Diagnostic Performance of Quantitative Perfusion Cardiac Magnetic Resonance Imaging in Patients with Prior Coronary Artery Disease

Abstract Aims The diagnostic performance of quantitative perfusion cardiac magnetic resonance (QP-CMR) imaging has scarcely been evaluated in patients with a history of coronary artery disease (CAD) and new onset chest pain. The present study compared the diagnostic performance of automated QP-CMR for detection of fractional flow reserve (FFR) defined hemodynamically significant CAD with visual assessment of first-pass stress perfusion CMR (v-CMR) and quantitative [15O]H2O positron emission tomography (PET) imaging in a true head-to-head fashion in patients with prior CAD. Methods and results This PACIFIC-2 substudy included 145 symptomatic chronic coronary symptom patients with prior myocardial infarction (MI) and/or percutaneous coronary intervention (PCI). All patients underwent dual-sequence, single bolus perfusion CMR and [15O]H2O PET perfusion imaging followed by invasive coronary angiography with three-vessel FFR. Hemodynamically significant CAD was defined as an FFR ≤0.80. QP-CMR, v-CMR and PET exhibited a sensitivity of 66%, 67%, and 80%, respectively, whereas specificity was 60%, 62%, and 63%. Sensitivity of QP-CMR was lower than PET (P=0.015), whereas specificity of QP-CMR and PET was comparable. Diagnostic accuracy and area under the curve (AUC) of QP-CMR (64% and 0.66) was comparable to both v-CMR (66% [P=NS] and 0.67 (P=NS]) and PET (74% [P=NS] and 0.78 [P=NS]). Conclusions In patients with prior MI and/or PCI, the diagnostic performance of QP-CMR was comparable to visual assessment of first-pass stress perfusion CMR and quantitative [15O]H2O PET for the detection of hemodynamically significant CAD as defined by FFR.