233 Prognostic implications of the automated detection of lipid core burden index at optical coherence tomography: post hoc analysis of the CLIMA study

Abstract Aims Plaque vulnerability features are associated with major coronary events and poor outcomes. However, routinary and reproducible manual assessment of plaque vulnerability features at optical coherence tomography (OCT) is still challenging. We recently developed and validated an OCT-derived automated approach that can identify the intra-plaque lipid core burden index (LCBI). Our aim was to investigate the association between the automated detection of OCT-derived LCBI and clinical events. Methods and results We conducted a post hoc analysis of the CLIMA study, a large prospective observational, multicentre registry recruiting all consecutive patients undergoing assessment of the proximal left anterior descending artery (LAD) segment by OCT in the context of clinically indicated coronary angiography. The automated detection of maximum LCBI was carried out in 4 mm of intervention-naïve proximal LAD segment (maxLCBI4mm) by using the validated software. The mean and median value of LCBI in all study population (n = 1003) was 407.6 and 411.1, respectively. Patients with higher LCBI (≥400) were more frequently male (P = 0.016) and affected by insulin-dependent diabetes mellitus (0.046). Furthermore, they showed more frequently at OCT analysis the vulnerable plaque characteristics investigated in the CLIMA study (Table 1). At Cox regression analysis, a maxLCBI4mm ≥400 predicted at 1 year both a hard endpoint of cardiac death and target-vessel myocardial infarction [hazard ratio (HR): 2.56, 95% confidence interval (CI): 1.2–5.3, P 0.011], as well as a composite endpoint of cardiac death, any myocardial infarction and target vessel revascularization (HR: 1.87, 95% CI: 1.1–3.1, P = 0.011). Conclusions In our study, the automated detection of LCBI at OCT was feasible and related to poorer clinical outcome at 1-year follow-up.