The Diagnostic Value of Bile Acids and Amino Acids in Differentiating Acute Coronary Syndromes

Abstract

Purpose: Acute coronary syndrome (ACS), comprising unstable angina and acute myocardial infarction, is the most dangerous and fatal form of coronary heart disease. This study evaluates serum bile acids (BAs) and amino acids (AAs) as potential predictors of AMI in UA patients.

Patients and methods: A total of 72 Non-Coronary Artery Disease (NCAD) patients, 157 UA patients, and 79 AMI patients were analyzed. Liquid chromatography-tandem mass spectrometry (LC-MS/MS) measured 15 bile acids and 19 amino acids. The data was split into training and validation sets (7:3). Univariate and multivariate analyses were performed. Diagnostic value and clinical benefits were assessed using receiver operating characteristic (ROC) curves, decision curve analysis, and metrics such as the area under the curve (AUC), integrated discrimination improvement (IDI), and net reclassification improvement (NRI).

Results: Orthogonal partial least squares discriminant analysis (OPLS-DA) of serum BAs and AAs effectively differentiated NCAD, UA, and AMI groups. The differences in serum BA and AA profiles between UA and AMI patients were primarily driven by four metabolites: deoxycholic acid (DCA), histidine (His), lysine (Lys), and phenylalanine (Phe). Together, they had an AUC of 0.830 (0.768 in the validation cohort) for predicting AMI in UA patients. After adjusting for multiple confounding factors, DCA, His, Lys, and Phe were independent predictors distinguishing UA from AMI. The results of AUC, IDI, and NRI showed that adding these four biomarkers to a model with clinical variables significantly improved predictive value, which was confirmed in the validation cohort.

Conclusion: These findings highlight the association of DCA, His, Lys, and Phe with AMI, suggesting their potential role in AMI pathogenesis.

Keywords: acute myocardial infarction; amino acid; bile acid; metabolomics; unstable angina.

Figure 1

Metabolomic profiling by targeted metabolomics in the training cohort. (A) OPLS−DA 3D models among NCAD, UA, and AMI groups. R2X = 0.425, R2Y = 0.287, and Q2Y = 0.196. (B) The 200−permutation test demonstrated no overfitting in the OPLS-DA model [Q2 = (0.0, −0.128)]. (C) OPLS−DA 3D models between UA and AMI groups. R2X = 0.416, R2Y = 0.492, and Q2Y = 0.263. (D) The 200−permutation test demonstrated no overfitting in the OPLS-DA model [Q2 = (0.0, −0.266)].

Figure 2

Markers of bile acid and amino acid differences between the UA and AMI.

Figure 3

Clustering Analysis and Correlation of Serum Metabolites and Clinical Indicators. (A) Clustering analysis of serum bile acids and amino acids levels in patients with non-coronary artery disease (NCAD), unstable angina (UA), and acute myocardial infarction (AMI). (B) Heatmap of Spearman correlation analysis between differential bile acids, amino acids, and clinical indicators. (A and B) *p<0.05, +p<0.01.