170 Chapter 8 CMR imaging analysis All CMR images were collected centrally and analyzed using cvi42 software, with a fully automated pixel-wise QP module (prototype version 5.13.8.2774, Circle Cardiovascular Imaging Inc., Calgary, Canada).13, 14 All images were anonymized before analyses. Two experienced readers (CV and SB), blinded to the group designations, visually assessed the images for scan quality and artifact detect (e.g., motion and/or dark-rim) according to standard interpretation guidelines15 and conducted QP analysis. The fully automated QP module corrected for in-plane respiratory motion of the heart and adjusted for surface coil-induced signal inhomogeneities using the proton density images. The endocardial and epicardial borders were automatically defined by the software, with no manual offset applied by the operator. All stress perfusion images were evaluated for a sufficient adenosine response, defined as presence of the splenic switch off phenomenon and a heart rate increase ≥ 10 BPM.16 Quantitative maps of myocardial blood flow (MBF) were generated using the acquired perfusion CMR data. QP analysis provided quantitative MBF (ml/g/min) during both rest and stress. Myocardial perfusion reserve (MPR), representing the ratio of MBF at stress over rest, was calculated by dividing the results at maximal vasodilation by those at rest.9 Both MBF and MPR values were computed for each myocardial segments, and mean MBF and MPR per coronary territory were determined by averaging across myocardial segments according the American Heart Association 17 segments model excluding apex.17 Additionally, global MBF and MPR were calculated as averages across all myocardial segments. A global MPR of < 2.19 was considered as abnormal.18 Coronary function test On the same day as the CMR, all patients underwent coronary angiography (CAG) to exclude obstructive CAD (defined as fractional flow reserve ≤ 0.80), with CFT to identify epicardial and/or microvascular spasm and CMD. CFT involved intracoronary administration of incremental doses of acetylcholine of 2, 20, 100 and 200 µg in the LAD to induce coronary spasm. An optional dose of 80 μg was administered in the right coronary artery (RCA) if no spasm occurred in the LAD. Nitroglycerin (200 µg) countered acetylcholine effects, followed by intracoronary adenosine (150 µg) to induce hyperemia for coronary flow reserve (CFR) and microvascular resistance assessment. Vasospasm was defined as recognizable chest pain and ischemic changes on electrocardiograph (ECG), with > 90% epicardial vasoconstriction on CAG distinguishing epicardial and microvascular spasm. CMD was defined as CFR ≤ 2.5 and/or a hyperemic microvascular resistance > 2.5 mmHg/cm/s. Microvascular angina (MVA) comprises microvascular spasm and/or CMD [19, 20]. Inconclusive CFT, or equivocal result, describes recognizable angina lacking conclusive criteria for vasospastic angina (with or without > 90% epicardial spasm) or evident ECG changes.21
RkJQdWJsaXNoZXIy MTk4NDMw