The aim of the study was to compare the detection of clinically significant coronary artery stenosis using a magnetic resonance 3-dimensional (3D) breath-hold true fast imaging with steady-state precession (trueFISP) sequence with conventional coronary angiography.

Twenty-nine patients who were scheduled to undergo or had recently undergone diagnostic conventional coronary angiogram participated in this study. The left main, left anterior descending, left circumflex, and right coronary arteries were studied using the 3D breath-hold trueFISP technique. Each artery was imaged with 1 acquisition in 16 to 32 heartbeats. The image quality, vessel length, and presence or absence of stenosis were evaluated. A score of 1 to 3 (1 = noninterpretable, 2 = good, 3 = excellent) was used to assess image quality. The sensitivity, specificity, accuracy, positive predictive value, and negative predictive value of magnetic resonance angiography in detecting significant coronary artery stenosis (>50% luminal narrowing) on conventional coronary angiogram were calculated.

Of 116 vessels, 108 were assessed in 29 patients. Twenty vessels that had image quality score of 1 were excluded from further evaluation. The mean lengths of vessels visualized were the following: left anterior descending 3.6 cm, left circumflex 3.4 cm, and right coronary 6.6 cm. The entire length of the left main coronary artery was visualized. For the visualized arterial segments, the sensitivity, specificity, accuracy, positive predictive value, and negative predictive value for detecting significant coronary artery stenosis were 92.8%, 95.3%, 95.0%, 68.4%, and 99.2%, respectively.

The 3D breath-hold trueFISP technique, which uses the endogenous contrast of blood, is potentially useful in ruling out significant coronary artery stenosis but not yet sensitive enough as a screening tool.