BACKGROUND: Vessel Fractional Flow Reserve (vFFR), a new angiography-derived method for the functional assessment of coronaries, was recently shown to have good correlation with invasive wire-derived FFR, when vFFR-specific image acquisition requirements were followed. We sought to investigate the feasibility of vFFR analysis and its correlation with FFR in the situation where angiography is completed in routine fashion, without intention for virtual analysis. METHODS: Utilizing an anonymized database maintained at our Cardiovascular Imaging Core Laboratory, we included angiographic images from patients that underwent pre- and post-PCI FFR. CAAS Workstation 8.1 software (Pie Medical Imaging) was used for vFFR evaluation. RESULTS: Out of 624 angiograms (312 pre-PCI and 312 post-PCI), vFFR was successfully analyzed in 219 (35.1%) (115 pre-PCI and 104 post-PCI). Reasons for vFFR analysis failure were: <2 angiographic projections (42.5%), table movement while acquisition (25.7%) and resolution incompatibility (15%). From 115 patients with analyzable pre-PCI vFFR, 74 (64.3%) showed agreement with the respective FFR results in terms of positive (≤0.80) vs negative (>0.80) FFR. Pearson's correlation coefficient between them was 0.449 (p < 0.0001). From 104 lesions with analyzable post-PCI vFFR, 94 had availability of FFR, 74 (78.7%) of which showed agreement between the vFFR and FFR. Pearson's correlation between the values was 0.115 (p = 0.2703). CONCLUSION: vFFR could be analyzed in about one-third of previously completed angiographies and a weak correlation was seen between vFFR and FFR. Our results show the importance of following the pre-specified requirements for vFFR analysis. Further studies are needed to validate the software in different settings.
BACKGROUND: Vessel Fractional Flow Reserve (vFFR), a new angiography-derived method for the functional assessment of coronaries, was recently shown to have good correlation with invasive wire-derived FFR, when vFFR-specific image acquisition requirements were followed. We sought to investigate the feasibility of vFFR analysis and its correlation with FFR in the situation where angiography is completed in routine fashion, without intention for virtual analysis. METHODS: Utilizing an anonymized database maintained at our Cardiovascular Imaging Core Laboratory, we included angiographic images from patients that underwent pre- and post-PCI FFR. CAAS Workstation 8.1 software (Pie Medical Imaging) was used for vFFR evaluation. RESULTS: Out of 624 angiograms (312 pre-PCI and 312 post-PCI), vFFR was successfully analyzed in 219 (35.1%) (115 pre-PCI and 104 post-PCI). Reasons for vFFR analysis failure were: <2 angiographic projections (42.5%), table movement while acquisition (25.7%) and resolution incompatibility (15%). From 115 patients with analyzable pre-PCI vFFR, 74 (64.3%) showed agreement with the respective FFR results in terms of positive (≤0.80) vs negative (>0.80) FFR. Pearson's correlation coefficient between them was 0.449 (p < 0.0001). From 104 lesions with analyzable post-PCI vFFR, 94 had availability of FFR, 74 (78.7%) of which showed agreement between the vFFR and FFR. Pearson's correlation between the values was 0.115 (p = 0.2703). CONCLUSION: vFFR could be analyzed in about one-third of previously completed angiographies and a weak correlation was seen between vFFR and FFR. Our results show the importance of following the pre-specified requirements for vFFR analysis. Further studies are needed to validate the software in different settings.