Prognostic Value of Quantitative Stress Perfusion CMR

Study Questions:

What are the prognostic implications of quantified myocardial perfusion reserve and ischemia by stress perfusion cardiac magnetic resonance (CMR) imaging?


Patients referred for stress CMR at a single tertiary hospital were prospectively consented. Only those who underwent adenosine stress perfusion CMR were included. Patients underwent imaging using a dual bolus technique in order to avoid signal saturation of the blood pool input function. Patients were followed up using electronic records for a primary composite endpoint of death from cardiovascular causes, nonfatal myocardial infarction, aborted sudden cardiac death, or late revascularization (>90 days after CMR study). Ischemia was visually assessed on three short-axis first-pass perfusion slices as present or absent across the standard American Heart Association 16 segments, each divided into an epicardial and endocardial half for a total of 32 segments per patient. Percent myocardium was computed by dividing ischemic segments by 32. Quantification of myocardial perfusion reserve was also performed using Fermi-constrained deconvolution of dynamic perfusion images at stress and rest and taking the ratio of stress blood flow divided by rest blood flow. Cross-validated Cox regression was used for survival modeling.


A total of 52 of 395 patients meeting inclusion/exclusion criteria experienced the primary endpoint after a median of 1.3 years (interquartile range, 0.5-2.4 years). Both visually quantified extent of perfusion and quantitative myocardial perfusion reserve were associated with outcomes with similar improvement of discrimination compared to a clinical model (c-statistics: clinical model 0.75, visual ischemia 0.84, and quantitative perfusion reserve 0.85). In addition, dichotomized ischemia (involving more than two segments of visually determined ischemia or ≥10% myocardium with quantitative myocardial perfusion reserve <1.5) also performed well for risk discrimination and reclassification.


Ischemia on stress perfusion CMR can be quantified based on visual assessment or as quantitative myocardial perfusion reserve in routine clinical practice, and these parameters offer improved risk stratification beyond clinical variables.


This study elegantly demonstrates that, like nuclear cardiology methods, stress perfusion CMR can also be used to quantify ischemia based on visual extent or based on quantification of myocardial blood flow. This is the largest study to date to have performed quantification of blood flow using CMR, although much larger studies have been published using PET methods.

There are a few limitations to this work. First, a double bolus technique was used, meaning a low-dose initial bolus at stress for quantification of blood pool immediately followed by a larger bolus also during stress for quantification of tissue perfusion. A similar double bolus is used at rest. This method is more cumbersome, but less prone to signal saturation than single bolus methods.

Second, it appears that after computing myocardial perfusion reserve, the proportion of myocardial with reserve <1.5 was computed and that was entered into models rather than the overall global myocardial reserve, as has usually been done in nuclear studies. This may be a reason why quantitative perfusion was not seen as adding to visual assessment. Combined visual and quantitative assessment may offer even greater value as has been consistently seen in the nuclear literature, although much larger studies may be required to demonstrate this.

Clinical Topics: Arrhythmias and Clinical EP, Noninvasive Imaging, SCD/Ventricular Arrhythmias, Magnetic Resonance Imaging, Nuclear Imaging

Keywords: Adenosine, Angina Pectoris, Death, Sudden, Cardiac, Diagnostic Imaging, Magnetic Resonance Imaging, Magnetic Resonance Spectroscopy, Myocardial Infarction, Myocardial Ischemia, Myocardial Perfusion Imaging, Myocardium, Perfusion, Tertiary Care Centers

< Back to Listings