Data_Sheet_1_Association Among Local Hemodynamic Parameters Derived From CT Angiography and Their Comparable Implications in Development of Acute Coronary Syndrome.docx

Background: Association among local hemodynamic parameters and their implications in development of acute coronary syndrome (ACS) have not been fully investigated. Methods: A total of 216 lesions in ACS patients undergoing coronary CT angiography (CCTA) before 1–24 months from ACS event were analyzed. High-risk plaque on CCTA was defined as a plaque with ≥2 of low-attenuation plaque, positive remodeling, spotty calcification, and napkin-ring sign. With the use of computational fluid dynamics analysis, fractional flow reserve (FFR) derived from CCTA (FFR CT ) and local hemodynamic parameters including wall shear stress (WSS), axial plaque stress (APS), pressure gradient (PG) across the lesion, and delta FFR CT across the lesion (ΔFFR CT ) were obtained. The association among local hemodynamics and their discrimination ability for culprit lesions from non-culprit lesions were compared. Results: A total of 66 culprit lesions for later ACS and 150 non-culprit lesions were identified. WSS, APS, PG, and ΔFFR CT were strongly correlated with each other (all p < 0.001). This association was persistent in all lesion subtypes according to a vessel, lesion location, anatomical severity, high-risk plaque, or FFR CT ≤ 0.80. In discrimination of culprit lesions causing ACS from non-culprit lesions, WSS, PG, APS, and ΔFFR CT were independent predictors after adjustment for lesion characteristics, high-risk plaque, and FFR CT ≤ 0.80; and all local hemodynamic parameters significantly improved the predictive value for culprit lesions of high-risk plaque and FFR CT ≤ 0.80 (all p < 0.05). The risk prediction model for culprit lesions with FFR CT ≤ 0.80, high-risk plaque, and ΔFFR CT had a similar or superior discrimination ability to that with FFR CT ≤ 0.80, high-risk plaque, and WSS, APS, or PG; and the addition of WSS, APS, or PG into ΔFFR CT did not improve the model performance. Conclusions: Local hemodynamic indices were significantly intercorrelated, and all indices similarly provided additive and independent ...