Despite lung protective ventilation, prone positioning, fluid management, neuromuscular blockers, and other therapies, mortality for acute respiratory distress syndrome (ARDS) is as high as 53%.1 In 2006, the conventional ventilatory support vs extracorporeal membrane oxygenation for severe adult respiratory failure (CESAR) trial demonstrated reduced mortality in ARDS to 37%.1 Recognition of these results and H1NI ARDS in 2009 led to veno-venous (VV) extracorporeal membrane oxygenation (ECMO) gaining acceptance as alternative treatment. Even for prolonged cannulation (>14 days), Posluszny demonstrated survival of 45.4%, with best outcomes in VV ECMO.2 Traditionally, VV ECMO required insertion of two venous cannula to facilitate drainage/flow and oxygenation. The dual-chamber Avalon cannula (Avalon Laboratories, Rancho Dominguez, CA) allows drainage from the inferior vena cava (IVC)/superior vena cava (SVC) and return through the right atrium across the tricuspid valve (TV). The cannula decreases recirculation and bleeding and allows for mobilization,3 and in our center, it is used in the majority of VV cases.
Acute cor pulmonale (ACP) is right heart failure resulting from disorder of the lung, not left ventricular failure, congenital disorders, or valvular pathology. In the 1990s, ACP in ARDS was 60%; however, as lung protective ventilation became accepted, the incidence dropped to 20–25%.4 Although improvements have been demonstrated, studies indicate that its onset is typically within 48–72 hours from the diagnosis or admission.4 ACP likely represents an independent risk factor for mortality in ARDS, as it does in sepsis and pulmonary embolism.5 We present three patients on VV ECMO for severe ARDS who developed ACP and propose potential mechanisms. The Cedars Sinai institutional review board (IRB) and privacy officer have both approved the case reports. In 2015, these cases represented three of seven total cases on VV ECMO for >14 days and comprised all cases of VV ECMO for >28 days at our institution.
Patient 1 was a 32 year old male, with no past medical history presenting with cough, fever, and pancytopenia. One week later, he was intubated for hypoxemia, hypercarbia, and tachypnea (intubated for a total of 39 days, tracheostomy placed 13 days post cannulation). Despite 5 days of lung protective ventilation, paralytics and inhaled nitric oxide plateau pressures (Pplat 25–30 cm H2O), PaCO2 (58 mm Hg) and oxygen requirements (PaO2 50–60 mm Hg/FiO2 1.0) continued to rise, and VV ECMO (right internal jugular [RIJ] Avalon catheter) was inserted. chest x-ray (CXR) demonstrated bilateral granular airspace opacification consistent with ARDS. Our ventilation strategy on ECMO includes TV under 6 ml/kg, positive end-expiratory pressure and respiratory rate set to 10, plateau pressures <25, and FiO2 titrated to PaO2 >60/SpO2 >88% on pressure controlled ventilation. Patients are typically sedated for the first week with subsequent lightening to a target Richmond Agitation and Sedation Score (RASS) of −1 to 0 in the absence of any cardiopulmonary instability. Echocardiography revealed normal right ventricular (RV) function and a normal TV with mild tricuspid regurgitation (TR). Left ventricular (LV) function was normal, ejection fraction (EF) of 55%, with the aortic valve absent of stenosis or insufficiency. Inflow/outflow cannula was in suitable position with flow across the TV at 4–5 L (we aim to achieve a cardiac index >2.5 L/min). Complications included recurrent GI bleeds and episodic hypoxia/hypercapnia associated with agitation or malpositioning of the cannula. To this point, imaging revealed stable left-/right-sided function. The only finding was a small posterior-lateral pericardial effusion without tamponade. First evidence of ACP was 6 weeks post cannulation. transthoracic echocardiogram (TTE) at that time showed development of moderate RV dysfunction associated with severe TR (RV/right atrium pressure gradient was 71 mm Hg). Pulmonary artery (PA) pressure was 86 mm Hg (estimated using modified Bernoulli’s equation), and the IVC was dilated with no collapse. The previously noted pericardial effusion was unchanged. LV function remained preserved with an EF of 57%. Over the next week, despite inotropic support and aggressive diuresis, RV function deteriorated until ECMO day 54, when the patient became hypotensive, leading to cardiac arrest and CPR. Echocardiography revealed a severely depressed RV with septal flattening in systole/diastole (indicative of RV pressure and volume overload), severe TR, and PA pressures to 108 mm Hg, with tamponade indicated by left atrial diastolic inversion. The patient was intubated and placed on inhaled nitric oxide, epinephrine, and milrinone. The patient’s right heart did respond to these therapies, and therefore, alternative cannulation (venoarterial (VA)/VAV- venoarterial-venous (VAV)) strategies were not required. After this event, the patient suffered anoxic brain injury that lead to the decision to change the goals of care to comfort measures after which the patient died.
