Accuracy of 64-slice CT angiography for the detection of functionally relevant coronary stenoses as assessed with myocardial perfusion SPECT

Abstract

CT angiography (CTA) offers a valuable alternative for the diagnosis of CAD but its value in the detection of functionally relevant coronary stenoses remains uncertain. We prospectively compared the accuracy of 64-slice CTA with that of myocardial perfusion imaging (MPI) using (99m)Tc-tetrofosmin-SPECT as the gold standard for the detection of functionally relevant coronary artery disease (CAD). MPI and 64-slice CT were performed in 100 consecutive patients. CTA lesions were analysed quantitatively and area stenoses > or =50% and > or =75% were compared with the MPI findings. In 23 patients, MPI perfusion defects were found (12 reversible, 13 fixed). A total of 399 coronary arteries and 1,386 segments was analysed. Eighty-four segments (6.1%) in 23 coronary arteries (5.8%) of nine patients (9.0%) were excluded owing to insufficient image quality. In the remaining 1,302 segments, quantitative CTA revealed stenoses > or =50% in 57 of 376 coronary arteries (15.2%) and stenoses > or =75% in 32 (8.5%) coronary arteries. Using a cut-off at > or =75% area stenosis, CTA yielded the following sensitivity, specificity, negative (NPV) and positive predictive value (PPV), and accuracy for the detection of any (fixed and reversible) MPI defect: by patient, 75%, 90%, 93%, 68% and 87%, respectively; by artery, 76%, 95%, 99%, 50% and 94%, respectively. Sixty-four-slice CTA is a reliable tool to rule out functionally relevant CAD in a non-selected population with an intermediate pretest likelihood of disease. However, an abnormal CTA is a poor predictor of ischaemia.