Clinical Management
Jongmin Seo, PhD
Assistant Professor
Kyung Hee University
Yongin-si, Giheung-gu, Kyonggi-do, Republic of Korea
Jongmin Seo, PhD
Assistant Professor
Kyung Hee University
Yongin-si, Giheung-gu, Kyonggi-do, Republic of Korea
Background: Large coronary artery aneurysms (CAAs) resulting from Kawasaki disease (KD) can pose severe risks such as thrombosis, coronary artery stenosis, occlusion, and even sudden death. Existing KD guidelines predominantly emphasize the aneurysm's size, neglecting considerations of CAA shape and associated hemodynamic factors. This study explores the correlation between computational fluid dynamics induced parameters and thrombus formation in KD patients with CAAs.
Methods: A retrospective analysis was conducted on 32 computed tomography (CT) scans from 30 KD patients diagnosed with CAA. Vessel models of coronary arteries with a Z-score of 2.0 or higher were reconstructed. Computational fluid dynamics simulations of coronary blood flows were performed using patient-specific data. CAAs were categorized into thrombosis and non-thrombosis groups, and hemodynamic factors were compared between them.
Results: Analysis included 48 coronary arteries with a diameter under 14 mm, with 4 vessels in the thrombosis group. Standard CAA metrics from the two groups were Z-score (17.60±10.93 vs 9.03±6.62, p< 0.1068), and the maximum diameter (Dmax) (9.79±3.28mm vs 6.32±2.77, p< 0.0364). Wilcox rank test uncovered significant differences hemodynamic parameters between two groups of vessels with and without thrombus formation: Residence time (RT) (4.422±4.041 vs 1.314±1.379 cycles, p< 0.0155), Time-averaged wall shear stress (TAWSS) (0.53±0.39 vs 8.94±7.72 dynes/cm2, p< 0.0028). Receiver-Operating-Characteristics (ROC) analysis revealed superior predictive performance on thrombus in hemodynamic parameters (AUCRT=0.87, AUCTAWSS=0.96), against the geometry-based metrics (AUCDmax=0.82, AUCZ-score=0.76), where AUC means area under the ROC curve.
Conclusion: ROC analysis on CFD parameters indicates better predictive accuracy for thrombus in hemodynamic factors compared to Z score of CAAs with diameters under 14 mm.