Acute coronary syndrome risk in coronary side branch versus main vessel lesions: Lumen, plaque, and hemodynamic insights from coronary CT angiography

Background: Risk assessment of coronary side branch (SB) lesions remains challenging because stenosis-based assessment alone often fails to identify clinically relevant lesions. Objectives: To characterize SB lesions associated with acute coronary syndrome (ACS), evaluate the performance of conventional main vessel (MV)-derived high-risk plaque criteria in SB lesions, and identify independent SB predictors of ACS. Methods: We analyzed 2451 coronary lesions (2011 MV, 440 SB) from the EMERALD-II study, in which coronary CT angiography was performed 1–36 months before ACS using an AI-assisted quantitative analysis platform. Lesion characteristics included stenosis severity, plaque burden, adverse plaque characteristics (APCs), and the change in CT-derived fractional flow reserve across the lesion (ΔFFR_CT). Diagnostic performance of conventional high-risk plaque criteria was compared between MV and SB lesions. Results: Culprit lesions showed greater stenosis severity, plaque burden, APC count, and ΔFFR_CT than non-culprit lesions in both vessels. However, these interrelationships were weaker in SB lesions. Among lesions meeting high-risk criteria, SB lesions were less often ACS culprits than MV lesions: ≥50% stenosis, 11.1% vs. 38.5%; plaque burden ≥70%, 8.6% vs. 23.8%; ≥2 APCs, 18.2% vs. 34.4%; and ΔFFR_CT ≥0.10, 19.4% vs. 49.4%. Positive predictive values and F1-scores were consistently lower for SB lesions. APC count and ΔFFR_CT were independent predictors of ACS in SB lesions. Conclusions: SB lesions had a substantially lower likelihood of subsequent ACS culprit status than equivalent MV lesions. These findings from AI-assisted CCTA analysis highlight the limitations of applying MV-derived thresholds to SB lesions and support the need for SB-specific risk assessment.