Galectin 3 is a powerful risk predictor for recurrent coronary events in acute coronary syndrome patients

Summary, in English

Background: Galectin 3 and ST2 are mediators and biomarkers of myocardial fibrosis and remodeling that have recently entered the clinical practice guidelines as prognostic factors in heart failure patients. Elevated galectin 3 and ST2 levels in acute myocardial infarction patients have also been associated with increased incidence of adverse events during follow-up. Purpose: We aimed to assess the comparative ability of Galectin 3 and ST2 to offer additional prognostic information for risk stratification in acute coronary syndrome (ACS) patients, on top of traditional cardiovascular risk factors and other established prognostic biomarkers. Methods: We measured the levels of galectin 3, ST2, N-terminal pro-B type natriuretic peptide (NTproBNP), high-sensitivity C-reactive protein (hsCRP), highsensitivity troponin T (TnT), cystatin C, and lipids in plasma collected from 524 ACS patients (STEMI, non-STEMI and unstable angina) on day 1 following the acute event. Biomarker levels were correlated with the risk to develop recurrent coronary events, in linear regression models adjusted for age, sex and traditional risk factors (smoking, diabetes mellitus, hypertension, LDL, HDL and triglycerides). Results: During a mean follow-up period of 2.13 years, 63 (12%) of the patients suffered a new coronary event. Baseline galectin 3, ST2, hsCRP, NTproBNP and cystatin C were significantly higher in these patients compared to the event-free controls. In a Cox proportional hazards model with forward step selection that included all biomarkers alongside traditional risk factors, age (HR per year of age 1.06, 95% CI 1.03-1.09), galectin 3 (HR per SD log increase 1.88, 95% CI 1.41- 2.51) and cystatin C (HR per SD log increase 1.39, 95% CI 1.12-1.74) were selected as the only independent predictors of recurrent events in the population. In receiver operating curve (ROC) analyses, addition of galectin 3 significantly improved the c-statistic of the model based on traditional risk factors alone (0.81 vs. 0.76, P