Abstract 16691: Racial Differences in Heart Failure With Preserved Ejection Fraction: A Proteomic Analysis

Abstract

Introduction: Heart failure with preserved ejection fraction (HFpEF) is common in African Americans (AAs). AAs with HFpEF tend to have lower natriuretic peptide levels, higher body mass index, worse hypertension, and higher risk for HF hospitalization than whites with HFpEF. Whether the underlying pathogenesis of HFpEF in AAs is different than whites is unclear. Hypothesis: Proteomic analysis could identify novel proteins that differ by race/ethnicity among HFpEF patients. Methods: 129 HFpEF patients (LVEF >50% and elevated LV filling pressures by invasive hemodynamics, echocardiography, or BNP > 100 pg/ml) were prospectively enrolled. Race/ethnicity was ascertained by self-report, and patients were stratified into 2 groups (AAs [n=23] and whites [n=94]), with the remaining patients (n=12) excluded. Biomarkers were quantified using the Olink antibody-based proximity extension assay technique (CVD2, CVD3, and inflammation panels). Each of the 276 biomarkers were compared between groups (AAs vs. whites) using multivariable logistic regression analysis adjusted for age, sex, BMI, hypertension, and creatinine. P-values were adjusted for multiple comparisons using false discovery rate (FDR). Principal component analyses (PCA) analyses were also used to analyze biomarker data. Results: Mean age was 70±12, 58% were female, and 20% were AA. History of HTN was more common in AAs compared to whites (96% vs. 72%, p=0.014); comorbidities were otherwise similar between the 2 groups. On multivariable analysis, 13/276 biomarkers were significantly different between groups after FDR correction: PSGL1, CXCL1, PRSS8, RAGE, CCL15, IL18, CPB1, CXCL5, BDNP, XCL1, THBS2, IL18, and CCL23. The top hit (PSGL1, P-selectin glycoprotein ligand 1), was significantly different between AAs and whites after multivariable adjustment (p=1.2 x 10 -9 ). PSGL1 is involved in the binding of cell adhesion molecules to the endothelium and white blood cells. Prior (non-HF) studies have found that RAGE (receptor for advanced glycation endproducts) levels are lower in AAs than whites, and low RAGE is associated with a number of chronic diseases. Of the top 10 principal components, 4 were significantly different between whites and AAs after multivariable adjustment, suggesting that groups of proteins differ between AAs and whites. Conclusions: On proteomic analyses, there are significant differences in circulating biomarkers between AAs and whites with HFpEF. If validated, these data may provide insight into racial differences in HFpEF pathogenesis and outcomes.