Stress-perfusion Cardiac Magnetic Resonance (CMR) at 3.0 tesla: diagnostic accuracy and appropriateness of CMR-driven coronary angiography. A single center experience

Purpose: Stress-perfusion cardiac magnetic resonance (CMR) has become an important non-invasive diagnostic tool for the functional assessment of patients with known or suspected coronary artery disease. However, little data are available about the appropriateness of CMR-driven coronary angiography (cath) in ischemia positive patients and about the diagnostic accuracy of stress-perfusion CMR at 3.0 Tesla. Methods: This was an observational single center study. All consecutive patients undergoing combined stress-perfusion/viability CMR on a 3.0 Tesla system at our institution between September 2011 and December 2012 were included. Left ventricular myocardial ischemia was quantified with a semi quantitative score by using a 16-segment model (score for each segment: 0-2, maximal score 32). The localization of myocardial ischemia was described based on the territories of the main epicardial coronary vessels. Territories with predominant transmural scar were excluded. The appropriateness of cath in ischemia positive patients and the diagnostic accuracy of stress-perfusion CMR to detect a significant coronary stenosis (≥ 70%) at subsequent, CMR-driven cath, was determined based on a coronary territory analysis. Results: 668 stress-perfusion/viability CMR were performed, of which 168 (20%) were positive for myocardial ischemia and 92 of such patients (55%) underwent cath after CMR at discretion of the treating cardiologist. Patients not undergoing cath had less extensive myocardial ischemia (mean ischemic score 2.7±1.8 vs. 4.4±3.1; p<0.01). In 80 patients (87% of those undergoing cath) either a significant coronary stenosis (≥ 70%) was detected or a revascularization was performed. All together 267 coronary territories (in 92 patients, 9 territories excluded because of predominant transmural scar) were compared with cath findings. Overall, sensitivity and specificity was of 90% and 87%. For the single coronary territories sensitivity and specificity was as follows: left anterior descending (n=88; 83%-85%), circumflex (n=91; 100%-79%), right coronary artery (n=88; 86%-94%). Conclusions: Stress perfusion CMR at 3.0 Tesla demonstrates good sensitivity and specificity for detecting significant coronary artery disease in a coronary territory analysis. CMR driven-cath in ischemia positive was shown to be appropriate in a large collective of unselected patients undergoing stress perfusion CMR for functional evaluation of coronary artery disease.