4. What implications do the present findings have for preclinical practice in HF?

In the past three decades, there have been few areas of medicine where significant progress has been made in the treatment of HF. However, progress was consistent only in chronic HF with reduced ejection fraction. In acute decompensated HF and HF with preserved ejection fraction, no treatment has been proven to improve survival. At present, several kinds of drugs on the market, such as ACE inhibitors, angiotensin II receptor blockers type I (ARB), beta blockers (BBS), aldosterone receptor blockers (ALRB), statins, as well as anticoagulant and anticoagulant therapy, have proved to be effective in achieving hard targets; however, the mortality reduction of most drugs is less than 30% and the residual cardiovascular risk is still high. Therefore, efforts to find new methods in the field of HF treatment have not stopped, and HF treatment drugs based on new mechanisms are constantly being discovered (Table 3 ). One of the most remarkable is the unexpected harvest of SGLT2 inhibitors in the treatment of HF. In May 2020, farxiga (empagliflozin) successfully passed the fast track qualification (FTD) and became the first SGLT2 inhibitor approved for the treatment of HF. In preclinical studies, the rate of cardiac remodeling in the non-diabetic pig model group with left anterior artery ischemia treated by empagliflozin decreased (Santos-Gallego et al., 2019). Although the precise mechanism of SGLT2 in treating HF is not clear, it put forward new demands for cardiovascular disease research, that is, to study on more complex disease models.

4.1 Study on animal model of HF induced by multiple stimulus

In view of the outstanding contribution of SGLT2 inhibitors in the treatment of heart failure, it suggests that these diseases have a common pathogenesis, However, many pharmaceutical companies and scientists currently focus on developing a single ”mechanism” therapy. As outlined in the previous section, in the past century, many research institutions and investigators worldwide have developed a large number of animal models and assessment methods, which cover different aspect that could resemble clinical characteristics HF in human. Therefore, the key to successful preclinical development in the future lies not in developing more HF models, but in modifying existing animal models to better imitate complex clinical situations. As initiating triggers for HF pathologies are highly diverse and the disease processes typically complex, each animal model of HF produced by single stimulus has obvious defects, so the combination strategy produced by multiple stimuli may be a better model to mimic disease conditions in human. There are already many ongoing efforts in this regard. Given the limitation of the reversible nature of rapid pacing model alone, Shen et al. combined rapid ventricular pacing with sequential coronary artery occlusion and banding aorta respectively, developed two multi-stimulus induced animal models (Y.-T. Shen et al., 2017). From which it was observed that banding aorta alone (which would cause left ventricular hypertrophy) did not reduce baseline cardiac function. But when myocardial ischemia or left ventricular hypertrophy is combined with pacing, decreased left ventricular dP/dt and increased left ventricular end diastolic pressure was observed. The combination of two different methods provides a better model to mimic the clinical phenotype of human heart failure, which would provide a unique opportunity for elucidation of cardiovascular disease mechanisms and also the evaluation of novel interventions. In some other studies, stimulus combined with transgenic animal models has become a common strategy for studying pathology and pharmacology of HF.

4.2 Highlighting the application of biomarker detection in animal model study

The frequency of positive Phase III clinical trials with investigational HF drugs has decreased over the past 20 years, while the rate of positive Phase II clinical trials with surrogate end points remains high. Although there are many compelling cases of biological plausibility, past experience strongly suggests that there is no reliable surrogate end point for the Phase III HF trial outcome (Greene et al., 2018). It has been found that the development of novel biomarkers may be more effective than the use of surrogate end points, and the evaluation of biomarkers is more repeatable and results more reliable in animal models than mimicking complex cardiovascular disease. Biomarkers can help detect the presence of heart failure, determine its severity, assess the risk of future events, and guide treatment, when in combination with clinical and physical assessments can provide greater diagnostic accuracy than physical assessments alone (Chow et al., 2017). Since the establishment of natriuretic peptide as a diagnostic criterion for heart failure in 2000, the evaluation of biomarkers has been included in a large number of preclinical studies (Januzzi & Felker, 2013). However, defining a “rule-in” cutoffs for HF is complicated because multiple factors influence natriuretic peptide levels for heart failure is complex because of a variety of factors that affect natriuretic peptide levels. Thus, other potential strategy explored a multimarker approach for predicting new-onset HF. In the subset of patients with high baseline risk determined by clinical parameters, the best models for predicting new onset HF include N-terminal pro-B-type natriuretic peptide (NT-proBNP), troponin T (TnT), and urinary albumin excretion(Chow et al., 2017). Introducing these indicators of biomarkers into animal model research not only can be used to distinguish the subtypes of HF of animal models with different clinical features, but also can help evaluate the pharmacological efficacy studies of drugs by combining the results from physiological and echocardiography parameters. Indeed, biomarkers have been used in association with many preclinical studies that have been completed or are ongoing (Table 2 ). Recently, a novel study demonstrated an increase in NT-proBNP levels in the dog I/R model, in contrast, LV mechanical unloading by the total support of transvascular LV assist device Impella could reduce LV end-diastolic pressure, increase LV end-systolic elastance and decrease NT-proBNP level, thereby preventing subsequent HF (Saku et al., 2018).