Introduction

Definition and importance of heart failure

Heart failure (HF) is a clinical syndrome characterized by the inability to send sufficient blood to the peripheral tissue as much as its metabolic requirement due to pathologies arising from the function or structure of the heart tissue, and it usually has a progressive course [1,2,3]. The complexity of HF has been challenging the scientific world for many years. Versatile and complex processes in the pathophysiology of HF have been investigated for a long time. Also, the heterogeneous structures of the disease negatively affect the speed of scientific developments in this field. The prevalence of HF continues to increase significantly today.

Prevalence and incidence

Unfortunately, HF is a leading cause of death worldwide [4, 5]. It affects more than 26 million people worldwide, with a prevalence of more than 12% over the age of 70 [6]. Although HF is generally seen in the elderly, it occurs at all ages [7, 8]. Currently, 6.5 million people in the USA are thought to have HF, and it is predicted that 8 million people in the USA will be diagnosed with HF by 2030 [8, 9]. HF is the leading cause of death worldwide, with approximately 50% of patients dying within the first 5 years after diagnosis [10]. HF not only is affected human health but also dramatically affects the social economy. It costs several billion dollars each year and is estimated to increase by around 70 billion costs by 2030 [10]. It is essential to analyze risk factors and develop new effective treatment methods for HF patients for all these reasons.

Pathophysiology, subclassification, and symptoms

The most important defining feature of HF is the heart’s inability to pump enough blood to the body and the resulting low quality of life in these patients [11]. HF is a complex clinical picture with many different etiologies. The main feature of HF has reduced contractile force or underfilling. Systemic inflammation, hypoxic environment, cardiomyocyte damage, mechanical stress, and other profibrotic cytokines, transforming fibroblasts into myofibroblasts at the injury site, are the essential pathophysiological processes underlying heart diseases [12]. Myofibroblasts have a more contractile structure by developing stress fibers. These induce cardiomyocyte hypertrophy by causing cytokines’ secretion, such as transforming growth factor-β, tumor necrosis factor-α, and angiotensin II [9, 13,14,15,16]. As a result of myofibroblasts’ activity, it also causes the formation of extracellular connective tissue and the formation of interstitial and perivascular fibrosis [17]. This fibrotic structure adversely affects the heart tissue’s physiological processes and causes myocardial stiffening and diastolic dysfunction [18]. These changes cause left ventricular hypertrophy, inhibition of the left ventricle’s relaxation ability, and diastolic dysfunction [19]. Developments such as ischemic damage in the heart lead to cardiomyocyte death, leading to decreased contraction of heart tissue and wall thickness [20]. As a result of these damages, the heart tissue’s cytosolic functions are impaired [21]. Pathological changes such as increased stiffness of the myocardial tissue reduced left ventricular contraction and heart tissue remodeling are the main hallmarks of HF [21]. The pathophysiology of systolic and diastolic heart failure and the differences between them are shown in Fig. 1. Left HF is often the leading cause of right HF and is associated with an increased risk of sudden death [22].

Fig. 1
figure 1

The pathophysiology of systolic and diastolic heart failure and the differences between them

HF is further divided into three subgroups: (1) HF with preserved EF (HFpEF, LV EF ≥ 50%), (2) HF with midrange EF (HFmrEF, 40–49% LV EF), (3) HF with reduced EF (HFrEF, LV EF < 40%). This sub-classification has an essential place in the difference in HF treatment protocols [23]. According to the analysis of the data obtained from the Framingham Heart Study, 56.2% of the patients were diagnosed with HFpEF, 31.1% with HFrEF, and 12.8% with HfmrEF [24]. Although the relative proportion of HFrEF decreases, the absolute number of HFrEF patients is predicted to increase in the coming years. HFpEF occurs when the ventricles fail to relax correctly and are clinically defined as normal EF or HF with diastolic dysfunction. About half of HF patients suffer from HFpEF and exhibit HF symptoms, including exercise intolerance, congestion, and edema associated with cardiac hypertrophy [25, 26]. HFpEF is more common in women than men and its incidence increases in older age [27]. In addition, it can be a secondary disease that occurs as a result of various chronic diseases such as hypertension and diabetes mellitus [9, 28].

As a result of clinical studies, it is known that there is a transition between HF subtypes. According to this, including HFrEF to HFmrEF/HFpEF or HFmrEF to HFpEF, it is referred to as “HF with enhanced LVEF” (HFiEF). These patients have transitioned to the lower HF phenotype from baseline. Including HFpEF to HFmrEF/HFrEF or HFmrEF to HFrEF, it is referred to as “HF with worsened LVEF” (HFdEF). The remainder of the enrolled patients were termed “HFuEF with unchanged LVEF” (HFuEF) [29]. Age, hypertensive and in some cases diabetes-related ventricular remodeling thus creates the slowly progressive substrate upon which HFpEF is formed, and recent evidence suggests that progression of a number of abnormalities in cardiovascular function may promote the transition to overt HFpEF, including loss of contractile reserve, diastolic reserve, chronotropy, vasodilation, and endothelial function. In contrast, HFrEF most commonly develops in response to distinct pathophysiologic perturbations leading to accelerated and larger-scale myocyte loss/dysfunction, with the most common etiologies including acute myocardial infarction, genetic abnormalities, myocarditis, or toxin effects (e.g., alcohol or chemotherapy) [30].

