The imortance of hyperemic contrast velocity assessment on image-based fractional flow reserve calculation

Image-based fractional flow reserve (FFR) calculations reported good agreement with FFR measured invasively.

The purpose of this study was to perform a retrospective analysis of the cases of a previous study on less invasive FFR calculation (simple FFR: FFRsim) as a simple calculation from hyperemic contrast flow data and three-dimensional coronary parameters.

We aimed to analyze the relations between the pressure wire-based FFR (FFRmeas) and fixed FFRsim: calculated from the fixed hyperemic velocity, rest FFRsim: calculated using the non-hyperemic frame count data to extrapolate the hyperemic velocity (based on the database used in the FAVOR1 study) hyp FFRsim: the hyperemic velocity derived from the frame count assessment during vasodilation.To calculate the frame count reserve (CFRFC) the resting frame count was divided by the hyperemic frame count; this value was then used to determine the CFRFC/FFRmeas ratio as an indicator of microvascular function in the corresponding myocardial area of the measured coronary vessel.

A total of 50 lesions with intermediate stenosis were investigated. Correlation between rest FFRsim (from the resting frame count extrapolated to the hyperemic velocity) and FFRmeas was lower than the correlation between hyp FFRsim and FFRmeas (r=0.761 vs. 0.824).

Based on ROC curve analysis for predicting the abnormal FFR of ≤0.80 the AUC were significantly higher for the hyperemia-based parameter than those calculated from resting frame counts. Significantly higher AUC were detected by the hyp FFRsim than by the rest FFRsim: 0.936 (95% CI: 0.828 to 0.985) vs. 0.862 (CI: 0.734 to 0.943); p=0.011.

Linear regression analyses between the FFRsim (either by fixed FFRsim or by rest FFRsim or by hyp FFRsim methods) and the FFRmeas showed higher intercepts and less steep of the slopes in the subgroups with presence of microvascular disease defined as CFRFC/FFRmeas <2 than in those without microvascular disease (CFRFC/FFRmeas >2); the difference reached significant level (p=0.019) when calculated by rest FFRsim.

Hyperemic challenge either by adenosine or regadenoson is required for exact image-based FFR calculation especially in cases of suspicion for microvascular coronary disease.

Type of funding source: None