Serum LRG1 as a non-invasive biomarker for identifying complex coronary anatomy in pre-angiography risk...

The SYNTAX Score provides a quantitative evaluation of the complexity of coronary anatomy and informs revascularization strategies; however, its reliance on invasive angiography limits its use in early non-invasive risk assessment. Leucine-rich alpha-2 glycoprotein-1 (LRG1), a novel serum biomarker associated with inflammation and vascular injury, has an unclear relationship with the SYNTAX Score. This study aims to investigate whether LRG1 is an independent risk factor for complex coronary lesions. Furthermore, we explored if a combined model with LDL-C could improve predictive performance.

Bioinformatics analyses were conducted on GEO datasets (GSE60993, GSE62646) to compare stable coronary artery disease (SCAD) with ST-elevation myocardial infarction (STEMI), as well as healthy individuals with acute myocardial infarction (AMI) patients. Serum levels of LRG1 were quantified using enzyme-linked immunosorbent assay (ELISA). Statistical approaches, including least absolute shrinkage and selection operator (LASSO) regression and Boruta feature selection, were employed to identify independent risk factors, which were subsequently incorporated into a logistic regression model. The diagnostic performance of the model was evaluated using receiver operating characteristic (ROC) curve analysis, calibration tests, and decision curve analysis.

GEO database analysis revealed significant overexpression of LRG1 in patients with AMI. This single-center observational study included 138 patients with confirmed coronary artery disease (CAD). Serum LRG1 levels were significantly elevated in AMI patients compared to non-AMI patients. Furthermore, individuals with SYNTAX Scores ≥ 23 demonstrated higher serum LRG1 concentrations than those with Scores < 23 [277.2 (122.8-393.6) μg/mL vs. 113.4 (65.8-248.9) μg/mL, P < 0.001]. Feature selection using the Boruta algorithm and LASSO regression identified LRG1 and low-density lipoprotein cholesterol (LDL-C) as independent predictors of elevated SYNTAX Score ≥ 23. The logistic regression model, depicted as a nomogram, exhibited moderate discriminatory ability (AUC = 0.771, 95% CI: 0.685-0.856), excellent calibration (Hosmer-Lemeshow test, P = 0.416), and demonstrated clear clinical net benefit.

This study provides preliminary evidence that serum LRG1, in conjunction with LDL-C, may serve as a non-invasive adjunct for identifying CAD patients with complex coronary anatomy (SYNTAX Score ≥23).