Is Discordance of Coronary Flow Reserve and Fractional Flow Reserve Due to Methodology or Clinically Relevant Coronary Pathophysiology?

Abstract

The purpose of this study was to determine whether observed discordance between coronary flow reserve (CFR) and fractional flow reserve (FFR) is due to methodology or reflects basic coronary pathophysiology. Despite the clinical importance of coronary physiological assessment, relationships between its 2 most common tools, CFR and FFR, remain poorly defined. The worst CFR and stress relative uptake were recorded from 1,500 sequential cardiac positron emission tomography cases from our center. From the literature, we assembled all combined, invasive CFR-FFR measurements, including a subset before and after angioplasty. Both datasets were compared with a fluid dynamic model of the coronary circulation predicting relationships between CFR and FFR for variable diffuse and focal narrowing. A modest but significant linear relationship exists between CFR and FFR both invasively (r = 0.34, p < 0.001) and using positron emission tomography (r = 0.36, p < 0.001). Most clinical patients undergoing CFR or FFR measurements have diffusely reduced CFR consistent with diffuse atherosclerosis or small-vessel disease. The theoretical model predicts linear relationships between CFR and FFR for progressive stenosis with slopes dependent on diffuse narrowing, matching observed data. Reported changes in CFR and FFR with angioplasty agree with model predictions of removing focal stenosis but leaving diffuse disease. Although CFR-FFR concordance is common, discordance is due to dominant or absent diffuse versus focal disease, reflecting basic pathophysiology. CFR is linearly related to FFR for progressive stenosis superimposed on diffuse narrowing. The relative contributions of focal and diffuse disease define the slope and values along the linear CFR and FFR relationship. Discordant CFR and FFR values reflect divergent extremes of focal and diffuse disease, not failure of either tool. With such discordance observed by invasive and noninvasive techniques and also fitting fluid dynamic predictions, it reflects clinically relevant basic coronary pathophysiology, not methodology.