1023Absolute quantification of cardiac magnetic resonance myocardial perfusion correlates significantly to positron emission tomography on a global and territorial basis in patients with chronic ischemic heart disease

Abstract

Objectives: Department of Clinical Physiology, Nuclear Medicine & PET, Rigshospitalet, University Hospital of Copenhagen and Faculty of Health Sciences, Copenhagen University, Blegdamsvej 9, 2100, Copenhagen, Denmark. Despite the increasing use of first-pass perfusion imaging with cardiac magnetic resonance (CMR), fully quantitative measurement of myocardial perfusion is still a challenge. We compared myocardial perfusion reserve and absolute global and territorial CMR myocardial perfusion with rubidium-82 positron emission tomography (PET) in patients with known severe chronic ischemic heart disease. Methods: Fourteen patients with angina and/or angina equivalent dyspnea due to at least one occluded coronary artery visualized on a recent invasive coronary angiography were included into the study. Rest and adenosine stress images were obtained by 1.5 Tesla MR scanner and mCT/PET 64 slice scanner on two separate days. The rest and stress CMR images were transferred into Matlab (MathWorks, Natick, MA) for the quantitative analysis and interpreted using the modified 17-segment American Heart Association model (minus the apical segment). Bland-Altman analysis and Pearson's correlation were used for the agreement with mean difference ± standard deviation and correlation between CMR and PET on a global and vessel territorial basis. Two-tailed probability values < 0.05 were considered significant. Results: The agreement between myocardial perfusion reserve for CMR and PET on a global and vessel territorial basis was modest and non-significant (global: 0.08 ± 0.98 mL/g/min, r = 0.484, p = 0.156; right coronary artery (RCA): 0.12 ± 1.19 mL/g/min, r = 0.352, p = 0.319; left anterior descending artery (LAD): 0.17 ± 0.97 mL/g/min, r = 0.586, p = 0.075; left circumflex artery (LCX): -0.11 ± 0.87 mL/g/min, r = 0.465, p = 0.175). However, the agreement between the absolute quantitative CMR myocardial perfusion and PET was stronger and significant on a global and vessel territorial basis (global: -0.89 ± 0.87 mL/g/min, r = 0.590, p = 0.003; RCA: -0.83 ± 0.81 mL/g/min, r = 0.670, p < 0.001; LAD: -0.81 ± 0.84 mL/g/min, r = 0.589, p = 0.003; LCX: -1.01 ± 0.95 mL/g/min, r = 0.529, p = 0.009). Conclusions: Absolute quantification of global and territorial CMR myocardial perfusion correlates significantly and stronger than myocardial perfusion reserve to PET findings.