[Value of fractional flow reserve derived from coronary computed tomographic angiography and plaque quantitative analysis in predicting adverse outcomes of non-obstructive coronary heart disease].

To investigate the value of coronary computed tomographic angiography (CCTA)-based fractional flow reserve (CT-FFR) and plaque quantitative analysis in predicting adverse outcomes in patients with non-obstructive coronary heart disease (CAD).Clinical data of patients with non-obstructive CAD who underwent CCTA at the Affiliated Hospital of Jiangnan University from March 2014 to March 2018 were retrospectively analyzed and followed up, and the occurrence of major adverse cardiovascular event (MACE) was recorded. The patients were divided into MACE and non-MACE groups according to the occurrence of MACE. The clinical data, CCTA plaque characteristics including plaque length, stenosis degree, minimum lumen area, total plaque volume, non-calcified plaque volume, calcified plaque volume, plaque burden (PB) and remodelling index (RI), and CT-FFR were compared between the two groups. Multivaritate Cox proportional risk model was used to evaluate the relationship between clinical factors, CCTA parameters and MACE. The receiver operator characteristic curve (ROC curve) was used to assess the predictive power of outcome prediction model based on different CCTA parameters.Finally 217 patients were included, of which 43 (19.8%) had MACE and 174 (80.2%) did not. The median follow-up interval was 24 (16, 30) months. The CCTA showed that patients in the MACE group had more severe stenosis than that in the non-MACE group [(44.3±3.8)% vs. (39.5±2.5)%], larger total plaque volume and non-calcified plaque volume [total plaque volume (mm3): 275.1 (197.1, 376.9), non-calcified plaque volume (mm3): 161.5 (114.5, 307.8) vs. 117.9 (77.7, 185.5)], PB and RI were larger [PB: 50.2% (42.1%, 54.8%) vs. 45.1% (38.2%, 51.7%), RI: 1.19 (0.93, 1.29) vs. 1.03 (0.90, 1.22)], CT-FFR value was lower [0.85 (0.80, 0.88) vs. 0.92 (0.87, 0.97)], and the differences were statistically significant (all P < 0.05). Cox regression analysis showed that non-calcified plaques volume [hazard ratio (HR) = 1.005. 95% confidence interval (95%CI) was 1.025-4.866], PB ≥ 50% (HR = 3.146, 95%CI was 1.443-6.906), RI ≥ 1.10 (HR = 2.223, 95%CI was 1.002-1.009) and CT-FFR ≤ 0.87 (HR = 2.615, 95%CI was 1.016-6.732) were independent predictors of MACE (all P < 0.05). The model based on CCTA stenosis degree+CT-FFR+quantitative plaque characteristics (including non-calcified plaque volume, RI, PB) [area under the ROC curve (AUC) = 0.91, 95%CI was 0.87-0.95] had significantly better predictive efficacy for adverse outcomes than the model based on CCTA stenosis degree (AUC = 0.63, 95%CI was 0.54-0.71) and the model based on CCTA stenosis degree+CT-FFR (AUC = 0.71, 95%CI was 0.63-0.79; both P < 0.01).CT-FFR and plaque quantitative analysis based on CCTA are helpful in predicting adverse outcomes in patients with non-obstructive CAD. Non-calcified plaque volume, RI, PB and CT-FFR are important predictors of MACE. Compared with the prediction model based on stenosis degree and CT-FFR, the combined plaque quantitative index can significantly improve the prediction efficiency of adverse outcomes in patients with non-obstructive CAD.