Differential risk of acute coronary syndrome in main vessels and side branches according to lumen, plaque, and haemodynamic characteristics

Background The management of coronary bifurcation lesions remains a challenge in interventional cardiology. While prior studies have demonstrated that upfront side branch (SB) stenting offers limited clinical benefit compared with a provisional approach, the optimal strategy for assessing SB lesion risk remains less well defined. Unlike main vessel (MV) lesions, where the interplay between luminal, plaque, and haemodynamic characteristics has been extensively studied, it remains uncertain whether the same high-risk features carry similar clinical significance in SB lesions. Purpose This study aimed to compare the associations among lumen, plaque, and haemodynamic characteristics in MV and SB lesions and to evaluate differences in their acute coronary syndrome (ACS) risk. Methods Using data from the EMERALD-II trial, we analysed coronary lesions from 351 patients who had undergone coronary computed tomography angiography (CCTA) between one month and three years prior to ACS onset. MV and SB lesions were identified, and the lesion characteristics including degree of stenosis, plaque burden, adverse plaque characteristics (APCs), and ΔFFRCT were assessed. Culprit lesions were determined based on invasive coronary angiography at the time of ACS. Results Among the 2,541 lesions analysed, 2,011 (82.0%) were MV lesions and 440 (18.0%) were SB lesions. The correlations between luminal stenosis, plaque burden, number of APCs, and ΔFFRCT were generally weaker in SB than in MV lesions. Among lesions with comparable high-risk features, the proportion of ACS culprits was consistently lower in SB than in MV lesions (≥50% stenosis: 38.5% vs. 11.1%; ≥70% plaque burden: 23.8% vs. 8.6%; ≥2 APCs: 34.4% vs. 18.2%; ΔFFRCT ≥0.1: 49.4% vs. 19.4%, all p < 0.05). Even when applying more stringent cutoffs for high-risk features in SB lesions, the proportion of ACS culprit lesions remained lower than in MV lesions with predefined high-risk features. Diagnostic performance of all four high-risk features, was lower in SB than in MV lesions, although ΔFFRCT ≥0.1 provided the highest predictive value in both MV and SB. Conclusions MV and SB lesions exhibit different lumen-plaque relationships, leading to differences in ACS incidence under identical conditions. These findings suggest that current MV-derived high-risk criteria may not be directly applicable to SB lesions, highlighting the need for a more tailored approach to SB lesion risk stratification.