Two-thirds of all HF patients fall under four main headings: ischemic heart disease, chronic obstructive pulmonary disease, hypertensive heart disease, or rheumatic heart disease [30]. Typical signs of HF and physical examination findings include dyspnea, weight gain, weakness, increased jugular venous pressure, pulmonary rales, and peripheral edema [31]. Symptoms of right and left-sided HF are shown in Fig. 2. These findings significantly affect the expected patient’s quality of life [32]. All these symptoms and signs occur due to increased cardiac hypertrophy, formation of fibrotic tissue, and decreased blood supply. Many cardiac problems, such as coronary artery disease, myocardial infarction, and cardiomyopathy cause HF [30]. HF causes the loss of the heart tissue’s physiological contractile ability by causing cardiomyocyte damage or death. All these pathological changes cause the insufficient pumping of blood to the systemic circulation, in other words, the emergence of systolic HF. This relationship showed that brain natriuretic peptide levels and mortality are higher in HFrEF patients than in HFpEF patients. The relative incidence of these three subgroups, a decrease in HFrEF, and an increase in HFpEF have been observed in recent years [27, 33].

Fig. 2
figure 2

Symptoms of right and left-sided HF

Possible therapeutic targets of HF

Despite significant improvements in cardiovascular mortality over the last decades, cardiovascular disease is the main reason for death in several countries. Cardiovascular therapies improved the survival of patients with cardiovascular disease but, at the same time, increased the number of subjects affected by chronic cardiovascular conditions such as HF. The development of new drugs for HF therapy must be necessarily focused on additional targets [34]. Possible therapeutic pathways and mechanisms in HF are shown in Table 1.

Table 1 New possible pharmacological targets and promising drugs for the treatment of HF

Experimental models of heart failure

It is well known that there are many risk factors associated with HF. Apart from risk factors, comorbidities such as obesity, hypertension, type 2 diabetes, and chronic kidney disease have revealed the necessity of developing new treatment options in HF [28, 62,63,64]. The pathogenesis of HFpEF and HFrEF is multifactorial. Therefore, it is complicated to distinguish underlying mechanisms that may be overlapping and interconnected. The complex mechanisms of all these variables make the disease difficult to understand and discover new treatment methods. One of these difficulties is the lack of ideal animal models similar to the pathophysiological features of human HF.

Various experimental animal models are used to analyze the causes of HF and to develop current treatment strategies [65]. Many small animal species are often preferred for this purpose, including mice, rats, and guinea pigs. Mice and rats are very similar to the human genome, with 30,000 protein-coding genes. The most important advantages of these experimental animal species are the short reproductive cycles and low housing costs. These models generally use genetic modifications and pharmacological and surgical approaches. Many animal models have provided significant progress in knowledge of HFrEF and HFpEF pathogenesis [26]. The HF models in small animals are shown in Fig. 3.

Fig. 3
figure 3

The HF models in small animals

Myocardial infarction (MI)

Ischemia/anoxia occurs when a part of the myocardium is not provided with adequate blood flow or if it is completely stopped due to coronary occlusion [66, 67]. Insufficient blood flow to the myocardium is one of the leading causes of HF [68,69,70].

Several animal models have been developed to mimic HF in humans with coronary artery disease [71, 72]. These models can help elucidate the pathophysiological mechanisms occurring in ischemic human hearts. The main model types used are coronary artery ligation, coronary artery embolization, hydraulic occluder or ameroid ring constrictor, and cryoinjury models. These interventions are generally used to narrow or occlude coronary vessels. This blockage can be used to trigger acute or chronic HF [71, 73]. Coronary artery ligation is the most commonly used model used to induce HF in many animals, from mice to pigs [74,75,76]. Coronary artery ligation has been performed not only in rats but also in mice [77, 78]. Since ischemic heart disease is the most important cause of HF in humans, coronary artery occlusion is the most common inducing acute myocardial injury in animal models.

Ischemic heart disease is the main cause of HF in humans [79]. Ligation of the left anterior descending (LAD) coronary artery or any of its branches is the most preferred and generally accepted method to induce HF [8, 16, 80, 81]. Coronary artery ligation is a widely used small animal HF model created by Pfeffer et al. in rats and later used by numerous groups [82]. After discovering the experimental model, Pfeffer et al. showed that the use of angiotensin-converting enzyme inhibitor (ACEI), captopril, increased the contractile function and survival of the ventricles after MI in rats [83]. These experimental studies laid the foundation for ACEIs traditionally used in MI patients today. This method also provides information about cardiac remodeling developed in the heart after MI and its pathophysiology [84]. Pathological changes occur due to the increased overload in the cardiac tissue and the increased fibrotic structure in the extracellular tissue [8, 85]. In addition, this condition is accompanied by neurohumoral activation [86]. Dilatation, necrosis, and apoptosis were observed in the cardiac tissue of the animals in which this experimental model was applied. Also, high ventricular filling pressures and decreased cardiac index were found. In addition to all these changes, the natriuretic peptide and renin-angiotensin system were activated [87, 88]. HF induced by LAD ligation is associated with myocardial hypertrophy, progressive myocardial enlargement, and late failure if 40–50% of the left ventricle is viable [14, 89]. There was no significant difference in cardiac output in rats 1 week after LAD ligation compared to the control group. However, a significant reduction in cardiac output was detected 3–5 weeks after LAD ligation [90]. A significant reduction in rats’ cardiac output, up to 25%, is observed 8 weeks after LAD ligation [91]. It is known that mice that survive progressively develop HF 4 weeks after surgery [92]. Infarction size varies between 10 and 45% and is directly related to left ventricular function damage; this ratio affects HF’s development time. In general, the infarct must affect at least 30% of the left ventricular mass to demonstrate the typical HF features and detect various biomarker levels [81].

