Abstract
Nonobstructive coronary artery disease (NOCAD) is increasingly recognized as a heterogeneous condition characterized by diverse pathophysiological mechanisms despite the absence of flow-limiting stenosis. We sought to establish a rule-based dominant imaging phenotype framework integrating functional, structural, and inflammatory dimensions derived from multiparametric coronary computed tomography angiography (CCTA). In this retrospective cohort of 485 patients with NOCAD, CT-derived fractional flow reserve (CT-FFR), quantitative plaque burden and high-risk plaque features, and perivascular fat attenuation index (FAI) were assessed. Using predefined percentile thresholds and hierarchical rules, patients were categorized into function-, structure-, inflammation-dominant, or low-risk phenotypes. During a median follow-up of 36 months, 56 patients (11.5%) experienced major adverse cardiovascular events (MACE). After multivariable adjustment, function dominance was associated with the highest risk (hazard ratio [HR] 4.054, 95% confidence interval [CI] 1.984-8.281; p < 0.001), followed by structure dominance (HR 3.129, 95% CI 1.410-6.944; p = 0.005), whereas isolated inflammation dominance did not show a statistically significant independent association with events, with wide confidence intervals indicating limited precision. These findings suggest a graded pattern of prognostic associations across functional and structural abnormalities in NOCAD and support a phenotype-oriented interpretation of CCTA metrics reflecting distinct biological axes of coronary pathology.