Work

Comparison of the Diagnostic Performance of Different Coronary Functional Assessment Methods in Moderate-to-Severe Stenosis: A Real-World Study

Year: 2026
Type: preprint
Study design: Observational Study
Abstract: Background Clinical management of moderate-to-severe coronary stenosis depends substantially on whether a lesion causes functional ischemia. Although invasive fractional flow reserve (FFR) serves as the reference standard for functional evaluation, its routine application is constrained by procedural invasiveness, examination time, drug tolerance, and resource requirements. By deriving CT-FFR from coronary computed tomography angiography (CCTA) and obtaining Murray law-based quantitative flow ratio (µQFR) from invasive coronary angiography (ICA), less invasive and more efficient functional assessment approaches become achievable. However, direct comparisons of different CT-FFR algorithms with ICA-based µQFR remain scarce for lesions with 50%–90% stenosis, especially when domestically developed CT-FFR software based on distinct computational techniques is considered. Methods Patients undergoing CCTA, ICA, and FFR measurement within a 30-day interval from January 2018 through December 2025 were retrospectively screened. After applying the predefined inclusion and exclusion criteria, 49 patients with 51 target vessels exhibiting visually assessed moderate-to-severe coronary stenosis (50%–90%) were ultimately enrolled. Using FFR as the benchmark, two CT-FFR algorithms and µQFR underwent evaluation and comparison for diagnostic performance. Results The two CT-FFR algorithms and µQFR were successfully evaluated in all 51 vessels. For CT-FFR1, sensitivity, specificity, and accuracy were 80.95% [95% confidence interval (CI), 58.09%–84.55%], 66.67% (95%CI, 47.19%–82.71%), and 72.55% (95%CI, 58.26%–84.11%), respectively. The corresponding values for CT-FFR2 were 61.91% (95%CI, 38.44%–81.89%), 83.33% (95%CI, 65.28%–94.36%), and 74.51% (95%CI, 60.37%–85.67%), while those for µQFR were 76.19% (95%CI, 52.83%–91.78%), 83.33% (95%CI, 65.28%–94.36%), and 80.39% (95%CI, 66.89%–90.18%), respectively. The area under the receiver operating characteristic curve (AUC) reached 0.73 (95%CI, 0.58–0.88) for CT-FFR1, 0.73 (95%CI, 0.60–0.87) for CT-FFR2, and 0.81 (95%CI, 0.69–0.94) for µQFR. None of the pairwise comparisons among the three methods reached statistical significance (all P > 0.05). Conclusions Each of the three imaging-derived functional assessment methods demonstrated diagnostic capability for detecting functional ischemia in moderate-to-severe coronary lesions (50%–90%). µQFR yielded the highest AUC and accuracy while maintaining relatively balanced sensitivity and specificity. The two CT-FFR software demonstrated similar overall diagnostic performance, although their individual diagnostic indices differed. Further validation in larger, multicenter prospective studies is needed. Trial registration Registration number: MR-33-24-025985 Date of registration: 16 April 2024
Language: English
Field: Medicine
Open Access status: green
OpenAlex ID: W7236367320
Cites: 28
Cited by:
Related to: 10