Studies have shown that female mice undergo less ventricular remodeling. Sex hormones are thought to underlie this difference. Female mice show less extensive ventricular remodeling than males, suggesting the role of sex hormones [93]. These differences should be taken into account when designing experiments. It must be taken into attention. Due to the MI experimental model in both mice and rats, the mortality rate ranges from 35 to 50%. Ventricular fibrillation usually occurs within 1 h after MI due to severe acute HF [94]. Although progressive, non-occlusive coronary artery occlusion is usually seen in clinical patients; this model is mostly caused by a normal coronary artery’s sudden occlusion. Ischemic heart disease in humans is mostly caused by non-occlusive coronary artery narrowing. This experimental model is created by reducing the left coronary artery’s inner lumen diameter by about 60% by occluding the vessel with a probe and then removing it in rats [95]. As a result of the decrease in coronary blood flow, a reduction in left ventricular performance and deterioration in HF are observed. Ischemic HF develops more slowly than this model of total occlusion, and the overall mortality rate is approximately 43%. Although coronary artery ligation is a reliable model for inducing tissue damage that leads to HF, it does not reflect HF development in patients. In sum, the coronary artery ligation-induced coronary artery test model can be established in a robust, reproducible manner in many species [96,97,98].

Temporary LAD occlusion models have been developed to induce human ischemia–reperfusion (I/R) injury. However, results vary due to the variability observed in the left coronary anatomical structure in mice [99]. The coronary artery is ligated for 180 min and then reperfused in the I/R method. Approximately 16% infarct area and an LVEF of roughly 40% were formed in this method parallel with increased LVEDP, LVESV, and N-terminal pro-BNP after 4 weeks from MI [100]. In a different version of these models, a temporary occlusion is achieved by imitating I/R, whereby the flow in the previously occluded coronary artery bed is allowed to recover [99]. Another method used to induce myocardial ischemia/infarction is coronary artery embolization. For this purpose, intracoronary embolizations, thrombin, and fibrinogen are used [101]. Coronary microembolization has been used extensively for a while and causes a decrease in ejection fraction (EF), an increase in LVEDP, and an increase in plasma norepinephrine levels [102]. This model causes a local infarction and global ischemia as a result of the development of the aneurysm. MI is created by serial injections of 90-micron diameter polystyrene microspheres in this model’s left coronary artery. Injections are made with a catheter inserted through the femoral artery [103]. Approximately 20,000 microspheres are injected into the artery with each injection. A total of 3 injections are made at 15-min intervals and repeated every week until the desired effect is observed. It is known that a total of 4–14 injections are usually required to obtain a chronic HF model. Injections are stopped when the ejection fraction (EF) is reduced to 35%. This microembolization-induced myocardial ischemia is an irreversible model of HF [104]. A study showed that this sheep model was stable for 6 months, and the mortality rate was shown to be between 30 and 50% [103]. The microembolization model is typically only applied to large animals to cause HF [105, 106].

This model has several advantages:

  1. 1.

    Lower risk of severe inflammatory complications than thoracotomy

  2. 2.

    This model is similar to the clinical situation in HF

However, it is difficult to understand the exact length and location of coronary artery occlusion. Arrhythmias during the experimental model may cause difficulties in interpreting the biological response [107]. It seems that I/R models were created to overcome this limitation, and transient LAD occlusion facilitated the investigation of molecular mechanisms and tissue damage [108,109,110]. It has been shown that ischemic injury in newborn mice is triggered by LAD ligation and complete recovery occurs after 3 weeks. The regenerative potential is reduced in aged mice due to the age-related decrease in the number of healthy cardiomyocyte cells [111]. Interestingly, in another study, rats were studied 4, 8, and 16 weeks after coronary artery ligation. According to the data obtained, hemodynamic and clinical signs indicated cup failure in only 16-week-old rats [112]. Essential advantages of LAD ligation include the cost and simplicity of the process. In addition, electrophysiology, coronary anatomy, physiology, and MI developmental processes in rodents differ when compared to humans. However, this experimental model is considered a significant disadvantage since some studies have shown that the mortality associated with this experimental model can reach up to 70% [81]. Despite its widespread use, there are significant drawbacks to this intervention [107]:

  1. 1.

    Infarctions seen in cases are relatively small (left ventricular mean ~ 21%).

  2. 2.

    Major hemodynamic changes are not seen due to a large number of pericardial collaterals.

  3. 3.

    Mortality rates of more than 50% are seen due to ventricular tachycardia.

Recent studies have shown that the branching pattern of coronary arteries in Lewis congenital rats is more consistent than in other strains. LAD ligation in Lewis inbred rats has been shown to have a lower mortality rate despite having a more extensive infarction than Sprague Dawley rats [95]. However, it should be kept in mind that the clinical manifestations of chronic HF and neurohumoral changes were not evaluated in this study. Therefore, this promising model needs to be studied in more detail. Studies show that recovery and termination processes begin faster in rats than in humans, which requires careful interpretation of the results. Another alternative MI model is the cryoinfarction, which causes damage to the epicardium of mice and rats [113]. However, this model has not attracted much attention from the scientific community and is no longer preferred. As a result of the study, it has been shown that myocardium wall thickness and fibrotic tissue due to infarction are less pronounced in mice than in rats [114].

