Abstract P121: Combining Extracellular Matrix Biomarkers Using Cluster Analysis More Accurately Predicts The Development Of Systolic Dysfunction In Acute Myocardial Infarction Compared To Single Biomarker Analysis

Abstract

Introduction: The extracellular matrix (ECM) is central to cardiac repair following acute myocardial infarction (AMI), and pathological ECM activity may result in adverse ventricular remodeling. Systolic dysfunction is a manifestation of adverse remodeling, and the utility of combined biomarker analysis for prognosis remains undetermined. Cluster analysis is a statistical methodology that can combine biomarkers, while also accounting for collinearity. Hypothesis: In this study, we assessed the hypothesis that combining biomarkers using cluster analysis may more accurately predict the development of systolic dysfunction in AMI patients when compared to single biomarker analysis. Methods: In a cohort of 120 AMI patients, plasma levels of matrix metalloproteinase (MMP) -3, -8, -9 and tissue inhibitor of matrix metalloproteinase-1 (TIMP-1) were measured using ELISA and multiplexing assays during hospital admission. All patients had an echocardiogram within 1 year of AMI (median [IQR], 145 [89 - 252] days). Patients were divided into impaired (n=37, LVEF <50%) and preserved (n=83, LVEF ≥50%) systolic function. Hierarchical clustering was performed using Ward’s method of minimum variance. Mann-Whitney U and Chi-Squared testing was performed for univariate analysis, and binary logistic regression was performed for multivariate analysis. Results: Upon univariate analysis, current smoking, prescription of ACE inhibitor at discharge, peak Troponin T > 610 ng/L (median), and MMP-8 levels were predictive of impaired systolic function. Cluster analysis partitioned patients into two clusters (Cluster One, n=31; Cluster Two, n=89). Cluster One comprised a higher proportion of patients with impaired systolic function when compared to Cluster Two (15 out of 31 [48.4%] versus 22 out of 89 [24.7%]). Upon multivariate analysis, Cluster One assignment (odds ratio [95% CI], 2.74 [1.04-7.23], p=0.04) remained an independent predictor of systolic dysfunction in combination with clinical variables, while MMP-8 levels did not (3.43 [0.32-36.68], p=0.308). Conclusion: Findings from our study demonstrate a combined biomarker approach outperforms single biomarker analysis for predicting the development of systolic dysfunction following AMI.