Diagnostic Accuracy of Three Different Protocols for 3.0T Coronary Magnetic Resonance Angiography

The usefulness of coronary magnetic resonance angiography (cMRA) has been reported, although the difference in the diagnostic accuracy of different protocols has not been established.We compared conventional coronary angiography (CAG) and cMRA, conducted within 6 months in 24 consecutive patients between September 2012 and July 2014. Three cMRA protocols were examined, cMRA1, free-breathing wholeheart coronary angiography (WHCA) without contrast; cMRA2, free-breathing WHCA with contrast; and cMRA3, breath-hold steady-state free precession with contrast using a 3.0 T scanner. Image quality was graded on a 4-point scale: 1) nonassessable; 2) assessable, fair vessel contrast; 3) assessable, good vessel contrast; and 4) assessable, excellent vessel contrast. Significant narrowing of the coronary arteries was visually assessed.Stenosis was observed in 34 segments, with a prevalence of 10.3%. For cMRA1, cMRA2, and cMRA3, the numbers of assessable segments were 245 (74.2%), 287 (87.0%), and 164 (49.7%), respectively (P < 0.001 by the McNemar test). For assessable segments, the sensitivity, specificity, positive predictive value, and negative predictive value were 89.3%, 99.1%, 92.6%, and 98.6% for cMRA1, 90.0%, 98.1%, 84.4%, and 98.8% for cMRA2, and 76.5%, 93.9%, 59.1%, and 97.1% for cMRA3, respectively. For the assessable segments, the image quality score was better with cMRA2 than with the other two protocols.cMRA is a useful modality to rule out coronary artery disease, especially the cMRA2 protocol, which performed better than the other two protocols.