Another method used to induce MI involves the use of a hydraulic occluder or ameroid ring constrictor. Procedures begin with a left anterolateral thoracotomy followed by a pericardial incision, while the left coronary artery branch is exposed and the hydraulic occluder is inserted. The occlusive is then inflated to cause partial stenosis or complete obstruction. An ultrasonic flow probe is positioned distal to the occluder to manage the degree of occlusion and record downstream flow from the left coronary artery [115]. Similarly, an ameroid ring constrictor is implanted. It will gradually shrink due to the peculiarity of the material. As a result of this method, complete or partial narrowing will occur in the coronary artery [116]. One of the essential factors in choosing a model is the variation in coronary variability and collateral vasculature between animals. Hedström et al. showed that in the region at risk, 50% infarction development time varies significantly between species [117]. Additionally, the type of anesthesia must also be considered when choosing a model affecting I/R [118]. HF model due to MI can be created in rats with the cryosurgery method. After intercostal thoracotomy, a 0.18 1.2-cm2 liquid nitrogen probe is applied to the left ventricular free wall 15 times for 20 s. Also, cryoinjury may not always induce a transmural lesion, and it should be noted that in this case, it can heal without fibrotic tissue forming and an aneurysm developing [119]. The cryoinjury model is reproducible and generally has a moderate resemblance to infarctions observed in the clinic. It has some disadvantages as with other surgical models because they are presented more acutely than clinically [120]. As a result, numerous animal models allow us to examine the effect of a drug in the field of HF. A model’s advantages must be carefully compared against disadvantages, and non-scientific factors such as cost, effort, and time must also be considered.

Chronic rapid cardiac pacing

Dilated cardiomyopathy caused by chronic tachycardia is a known clinical picture [121]. Whipple et al. showed that atrial pacing above 330 beats/min can induce physical signs of HF and cardiomyopathy due to experimental tachycardia that was first shown in 1962 [122]. The experimental model is simple and neurohumoral activation develops after about 8 weeks. Increased left ventricular filling pressures after 1–2 months of pacing are associated with pulmonary wedge pressure and right atrial pressure, ascites, pulmonary congestion, and reduced EF [123]. After pacing cessation, hemodynamic changes usually correct within 4 weeks. This is a unique feature of this model [124]. The interaction between HF and arrhythmia is well established to increase the risk of developing HF and morbidity or mortality [125]. Sudden cardiac death is one of the major causes of episodes in HF [126]. This situation can develop after arrhythmia. For example, arrhythmias such as atrial fibrillation significantly increase the risk of HF. Experimental models of arrhythmia-induced HF are characterized by chronic rapid pacing periods indicated by the pacemaker’s anatomical location [127, 128]. In tachycardia models, the pacemaker is implanted in the right or left ventricle. The right ventricle was paced at a rate ranging from 180 to 240 bpm, causing congestive HF within 3 to 4 weeks [129]. While the right atrium was paced at a rate of 400 bpm, the left and right ventricles were dilated without hypertrophy [130]. In dogs, pigs, and sheep, 128–130 rapid pacing of the atrium or ventricle for at least 3–4 weeks is known to cause progressive HF and is partially reversible upon cessation of stimulation [131].

According to experimental studies on dogs, it is known that atrial pacing above 330 beats/minute can cause symptoms of HF. In general terms, this model has been shown in creatures such as dogs, pigs, sheep, and rabbits [132]. It is known that the pacing-induced HF model is also suitable for rodents. The rapid right ventricular pacing model has been shown to cause a progressive biventricular dilatation, including decreased cardiac output and increased peripheral vascular resistance and neurohumoral activation within a few weeks [133]. Tachycardia causes a decrease in systolic, diastolic function and cardiac output. In addition, left ventricular end-diastolic pressure, mean arterial pressure, pulmonary artery pressure, and wall stress increase in this model approximately 24 h later. Ventricular dysfunction is seen after 3–5 weeks and results in HF. However, no significant changes are observed in the structure, physiology, and hypertrophy of the left ventricle. Indeed, cardiac widening is accompanied by very little cardiac hypertrophy [134, 135]. Therefore, the chronic pacing tachycardia experimental model has been used to generate dilated cardiomyopathy and chronic HF.

In the neurohumoral activity, early sympathetic activation and reduced parasympathetic activity are seen. Atrial natriuretic peptide and brain natriuretic peptide levels in plasma increase before damage to the left ventricle. This was thought to be a protective mechanism, but exogenous atrial natriuretic peptide administration did not improve hemodynamic and renal responses [136]. It has been shown that cytokines such as endothelin-1 and TNF-α increase in plasma in HF induced by this model [137]. HF created with this model develops due to left ventricular dysfunction similar to HF in humans. Accordingly, catecholamines’ plasma levels first increase in the early period and then reach the plateau [137]. Increases in plasma endothelin levels, beta-receptor density changes, and function are similar to patients with HF [138]. However, increased levels of endothelin and renin in plasma are seen in advanced stages of left ventricular dysfunction [139]. This experimental model has some essential features in terms of being similar to HF. First, major surgical traumas such as thoracotomy and pericardiectomy that may affect hemodynamics are not required. Second, HF occurs in as little as a few weeks in this experimental model, allowing sequential observations. Third, calibration can be achieved using a pacemaker to stimulate the heart rate. Fourth, rapid pacing leads to the emergence of well-defined signs of biventricular failure with cardiomegaly, hypoperfusion, pulmonary congestion, cachexia, and ascites. Finally, the HF model created by this experimental model is reversible [136, 140].

