Historical perspective
One of the most informative early animal models of heart failure was developed in rats in which the left coronary artery undergoes surgical ligation, leading to myocardial infarction, ventricular remodeling, and systolic failure. This animal model was exploited by Pfeffer et al. in 1979 to explore the correlation between cardiac infarct size and myocardial dysfunction 24 . These studies led to the development of novel therapies using angiotensin converting enzyme inhibitors to prevent adverse ventricular remodeling, and had a major impact on heart failure therapy and saved many lives22-25 . Yet while this experimental MI model is an informative approach to investigate mechanisms of cardiac dysfunction and remodeling, a substantial fraction of the experimental animals may die from arrhythmia or cardiac rupture following coronary ligation, and the cardiac remodeling response in the surviving animals is variable and not consistent. Moreover, each animal requires an individual surgical intervention, which undermines even medium-throughput screening approaches. Another commonly used model of heart failure is the pressure-overload model, which can be used in either mice or rats. In this model, a ligature is tightened around the animal’s aorta, which leads first to left ventricular hypertrophy, followed by ventricular dilation and cardiac dysfunction. This model does mimic aspects of the heart failure progression observed in human patients26 , but even subtle differences in surgical techniques and between mouse strains can lead to different outcomes. With the growing incidence of cardiovascular diseases observed in patients with diabetes and obesity/metabolic syndrome, animal models of heart failure have been developed in which animals are exposed to specific diets (e.g. high-fat/high-sucrose feeding) in various models of type II diabetes mellitus. However, the cardiac dysfunction that is induced by these metabolic derangements may develop only after many months of treatment, and even then may often lead only to a subtle cardiac phenotype 8 . There are also many “knockout” mouse models in which individual genes that cause cardiomyopathy in humans have been targeted; these models have been quite informative, but the relevance of some of these monogenic cardiomyopathies to most forms of human heart failure is less clear2 . A recent report 31described a “two hit” approach in mice involving both a dietary treatment (high fat feeding) along with a pharmacological intervention (nitric oxide synthase inhibition) that led to the development of a novel heart failure phenotype. Each of the animal models of heart failure has its strengths and weaknesses, and of course no model system can faithfully recapitulate the complex pathways and protean manifestations of heart failure in humans. There remains an urgent need for new and more tractable heart failure models that recapitulate key features of human heart failure.