Progression toward decompensated right ventricular failure in the ovine pulmonary hypertension model

Abstract

Decompensated right ventricular failure (RVF) in patients with pulmonary hypertension (PH) is fatal, with limited treatment options. Novel mechanical circulatory support systems have therapeutic potential for RVF, but development of these devices requires a large animal disease model that replicates the pathophysiology observed in humans. We previously reported an effective disease model of PH in sheep through ligation of the left pulmonary artery (PA) and progressive occlusion of the main PA. Here, we present a case of acute on chronic RVF with this model. Gradual PA banding raised the RV pressure (maximum RV systolic/mean pressure = 95 mmHg/56 mmHg). Clinical findings and laboratory serum parameters suggested appropriate physiologic compensation for 7 weeks. However, mixed venous saturation declined precipitously on week 7, and creatinine increased markedly on week 9. By the 10^th week, the animal developed dependent, subcutaneous edema. Subsequently, the animal expired during induction of general anesthesia. Post-mortem evaluation revealed several liters of pleural effusion and ascites, RV dilatation, eccentric RV hypertrophy, and myocardial fibrosis. The presented case supports this model’s relevance to human pathophysiology of RVF secondary to PH and its value in the development of novel devices, therapeutics, and interventions.

Introduction

Pulmonary vascular disease causes chronic pulmonary hypertension (PH) and progressive strain on the right ventricle (RV), leading to eventual RV dysfunction and failure.^1,2 To prevent this fatal trajectory, PH patients who fail medical therapy may benefit from durable devices that provide both hemodynamic and oxygenation support, such as an implantable mechanical circulatory support device with a low-impedance oxygenator. Developing these technologies requires a large animal model that recapitulates the complex pathophysiology of chronic RV failure secondary to PH (PH-RVF) in humans.

Previously, our group published a large animal PH model which uses a combined approach of left pulmonary artery (PA) ligation and progressive occlusion of the main PA.³ Here, we report the fatal trajectory of this model over a prolonged course of chronic RV pressure overload and strain. The pathophysiology and outcome observed in this large animal case parallels that of patients with advanced PH.

Method / Case Presentation

Progressive PH

The Institutional Animal Care and Use Committee at Vanderbilt University Medical Center approved the protocol. The described procedures were conducted in accordance with the Animal Welfare Act Regulations and the U.S. National Research Council’s Guide for the Care and Use of Laboratory Animals, 8th edition.

The surgical technique for left PA ligation and PA banding have been previously described.³ Briefly, a 70-kg Dorset-cross wether underwent a mini left-thoracotomy for implantation of a 16-mm heavy duty main PA occluder. Pressure tubing was placed into the RV outflow tract for hemodynamic monitoring and blood sampling, and the left PA was ligated with umbilical tape. The free ends of the occluder and RV pressure tubing were tunneled to the left dorsum and connected to individual subcutaneous ports. The animal showed no major abnormalities at baseline (Supplementary Table 1). Transthoracic echocardiography demonstrated normal biventricular function at baseline prior to left PA ligation and progressive PA inflation (Supplementary Video 1). Furthermore, there were no immediate perioperative complications.

The progressive development of PH proceeded for the next 10 weeks after its initial surgery, as previously described.³ Every 2–4 days, the subcutaneous ports were accessed to tighten the PA occluder and to acquire RV pressure measurements (Figure 1A). By week 7, a mean internal cuff pressure of 1400–1500 mmHg was achieved. RV pressure increased over time in response to cuff inflation, reaching RV systolic pressure of 95 mmHg and mean pressure of 56 mmHg by week 9 (Figure 1B). Furthermore, both heart rate and respiratory rate increased as the cuff was inflated, reflecting the animal’s compensatory physiological response to progressive PH (Figure 1C, D).

