Selective microembolization of the circumflex coronary artery in an ovine model: dilated, ischemic cardiomyopathy and left ventricular dysfunction

Coronary microembolization sheep model by adjusting the number of microspheres based on coronary blood flow

BACKGROUND

A heart failure (HF) model using coronary microembolization in large animals is indispensable for medical research. However, the heterogeneity of myocardial response to microembolization is a limitation. We hypothesized that adjusting the number of injected microspheres according to coronary blood flow could stabilize the severity of HF. This study aimed to evaluate the effect of microsphere injection based on the left coronary artery blood flow in an animal model.

METHODS

Microembolization was induced by injecting different numbers of microspheres (polystyrene, diameter: 90 μm) into the left descending coronary artery of the two groups of sheep (400 and 600 times coronary blood flow [ml/min]). Hemodynamic parameters, the pressure-volume loop of the left ventricle, and echocardiography findings were examined at 0.5, 1.5, 3.5, and 6.5 h after microembolization.

RESULTS

End-diastolic pressure and normalized heart rate increased over time, and were significantly higher in 600 × coronary blood flow group than those in 400 × coronary blood flow group (p = 0.04 and p < 0.01, respectively). The maximum rate of left-ventricular pressure rise and normalized stroke volume decreased over time, and were significantly lower in 600 × coronary blood flow group than those in 400 × coronary blood flow group (p < 0.01 and p < 0.01, respectively). The number of microspheres per coronary blood flow was significantly correlated with the decrease in stroke volume and the maximum rate of left ventricular pressure rise in 6.5 h (r = 0.74, p = 0.01 and r = 0.71, p = 0.02, respectively).

CONCLUSIONS

Adjusting the number of injected microspheres based on the coronary blood flow enabled the creation of HF models with different degrees of severity.

Large Animal Models of Heart Failure: Reduced vs. Preserved Ejection Fraction

Heart failure (HF) is the final common end point of multiple metabolic and cardiovascular diseases and imposes a significant health care burden worldwide. Despite significant improvements in clinical management and outcomes, morbidity and mortality remain high and there remains an indisputable need for improved treatment options. The pathophysiology of HF is complex and covers a spectrum of clinical presentations from HF with reduced ejection fraction (HFrEF) (≤40% EF) through to HF with preserved EF (HFpEF), with HFpEF patients demonstrating a reduced ability of the heart to relax despite an EF maintained above 50%. Prior to the last decade, the majority of clinical trials and animal models addressed HFrEF. Despite growing efforts recently to understand underlying mechanisms of HFpEF and find effective therapies for its treatment, clinical trials in patients with HFpEF have failed to demonstrate improvements in mortality. A significant obstacle to therapeutic innovation in HFpEF is the absence of preclinical models including large animal models which, unlike rodents, permit detailed instrumentation and extensive imaging and sampling protocols. Although several large animal models of HFpEF have been reported, none fulfil all the features present in human disease and few demonstrate progression to frank decompensated HF. This review summarizes well-established models of HFrEF in pigs, dogs and sheep and discusses attempts to date to model HFpEF in these species.

Large Animal Models of Heart Failure: A Translational Bridge to Clinical Success

•

Preclinical large animal models play a critical and expanding role in translating basic science findings to the development and clinical approval of novel cardiovascular therapeutics.

•

This state-of-the-art review outlines existing methodologies and physiological phenotypes of several HF models developed in large animals. A comprehensive list of porcine, ovine, and canine models of disease are presented, and the translational importance of these studies to clinical success is highlighted through a brief overview of recent devices approved by the FDA alongside associated clinical trials and preclinical animal reports.

•

Increasing the use of large animal models of HF holds significant potential for identifying new mechanisms underlying this disease and providing valuable information regarding the safety and efficacy of new therapies, thus, improving physiological and economical translation of animal research to the successful treatment of human HF.

Preclinical large animal models of heart failure (HF) play a critical and expanding role in translating basic science findings to the development and clinical approval of novel therapeutics and devices. The complex combination of cardiovascular events and risk factors leading to HF has proved challenging for the development of new treatments for these patients. This state-of-the-art review presents historical and recent studies in porcine, ovine, and canine models of HF and outlines existing methodologies and physiological phenotypes. The translational importance of large animal studies to clinical success is also highlighted with an overview of recent devices approved by the Food and Drug Administration, together with preclinical HF animal studies used to aid both development and safety and/or efficacy testing. Increasing the use of large animal models of HF holds significant potential for identifying the novel mechanisms underlying the clinical condition and to improving physiological and economical translation of animal research to successfully treat human HF.

