Chronic high-rate pacing induces heart failure with preserved ejection fraction-like phenotype in Ossabaw swine

Abstract

The lack of pre-clinical large animal models of heart failure with preserved ejection fraction (HFpEF) remains a growing, yet unmet obstacle to improving understanding of this complex condition. We examined whether chronic cardiometabolic stress in Ossabaw swine, which possess a genetic propensity for obesity and cardiovascular complications, produces an HFpEF-like phenotype. Swine were fed standard chow (lean; n = 13) or an excess calorie, high-fat, high-fructose diet (obese; n = 16) for ~ 18 weeks with lean (n = 5) and obese (n = 8) swine subjected to right ventricular pacing (180 beats/min for ~ 4 weeks) to induce heart failure (HF). Baseline blood pressure, heart rate, LV end-diastolic volume, and ejection fraction were similar between groups. High-rate pacing increased LV end-diastolic pressure from ~ 11 ± 1 mmHg in lean and obese swine to ~ 26 ± 2 mmHg in lean HF and obese HF swine. Regression analyses revealed an upward shift in LV diastolic pressure vs. diastolic volume in paced swine that was associated with an ~ twofold increase in myocardial fibrosis and an ~ 50% reduction in myocardial capillary density. Hemodynamic responses to graded hemorrhage revealed an ~ 40% decrease in the chronotropic response to reductions in blood pressure in lean HF and obese HF swine without appreciable changes in myocardial oxygen delivery or transmural perfusion. These findings support that high-rate ventricular pacing of lean and obese Ossabaw swine initiates underlying cardiac remodeling accompanied by elevated LV filling pressures with normal ejection fraction. This distinct pre-clinical tool provides a unique platform for further mechanistic and therapeutic studies of this highly complex syndrome.

Keywords: Cardiac function; Coronary blood flow; Hemorrhage; Obesity.

Fig 1.

Representative images of transverse sections of hearts and baseline pressure-volume relationships from Lean Control and Chronic Paced swine (a). Individual and average data of LV end diastolic volume (b), stroke volume (c), ejection fraction (d), and LV end diastolic pressure (e) from Lean (n = 8), Lean HF (n = 5), Obese (n = 8), and Obese HF (n = 8). Yellow scale bar = 1 cm. P value provided for overall group comparison between control vs. paced swine. *P < 0.05 versus Lean, †P < 0.05 versus Obese.

Fig 2.

Representative images (a) of cardiac interstitial fibrosis (picrosirius red) and immunofluorescent staining (wheat-germ agglutinin) of myocardium from lean (n = 5–8), lean HF (n = 5) obese (n = 5), and obese HF (n = 5). Staining for cross-sectional area (red), isolectin staining for capillaries (green), and DAPI nuclear stain (blue); scale bar = 20 μm. Individual and average data of interstitial fibrosis (b), myocardial capillary density (c), and myocyte cross-sectional area (d). P value provided for overall group comparison between control vs. paced swine. *P < 0.05 versus Lean, †P < 0.05 versus Obese.

Fig 3.

Representative traces of LV pressure over time in Lean vs. Lean HF (a) and Obese and Obese HF (b) swine. LV pressure at the beginning (B) and end (E) of diastole in Lean (n = 8) vs. Lean HF (n = 5) (c) and in Obese (n = 8) vs. Obese HF (n = 8) (d). Relationship between LV pressure and LV volume at the beginning and end of the diastolic filling period (e) and LV stroke volume and LV end diastolic volume (f). Regression fits are represented by: Lean = solid turquoise line; Lean HF = solid navy line; Obese = dashed turquoise line; Obese HF = dashed navy line. Shared fits between groups are denoted by side by side regression lines. *P < 0.05 versus Lean, †P < 0.05 versus Obese.

Fig 4.

Average hemodynamic and metabolic data at rest and in response to acute hemorrhage in Lean (n = 6), Lean HF (n = 4), and Obese HF (n = 5) swine. Effect of reductions in blood pressure on heart rate (a), cardiac index (b), LV end diastolic pressure (c), Pressure Work Index (d), and MVO₂ (e). Assessment of cardiac efficiency in the relationship between Pressure Work Index and MVO₂ (f). Regression fits are represented by: Lean = solid turquoise line; Lean HF = solid navy line; Obese = dashed turquoise line; Obese HF = dashed navy line. Shared fits between groups are denoted by side by side regression lines.

Fig 5.

Relationship between LV end diastolic pressure and end diastolic volume at rest and in response to hemorrhage in Lean (n = 6), Lean HF (n = 4), Obese HF (n = 5) swine. Regression fits are represented by: Lean = solid turquoise line; Lean HF = solid navy line; Obese = dashed turquoise line; Obese HF = dashed navy line. Shared fits between groups are denoted by side by side regression lines.

Fig 6.

Effect of reductions in blood pressure on myocardial oxygen delivery (a), coronary venous PO₂ (b), and myocardial lactate uptake (c) in Lean (n = 6), Lean HF (n = 4), and Obese HF (n = 5) swine. Relationship between myocardial oxygen delivery (d), coronary venous PO₂ (e), and myocardial lactate uptake (f) vs. MVO₂ in Lean (n = 6), Lean HF (n = 4), and Obese HF (n = 5) swine. Regression fits are represented by: Lean = solid turquoise line; Lean HF = solid navy line; Obese = dashed turquoise line; Obese HF = dashed navy line. Shared fits between groups are denoted by side by side regression lines.

Fig 7.

Effect of reductions in blood pressure on Endocardial to Epicardial flow ratio (Endo:Epi) in the left anterior descending coronary artery (LAD) LV perfusion territory (a), in the left circumflex coronary artery LV perfusion territory (b), and in the RV free wall (c) in Lean (n = 6), Lean HF (n = 4), and Obese HF (n = 5) swine. Regression fits are represented by: Lean = solid turquoise line; Lean HF = solid navy line; Obese = dashed turquoise line; Obese HF = dashed navy line. Shared fits between groups are denoted by side by side regression lines.