Accuracy of 64-Slice Multidetector Computed Tomography for Diseased Coronary Artery Graft Detection

Background. Sixty-four-slice multidetector computed tomography (64-MDCT) has been shown to be a feasible modality for diagnosing coronary artery disease. We studied the accuracy of 64-MDCT in the detection of diseased grafts and also evaluated its limitations. Methods. This study comprised 19 patients who underwent coronary artery bypass grafting and both invasive coronary angiography (ICA) and 64-MDCT. The 64-MDCT images were analyzed for bypass graft occlusion and significant stenosis (>50%) of the anastomosis, and the results were compared with those of ICA. Results. A total of 90 anastomoses, including 25 proximal anastomoses, were evaluated. Of 65 distal anastomoses, including 5 previously occluded grafts in redo cases, 12 distal anastomoses were identified by 64-MDCT as occluded. In comparison, only 10 grafts were identified as occluded by ICA. The sensitivity, specificity, positive predictive value, and negative predictive value for patency were 100% (10 of 10), 96.5% (55 of 57), 83.3% (10 of 12), and 100% (55 of 55), respectively. The ICA patent grafts were evaluated with respect to stenosis. Invasive coronary angiography identified significant stenosis at only 1 site, whereas 64-MDCT showed significant stenosis at 6 sites. The sensitivity, specificity, positive predictive value, and negative predictive value for stenoses were 100% (1 of 1), 93.1% (67 of 72), 16.7% (1 of 6), and 100% (67 of 67), respectively. Conclusions. Although 64-MDCT demonstrated diagnostic accuracy in evaluating bypass grafts, limitations of this method include false positive results in cases of competitive flow between the graft and the native coronary artery. (Ann Thorac Surg 2010; 89: 1906-11) (C) 2010 by The Society of Thoracic Surgeons