Chronic stable heart failure model in ovine species

Establishing a chronic heart failure (HF) model is challenging, particularly in the ovine model. The aim of this study was to establish a reproducible model of HF in an ovine model. Seventeen sheep were operated using the left thoracotomy approach. Chronic HF was induced through ligation of the diagonal and marginal branches only. Perioperative hemodynamic and echocardiographic parameters were compared. A total of (3 ± 1) coronary ligations were used. Thirteen animals survived the procedure and were followed up for (15 ± 5) days. The mean arterial pressure, heart rate (HR), mean pulmonary artery pressure (mPAP), central venous pressure, and cardiac output at baseline and prior to animal sacrifice was (75 ± 14 mmHg) and (68 ± 16 mmHg) P = .261; (72 ± 9 bpm), (100 ± 28 bpm) P = .01; (15 ± 4 mmHg) and (18 ± 5 mmHg) P = .034; (10 ± 6 mmHg) and (8 ± 4 mmHg) P = .326; (3.4 ± 1 L/min) and (3.9 ± 1 L/min) P = .286, respectively. The LVEF at baseline and prior to animal sacrifice was (63 ± 13%) and (43 ± 6%) P = .012. Twelve surviving animals were supported with LVAD in a follow-up procedure. Chronic stable HF in sheep was successively established. Clinical symptoms and drastic increase in the mPAP and HR as well as echo findings were the most sensitive parameters of HF. This reproducible ovine model has proven to be highly promising for research regarding HF.

Decreased atrioventricular plane displacement after acute myocardial infarction yields a concomitant decrease in stroke volume

Acute myocardial infarction (AMI) can progress to heart failure, which has a poor prognosis. Normally, 60% of stroke volume (SV) is attributed to the longitudinal ventricular shortening and lengthening evident in the atrioventricular plane displacement (AVPD) during the cardiac cycle, but there is no information on how the relationship changes between SV and AVPD before and after AMI. Therefore, the aim of this study was to determine how SV depends on AVPD before and after AMI in two swine models. Serial cardiac magnetic resonance imaging was carried out before and 1–2 h after AMI in a microembolization model (n = 12) and an ischemia-reperfusion model (n = 14). A subset of pigs (n = 7) were additionally imaged at 24 h and at 7 days. Cine and late gadolinium enhancement images were analyzed for cardiac function, AVPD measurements and infarct size estimation, respectively. AVPD decreased (P < 0.05) in all myocardial regions after AMI, with a concomitant SV decrease (P < 0.001). The ischemia-reperfusion model affected SV to a higher degree and had a larger AVPD decrease than the microembolization model (−29 ± 14% vs. −15 ± 18%; P < 0.05). Wall thickening decreased in infarcted areas (P < 0.001), and A-wave AVPD remained unchanged (P = 0.93) whereas E-wave AVPD decreased (P < 0.001) after AMI. We conclude that AVPD is coupled to SV independent of infarct type but likely to a greater degree in ischemia-reperfusion infarcts compared with microembolization infarcts. AMI reduces diastolic early filling AVPD but not AVPD from atrial contraction. These findings shed light on the physiological significance of atrioventricular plane motion when assessing acute and subacute myocardial infarction.

NEW & NOTEWORTHY The link between cardiac longitudinal motion, measured as atrioventricular plane displacement (AVPD), and stroke volume (SV) is investigated in swine after acute myocardial infarction (AMI). This cardiac magnetic resonance study demonstrates a close coupling between AVPD and SV before and after AMI in an experimental setting and demonstrates that this connection is present in ischemia-reperfusion and microembolization infarcts, acutely and during the first week. Furthermore, AVPD is equally and persistently depressed in infarcted and remote myocardium after AMI.

