# A metabolomics study reveals potential plasma biomarkers for predicting post-infarction left ventricular remodeling: findings from the metabolights database

## TL;DR

This study identifies potential blood biomarkers that could predict heart remodeling after heart attacks, using a public metabolomics database.

## Contribution

The study identifies threonine, glycine, and histidine as potential novel biomarkers for post-STEMI left ventricular remodeling.

## Key findings

- Lower plasma levels of threonine, glycine, and histidine were associated with left ventricular remodeling after STEMI.
- These metabolites showed moderate predictive accuracy with AUCs ranging from 0.745 to 0.785.
- Pathway analysis linked these metabolites to amino acid metabolism and energy regulation, possibly involving oxidative stress and inflammation.

## Abstract

Left ventricular remodeling (LVR) after ST-segment elevation myocardial infarction (STEMI) is a major determinant of adverse prognosis, yet reliable early biomarkers remain limited.

This secondary analysis of publicly available metabolomic data from the MetaboLights database aimed to explore metabolic signatures associated with left ventricular remodeling (LVR) following STEMI patients. Key differential metabolites were identified through an integrated approach combining multivariate screening (PLS-DA with VIP >1.0) and univariate validation (Log2FC, FDR, Cohen's d). The robustness of principal findings was further verified by bootstrap resampling. These metabolites subsequently underwent comprehensive evaluation through univariate testing, ROC analysis, and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment.

Exploratory analysis revealed distinct differences in plasma metabolic profiles between patients with and without LVR. Notably, levels of threonine, glycine, and histidine were observed to be lower in the LVR group, with their respective AUCs of 0.785, 0.748, and 0.745 suggesting potential predictive value. Pathway analysis suggested an enrichment of these metabolites in processes such as amino acid metabolism and energy regulation, which might be linked to disruptions in oxidative stress, inflammatory, and fibrotic pathways.

In summary, our preliminary findings suggest that threonine, glycine, and histidine may merit further investigation as potential biomarkers for post-STEMI LVR, and their association with key pathophysiological pathways hints at possible mechanistic roles. Given the absence of clinical and demographic data, combined with the small and imbalanced sample (6 LVR vs. 55 Non-LVR patients), our findings should be interpreted as exploratory and hypothesis-generating. Their clinical predictive utility requires validation in larger, independent prospective cohorts.