Predictive models for cholesterol crystals and plaque vulnerability in acute myocardial infarction: Insights from an optical coherence tomography study

Abstract

BackgroundCholesterol crystals (CCs) are recognized as a risk factor for vulnerable atherosclerotic plaque rupture (PR) and major adverse cardiovascular events. However, their predictive factors and association with plaque vulnerability in patients with acute myocardial infarction (AMI) remain insufficiently explored. Therefore, This study aims to investigate the association between CCs and plaque vulnerability in culprit lesions of AMI patients, identify the factors influencing CCs formation, and develop a predictive model for CCs. MethodsA total of 431 culprit lesions from AMI patients who underwent pre-intervention optical coherence tomography (OCT) imaging were analyzed. Patients were divided into groups based on the presence or absence of CCs and PR. The relationship between CCs and plaque vulnerability was evaluated. A risk nomogram for predicting CCs was developed using the least absolute shrinkage and selection operator method and logistic regression analysis. ResultsCCs were identified in 64.5 % of patients with AMI. The presence of CCs was associated with a higher prevalence of vulnerable plaque features, such as thin-cap fibroatheroma (TCFA), PR, macrophage infiltration, neovascularization, calcification, and thrombus, compared to patients without CCs. The CCs model demonstrated an area under the curve (AUC) of 0.676 for predicting PR. Incorporating CCs into the TCFA model (AUC = 0.656) significantly enhanced predictive accuracy, with a net reclassification improvement index of 0.462 (95 % CI: 0.263–0.661, p < 0.001) and an integrated discrimination improvement index of 0.031 (95 % CI: 0.013–0.048, p = 0.001). Multivariate regression analysis identified the atherogenic index of plasma (odds ratio [OR] = 2.417), TCFA (OR = 1.759), macrophage infiltration (OR = 3.863), neovascularization (OR = 2.697), calcification (OR = 1.860), and thrombus (OR = 2.430) as independent risk factors for CCs formation. The comprehensive model incorporating these factors exhibited reasonable discriminatory ability, with an AUC of 0.766 (95 % CI: 0.717–0.815) in the training set and 0.753 (95 % CI: 0.704–0.802) in the internal validation set, reflecting good calibration. Decision curve analysis suggested that the model has potential clinical utility within a threshold probability range of approximately 18 % to 85 %. ConclusionsCCs were associated with plaque vulnerability in the culprit lesions of AMI patients. Additionally, this study identified key factors influencing CCs formation and developed a predictive model with potential clinical applicability.