Clinical and plasma proteomic characterization of heart failure with supranormal left ventricular ejection fraction: An emerging entity of heart failure

Abstract

Aims: The clinical guidelines categorize heart failure (HF) based on left ventricular ejection fraction (LVEF). However, the current LVEF cutoffs, 40% and 50%, may not fully address the underlying characteristics and cardiovascular risk of HF, particularly for HF with higher LVEF. This study aimed to characterize HF with supranormal ejection fraction (HFsnEF) using different LVEF cutoffs (35%, 55%, and 70% for men, and 40%, 60%, and 75% for women).

Methods and results: This study divided 442 patients from the CHART-Omics study into four groups: HF with reduced ejection fraction (HFrEF) (n = 55, 65.5 years), HF with mildly reduced ejection fraction (HFmrEF) (n = 125, 69.3 years), HF with normal ejection fraction (HFnEF) (n = 215, 69.0 years) and HFsnEF (n = 47, 67.1 years). When clinical backgrounds were adjusted and HFnEF served as the reference, HFsnEF carried an increased hazard ratio (HR) for the composite of cardiovascular death and HF hospitalization of 2.71 (95% confidence interval [CI] 1.10-6.66, p = 0.030), while HFrEF had a HR of 3.14 (95% CI 1.36-7.23, p = 0.007). HFsnEF was characterized by an increase in relative left ventricular wall thickness and a decrease in left ventricular dimensions, whereas increased left ventricular mass and dimensions characterized HFrEF. Quantitative analysis of 4670 plasma proteins showed essential differences between HFsnEF and HFrEF, for example, 'protein synthesis' versus 'cell morphology', 'cellular assembly and organization' and 'nucleic acid metabolism' for underlying pathophysiology, and 'energy production' versus 'connective tissue disorders' and 'cell-to-cell signalling and interaction' for prognostication.

Conclusions: Heart failure with supranormal ejection fraction, an unnoticed but emerging entity in HF, carries a similarly increased cardiovascular risk as HFrEF but has unique structural and plasma proteomic characteristics.

Keywords: Cardiac structure; Cardiovascular risk; Classification; Heart failure; Left ventricular ejection fraction; Proteomics.

Figure 1

Patient selection, classification, and prognosis. (A) The patient selection and classification flow. (B) Distribution of the participants in each HF group in men (upper panel) and women (lower panel). (C) Incident curves of the first HF hospitalization or CV death for HFrEF, HFmrEF, HFnEF, and HFsnEF. (D) Forrest plots showing the impacts of HFrEF, HFmrEF, and HFsnEF on the first HF hospitalization or CV death when HFnEF served as the reference. (E) Incident curves of the first HF hospitalization or CV death for HFrEFcon, HFmrEFcon, HFnEFcon, and HFsnEFcon. (F) Forrest plots showing the impacts of conventionally used cutoffs of LVEF (HFrEF_con, HFmrEF_con, and HFsnEF_con) on the first HF hospitalization or CV death when HFnEF_con served as the reference. CV, cardiovascular; EF, ejection fraction; HF, heart failure; HFmrEF, heart failure with mildly reduced ejection fraction; HFnEF, heart failure with normal ejection fraction; HFrEF, heart failure with reduced ejection fraction; HFsnEF, heart failure with supranormal ejection fraction; HR, hazard ratio; LVEF, left ventricular ejection fraction.

Figure 2

Two‐ and three‐dimensional PC analysis of the spatial relationship of the plasma proteome profiles of the heart failure groups. (A) Two‐dimensional PC analysis plot with PC1 and PC2. (B) Two‐dimensional PCA plot with PC1 and PC3. (C) Two‐dimensional PCA plot with PC2 and PC3. (D) Three‐dimensional PCA plot for the centroids of HFrEF, HFmrEF, HFnEF, and HFsnEF. (E) Distances between each centroid of HFrEF, HFmrEF, HFnEF, and HFsnEF. HFmrEF, heart failure with mildly reduced ejection fraction; HFnEF, heart failure with normal ejection fraction; HFrEF, heart failure with reduced ejection fraction; HFsnEF, heart failure with supranormal ejection fraction; PC, principal component.

Figure 3

Canonical pathways and functions associated with underlying pathophysiology of heart failure. (A) Left panel: Canonical pathways and functions significantly modulated in plasma levels of HFrEF, HFmrEF, and HFsnEF as compared to HFnEF. Right panel: The green, red, and blue lines indicate p‐values 0.1, 0.05, and 0.01, respectively. The top 1, 8, and 7 pathways were statistically significant at p < 0.01 for HFrEF, HFmrEF, and HFsnEF as compared to HFnEF, respectively. (B) Heatmaps showing average standardized levels of plasma proteins involved in representative functions linked to HFrEF. Red arrows indicate the pathways activated. A blue arrow indicates the pathway inactivated. ※ (black), pathways shared between HFrEF and HFmrEF; ※ (red), a pathway shared between HFmrEF and HFsnEF; @, pathways no functions were indicated. HFmrEF, heart failure with mildly reduced ejection fraction; HFnEF, heart failure with normal ejection fraction; HFrEF, heart failure with reduced ejection fraction; HFsnEF, heart failure with supranormal ejection fraction.

Figure 4

Proteins associated with risk of HF hospitalization or CV death. (A) Volcano plots for proteins associated with first HF hospitalization or CV death in overall patients. (B) Top 10 proteins associated with risk of first HF hospitalization or CV death in overall patients. (C) Volcano plots for proteins associated with first HF hospitalization or CV death in each HF group. (D) Venn diagram showing overlaps among proteins associated with the first HF hospitalization or CV death across HFrEF, HFmrEF, HFnEF, and HFsnEF. (E) Forrest plots for the representative proteins that were specifically associated with the first HF hospitalization or CV death in HFsnEF, but not in HFrEF, HFmrEF, and HFnEF, showing a significant interaction with each of HFrEF, HFmrEF, and HFnEF. p values for interaction versus HFsnEF are shown for each heart failure group. CI, confidence interval; CV, cardiovascular; FDR, false discovery rate; HF, heart failure; HFmrEF, heart failure with mildly reduced ejection fraction; HFnEF, heart failure with normal ejection fraction; HFrEF, heart failure with reduced ejection fraction; HFsnEF, heart failure with supranormal ejection fraction; HR, hazard ratio.

Figure 5

Canonical pathways and functions associated with CV risk for respective HF groups. (A) Left panel: Canonical pathways and functions significantly modulated in plasma levels of HFrEF, HFmrEF, and HFsnEF as compared to HFnEF. Right panel: The green, red, and blue lines indicate p‐values 0.1, 0.05, and 0.01, respectively. The top 5, 2, 0, and 12 pathways were statistically significant at p < 0.01 for HFrEF, HFmrEF, HFnEF, and HFsnEF, respectively. (B) Heatmaps showing average standardized levels of plasma proteins involved in representative functions linked to HFrEF. Red arrows indicate the pathways activated. Blue arrows indicate the pathways inactivated. ※ (purple), pathways shared between HFrEF and HFmrEF; ※ (blue), a pathway shared between HFrEF and HFsnEF; ※ (red), a pathway shared between HFmrEF and HFsnEF; @, a pathway no functions were indicated. CV, cardiovascular; HF, heart failure; HFmrEF, heart failure with mildly reduced ejection fraction; HFnEF, heart failure with normal ejection fraction; HFrEF, heart failure with reduced ejection fraction; HFsnEF, heart failure with supranormal ejection fraction.