CTCA-Based Pericoronary Fat and Anatomy–Flow Signatures Identify Future Culprit Lesions in Moderate Stenoses

Background A substantial proportion of coronary events originate from angiographically moderate lesions, indicating that stenosis severity alone does not reflect lesion biomechanical risk. Objectives To test whether adding lesion-adjacent pericoronary adipose tissue (PCAT) and CTCA-derived anatomy–flow descriptors to quantitative plaque assessment improves identification of future culprit lesions, with a prespecified focus on moderate stenosis. Methods We performed a within-patient, lesion-level case–control analysis in the GeoCAD cohort, including patients undergoing coronary revascularisation during follow-up. Culprit lesions were identified from longitudinal CTCA. Stenosis severity, quantitative plaque composition, and PCAT volume were quantified (MEDIS), and vessel centreline geometry and lesion haemodynamics derived using computational modelling. Incremental prognostic value was assessed using Cox models with drop-one and stepwise workflow analyses, including a prespecified subgroup analysis of moderate stenosis lesions (25–49% diameter stenosis). Results Among 46 patients (212 lesions; 55 culprit), percent area stenosis (%AS) dominated culprit lesion discrimination (HR: 2.01; 95% CI: 1.54–2.62; p<0.001). In 82 moderate-stenosis lesions (30 culprit), %AS provided minimal discrimination (ΔC-index: 0.01; p=0.895). Culprit lesions were characterised by greater PCAT volume (HR: 1.75; 95% CI: 1.29–2.37; p<0.001), higher helical flow intensity (HR: 1.35; 95% CI: 1.16–1.57; p<0.001), and lower torsion (HR: 0.50; 95% CI: 0.29–0.84; p=0.009). Adding anatomy– flow descriptors improved risk stratification for moderate lesions beyond CTCA stenosis and plaque/PCAT features (p=0.007). Conclusions In moderate stenosis, lesion-adjacent PCAT and anatomy–flow descriptors provided incremental prognostic information beyond luminal narrowing and plaque composition, supporting integrated CTCA phenotyping to identify high-risk nonobstructive coronary lesions. CONDENSED ABSTRACT A substantial proportion of coronary events arise from nonobstructive, limiting stenosis-based risk assessment on CTCA. In a lesion-level analysis of the GeoCAD cohort, we tested whether adding lesion-adjacent pericoronary adipose tissue (PCAT) and CTCA-derived anatomy–flow descriptors improves identification of culprit lesions. Among 212 lesions, stenosis severity dominated discrimination overall but showed minimal value in 82 moderately stenotic lesions (25–49% diameter stenosis). In this subgroup, PCAT volume, helical flow intensity, and centreline torsion provided stronger prognostic value, and their integration improved risk stratification beyond CTCA stenosis and plaque features, supporting integrated CT phenotyping of high-risk nonobstructive lesions.