Impaired Baroreflex Function in an Ovine Model of Chronic Heart Failure Induced by Multiple Coronary Microembolizations

Testing new therapies in heart failure (HF) requires a chronic stable model of HF in large animals. Microembolization of the coronary arteries has been used to model HF previously; however, neural control has not been previously explored in this model. Thus the aim of this study was to further characterize neural control in this model of HF. HF was induced by infusion of microspheres (45 micron; 1.3 ml) into the proximal left coronary artery or left descending coronary arteries, with three sequential embolizations over 3 weeks. Twelve to 14 weeks after the final embolization, and when ejection fraction had decreased below 45%, animals were instrumented to record blood pressure and heart rate. Baroreflex control of heart rate was investigated in conscious animals. Additionally, pressure-volume loops were constructed under anesthesia. Embolization-induced HF was associated with a decrease in mean arterial pressure (67 ± 2 vs. 85 ± 4 mmHg, p < 0.05), an increase in heart rate (108 ± 4 vs. 94 ± 4 bpm, p < 0.05), and a significant increase in left ventricular end-diastolic pressure (11.4 ± 2 vs. 6.2 ± 1 mmHg, p < 0.01). Under conscious conditions, there was a significant decrease in the gain (-8.2 ± 2 vs. -4.1 ± 1 beats/min/mmHg, p < 0.05) as well as the lower plateau of the baroreflex in HF compared to control animals. HF was also associated with significantly increased respiratory rate (107 ± 4 vs. 87 ± 4 breaths/min, p < 0.01) and incidence of apneas (520 ± 24 vs. 191 ± 8 apnea periods >4 s, p < 0.05), compared to control sheep. The microembolization model of heart failure is associated with an increase in left ventricular end-diastolic pressure, impaired cardiac function, and altered baroreflex control of the heart. These findings suggest this chronic model of HF is appropriate to use for investigating interventions aimed at improving neural control in HF.

A Novel Large Animal Model of Thrombogenic Coronary Microembolization

Coronary microembolization is one of the main causes of the “no-reflow” phenomenon, which commonly occurs after reperfusion of an occluded coronary artery. Given its high incidence and the fact that it has been proven to be an independent predictor of cardiac morbidity and mortality, there is an imperative need to study its underlying mechanisms and pathophysiology. Large animal models are essential to perform translational studies. Currently there is no animal model that recapitulates a clinical scenario of thrombogenic microembolism with preceding myocardial ischemia. Therefore, the goal of this study was to develop and characterize a novel pig model of coronary microembolization using autologous thrombus injection (CMET). Twenty-three pigs underwent myocardial infarction through percutaneous balloon occlusion of the left anterior descending artery (LAD). Each animal was enrolled in one of two groups: (1) the CMET group, in which the LAD occlusion was followed by delivery of autologous clotted blood in the LAD (distal to the balloon occlusion) and reperfusion; (2) the ischemic reperfusion (I/R) group, in which the LAD ischemia was followed by reperfusion. Surviving animals underwent functional and morphological characterization at 1-week post-procedure. Three sham operated animals were used as a control. CMET resulted in impaired left ventricular function compared to I/R pigs at 1 week. Three-dimensional echocardiography demonstrated reduced ejection fraction in the CMET group (CMET vs. I/R: 35.6 ± 4.2% vs. 47.6 ± 2.4%, p = 0.028). Invasive hemodynamic measurements by Swan-Ganz and left ventricular pressure-volume catheters revealed that CMET impaired left ventricular contractility and diastolic function. This was confirmed by both load-dependent indices including cardiac output (CMET vs. I/R: 2.7 ± 0.2 l/min, vs. 4.0 ± 0.1 l/min, p = 0.002) and load independent indices including preload-recruitable stroke work (CMET vs. I/R: 25.8 ± 4.0 vs. 47.5 ± 6.5 mmHg, p = 0.05) and end-diastolic pressure-volume relationship (slope, 0.68 ± 0.07 vs. 0.40 ± 0.11 mmHg/ml, p = 0.01). Our unique closed-chest model of coronary microembolization using autologous thrombus injection resembles the clinical condition of thrombogenic coronary microembolization in I/R injury. This model offers opportunities to conduct translational studies for understanding and treating coronary microembolization in myocardial infarction.

