Diagnostic performance of on-site computed CT-fractional flow reserve based on fluid structure interactions: Comparison with invasive fractional flow reserve and instantaneous wave-free ratio

Abstract

Aims We evaluated diagnostic accuracy of CT-fractional flow reserve (CT-FFR) computed on-site with a new vendor workstation, against invasive FFR as the reference standard. Methods and results Retrospective analyses compared CT-FFR of 104 vessels with 30-90% diameter stenosis in 75 patients imaged using single-rotation 320 detector-row coronary CT angiography (CCTA) with invasive FFR performed within 90 days. Prospective ECG-gated CCTA included exposure of 70-99% of the R-R interval. CT-FFR was computed on-site within the same physical space as the CT scanner and reading room. The diagnostic accuracy of CCTA >50% and CT-FFR ≤0.8 to detect hemodynamically significant stenosis, defined as FFR ≤0.8, was determined, as was the correlation of CT-FFR to FFR and instantaneous wave-free ratio (iFR). Forty-four vessels (42.3%) had an invasive FFR ≤0.8. The sensitivity, specificity, positive, and negative predictive value of CT-FFR ≤0.8 vs. CCTA >50% to detect hemodynamically significant stenosis defined as FFR ≤0.8 were 90.9% vs. 70.5%, 78.3% vs. 43.3%, 75.5% vs. 47.7%, and 92.2% vs. 66.7%, respectively. Area under the curve of CT-FFR was significantly higher than CCTA >50% [0.85, 95% confidence interval (CI): 0.76-0.91 vs. 0.57, 95% CI: 0.47-0.67; P < 0.0001]. The correlation coefficient between CT-FFR and iFR was r = 0.62 (95% CI: 0.40-0.77, P < 0.0001) and that between CT-FFR and invasive FFR was r = 0.52 (95% CI: 0.28-0.70, P = 0.0001). CT-FFR inter- and intra-observer correlations were excellent (r = 0.83 and r = 0.82, respectively). Conclusion Locally computed CT-FFR based on fluid structure interaction has excellent diagnostic accuracy to detect a significant FFR ≤0.8 compared with conventional CCTA and high reproducibility.