Abstract:Background: Angiography-derived fractional flow reserve (virtual FFR) has shown excellent diagnostic performance compared with wire-based FFR. However, virtual FFR pullback curves have not been validated yet. Objectives: To validate the accuracy of virtual FFR pullback curves compared to wire-based FFR pullbacks and to assess their clinical utility using patient-reported outcomes. Methods: Pooled analysis of two prospective studies, including patients with hemodynamically significant (FFR ≤ 0.80) coronary artery disease (CAD). Virtual and wire-based FFR pullbacks were compared to assess the accuracy of virtual pullbacks to characterize CAD as focal or diffuse. Pullbacks were analyzed visually and quantitatively using the pullback pressure gradient (PPG). Patients underwent PCI, and the Seattle Angina Questionnaire (SAQ) was administered at 3-month follow-up. Results: A total of 298 patients (300 vessels) with both virtual and wire-based pullbacks who underwent PCI were included in the analysis. The mean age was 61.8 ± 8.8, and 15% were female. The agreement on the visual adjudication of the CAD pattern was fair (Cohen's Kappa: 0.31, 95% confidence interval: 0.18-0.45). The mean PPG were 0.65 ± 0.18 from virtual pullbacks and 0.65 ± 0.13 from wire-based pullbacks (r = 0.68, mean difference 0, limits of agreement -0.27 to 0.28). At follow-up, patients with high virtual PPG (>0.67) had higher SAQ angina frequency scores (i.e., less angina) than those with low virtual PPG (SAQ scores 92.0 ± 14.3 vs. 85.5 ± 23.1, p = 0.022). Conclusion: Virtual FFR pullback curves showed moderate agreement with wire-based FFR pullbacks. Nonetheless, patients with focal disease based on virtual PPG reported greater improvement in angina after PCI.

Validation of virtual fractional flow reserve pullback curves