Large Animal Models of Heart Failure With Reduced Ejection Fraction (HFrEF)

Heart failure with reduced ejection fraction (HFrEF) is defined by an ejection fraction (EF) below 40%. Many distinct disease processes culminate in HFrEF, among them acute and chronic ischemia, pressure overload, volume overload, cytotoxic medication, and arrhythmia. To study these different etiologies the development of accurate animal models is vital. While small animal models are generally cheaper, allow for larger sample sizes and offer a greater variety of transgenic models, they have important limitations in the context of HFrEF research. Small mammals have much higher heart rates and distinct ion channels. They also have much higher basal metabolic rates and their physiology in many ways does not reflect that of humans. The size of their organs also puts practical constraints on experiments. Therefore, large animal models have been developed to accurately simulate human HFrEF. This review aims to give a short overview of the currently established large animal models of HFrEF. The main animal models discussed are dogs, pigs, and sheep. Furthermore, multiple approaches for modeling the different etiologies of HF are discussed, namely models of acute and chronic ischemia, pressure overload, volume overload as well as cytotoxic, and tachycardic pacing approaches.

Bovine model of chronic ischemic cardiomyopathy: implications for ventricular assist device research

Ventricular assist devices (VADs) have emerged as a successful treatment option for advanced heart failure. The objective of this study was to develop a clinically relevant model of chronic ischemic cardiomyopathy to investigate functional, histological, and molecular changes during mechanical circulatory support. In calves (n = 17, 94 ± 7 kg), 90 μm microspheres were injected percutaneously into the left coronary artery. Serial echocardiography was performed weekly to evaluate cardiac function. Sixty days after coronary microembolization, a terminal study was performed via thoracotomy to measure hemodynamics. Regional myocardial and end-organ blood flows were quantified with 15-μm fluorescent-labeled microspheres. Myocardial fibrosis, myocyte size, and myocardial apoptosis were quantified with histological stains. Eleven animals survived coronary microembolization and exhibited clinical and statistically significant echocardiographic and hemodynamic signs of severe systolic dysfunction. Statistically significant decreases in regional myocardial blood flow and increases in myocardial fibrosis, myocyte size, total myocardial apoptosis, and cardiac myocyte-specific apoptosis were observed. End-organ hypoperfusion was observed. Coronary microembolization induced stable and reproducible chronic left ventricular failure in calves. The anatomical size and physiology of the bovine heart and thorax are appropriate to study novel interventions for the clinical management of heart failure. This model is an appropriate physiological substrate in which to test VAD and adjunctive biological therapies.

Ovine models for chronic heart failure

PURPOSE

Testing and optimizing of surgical therapies for chronic heart failure (CHF) requires large animal models. CHF has been induced in several large animal species. Sheep have modest body mass increase and demonstrate docile behavior and are therefore a preferred species in research on surgical therapies for CHF METHODS: A literature search for existing ovine CHF models was performed, using search terms "sheep" and "heart failure". Relevant secondary references were traced.

RESULTS

Rapid ventricular pacing produces rapid-onset CHFE Its severity ranges from moderate left ventricular failure to severe biventricular failure, depending on length and frequency of pacing. Its counterpart in human CHF is tachycardia-induced HF since it is reversible upon cessation of pacing. Myocardial damage models include CHF induced by cardiototoxic drugs and ischemia. Ischemia-based models include coronary microembolization, occlusion and ischemia/reperfusion models. The microembolization model is relevant to diabetic cardiomyopathy. Coronary occlusion models exhibit variable functional impairment, some with aneurysm formation, and some with mitral valve regurgitation, depending on occlusion localization. They are relevant to CHF following non-reperfused myocardial infarction. Coronary occlusion/reperfusion models are relevant to the occurrence of human ãã despite coronary artery recanalization. Pressure overload of left and right ventricle is induced by aortic and pulmonary artery banding, respectively. Hypertrophy precedes CHF as in patients with valve stenosis and hypertension. Volume overload is induced by valve damage or shunt creation. Atrioventricular valve regurgitation is the most important clinical counterpart.

CONCLUSION

Several ovine CHF models exist. Since they exhibit important cardiac pathology differences, the choice of model should be based on the specific experimental question.

