Proteomic Correlates and Prognostic Significance of Kidney Injury in Heart Failure With Preserved Ejection Fraction

Abstract

Background: Kidney disease is common in heart failure with preserved ejection fraction (HFpEF). However, the biologic correlates and prognostic significance of kidney injury (KI), in HFpEF, beyond the estimated glomerular filtration rate (eGFR), are unclear.

Methods and results: Using baseline plasma samples from the TOPCAT (Treatment of Preserved Cardiac Function Heart Failure With an Aldosterone Antagonist) trial, we measured the following KI biomarkers: cystatin-C, fatty acid-binding protein-3, Beta-2 microglobulin, neutrophil gelatinase-associated lipocalin, and kidney-injury molecule-1. Factor analysis was used to extract the common variability underlying these biomarkers. We assessed the relationship between the KI-factor score and the risk of death or HF-related hospital admission in models adjusted for the Meta-Analysis Global Group in Chronic Heart Failure risk score and eGFR. We also assessed the relationship between the KI factor score and ~5000 plasma proteins, followed by pathway analysis. We validated our findings among HFpEF participants in the Penn Heart Failure Study. KI was associated with the risk of death or HF-related hospital admission independent of the Meta-Analysis Global Group in Chronic Heart Failure risk score and eGFR. Both the risk score and eGFR were no longer associated with death or HF-related hospital admission after adjusting for the KI factor score. KI was predominantly associated with proteins and biologic pathways related to complement activation, inflammation, fibrosis, and cholesterol homeostasis. KI was associated with 140 proteins, which reproduced across cohorts. Findings regarding biologic associations and the prognostic significance of KI were also reproduced in the validation cohort.

Conclusions: KI is associated with adverse outcomes in HFpEF independent of baseline eGFR. Patients with HFpEF and KI exhibit a plasma proteomic signature indicative of complement activation, inflammation, fibrosis, and impaired cholesterol homeostasis.

Keywords: HFpEF; cardiorenal interactions; chronic kidney disease; kidney injury; outcomes; proteomics.

Figure 1. Hierarchical clustering and factor loadings maps.

A, Hierarchically clustered heatmap showing the association of CST‐C, FABP3, B2M, NGAL, and KIM1 with one another in TOPCAT. B, Factor loadings map showing the association of our factor with different kidney injury markers in TOPCAT. C, Hierarchically clustered heatmap showing the association of CST‐C, FABP3, B2M, NGAL, and KIM1 with one another in PHFS. D, Factor loadings map showing the association of our factor with different kidney injury markers in PHFS. B2M indicates Beta‐2 microglobulin; CST‐C, cystatin‐C; FABP3, fatty acid binding protein 3; KIM1, kidney injury molecule‐1; NGAL, neutrophil gelatinase‐associated lipocalin; PHFS, Penn Heart Failure Study; and TOPCAT, Treatment of Preserved Cardiac Function Heart Failure With an Aldosterone Antagonist.

Figure 2. Kaplan–Meier plots for death or heart failure‐related hospital admission across tertiles of the kidney injury factor score in (A) TOPCAT and (B) PHFS.

DHFA indicates death or heart failure‐related hospital admission; H.Tertile, higher tertile; L.Tertile, lower tertile; M.Tertile, Middle Tertile; PHFS, Penn Heart Failure Study; and TOPCAT, Treatment of Preserved Cardiac Function Heart Failure With an Aldosterone Antagonist.

Figure 3. Box and whisker charts of Cox proportional hazards regression models for death or heart failure‐related hospital admission of (A) nonadjusted KI factor in TOPCAT, (B) KI factor adjusted for MAGGIC risk score and eGFR in TOPCAT, (C) nonadjusted KI factor in PHFS, (D) KI factor adjusted for MAGGIC risk score and eGFR in PHFS.

eGFR indicates estimated glomerular filtration rate; KI, kidney injury; MAGGIC risk score, Meta‐Analysis Global Group in Chronic Heart Failure Risk Score; PH, proportional hazards; PHFS, Penn Heart Failure Study; and TOPCAT, Treatment of Preserved Cardiac Function Heart Failure With an Aldosterone Antagonist.

Figure 4. Volcano plots showing proteins that are significantly associated with the kidney injury factor score in TOPCAT (A) and PHFS (B).

The lower faint solid line represents a nominal alpha level of 0.05. The upper solid line represents an alpha level of 0.05 corrected for multiple comparisons. PHFS indicates Penn Heart Failure Study; and TOPCAT, Treatment of Preserved Cardiac Function Heart Failure With an Aldosterone Antagonist.

Figure 5. Pathways significantly correlated with the kidney injury factor score in TOPCAT (A) and PHFS (B).

Pathways in gray indicate a significant association with indeterminate direction. Blue bars indicate a significant positive association. Red bars indicate a significant negative association. Numbers at the end of the bars indicate the Z score corresponding to direction and strength of association. IL‐4 indicates interleukin‐4; LXR/RXR, liver‐X‐receptor/retinoid‐X‐receptor; NFAT, nuclear factor of activated T cells; PHFS, Penn Heart Failure Study; and TOPCAT, Treatment of Preserved Cardiac Function Heart Failure With an Aldosterone Antagonist.

Figure 6. Concordance plot showing the proteins that were concordant across TOPCAT and PHFS cohorts.

PHFS indicates Penn Heart Failure Study; and TOPCAT, Treatment of Preserved Cardiac Function Heart Failure With an Aldosterone Antagonist.