This experimental model is characterized by increasing cardiac volume with fluid retention, increasing catecholamines, atrial natriuretic peptide, renin-angiotensin, aldosterone, endothelin-1, and TNF-α [141]. Due to chronic tachycardia, calcium abnormalities, and deterioration of the extracellular matrix through the activation of myocardial matrix, metalloproteases, gelatinase, and other cytokines are seen [142,143,144,145]. After pacing for 1 week, apoptosis is seen due to myocyte damage. Both the heart’s systolic and diastolic function recovered within approximately 2–3 weeks after the cessation of stimulation [71].

As a result of all these reasons, plasma markers of neurohormonal activation in pacing models are very similar to patient populations. In terms of calcium balance, the sarcolemmal Na/Ca exchanger is known to increase, while SERCA levels are lower [141, 145, 146]. As a result of the analyses performed, it was shown that there was a significant loss in myocyte count, and this was thought to be the main component of cardiomyopathy. Increased p53 DNA binding activity to the Bax promoter, increased Bax protein expression, and decreased Bcl-2 have been shown in pacing-induced HF models [147].

The most important advantages of the model are its predictability, reproducibility, and resemblance to the hemodynamic and mechanical phenotype of HF due to dilated cardiomyopathy in humans. One of the essential advantages of this model is that the model is simple and requires simple instrumentation. This HF model closely resembles human HF due to dilated cardiomyopathy in terms of mechanical, structural, neurohormonal, and myocyte functional changes [11, 131]. However, the return of mechanical and neurohormonal changes within a few days after pacing is stopped is a limitation of this model. Limitations include the absence of myocardial hypertrophy and fibrosis and the reversibility of this myopathy [148]. Approximately 48 h after pacing is stopped, hemodynamic variables return to normal, and left ventricular EF normalizes after 1–2 weeks. During the first hours after stopping pacing, the circulating atrial natriuretic peptide level drops by 60%. Unlike hemodynamic dysfunction and neurohormonal activation, ventricular dilation continues even after pacing is terminated. Another limitation is that events due to mechanical and electrical dyshomogeneity of the heart due to the homogeneity of myocardial damage induced by rapid pacing can not be shown in this model. Unlike in humans, hypertrophy does not occur in the HF model created by chronic rapid pacing. The absence of fibrotic tissue contributes to ventricular remodeling. Therefore, it may not reflect all the features of HF [132]. The enlargement of both ventricles and mitral valve annulus shown in the experimental model has been associated with valve insufficiency [129].

Rapid pacing models have been used to evaluate intracellular and extracellular changes and develop new pharmacological treatment methods. Rapid pacing models have also been used to assess the different surgical procedures required for HF [149, 150]. It is predicted that this experimental HF model, which is used in many different animal species, may help to elucidate the mechanisms related to HF in the future.

Pressure overload

Chronic left ventricular pressure-overload is an experimental model in mice and rats that causes HF that mimics adaptations associated with hypertension and aortic valve stenosis in patients [151,152,153,154]. Although the pressure-overload model is used more in rodent studies, they are less used in large animal experiments than ischemic models [72, 155, 156]. Similar to rodents, methods such as aortic taping can also be applied to large animals [157]. Aortic constriction causes an increase in pressure in the left ventricle and hypertrophy. While banding does not initially significantly affect the contraction, gradually, the constriction’s relative intensity increases as the animal grows, resulting in cardiac hypertrophy [155].

Overpressure-induced left ventricular hypertrophy and HF have been produced in rodents by transverse aortic constriction (TAC) and abdominal aortic constriction (AAC) [12, 158, 159]. Rats exposed to pressure-overload for more than 8 weeks have impaired systolic function, fractional shortening, and EF values. Significant increases in left ventricular posterior wall thickness, left ventricular inner size, and ventricular septal wall-size indicate concentric cardiac hypertrophy [10, 143]. A decrease in E, A wave, and E/A ratio on echocardiography indicates diastolic dysfunction [160]. All results show a significant increase in left ventricular end-diastolic dimension, left ventricular end-systolic dimension, and left ventricular posterior wall thickness. Also, a 60% reduction in the maximum pressure development rate (+dP/dt) indicates diastolic dysfunction [154]. Various surgical approaches have been developed to examine HF caused by chronic overpressure on the left ventricle and to mimic the protection mechanisms associated with hypertension in patients. These experimental models cause significant macroscopic and microscopic hypertrophy, but systolic dysfunction appears to progress much more slowly than rodent models [161].

The development of animal models with preserved EF is difficult due to the complexity of the pathophysiology of the disease [162, 163]. Pressure-overload is directly related to the development of left ventricular hypertrophy seen in HF. Myocardial hypertrophy is a protective mechanism that occurs due to increased myocardial cell tension [164]. Although it is beneficial, it may cause the myocardium to decrease its elasticity and become fibrotic after a certain level. Also, it gradually surpasses all post-loading protection mechanisms. Then, despite the left ventricle’s enlargement, the load reserve is depleted, and the basal contraction becomes incompatible with the post-load level [165].

