Prospective evaluation of ultrafast cardiac computed tomography for determination of coronary bypass graft patency.

Twenty-five consecutive patients with 68 independent (single distal anastomosis) saphenous vein aortocoronary and 12 internal mammary bypass grafts (27 to left anterior descending, 10 to diagonal, 23 to left circumflex, 20 to right coronary artery) entered a reader-blinded, prospective, standardized study to establish the accuracy of ultrafast (cine) cardiac computed tomography (CT) for determining graft patency compared with invasive angiography. All patients underwent imaging after injection of 35 to 45 ml of meglumine diatrizoate (Renografin-76; 7 to 9 ml/sec for 5 sec) into an arm vein. Electrocardiographically triggered images were acquired over eight to 16 tomographic levels at 1 cm intervals from aortic arch to mid left ventricle. Criteria for graft patency were contrast opacification on at least two noncontinguous levels and contrast density-time curves morphologically similar to that of the aorta. Ultrafast CT correctly determined that 46 of 48 bypass grafts were patent and 31 of 32 were occluded (sensitivity, specificity, and accuracy 96%, 97%, and 96%); there were no interpretation errors in 23 (92%) of the 25 patients. Accuracy was independent of vessel bypassed and not different for saphenous veins (96%) compared with internal mammary bypasses (100%). This study establishes a 20 min outpatient intravenous injection technique that is highly accurate for determining patency of coronary artery bypass grafts.