Analysis in Coronary Artery Bypass Grafting

Abstract Background and Objective: Coronary artery bypass grafting (CABG) is the primary treatment for severe coronary artery disease. Competitive flow and anastomotic stenosis are two core etiologies leading to low graft flow. However, there is a lack of effective clinical methods to accurately distinguish between these two causes. This study aims to explore a novel approach to differentiate the etiologies of low-flow grafts after CABG based on fractional flow reserve (FFR) and intraoperative transit-time flow measurement (TTFM) waveform characteristics, and to quantitatively analyze the impact of competitive flow on the graft hemodynamic microenvironment. Methods : This retrospective study included data from 157 patients who underwent CABG and had intraoperative TTFM measurements. Patient-specific 0D-3D models were constructed based on postoperative CTA images and physiological parameters, and FFR and hemodynamic parameters were computed non-invasively using computational fluid dynamics (CFD). After excluding the gray zone (0.75–0.80), low-flow grafts were classified into a competitive flow group(FFR > 0.80) and an anastomotic stenosis group (FFR < 0.75) based on postoperative FFR. Binary logistic regression was performed to evaluate the predictive value of preoperative FFR, pulsatility index (PI), and diastolic filling fraction (DF). Hemodynamic parameters, including wall shear stress (WSS) and relative residence time (RRT), were compared between the two groups. Results : After initial screening by TTFM, a total of 80 low-flow left internal mammary artery (LIMA) grafts and 80 low-flow saphenous vein graft (SVG) grafts were included. For LIMA grafts, preoperative FFR was an independent predictor of competitive flow (OR = 1.418, 95%CI:1.086–1.851, P = 0.010), with an area under the ROC curve (AUC) of 0.703. The competitive flow group exhibited characteristic systolic retrograde flow with a high-pulsatility waveform, along with significantly decreased mean WSS (WSS-mean, P = 0.004), significantly decreased wall shear stress gradient (WSSG, P = 0.019), and significantly prolonged RRT (P = 0.005). For SVG grafts, DF was an independent risk factor for identifying anastomotic stenosis (OR = 0.912, 95% CI: 0.844–0.985, P = 0.020), with a model AUC of 0.768. However, no significant differences in hemodynamic parameters were found between the two groups, suggesting that venous grafts have relative tolerance to the adverse hemodynamic microenvironment induced by competitive flow. Conclusion : This study establishes a diagnostic system for low-flow graft etiology differentiation based on FFR and TTFM parameters, confirming that preoperative FFR independently predicts competitive flow in LIMA, while DF is more valuable for SVG, revealing significant heterogeneity in hemodynamic responses to competitive flow between arterial and venous grafts, providing a theoretical basis for individualized risk assessment and precise management after CABG.