Accuracy of MR coronary angiography in the evaluation of coronary artery stenosis.

PURPOSE To examine the accuracy of magnetic resonance (MR) coronary angiography for the evaluation of coronary artery stenoses, and to compare the results with conventional angiography as the reference standard. MATERIALS AND METHODS Eighteen patients were examined. The pulse sequence was an ECG-triggered, T2-prepared, fat-suppressed 3D true fast imaging with steady state precession sequence with navigator respiratory gating. No contrast material was used. The 3D slabs were oriented axially, covering the whole heart. The MR images were evaluated for the presence of stenoses exceeding 50% luminal narrowing. The diagnostic accuracy of MR angiography was calculated, and was compared with conventional coronary angiography. All patients gave informed consent, and the study was approved by the institutional ethics committee. RESULTS In conventional angiography, 16 stenoses were detected in 11 patients, and 7 patients were normal. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of MR coronary angiography for detecting coronary stenosis on segment-based analysis were 75%, 98%, 80%, 97%, and 96%, respectively. McNemar test demonstrated no significant difference between MR and conventional angiography (P = 0.62). CONCLUSION MR coronary angiography is a non-invasive diagnostic method currently in a state of evolution. It has limitations, but can be regarded as an alternative noninvasive modality to detect coronary artery disease in selected patient in whom invasive angiography may be hazardous.