Coronary Flow Reserve by Cardiac MR

Study Questions:

What is the prognostic value of coronary flow reserve (CFR) by cardiac magnetic resonance imaging (MRI) as compared to standard stress perfusion MRI?


Patients with known or suspected coronary artery disease (CAD) underwent stress cardiac MRI using adenosine triphosphate to induce vasodilator stress. In addition to standard cine imaging for cardiac function and volumes, first pass perfusion imaging was performed at peak stress. Phase contrast imaging was performed to quantify blood flow in the coronary sinus at rest and at peak stress. Late gadolinium enhancement (LGE) imaging was performed to identify scar. Patients were followed-up for a composite primary endpoint of cardiovascular death, acute myocardial infarction, unstable angina, hospitalization for heart failure, or ventricular tachyarrhythmia requiring defibrillation.


A total of 709 patients were approached, although 33 were excluded from analyses due to claustrophobia (n = 6), severe symptoms during vasodilation (n = 4), poor image quality (n = 9), or loss to follow-up (n = 14). CFR did not differ significantly between patients with known or suspected CAD (2.69 ± 0.60 vs. 2.87 ± 0.61, p = 0.22). Rest myocardial blood flow (MBF) was significantly higher in those who experienced major adverse cardiac events (MACE), regardless of whether they had prior CAD or not (p < 0.001). Over a median follow-up of 2.3 years, 6.9% of patients experienced MACE. Both reduced CFR (<2.0) and >10% ischemia on stress perfusion CMR were associated with MACE in both patients with known CAD (hazard ratio [HR] 5.2 and 5.1, respectively) and suspected CAD (HR 14.2 and HR 6.5, respectively). Similar areas under the curve were observed for both CFR and perfusion CMR ischemia extent (0.773 vs. 0.731, p = 0.58 among those with CAD and 0.885 vs. 0.776, p = 0.06 among those with suspected CAD). Overall, CFR was less sensitive for MACE (64% vs. 82% with established CAD and 65% vs. 72% with suspected CAD), but more specific than stress perfusion defect (91% vs. 59% with established CAD and 99% vs. 83% with suspected CAD).


Overall prognostic value of stress perfusion CMR and CFR by phase contrast MRI were similar for MACE.


This study shows that both relative perfusion defects and quantitative measures of perfusion are related to cardiac outcomes. The importance of this study is that it extends previous findings from positron emission tomography (PET)–derived parameters to MRI. One intriguing difference is that CFR in PET literature at a threshold of <2.0 is generally more sensitive than specific. This paper finds the opposite for MRI-derived CFR—it is more specific than sensitive. Given that CFR is disturbed even by diffuse disease and microvascular dysfunction, this is somewhat surprising and suggests that the appropriate cut points for PET-derived CFR and phase contrast MRI–derived CFR may be different.

Clinical Topics: Arrhythmias and Clinical EP, Heart Failure and Cardiomyopathies, Noninvasive Imaging, SCD/Ventricular Arrhythmias, Atrial Fibrillation/Supraventricular Arrhythmias, Acute Heart Failure, Computed Tomography, Magnetic Resonance Imaging, Nuclear Imaging

Keywords: Adenosine Triphosphate, Angina, Unstable, Coronary Artery Disease, Coronary Sinus, Cardiac Imaging Techniques, Gadolinium, Heart Failure, Ischemia, Magnetic Resonance Imaging, Perfusion Imaging, Positron-Emission Tomography, Magnetic Resonance Spectroscopy, Myocardial Infarction, Phobic Disorders, Tachycardia, Vasodilation

< Back to Listings