Figure 1:

The animal’s physiologic parameters during the disease development: A) PA cuff internal pressure, B) RV mean (black diamond) and systolic (blue square) pressure measurements, C) respiratory rate, D) heart rate

Using the RV port, blood samples were collected to measure complete blood count, complete metabolic panel, cardiac enzymes, and S_vO₂. Despite the increased RV pressure during PH development, neither cardiac troponin I nor N-type proBNP (NT-proBNP) levels changed substantially from their starting values (Figure 2A). However, atrial natriuretic peptide (ANP) increased modestly at week 6 to 1958 pg/mL from its baseline value of 1328 pg/mL (Figure 2A). The serum level of alanine aminotransferase (ALT) and gamma-glutamyl transferase (GGT) remained stable. While aspartate aminotransferase (AST) remained within normal limits throughout the model duration, it increased from a baseline of 82 IU/L to 122 IU/L at week 9 (Figure 2B). Blood urea nitrogen (BUN) remained stable, while serum creatinine increased from 1.23 mg/dL at week 8 to 1.82 mg/dL at week 9 (Figure 2C). White blood cell count was mostly stable but increased at week 9 (Figure 2D). S_vO₂ slowly declined over the first 6 weeks and then dropped precipitously to 40.7% by week 7 and to 27.1% by week 8 (Figure 2E). Hemoglobin, while it remained within normal range, showed an increase at week 5 (Figure 2E).

Figure 2:

The animal’s blood laboratory values during the disease development: A) plasma levels of cardiac troponin I (CTnI), N-terminal pro-B-type natriuretic peptide (NT-proBNP), and atrial natriuretic peptide (ANP); B) serum levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), and gamma-glutamyl transferase (GGT); C) blood levels of creatinine and blood urea nitrogen (BUN); D) white blood cell count. E) Mixed venous oxyhemoglobin saturation (S_vO₂) and hemoglobin level. Week 0 values were measured before initiating PH development.

Right Ventricular Decompensation and Sudden Death

On week 10, subcutaneous edema on the right and left aspects of the ventral thorax (Figure 3A) suddenly developed within a single day. The acute onset suggested a transition from compensated PH to decompensated RVF. On this same day, the animal was tachypneic with a respiratory rate of 112 breaths/min and tachycardic with a heart rate of 146 beats/min. In response to these clinical findings, the PA cuff was partially deflated to alleviate the RV pressure overload, and 10 mcg/kg buprenorphine was administered intramuscularly to alleviate any discomfort.

Figure 3:

Clinical presentation and gross pathology of the animal near the end of progressive PH development: A) Brisket edema formation observed on the 69^th day of the disease model; B) Necropsy photographs of the heart (top left), lungs (top right), and liver (bottom)

On the following day, the animal was scheduled for an acute, non-survival surgery. The animal underwent anesthesia induction with 4 mg/kg tiletamine/zolazepam and immediately showed worsening dyspnea. As the endotracheal tube was placed, the respiratory rate slowed. Upon initiation of mechanical ventilation, the animal was noted to be apneic, asystolic, and pulseless. Despite attempted cardiac resuscitation with a bolus dose of 1 mg epinephrine and cardiac defibrillation, the sheep expired.

Postmortem Examination

A necropsy was immediately performed. Severe volume overload was observed and manifested as subcutaneous and diffuse tissue edema, 4 liters of pleural effusion, and 4–5 liters of ascites. Total heart weight was 439 grams (Figure 3B). The RV cavity was enlarged, the septum was flattened, and the RV free wall was hypertrophied. Fulton’s index, the weight ratio of RV free wall (142 grams) to combined left ventricle (LV) and interventricular septum (183 gram), was 0.78, markedly higher than that of healthy sheep (average Fulton’s Index of 0.35)^4,5 and thereby indicating extensive RV hypertrophy. The interventricular septum thickness was approximately 12 mm; RV free wall, 8 mm; and LV free wall, 12 mm. Tissue samples were collected from both ventricles, lungs, liver, and kidneys for histology. Mild to moderate myocardial fibrosis was present in both ventricles but greater on the right side (Figure 4A). By gross examination, the lungs were dark red-purple and firm (Figure 3B). Microscopic histology of the lungs revealed atelectasis, alveolar edema, and hemorrhage, which were more severe in the left lobes than the right (Figure 4A). The liver was diffusely firm and dark (Figure 3B). Meanwhile, centrilobular sinusoids were dilated and congested with erythrocytes. The liver was necrotic with loss of zone 3 hepatocytes. These histologic findings in the liver suggest acute right-sided heart failure (Figure 4B). The kidneys did not exhibit any gross or microscopic abnormalities (Supplementary Figure 1). Overall, the post-mortem findings suggest adaptive changes to progressive PH and RV pressure overload that ultimately resulted in cardiopulmonary collapse.

