Abstract
OBJECTIVES: The aims of this study were (1) to compare computed tomography-derived fractional flow reserve (CT-FFR) values with graft patency and (2) to establish the cut-off value of CT-FFR for predicting competitive graft flow after coronary artery bypass grafting (CABG). METHODS: Of the 77 patients who underwent isolated CABG with an in situ internal thoracic artery (ITA)-based composite graft and who were also evaluated by preoperative cardiac CT, CT-FFR values were obtained in 74 patients. Early postoperative angiograms were performed in all 74 patients. Angiograms were performed to evaluate the grafts as well as the native coronary arteries to find any competitive flow present. Postoperative angiographic findings of graft flow were categorized as perfectly patent, bidirectionally competitive, unidirectionally competitive, and occluded. Receiver operating characteristic curve analysis of preoperative CT-FFR values for predicting postoperative angiographic competition was performed, and cutoff values of CT-FFR and area under the curve were identified. RESULTS: In total, 234 anastomoses were performed in 74 patients (median 3 [interquartile range, 2, 4] anastomoses per patient). Postoperative (median 1 [interquartile range, 1, 2] day) angiograms showed that 196 (83.8%) anastomoses were perfectly patent, 25 (10.7%) anastomoses were bidirectionally competitive, 12 (5.1%) anastomoses were unidirectionally competitive, and 1 (0.4%) anastomosis was occluded. Median CT-FFR values of the coronary arteries with perfectly patent, bidirectionally competitive, and unidirectionally competitive grafts were 0.658 (interquartile range, 0.500, 0.725), 0.809 (interquartile range, 0.789, 0.855), and 0.849 (interquartile range, 0.833, 0.865), respectively. The cutoff value of CT-FFR predicting competitive graft flow was 0.774 (sensitivity, 97.4%; specificity, 98.5% [area under the curve 0.977; P < .001]). CONCLUSIONS: The diagnostic accuracy of CT-FFR for predicting competitive graft flow after CABG was high, and CT-FFR could be used as a guide for predicting functional coronary artery stenosis in surgical revascularization.