AHA Scientific Sessions 2026 Abstract · Nov 2, 2026
Phenotypic and Transcriptomic Evaluation of Rodent Heart Failure with Preserved Ejection Fraction Models to Improve Predictive Validity in Drug Discovery
By Pharaoh G, Muraoka N, Berkman T et al.
Background: Developing effective heart failure with preserved ejection fraction (HFpEF) therapeutics requires preclinical models with high translational fidelity. While ZSF1 obese rats and mice treated with high-fat diet and L-NAME (HFD+L-NAME) are widely used, their alignment with human pathology is poorly defined, driving high failure rates in clinical translation. Hypothesis: Comparative profiling and response to standard of care will expose HFpEF model-specific variations and identify models most translatable to human disease. Methods: Three cardio-metabolic HFpEF models were characterized phenotypically and transcriptomically compared to age-matched controls (WKY/ZSF1 lean rats, healthy mice) and human transcriptomic data: 1. Young (10-week-old) male C57BL/6N mice (HFD+L-NAME, treated up to 40 weeks). 2. Aged (17-month-old) male and female C57BL/6N mice (HFD+L-NAME, treated up to 40 weeks). 3. Male and female ZSF1 obese rats (12-26 weeks old). ZSF1 Obese rats received standard-of-care (SoC) therapies: empagliflozin (empa; 3-30 mg/kg/day), sacubitril/valsartan (68 mg/kg/day), or finerenone (10 mg/kg/day). Characterization included echocardiography, invasive hemodynamics, exercise tolerance, plasma NT-proBNP, blood glucose, and left ventricular bulk or single-nuclei RNA-sequencing. Results: All models developed obesity, hypertension, cardiac hypertrophy, exercise intolerance, elevated NT-proBNP, and diastolic dysfunction (elevated LV end-diastolic pressure [EDP], prolonged tau, increased isovolumic relaxation time [IVRT]). Ejection fraction (EF) was preserved except in aged male mice treated with HFD+L-NAME, where it was mildly reduced. Male ZSF1 obese rats displayed the most prominent hyperglycemia. The transcriptomic shift of human HFpEF patients was most closely captured by ZSF1 obese rats, followed by aged HFD+L-NAME mice, then young HFD+L-NAME mice. In ZSF1 obese rats, empa increased glycosuria in both sexes but decreased fasting blood glucose only in males. Empa maintained EF, reduced LV mass, and improved IVRT in both sexes. This therapeutic response was more pronounced in males, with greater reductions in fasting glucose and LV mass alongside improved diastolic function (tau and LV EDP). Exercise tolerance and blood pressure were unchanged suggesting hypertrophy attenuation was not driven by reduction in afterload. Consequently, male ZSF1 obese rats were selected for subsequent evaluation with sacubitril/valsartan and finerenone. Conclusions: Rodent HFpEF models display distinct age- and sex-specific phenotypes and transcriptomic shifts. The ZSF1 obese rat provides a robust metabolic-HFpEF phenotype with the greatest overlap in transcriptomic shift with human HFpEF patients and HFpEF SoC responsiveness. Overall, male ZSF1 rats demonstrate the greatest potential for predictive validity in clinical translation.