Urinary Proteomic Panel Distinguishes Acute Myocarditis from Acute Myocardial Infarction

Abstract

Reliable noninvasive discrimination of acute myocarditis (AM) from acute myocardial infarction (AMI) in patients presenting with acute chest pain remains an unmet clinical need. We aimed to identify urinary proteomic markers using mass spectrometry and LASSO regression. Patients with suspected AM or AMI were prospectively enrolled. Urinary proteomes were profiled by data-independent acquisition mass spectrometry. The discovery cohort was stratified by age and sex and randomly divided into training and internal validation sets at a 3:1 ratio. Differentially expressed proteins were screened, and a diagnostic model was constructed using LASSO with 1000 bootstrap resamplings in the training set. The model was validated internally and further verified in an independent cohort using ELISA. A total of 42 AM and 80 AMI patients were included in the discovery cohort. Differential expression analysis in the training set identified 880 and 1099 differentially expressed proteins in AM and AMI, respectively, compared with controls. A three-protein panel comprising TNFRSF1A, F9, and PRDX3 achieved an AUC of 0.954 in the training set and exceeded the clinical models (all DeLong’s P < 0.05). Addition of the protein panel to either clinical model further improved diagnostic performance (both P < 0.001). Similar performance was confirmed in the validation cohorts. A urinary proteomic panel comprising TNFRSF1A, F9, and PRDX3 may provide a noninvasive approach for differentiating AM from AMI.