Abstract 17111: Genetic Influences on Human Plasma Metabolites That Determine Mortality in Heart Failure Patients

Abstract

Background: Heart failure with reduced ejection fraction (HFrEF) continues to be a global health burden. We recently reported a strong association of plasma metabolite profile (thirteen different metabolites) with survival among patients with HFrEF but the mechanism of this association is unclear. We studied the genetic associations of plasma metabolite to try to illuminate the disease mechanisms at work. Methods: A prospective genetic registry including 1032 HFrEF with genome-wide genotyping and imputation of single nucleotide polymorphisms (SNP; ~7 millions) and targeted metabolomics profiling. Genome-wide (GW) association analyses were performed to investigate the genetic impact of each SNP on the 13 metabolites of interest. Analyses were performed using linear models stratified by race (European Americans [N=516] and African Americans [N=516] and including the top five principal components (PC) for each population to avoid spurious associations due to population stratification. A fixed-effect meta-analysis was then used to combine results from the two populations. Threshold for GW significance was p<5E-8. As a second step, Cox proportional hazard regression models of overall survival were tested for each SNP that met GW significance. Results: We found at least one SNP reaching GW significance for 6 metabolites metabolites (a-KG, C5.Isovaleryl, Fumarate, Leucine, Succinate and X3.HBA). The most significant SNP (rs77836784, an intronic variant within PACRG) was associated with X3.HBA in African American HF patients (Beta estimate=22.82, P-value=7.14E-14). A cross-checking of SNP associations on both metabolites and HF survival identified 13 independent SNPs that may partially explain the contribution of the metabolites on HF mortality (Table 1) and seem to have biologic plausibility including for example a G-protein coupled receptor (GPR68) and protein kinase C (PRKC1). Conclusion: Genetic variants are associated with predictive plasma metabolites and survival. This may be useful to understand the underlying pathways connecting metabolites to HF progression. A complete investigation of genomic and metabolomics determinants including causal mediation and pathway analyses are on-going.