Patient 2 was a 59 year old male with asthma, hypertension, chronic kidney disease, and hyperlipidemia who presented with small bowel obstruction. During induction of general anesthesia, the patient had massive vomitus of gastric contents and aspiration. Post operatively, he developed ARDS with bilateral pulmonary opacification presumably from severe aspiration pneumonitis. Ten days later, he was placed on VV ECMO (RIJ Avalon catheter) in the setting of refractory hypoxemia/hypercarbia (PaO2 54 mm Hg/PaCO2 59), despite maximal ventilatory support and neuromuscular blockade (intubated for a total of 55 days with ventilator settings in accordance to our clinical goals discussed in patient 1). Initial echocardiograms revealed hyperdynamic LV systolic function (EF >60%) with normal function of the RV with no valvular abnormalities. In this case, severe RV pressure and volume overload developed 1 month into his ECMO run. The RV was dilated with flattening of the septum in systole/diastole. PA pressure was 75 mm Hg with mild TR and dilation of the IVC. LV function was preserved with an EF of 65%, and the position of the cannula was satisfactory with adequate flow by color Doppler. Inotropes and aggressive diuresis were initiated, and in view of the fact that the patient responded to these interventions, alternative variations of cannulation (VA/VAV) were not required. Two weeks after ACP, the patient became hypotensive, necessitating escalation of vasopressors and inotropes (epinephrine/dobutamine). He was taken to interventional radiology for pelvic angiogram with suspicion of retroperitoneal bleed. On the same day, despite maximal inotropes, inhaled nitric oxide, and mechanical support, the patient died.
Patient 3 was a 23 year old male with no past medical history transferred from an outside hospital for hypoxemia refractory to mechanical ventilation diagnosed with severe ARDS. Etiology was suspected to be viral; however, despite extensive workup, there was no confirmation. Two days after transfer, the patient was placed on veno-venous ECMO (RIJ Avalon catheter) for worsening oxygenation (PaO2 57 mm Hg) on maximal ventilatory support (intubated for a total of 97 days with ventilator settings in accordance to our clinical goals discussed in patient 1). CXR demonstrated bilateral alveolar opacification. Initial echocardiogram revealed an LV EF of 70%, normal RV size and function, and no valvular abnormalities. The course was complicated by GI hemorrhage (clipped duodenal ulcers), seizures, cholecystitis, and dysrhythmias. ACP was encountered 6 weeks post ECMO. Multiple echocardiograms revealed severe RV dilation, with systolic dysfunction and TR. PA pressure was >60 mm Hg, and the septum was flattened with dilation of the IVC. No abnormalities were seen with the LV, and EF remained >60%. With inotropic support, aggressive diuresis, and nitric oxide, the patient’s RV recovered, and there was no need for the addition of veno-arterial extracorporeal support. The patient was maintained on ECMO for an additional 21 days during which lung transplant evaluation occurred; however, he ultimately died from septic shock (Figures 1–3).
Despite variations in history, presentation, and course, ACP became prevalent in these cases of VV ECMO 4 to 6 weeks after Avalon cannulation. Although no studies account for the rates of ACP in VV ECMO, ACP occurs in 25% to 50% of ARDS patients.6 In ARDS without ECMO, mechanisms include hypoxia/hypercapnia and lung fibrosis. For patients on ECMO, one must consider pulmonary emboli and the consequences of ongoing nonphysiologic flow across the TV and right heart. In studies demonstrating ACP in ARDS without ECMO, patients frequently have abnormal echocardiographic findings within 48–72 hours.4 In our experiences with VV ECMO, we have yet to encounter early onset ACP, instead noting ACP after 4 to 6 weeks of support. Although we cannot exclude these patients who simply developed ACP as a result of ARDS, the late onset raises the potential for ACP in these cases to be related to the provision of ECMO. Additionally, it is possible that ECMO protects the RV initially by improving gas exchange and allowing less injurious ventilator settings only for some patients to succumb to late ACP caused by overwhelming progression of disease or long-term ECMO specific causes of ACP. All patients described here responded to standard therapies for acute right heart failure (inotropes, diuretics, and inhaled nitric oxide); however, alternative variations of cannulation (VA/VAV) that decompress, bypass, or otherwise protect the right heart should be considered in refractory cases.
