MR perfusion looks even better for detecting CAD

Myocardial MR perfusion may deserve a place in clinical practice for assessing patients for coronary artery disease (CAD). Using a gold standard as reference, researchers gave MR perfusion’s diagnostic ability high marks.

MR perfusion offers an attractive option to existing imaging modalities because it doesn’t require radiation, has high spatial resolution and no attenuation artifacts, Min Li, MD, of Jinan Military General Hospital in Jinan, China, and colleagues, wrote in a study published online Oct. 8 in the Journal of the American College of Cardiology: Cardiovascular Imaging. Previous meta-analyses have compared perfusion MR favorably to quantitative coronary angiography, but fractional flow reserve (FFR) is considered the better approach for detecting stenotic lesions.

Li and colleagues conducted a meta-analysis that looked at perfusion MR with FFR as a reference standard to better understand its ability to detect ischemic CAD. Using PubMed and Embase, they identified 14 studies that had either patient-level or artery/territory-level data.

At the patient level the pooled sensitivity and specificity of MR perfusion was 0.87 and 0.87. Eliminating one outlier study changed both results to 0.9. At the artery/territory-level, the pooled sensitivity and specificity were 0.89 and 0.86, respectively.

The summary receiver-operating characteristic curve for patient- and artery/territory-levels was 0.95 and 0.93, respectively. No one study influenced the results in a sensitivity analysis.

They noted that with quantitative coronary angiography as a reference, MR perfusion had a sensitivity of 0.89 and a specificity of 0.76.

“The present analysis indicates that with FFR as the standard reference, the specificity (0.87) of MR perfusion increases whereas the sensitivity (0.90) is similar,” they observed. “The improved specificity may stem from a lower rate of false-positive reports of arteries with no functional flow limitation despite appearing stenotic on QCA [quantitative coronary angiography].”  

MR perfusion may not be the best option for patients with low pretest probability of CAD, given the time needed for analyses and the use of adenosine during stress tests, they wrote. Nor may it make sense in cases of high pretest probability, since those patients likely would undergo invasive testing anyhow.

“MR perfusion appears most clinically useful in patients with an intermediate pre-test probability of CAD, as both a positive and negative test can provide a relatively acceptable post-test probability of CAD,” they proposed.

They recommended more multicenter studies with long-term follow-up to determine MR perfusion’s usefulness in detecting ischemia and improving patient outcomes.