Pressure-overload usually occurs in conditions such as left ventricular outlet obstruction such as hypertension or aortic stenosis. As a result of these experimental models, it has been shown that hypertrophy and fibrotic tissue develop in the myocardium of experimental animals. Also, impaired relaxation, increased left ventricular stiffness, and left atrial remodeling were also observed [166]. According to the Framingham study, left ventricular hypertrophy is seen in many serious diseases such as sudden cardiac death, HF, myocardial infarction, and stroke. Therefore, left ventricular hypertrophy is considered to be an independent cardiac risk factor. However, reverse remodeling with normalization of the left ventricular mass can be expected after treatment. Reverse remodeling is determined clinically as a result of echocardiographic studies [167]. Despite scientific advances, cellular mechanisms in myocardial tissue are not clearly known. The experimental model of pulmonary artery banding induces right ventricle overpressure. The supravalvular aortic stenosis model is widely used in rats.

New experimental models are being developed to induce metabolic diseases in large animals to induce HF with preserved ejection fraction [157]. Transverse aortic stenosis in mice was first described by Rockman et al. [168]. Today, it has been the most used experimental model to examine HF caused by excessive left ventricular pressure. The TAC model leads to concentric hypertrophy, interstitial fibrosis, increased left ventricular stiffness, and eventually systolic HF due to increased pressure in the left ventricle [169, 170]. The TAC model causes concentric cardiac hypertrophy and subsequent systolic HF by increasing left ventricular afterload. Evaluation of the TAC procedure and comparison with the control group is done by doppler [171]. The TAC model has enabled discovering the cause of many fundamental pathological changes seen in HF, especially left ventricular remodeling [159, 172, 173]. Among this model’s advantages are the lower mortality rate (5–10%) and the occurrence of left ventricular hypertrophy [132]. Another advantage of this experimental model is that it can manipulate the degree of overpressure load by changing the contraction intensity [174]. It is also an essential advantage that thoracic aortic constrictions resemble those in humans, especially aortic stenosis [175]. The major disadvantage of the TAC procedure is that it is dependent on the experimenter. Repeating the experiment is a complicated and technically demanding model. It also has the disadvantage of large variability between individuals [171].

As a result of studies conducted in rats, it was shown that AAC initially caused an increase in contractility due to the sympathetic nervous system [176]. Only after 8 weeks of concentric hypertrophy becomes apparent, as well as systolic and diastolic dysfunction [175]. The hypertrophic response and progression to HF resulting from the TAC model depend on sex, weight, age, and genetics. Approximately 1 month after adrenal AAC in mice, cardiac hypertrophy is seen, and the degree of this narrowing causes HF after 15–21 weeks [177]. Also, as a result of the studies, the mortality rate of TAC was found to be between 6 and 45%, and it was found to be very variable [178, 179].

As a result of recent developments, the double loop-clip technique has been developed. They measured the lumen diameter of the middle aortic arch during preoperative echocardiography to calculate the inter-knot span of the suture for the double loop-clip technique modified by Merino et al. This new procedure reduced the mortality rate in experimental animals and resulted in reproducible aortic stenosis [180]. The application of this model results in marked left ventricular hypertrophy and HF associated with increased β-myosin heavy chains. Also, a decrease in EF is observed. The hemodynamic function is associated with decreased sarco(endo)plasmic reticulum calcium ATPase (SERCA) expression, glucose uptake, reduced amount of adenosine in the coronary artery, and increased cardiomyocyte microtubule density [79, 145]. A decrease in SERCA expression in the left ventricle was observed 20 weeks after the experimental model. This suggests that SERCA may be a marker in the process between hypertrophy and HF [63, 181]. In terms of changes in the neurohormonal level, when hypertrophy occurs, plasma catecholamine levels are normal, and local myocardial RAS activation is increased [182]. The number of ANP increases due to HF. Also, as a result of the studies performed, it was determined that the PTHrP/PTH bioregulatory system, which is locally expressed in the ventricular myocardium, is associated with pressure-overload hypertrophy [183]. It has been shown that not only myosin heavy chain and atrial natriuretic peptide (ANP), but also cytokines such as interleukin-1, interleukin-6, and TNF-α are increased in HF induced by this experimental model [184]. Apoptotic processes, collagen, and matrix remodeling are seen in this experimental model. Also, Ca2+ is an ion required for cell growth and the continuation of vital cell functions. In response to growth stimuli, the cytosolic Ca2+ level is increased, and calcineurin is indirectly activated. These changes also cause dephosphorylation of transcription factors, which in turn regulates the expression of specific genes. Although not well known, calcineurin is thought to be a mediator of myocardial hypertrophy [181]. Changes in the nitric oxide (NO) pathway in pressure-overloaded heart tissue are important [185]. Studies have shown that sildenafil, a phosphodiesterase-5 inhibitor, reduced the left ventricular hypertrophy in the TAC model created in mice [186, 187]. Also, exogenous administration of the NO synthase cofactor (BH4) improved this left ventricular hypertrophy in mice [188]. Aortic constriction has been demonstrated in both the infrarenal and adrenal position in rats.

