Proteomic Associations of Adverse Outcomes in Human Heart Failure

Abstract

Background Identifying novel molecular drivers of disease progression in heart failure (HF) is a high‐priority goal that may provide new therapeutic targets to improve patient outcomes. The authors investigated the relationship between plasma proteins and adverse outcomes in HF and their putative causal role using Mendelian randomization. Methods and Results The authors measured 4776 plasma proteins among 1964 participants with HF with a reduced left ventricular ejection fraction enrolled in PHFS (Penn Heart Failure Study). Assessed were the observational relationship between plasma proteins and (1) all‐cause death or (2) death or HF‐related hospital admission (DHFA). The authors replicated nominally significant associations in the Washington University HF registry (N=1080). Proteins significantly associated with outcomes were the subject of 2‐sample Mendelian randomization and colocalization analyses. After correction for multiple testing, 243 and 126 proteins were found to be significantly associated with death and DHFA, respectively. These included small ubiquitin–like modifier 2 (standardized hazard ratio [sHR], 1.56; P<0.0001), growth differentiation factor‐15 (sHR, 1.68; P<0.0001) for death, A disintegrin and metalloproteinase with thrombospondin motifs–like protein (sHR, 1.40; P<0.0001), and pulmonary‐associated surfactant protein C (sHR, 1.24; P<0.0001) for DHFA. In pathway analyses, top canonical pathways associated with death and DHFA included fibrotic, inflammatory, and coagulation pathways. Genomic analyses provided evidence of nominally significant associations between levels of 6 genetically predicted proteins with DHFA and 11 genetically predicted proteins with death. Conclusions This study implicates multiple novel proteins in HF and provides preliminary evidence of associations between genetically predicted plasma levels of 17 candidate proteins and the risk for adverse outcomes in human HF.

Keywords

• heart failure
• HFrEF
• Mendelian randomization
• proteomics