Figure 4:

Histological findings in the lungs and liver are consistent with decompensated RV failure by the ovine PH model: A) H&E and Masson’s trichrome stains of left and right ventricles (40x magnification); B) H&E stains of left lung, right lung, and liver (left images at 20x magnification, right images at 100x (lung) and 200x (liver) magnification).

Discussion

Here we report a case of decompensated RV failure and sudden death upon anesthetic induction in an ovine model of advanced PH. While chronic PA banding has been used to induce PH in sheep,⁶ challenges remain with replicating the complex chronic disease state of PH, such as eccentric RV hypertrophy, hepatic and pleural congestion, and edema. To our knowledge, we are the first to report the fatal trajectory of chronic PH in a large animal model. Our clinical observations and post-mortem evaluation are consistent with clinical findings seen in patients with PH-RVF.^1,2 The RV can accommodate to elevated pulmonary arterial pressure by adaptive hypertrophy, increased contractility, and compensatory tachycardia to an extent, but if left untreated, RV decompensation and rapid deterioration may ensue.^1,2 The presented case therefore demonstrates that this large animal model replicates the human pathophysiology of PH-RVF.

In future studies, we intend to control more precisely the degree of PA occlusion to reduce the risk for acute fatal decompensation. In particular, the cuff pressure will be increased at a slower rate, and its target pressure will be set below 1400–1500 mmHg (Figure 1A). In our more recent cohort, this approach has helped reduced the rate of irrecoverable RV decompensation (unpublished data). As the PA cuff is inflated, it is also important to trend physiologic and biochemical parameters to identify impending decompensation. In this case report, RV pressure increased over time, and common biomarkers also remained close to their baselines, suggesting compensated PH. However, some notable changes towards the end of PH development suggest onset of RV decompensation. Creatinine increased by 48% between weeks 8 and 9 (Figure 2C). Following a slow gradual decline, S_vO₂ dropped precipitously throughout weeks 7 and 8 (Figure 2E). The onset of these findings likely denotes crossing pathophysiologic tipping points of venous hypertension and cardiac output. Two recent studies by Khirfan et al.⁷ and Shah et al.⁸ demonstrated that S_vO₂ and serum creatinine, respectively, prognosed mortality and adverse outcomes in pulmonary arterial hypertension patients. These markers may be important parameters to detect the transition from compensated PH to decompensated RV failure, systemic congestion, and multi-organ failure. Clinical findings (Figure 3A) during this period correlated with these laboratory results, corroborating the efficacy of this PH-RVF model. The observed subcutaneous edema and the postmortem findings of severe hypervolemia suggest progression to decompensated RV failure, which may also be identified in the future through trending the animal’s weight. We will continue to evaluate these longitudinal trends in future experiments to characterize the onset of RV decompensation, and seek novel biomarkers that can help better detect this transition.

Conclusion

We report a case of decompensated RV failure and sudden death in an ovine model of PH-RVF. This case demonstrates the ability of this large animal model to replicate the pathophysiology of patients with PH-RVF. This model can serve as a clinically relevant platform for future investigations in biochemical and metabolic markers, tissue morphology, cardiopulmonary response, as well as pharmacologic and mechanical interventions for PH and RV failure.

Supplementary Material

Supplemental Video File, echocardiography

Supplementary Video 1: Transthoracic echocardiography shows the parasternal short-axis view of the sheep’s heart before the disease model.