Improved Cardiac Risk Assessment With Noninvasive Measures of Coronary Flow Reserve

Study Questions:

What is the incremental value of coronary flow reserve (CFR) determined from rest and stress positron emission tomography (PET) compared to clinical risk factors?


PET was performed for myocardial perfusion and CFR in 2,783 consecutive patients at a single center and analyzed for perfusion defects as well as rest and stress coronary flow, from which CFR was calculated as stress divided by rest myocardial blood flow. CFR was calculated globally for the entire left ventricular (LV) myocardium. Patients were followed for 0.7-3.2 years (median 1.4) for cardiac death.


There were 137 cardiac deaths in the total patient cohort. Patients experiencing cardiac death were older (74.6 vs. 54.8) than those without cardiac death and more often male (66.4% vs. 46.9%, both p < 0.001). Patients experiencing cardiac death were also more likely to have had a diagnosis of prior coronary artery disease (CAD) (76.6% vs. 40.4%) and prior revascularization than those without subsequent cardiac death.

Rest LV ejection fraction (EF) was 39% versus 59% in those without cardiac death and the percent of scar plus ischemic myocardium was 16.2% versus 0%; ischemic myocardium was 4.4% versus 0%. Global CFR was 1.76 in those without cardiac death and 1.31 in those with cardiac death (all comparisons p < 0.0001). The lowest tertile (CFR <1.5) was associated with a 16-fold increased risk compared to the highest CFR tertile (CFR >2.0). Other univariate predictors of increased risk included age, male sex, dialysis, and a diagnosis of prior CAD. Cardiac death occurred in 4.9% of the total population and was noted in 0.8%, 4.0%, and 10.0% of patients in the upper, middle, and lower tertiles of CFR (p < 0.0001). Multiple multivariable survival analysis models were created, which included traditional demographics and risk factors as well as incrementally LVEF, ischemia plus scar percent, LVEF reserve, and CFR lower tertile. Global χ2 increased from 155 to 210 to 227 to 227 to 253 after addition of CFR data (p < 0.001 with addition of CFR). Subgroup analysis revealed that CFR impacted the risk of cardiac death in 29 of 33 evaluated subgroups including gender, age dichotomized at 65 years, hypertensives, diabetes, smoking, prior CAD, among others. Inclusion of CFR data to risk assessment models reclassified 15%, 48%, and 12% of patients predicted to be at low, intermediate, and high cardiac risk without CFR data.


The authors concluded that noninvasive quantitation of CFR with PET provides additional prognostic information in patients with known and suspected coronary disease with respect to both cardiac death and all-cause mortality.


This study in 2,783 consecutive patients nicely demonstrates the incremental value of determining CFR in patients with known or suspected coronary disease for enhanced prognostication of cardiac death. While smaller studies have suggested an independent impact of CFR, this study is substantially larger and provides subgroup analysis demonstrating its incremental value in virtually all patient subgroups. Multiple reasons for the link between a low CFR and outcome are probably present including the fact that a reduced CFR is a marker of obstructive epicardial CAD, which obviously impacts prognosis. Additionally, CFR will reflect not only flow restriction due to epicardial disease, but also microvascular reserve, which may be abnormal in a number of patients without discrete fixed obstruction. The enhanced risk stratification in the patients with visually normal PET scans suggests a role for diffuse atherosclerosis, vascular remodeling, and microvascular dysfunction.

Clinical Topics: Noninvasive Imaging, Atherosclerotic Disease (CAD/PAD), Computed Tomography, Nuclear Imaging

Keywords: Coronary Artery Disease, Risk Factors, Risk Assessment, Positron-Emission Tomography

< Back to Listings