Right ventricular remodeling in restrictive ventricular septal defect

Restrictive ventricular septal defect (rVSD) presents with little/no hemodynamic aberrations despite a patent septal defect. Clinically, these patients are observed with the hope that the defect will functionally close over time without the need for surgical repair and development of heart failure. Without evidence supporting a definitive therapeutic strategy, rVSD patients may have increased risk of a poor outcome. We tested the hypothesis that rVSD results in subclinical RV diastolic dysfunction and molecular remodeling. Five pigs underwent surgical rVSD creation. Echocardiography, hemodynamics, myocyte contractility experiments, and proteomics/Western blot were performed 6-weeks post-rVSD and in controls. *p<0.05. LV and RV hemodynamics in rVSD were comparable to controls. The tricuspid valve early/late diastolic inflow velocity ratio (TV E/A ratio) decreased from 1.6+/-0.05 in controls to 1.0+/-0.08* in rVSD, indicating RV diastolic dysfunction. rVSD RV myocytes showed abnormalities in contraction (departure velocity (Vd) -51%*, Vd time +55%*) and relaxation (return velocity (Vr) -50%*, Vr time +62%*). Mitochondrial proteins (fatty acid, TCA cycle) increased 2-fold*, indicating heightened RV work. Desmin protein upregulated 285%* in rVSD RV myocardium, suggesting cytoskeletal remodeling. rVSD causes RV diastolic dysfunction, myocyte functional impairment, and mitochondrial/cytoskeletal protein upregulation in our model. Desmin upregulation may hinder sarcomeric organization/relaxation, representing a key subclinical early marker for future RV dysfunction. TV E/A measurements are a non-invasive modality to assess rVSD patients for diastolic dysfunction. Translational research applications may lead to fundamental changes in the clinical management of rVSD by providing evidence for early repair of the defect.

Left Ventricular Assist Device Support Induces Acute Changes in Myocardial Electrolytes in Heart Failure

The regulation of myocardial electrolyte concentrations is critical to proper cardiac function. Myocardial ischemia is associated with deranged ion transport. Left ventricular assist device (LVAD) therapy improves myocyte bioenergetics in chronic heart failure (CHF), which may manifest as electrolyte alterations; however, rapid electrolyte shifts may place critically ill patients at risk for arrhythmias upon initiation of LVAD support. We examine the effect of incremental increases in LVAD support on acute changes in myocardial arteriovenous electrolytes in CHF. CHF was induced in sheep via coronary microembolization. Four months later, sheep underwent acute LVAD implantation. LVAD support was incrementally increased (0%, 25%, 50%, 75% support). Paired arterial and coronary sinus blood samples were obtained at each increment and analyzed for K+, Ca2+, and Na+ concentrations. Arteriovenous electrolyte concentrations (mmol/l) were inverted in CHF before LVAD support: K+ (-0.08), Ca2+ (-0.04), and Na+ (0.04). These imbalances were corrected within 20 minutes and with as little as 25% LVAD support: K+ (0.06), Ca2+ (0.012), and Na+ (-0.80). The arteriovenous differences further widened as LVAD support was increased. In conclusion, LVAD support in CHF induces acute alterations in myocardial electrolytes. Rapid shifts myocardial arteriovenous electrolyte balances during LVAD support may in part explain the incidence of post-LVAD arrhythmias observed clinically in humans.

Partial Support with a Centrifugal Left Ventricular Assist Device Reduces Myocardial Oxygen Consumption in Chronic, Ischemic Heart Failure

BACKGROUND

Left ventricular assist devices (LVAD) are increasingly used for heart failure (CHF); however, the level of optimal support has not been elucidated. We hypothesize that partial LVAD support in an ovine model of microinfarction-induced CHF significantly reduces left ventricular myocardial oxygen consumption (LVVO2).

METHODS AND RESULTS

Microembolization of the circumflex coronary artery was used to induce CHF in 5 sheep (ejection fraction 28 +/- 2%). Four months later, animals underwent implantation of a centrifugal LVAD. LVAD flow was incrementally increased from 0% (baseline) to 25%, 50%, and 75% support of the LV. LVVO2 and stroke work (SW) were calculated at each increment. At baseline, LVVO2 (microL/100 g LV/beat) measured 43.2 +/- 3.4. LVVO2 decreased to 26.5 +/- 8.2,* 20.3 +/- 8.9,* and 12.6 +/- 6.3* at 25%, 50%, and 75% support (*P < .05). SW (mm Hg/mL) measured 1933.0 +/- 275.7 at baseline and decreased to 1588.0 +/- 204.1, 1181.0 +/- 157.2,* and 764.5 +/- 171.7* at 25%, 50%, and 75% support. Cardiac output, heart rate, and left main coronary artery blood flow were unaffected with partial support.

CONCLUSION

Complete support with a centrifugal LVAD is not necessary for achieving significant reductions in LVVO2 . Partial support of as little as 25% significantly reduces LVVO2 in CHF through comparatively minor reductions in cardiac work. This is the first study to examine partial LVAD support in a CHF model.