Background
Fractional flow reserve (FFR) and the instantaneous wave-free ratio (iFR) identify arteries that benefit from percutaneous coronary intervention (PCI). FFR or iFR gradients on pullback are often used to predict the physiological result (FFRΔ or iFRΔ), but this approach is unvalidated.
Objectives
The aim of this study was to compare the accuracy of FFRΔ, iFRΔ and FFRcalc (a mathematical solution incorporating interaction between lesions) for predicting post-PCI physiology in serial or diffuse disease.
Methods
Patients with a focal target lesion and either a second focal lesion or a diffusely diseased segment in the same vessel were randomized to FFR- vs iFR-guided PCI (ISRCTN18106869). FFR and iFR pullbacks were performed, with operators blinded to one modality. Following target lesion PCI, FFR and iFR were remeasured. The primary outcome was the error in predicted post-PCI physiology compared with actual values.
Results
A total of 87 patients were randomized to FFR (n = 45) or iFR (n = 42). Median FFR and iFR were 0.70 (Q1-Q3: 0.62 to 0.78) and 0.81 (Q1-Q3: 0.68 to 0.90) at baseline and 0.82 (Q1-Q3: 0.74 to 0.87) and 0.89 (Q1-Q3: 0.83 to 0.93) after target lesion PCI. The predictive errors were 12% (6% to 17%) for FFRΔ, 4% (0% to 9%; P < 0.001) for iFRΔ, and −5% (−18% to 8%; P = 0.427) for FFRcalc. Significant residual disease was missed in 36% of cases with FFRΔ, 34% with iFRΔ, and 14% with FFRcalc.
Conclusions
FFR and iFR pullback gradients overestimate the benefit of target lesion PCI and can miss residual ischemia in one-third of patients. FFR or iFR should be routinely repeated post-PCI in serial disease.