The experimental model of aortic constriction causes hypoperfusion, hypertension, and left ventricular hypertrophy. The AAC model contributes to the slower development of HF [160, 189]. When aortic insufficiency is combined with aortic narrowing, HF develops more rapidly. Ezzaher et al. reported that aortic insufficiency is produced by destroying the aortic valve with a catheter introduced through the carotid artery. After 14 days, aortic narrowing was performed just below the diaphragm. HF developed approximately 1 month after the first surgical procedure [190]. Although the protein and mRNA levels of the Na+/Ca2+ exchanger increased significantly in the HF model, the Ca2+/ATPase level in the sarcoplasmic reticulum did not change significantly [79]. This model mimics the changes in myocardial function observed in the late human myocardium. Therefore, this experimental model may be very suitable for studying HF after hypertrophy. Hyperreninemia develops in less than 4 days as a result of adrenal aortic coarctation. After several weeks, ventricular ACE activity may return to typical values, which may be associated with increased hypertrophy and normalization of wall stress [191]. As a result of this experimental model, left ventricular hypertrophy and HF were associated with increased β-myosin heavy chain mRNA and atrial natriuretic factor mRNA. Interestingly, after 20 weeks of banding, a decrease in SR-Ca2+/ATPase mRNA levels was seen by the polymerase chain reaction (PCR), while not in the group that did not develop HF. These data indicate that the decrease in SR-Ca2+/ATPase mRNA levels may indicate compensatory hypertrophy’s transition to failure in these animals [181]. During compensated hypertrophy, although catecholamine levels are normal, there is a local myocardial increased renin-angiotensin system. Also, with the development of HF, catecholamine levels in plasma may increase [192]. Therefore, this model should be preferred to examine the pathophysiological changes in the transition from hypertrophy to HF at the myocardial level. It is also a significant advantage that it can elucidate the molecular mechanisms involved in reverse remodeling of left ventricular hypertrophy [193].

Volume overload

The HF model induced by volume overload can be created by many different methods such as the arteriovenous fistula, aortic valve insufficiency, or destruction of the mitral valve [154, 194, 195]. It can cause HF in hyperdynamic conditions such as hyperthyroidism, beriberi (vitamin B1 deficiency), and severe anemia [154]. A shunt is created surgically to induce HF by causing an increase in volume. The shunt is usually formed between the aorta and the vena cava, femoral artery and vein, carotid artery, and internal jugular vein. Mice and rats are generally used in the HF model induced by volume overload [26]. Although HF progression varies, these models investigate disorders in fluid balance, electrolytes, and hormones common in HF. Increasing the heart’s volume load causes an increase in left ventricular end-diastolic pressure and results in expanding all the chambers in the heart. Although the systolic function is not impaired until the last stage of HF, the heart enlarges, and eccentric hypertrophy develops [196]. A catheter creates aortic valve perforation and chronic severe aortic regurgitation in a study on rabbits. Although eccentric left ventricular hypertrophy is observed first, it is often followed by systolic dysfunction [197]. However, it is difficult to induce aortic regurgitation in an experimentally reproducible manner, so most animal models have focused on inducing mitral regurgitation [198]. While arteriovenous fistula formation is a common volume overload model in small animals, it is rarely used in large animal models [199].

Surgical or percutaneous cutting of the mitral valve chordae tendineae causes mitral regurgitation and leads to chronic HF. However, this model is known to have a high mortality rate of up to 50% [200]. The reason for high mortality is that the degree of experimental HF created is difficult to control. One method of controlling this procedure is to place a graft between the left ventricle and the left atrium [201]. Although implantation is technically difficult, clamping the graft allows the degree of HF to be controlled. Inferior vena cava filters can be placed to keep the degree of insufficiency under control in the experimental HF model created due to mitral valve insufficiency [201]. Mitral valve insufficiency can also be observed in patients with postmyocardial infarction, and this circumstance is a poor prognosis. This model is usually done in sheep and pigs by infarctions the posterior wall of the left ventricle. The prognosis of mitral valve insufficiency after ischemia is quite poor [202].

Arteriovenous fistula models are another HF model used to create volume-overload HF in large animals [203]. The degree of shunt and proximity to the heart are among the factors that determine the severity of HF. The internal jugular vein and left carotid artery are exposed in the cervical arteriovenous shunt procedure. Approximately 30 min after the opening of the shunt, a 40% increase in cardiac output is observed. The left ventricular end-diastolic diameter and left ventricular end-diastolic volume increase after 8 weeks [132]. Although the arteriovenous fistula created between the carotid artery and the jugular vein caused an increase in left ventricular end-diastolic volume by approximately 50%, it did not cause a change in the end-diastolic pressure value [204]. Pinsky et al. performed a shunt between the infrarenal aorta and the superior vena cava to induce volume overload in 1979, whereas Tessier et al. have also combined this method with doxorubicin administration with a similar approach in goats [199].

Apart from this, mitral valve regurgitation was induced in dogs causing left ventricular enlargement. While left ventricular mass increased, there was not much change in the right ventricle. Asymmetric left ventricular dilatation was observed. Septum increased its contribution to the left ventricular stroke volume [205]. Mitral valve insufficiency models do not fully reflect HF. These models do not have pathophysiological changes in the myocardial structure due to ischemia or hypertrophy and are observed in congestive HF. The most important advantage of this model is evaluating the aperture of the shunt by palpating the neck. Also, ultrasound can be used for this procedure. A simple technique, the needle technique, is used to create an aortocaval shunt [206]. Compensated hypertrophy occurs approximately 2–8 weeks after the aortocaval fistula is formed. Left ventricular end-diastolic pressure increases around five times in the first week [206]. Arteriovenous shunts cause HF due to dilated cardiomyopathy by induction of volume overload in rodents. Femoral arteriovenous fistula causes HF, although it has a high mortality rate (over 25%) [207]. Also, the HF model’s essential advantages created with an aortocaval shunt are low mortality rate, simple, rapid, and no need for thoracotomy [208]. Studies in rats showed that cardiac hypertrophy developed and left ventricular diastolic pressure increased 4 weeks after shunt induction [209].

