1Friedrich-Alexander-Universität Erlangen-Nürnberg Medizinische Klinik 2 Erlangen, Deutschland
BACKGROUND:
FFRangio – a simulation of FFR based on anatomy derived from invasive coronary angiograms – can be used to assess the hemodynamic relevance of coronary stenoses and has shown accuracy compared to invasive FFR. The specific performance of FFRangio in particular anatomic situations including bifurcation lesions has not yet been studied. We therefore evaluated the diagnostic accuracy of FFRangio in coronary artery bifurcation lesions and validated FFRangio results against invasively measured FFR.
METHODS:
Consecutive patients with bifurcation stenosis in invasive coronary angiography requiring invasive FFR measurement were prospectively included in this analysis. Exclusion criteria were moderate to severe aortic stenosis, more than mildly reduced left ventricular ejection fraction (LV-EF < 45%), in-stent-stenosis, previous PCI in the target vessel in the last 12 months and less than TIMI III flow in the target vessel. Image acquisition rate of coronary angiography was 10-15 frames per second. The target lesion was visualized in 3 angulations at least 30° apart. FFRangio was calculated based on end-diastolic frames using a dedicated workstation with proprietary software. Invasive FFR was determined after intracoronary adenosine injection with a pressure wire placed downstream to the stenosis. Results of FFRangio were compared to invasive FFR; lesions with invasively measured FFR ≤ 0.80 were classified as hemodynamically significant.
RESULTS:
46 lesions in 44 patients (mean age 68±9 years) were analyzed. By visual estimation, the mean degree of luminal stenosis was 62%. The median invasive FFR value across all 46 vessels was 0.81 (IQR 0.13) whereas the median FFRangio value was 0.84 (IQR 0.16) (p=0.03). Invasive FFR revealed hemodynamic significance in 23 lesions. FFRangio showed a close correlation with invasive FFR (r=0.69, p<0.001). Using a cut-off value of ≤0.80, FFRangio displayed a sensitivity of 65% and a specificity of 100% to detect hemodynamically significant stenoses compared to invasive FFR, resulting in a diagnostic accuracy of 83%. ROC-Curve analysis showed an AUC of 0.94. Bland-Altman analysis showed systematically higher values for FFRangio compared to invasive FFR (mean difference of -0.03, 95% limits of agreement from 0.15 to -0.21).
CONCLUSION:
In coronary lesions involving a bifurcation, simulation of FFR derived from coronary angiograms is challenging and may underestimate hemodynamic significance compared to invasive FFR.