Metabolomics integrated with transcriptomics analysis: identifying biomarkers for 4-year all-cause mortality in heart failure patients with preserved ejection fraction in the Chinese Han population
Y LiAbstract
Background
For patients with heart failure with preserved ejection fraction (HFpEF), the risk prediction of subsequent death events remains an unmet clinical issue. A metabolomics strategy, assisted by transcriptomic analysis, was conducted to identify novel serum biomarkers linked to 4-year all-cause death in patients with HFpEF.
Methods
We characterized the plasma metabolomic profile of 500 hospitalized patients with HFpEF. The dataset was randomly divided at a ratio of 1:1 into a training set and a validation set. Differential metabolites were first identified by logistic regression, then refined with the Boruta algorithm and LASSO regression. Meanwhile, weighted gene co-expression network analysis (WGCNA) was performed on mRNA microarray data from 291 HFpEF patients and yielded key gene modules and their associated biological processes. By integrating metabolomic signatures with transcriptomic modules, we assess the potential regulatory mechanisms for the progression of HFpEF.
Results
A total of ten metabolites were selected as the biomarkers to establish risk stratification model (metabolites model). The metabolites model yielded a satisfactory capability for 4-year all-cause mortality with an area under the curve (AUC) of 0.878 (95% CI 0.835-0.921), which outperformed the clinical risk score for 4-year all-cause mortality (AUC 0.707, 95% CI 0.642-0.772), with net reclassification improvement (NRI) of 0.528 (95% CI 0.362-0.693; P < 0.001) and integrated discrimination improvement (IDI) of 0.301 (95% CI 0.187-0.415; P < 0.001). Individuals in the medium and the highest tertile of the metabolites model had a 2.8-fold (HR = 2.78, 95% CI 1.32-5.86) and 23-fold (HR = 22.95, 95% CI 11.07-57.54) increased risk of mortality compared to those in the lowest tertile, respectively. Multi-omics enrichment analysis has revealed unique metabolic patterns, such as glyceride metabolism, tyrosine metabolism, the AMPK signaling pathway and TCA cycle. These pathways are likely to be mutually regulated, which in turn can impact the prognosis of patients with HFpEF.
Conclusions
This study established a risk stratification model consisting of 10 metabolites, which demonstrated a satisfactory ability to predict 4-year all-cause death in patients with HFpEF and enhanced the understanding of the mechanism underlying the prognosis of HFpEF patients.Figure 1Figure 2