Evaluation of coronary allograft vasculopathy using multi-detector row computed tomography: a systematic review

Coronary allograft vasculopathy (CAV) is a significant cause of morbidity and mortality after cardiac transplantation and requires frequent surveillance with catheter-based coronary angiography (CCA). Multi-detector row computed tomography (MDCT) has been shown to be effective in assessing atherosclerosis in native coronary arteries. This article systematically reviews the literature to determine the accuracy of MDCT in CAV assessment. An English-language literature search was performed using EMBASE, OVID, PubMed, and Cochrane Library databases. Studies that directly compared MDCT with CCA and/or IVUS for the detection of coronary artery stenosis or significant intimal thickening in cardiac transplant patients were analyzed. Data were pooled to obtain weighted sensitivities, specificities, and diagnostic accuracies. Negative and positive predictive values (NPV/PPV) were calculated. A total of seven studies with a sum of 272 patients were included in this review. There were three studies examining 16-slice MDCT and four studies looking at 64-slice MDCT in CAV. Using per-segment analysis, MDCT assessed between 91% and 96% of all coronary segments when evaluating for stenosis. Pooled estimates for sensitivity and specificity for MDCT ranged from 82% to 89% and 89% to 99%, respectively, while NPV was 99%. Per-patient analysis revealed a sensitivity of 87-100% and NPV of 96-100%. PPV was less than 50% for 64-slice MDCT in both per-segment and per-patient analysis. When compared with IVUS, MDCT had a sensitivity of 74-96% and specificity of 88-92% in assessment of intimal thickening. NPV and PPV were 80-81% and 84-98%, respectively. The high sensitivity and NPV of MDCT suggest that it may be a useful, noninvasive screening tool to rule out CAV.