Abstract 16946: Proteomics Identifies Inflammatory, Lipid, and Cell Proliferative Pathways in Diastolic Dysfunction in HIV: A CHART Substudy

Background: Anti-retrovirals have shifted the epidemiology of heart failure in HIV from one of systolic dysfunction to that of diastolic dysfunction (DD), but the underlying mechanisms are incompletely understood. We hypothesized that a proteomics analysis could identify novel biomarkers of DD in HIV. Methods: The NIH-funded Characterizing Heart Function on Anti-Retroviral Therapy (CHART) study enrolled HIV-infected individuals through the Heart Failure Network (HFN) with the goal of identifying clinical, imaging, and molecular markers of DD in HIV. Using the Olink platform (proximity extension assay technology), we profiled 977 proteins in frozen plasma samples from the 195 individuals enrolled in CHART. Logistic regression was used to compare protein levels between DD cases (N=94) and controls without DD (N=101). Results: We observed differences in many proteins between DD+ and DD- including inflammation (EDA2R, TNF-R1, TNFRSF19, VSIG4), lipid metabolism (leptin, FABP4, CLMP, PON3, PLIN1), cell proliferation and cell-cell adhesion (IGFBP6, TGFBR2, NOV, INHBC), and extracellular matrix (COL18A1, PRELP, HSPG2, COL1A1) proteins (p= 1.4x10 -3 – 1.2x10 -4 ). Of these, eight met false discovery rate (FDR) adjustment (EDA2R [tumor necrosis factor receptor superfamily member], leptin, FABP4 [fatty acid binding protein 4], CLMP [cell-cell adhesion, adipocyte differentiation], IGFBP6 [IGF binding protein involved in cell proliferation], TGFBR2 [transforming growth factor binding receptor 2], NOV [cell proliferation and adhesion], INHBC [member of TGF-β superfamily involved in cell development], and adrenomedullin, all FDR p<0.05). Adjustment for BMI, SBP, age, creatinine attenuated results (nominal p=0.24 – 1.3x10 -3 ). Pathway analysis identified extracellular space, TNF activated receptor activity, cytokine-cytokine receptor interaction, collagen trimer, glycosaminoglycan binding, and response to chemical stimulus pathways (FDR P=5x10 -4 - 5x10 -7 ). Conclusions: Using a high throughput proteomics platform we have identified proteins that differentiate HIV-infected individuals with DD from those without DD, highlighting potential biomarkers and molecular mechanisms mediating DD in HIV for further investigation.