Volume overload resulting from aortocaval fistula causes a decrease in left ventricular function. Subsequently, hypertrophy develops as a mechanism of protection, leading to near-normal function at 4 weeks [210]. The decompensated hypertrophic HF and decreased systolic and diastolic function are seen approximately 8–16 weeks after the surgical procedure [211]. After a large aortocaval fistula is opened, severe volume overload occurs. Initially, the left ventricular function is suppressed, and hypertrophy develops as a protective mechanism. Hemodynamic functions are almost the same as normal after 4 weeks [212]. It has been determined that in about 7% of cases, the shunt closes spontaneously, so the fistula opening must be confirmed at the end of the experiment. It is known that left ventricular end-diastolic pressure increase is caused only by a large shunt and develops after at least 4 weeks [213].

Aortic valve insufficiency also causes HF as a result of increased volume in rats [214]. An increase in beta-receptor density was observed in heart tissues with HF in the 16th week. Progressive neurohormonal activation derived from the chronic mitral valve regurgitation model allowed investigation of the effects of angiotensin II type 1 and beta-receptor blockade on left ventricular failure [215]. For example, Tsutsui et al. showed in a study performed by bridges that the beta-receptor blockade provides an improvement in contractile function in the left ventricle compared to the control group [216]. It was observed that basal adenylyl cyclase activity increased, and beta-adrenoceptor signal transduction was similar to hypertrophy by induced volume overload [210]. According to recent studies, it has been shown that the synthesis of ANP expression increases due to increasing volume rather than pressure [217]. Also, induced myocyte elongation, changes in myofilament structure, and decreases in myocyte contractile function were observed [218]. It has been shown that SERCA2α expression can prevent systolic and diastolic dysfunction and left ventricular remodeling [219]. Volume overload in mice causes minimal apoptosis in the absence of pathological remodeling than the TAC model [26]. Surgical-induced HF models are presented in the Table 2.

Table 2 Characteristics of surgical-induced HF models

Drug-induced

Drugs induce HF model in different animals to study the etiology of HF that includes chemotoxicity, hypertension, kidney damage, and liver damage [226, 227]. Drug-induced HF models are presented in the Table 3. In summary, HF can be caused by the administration of chemical agents. However, in general, the advantages and disadvantages should be evaluated carefully, and the appropriate experimental model should be selected.

Table 3 Characteristics of drug-induced HF models

Hypertension

Hypertension is one of the essential conditions leading to diastolic HF in humans [144, 240]. Hypertension causing extensive inflammation and metabolism changes can lead to myocardial stiffness and diastolic dysfunction [241, 242]. Ventricular dysfunction develops due to the proliferation of fibroblasts, hypertrophy of vascular smooth muscle cells, and pathological accumulation of interstitial collagen [143, 243]. Hypertension-induced HF models are presented in the Table 4.

Table 4 Characteristics of hypertension-induced HF models

High fat diet + L-NAME

It is known that the comorbidities that increase the risk of HFpEF are diabetes, obesity, and hypertension [249]. Recently, Hill et al. proposed a “two-hit” mouse model of HFpEF that mimics the concomitant metabolic and hypertensive stress in mice [250]. In this model, a high-fat diet (HFD) induces metabolic stress (obesity, glucose intolerance, and metabolic syndrome), and a drug called Nω-nitro-l-arginine methyl ester (L-NAME), which inhibits nitric oxide synthase, also causes hypertension. This model recapitulates numerous systemic and cardiovascular features of HFpEF, including impaired cardiac filling, cardiac hypertrophy, cardiac fibrosis, decreased myocardial capillary density, pulmonary hyperemia, decreased exercise tolerance, myocardial capillary rarefaction, and increased levels of inflammatory markers [250]. Diastolic dysfunction was associated with cardiac hypertrophy and fibrosis [251]. This experimental model uses 60% kilocalories from fat (lard) and drinking water with 0.5 g/L of L-NAME [252].

Conclusion

Pre-clinical HF models are useful in the studying pathophysiology of complex cardiovascular diseases. Experimental animal models are becoming less relevant due to modern techniques to determine HF’s hemodynamics in humans. However, pre-clinical studies have great significance in the experimental HF studies currently. Also, pre-clinical studies have elucidated molecular and biochemical modifications and remodeling processes in human HF myocardium. Animal HF models may be suitable for testing the effect of new treatment modalities on hemodynamics, neurohumoral activation, and survival under preclinical conditions. Today, transgenic models provide a better understanding of HF pathophysiology at the molecular level. It enables the identification of genes that cause HF and elucidate their role in HF. It also enables it to be characterized in the molecular mechanisms responsible for HF development and progression.

Finally, various experimental animal models that mimic human HF features will play a vital role in elucidating the pathophysiology in HF and in solving molecular techniques. These studies are invaluable and provide knowledge on HF’s underlying mechanisms and contribute significantly to creating new therapeutic strategies for HF. Experimental HF models will have mechanistic perspectives that will contribute to new classes of therapy in the future. Animal HF models are now becoming excellent tools for creating new goal settings for new HF management agents.