Abstract Background Fractional flow reserve (FFR) remains the gold standard for evaluating coronary functional stenosis but is limited by its invasive requirements for pressure wires and vasodilators. While optical coherence tomography (OCT) provides high-resolution anatomical insights, conventional OCT parameters demonstrate only weak-to-moderate correlation with FFR. To address this, virtual FFR derived from OCT (VFRoct) has emerged, offering combined anatomical and physiological lesion assessment without additional instruments or pharmacotherapy. However, its diagnostic accuracy remains to be determined. Purpose This study aimed to compare the diagnostic performance of VFRoct against invasive FFR in identifying haemodynamically significant coronary lesions. Methods In this prospective, multicenter, single-arm study, 210 patients with indications for percutaneous coronary intervention underwent simultaneous FFR and OCT assessments across four Chinese centres. VFRoct was calculated through OCT-based 3D vascular reconstruction, computational fluid dynamics, and real-time haemodynamic modelling, with blinded analysis by an independent core laboratory. The primary outcome was the diagnostic accuracy of VFRoct. Results The per-protocol analysis included 200 lesions in 200 patients. The mean diameter stenosis was 67.4%±9.0%, and the mean FFR was 0.85±0.08. Using FFR ≤0.80 as the reference standard, the accuracy, sensitivity, specificity, and positive predictive value of VFRoct versus FFR were 98%, 93.6%, 99.3%, 97.8%, respectively. Within the FFR grey zone (0.75-0.85), accuracy remained high at 95.9% (89.7% sensitivity and 100% specificity). VFRoct showed strong correlation (r=0.95, p0.001) and agreement (mean difference -0.001±0.026) with FFR. Compared to OCT anatomical parameters (minimum lumen area, diameter stenosis and area stenosis), VFRoct achieved superior diagnostic performance (AUC=0.99, p0.01). Conclusions VFRoct exhibits excellent concordance with invasive FFR for functional coronary assessment. This technology enables simultaneous anatomical-physiological lesion evaluation, potentially optimising revascularisation strategies in clinical practice.Diagnostic accuracy ROC curves
Jiang et al. (2025) studied this question.