Beschreibung:
Background: Coronary artery tree length and lumen volume follow an allometric power-scaling relationship via Length = K LV * Volume 0.78 . Derangement from this morphometric relationship may indicate atherosclerosis. Methods: This study was a retrospective analysis of the CREDENCE trial, a large multicenter international study. In CREDENCE, patients with suspected CAD underwent CCTA followed by invasive coronary angiography and fractional flow reserve (FFR). For this study, 315 patients with invasive FFR measurements were selected. There were 945 vessel territories, with 681 nonischemic territories (FFR > 0.80) and 264 ischemic territories (FFR ≤ 0.80). Following AI-enabled quantitative CT analysis (AI-QCT) (Cleerly; Denver, CO), coronary territories were split into “stem-crown” units, every 1 cm along each territory. Territories were truncated at 1.5 mm lumen diameter. A single parameter allometric scaling model was employed using a previously validated exponential term (L CROWN = K LV * V CROWN 0.78 ). After log-transformation, K LV was determined via fitting with generalized estimating equations. Results: The mean K LV for ischemic and nonischemic territories was 28.92±4.77 and 25.02±4.84 (p <0.001). The ROCAUC for K LV , diameter stenosis (DS AI-QCT ), and total plaque volume was 0.72 [0.69, 0.76], 0.85 [0.83, 0.88], and 0.70 [0.66, 0.73], respectively, for the diagnosis of ischemia (K LV vs. diameter stenosis: p-value <0.001; K LV vs. total plaque volume: p-value=0.400; DS AI-QCT vs. total plaque volume: p-value <0.001). Multivariable ROCAUC analysis of K LV and diameter stenosis (model 2), and K LV , diameter stenosis, and total plaque volume (model 3) was 0.87 [0.85, 0.90], 0.87 [0.85, 0.90], respectively. Conclusion: Ischemic coronary territories demonstrated significantly larger K LV than nonischemic territories. Moreover, using K LV and DS AI-QCT provided incremental improvement in the diagnosis of ischemia.