Despite early ECMO, periods of hypoxia/hypercapnia are still prevalent as a result of malpositioning, shunting, or clot burden in the circuit/oxygenator. As alveolar PaO2 decreases, smooth muscle in the pulmonary vasculature undergoes hypoxic pulmonary vasoconstriction,7 increasing pulmonary vascular resistance. To a lesser effect, hypercapnia results in acidosis, leading to rises in PA pressure and RV dysfunction.7
The histologic features of ARDS evolve with the duration of disease.8 These include inflammation, exudation, proliferation, and fibrosis. Lung scarring and vascular constriction—from direct/indirect lung injury—occurs within 3 weeks of disease onset.8 The proposed mechanism involves myo-intimal thickening and mural fibrosis.7 Our findings of late ACP mean that these changes had already occurred and did not contribute to ACP and that its etiology was related to ECMO, or that the provision of ECMO delayed these changes because they allowed for lung rest and protection with less ventilator induced lung injury.9
Thromboembolic phenomena have been documented in ECMO. Auvil et al10 retrospectively demonstrated that 37% of patients developed thromboembolism when heparin was held for >24 hours. Contributing variables include exposure to artificial surfaces from tubing, connectors, and oxygenator and endothelial injury from cannulation. The extent of hemodynamic response is related to the clot burden in the pulmonary vasculature. Determinants of course include size/location of embolus, pre-existing RV function, and severity of inflammation driven by vasoreactive substances. Without compensatory mechanisms, RV intolerance of acute increases in afterload may result in ACP and cardiovascular collapse.6
With ECMO support, patients are exposed to 4–5 L/min of nonpulsatile blood flow from the outflow port across the TV and into the right heart. In a recent report, Lee et al11 postulated a causal mechanism related to long durations of nonpulsatile flow, leading to inhibition of ventricular recoil and therefore function. As experience grows with prolonged ECMO, a relationship between nonphysiologic flow and ACP should be considered.2 Investigation is needed to confirm this hypothesis and, if correct, may create rationale for alternative cannulation strategies.
In conclusion, we present three cases of prolonged ECMO and normal RV function upon cannulation, who developed ACP later in the course than is usually described for non-ECMO ARDS patients. Potentially modifiable causes include thromboembolic burden to the pulmonary vasculature, hypoxemia, and acidosis. Nonmodifiable causes of ACP in this setting may include the pathologic progression of ARDS and chronic nonphysiologic flow to the right heart. Greater awareness of late onset ACP on ECMO may lead to the discovery that the phenomenon is more common than recognized and hopefully a better understanding of the etiology so as to allow for modifiable interventions.
1. Peek GJ, Mugford M, Tiruvoipati R, et al.; CESAR trial collaboration: Efficacy and economic assessment of conventional ventilatory support versus extracorporeal membrane oxygenation
for severe adult
respiratory failure (CESAR): a multicentre randomised controlled trial. Lancet 2009.374: 13511363.
2. Posluszny J, Rycus PT, Bartlett RH, et al.; ELSO Member Centers: Outcome of adult
respiratory failure patients receiving prolonged (≥14 days) ECMO. Ann Surg 2016.263: 573581.
3. Javidfar J, Brodie D, Wang D, et al. Use of bicaval dual-lumen catheter for adult
venovenous extracorporeal membrane oxygenation
. Ann Thorac Surg 2011.91: 17631768; discussion 1769.
4. Repessé X, Charron C, Vieillard-Baron A. Acute cor pulmonale in ARDS: rationale for protecting the right ventricle. Chest 2015.147: 259265.
5. Romero-Bermejo FJ, Ruiz-Bailen M, Gil-Cebrian J, Huertos-Ranchal MJ. Sepsis-induced cardiomyopathy. Curr Cardiol Rev 2011.7: 163183.
6. Vieillard-Baron A, Schmitt JM, Augarde R, et al. Acute cor pulmonale in acute respiratory distress syndrome
submitted to protective ventilation: incidence, clinical implications, and prognosis. Crit Care Med 2001.29: 15511555.
7. Aaronson PI, Robertson TP, Knock GA, et al. Hypoxic pulmonary vasoconstriction: mechanisms and controversies. J Physiol 2006.570(pt 1): 5358.
8. Thille AW, Esteban A, Fernández-Segoviano P, et al. Chronology of histological lesions in acute respiratory distress syndrome
with diffuse alveolar damage: a prospective cohort study of clinical autopsies. Lancet Respir Med 2013.1: 395401.
9. Slutsky AS, Ranieri VM. Ventilator-induced lung injury. N Engl J Med 2013.369: 21262136.
10. Auvil B, Mattiola R, Pechulis R, Wu J. Outcomes of patients on extracorporeal membrane oxygenation
(ECMO) for periods of time without anticoagulation at LVHN in the past 3 years. Poster presented at LVHN Research Scholar Program Poster Session, Lehigh Valley Health Network, Allentown, PA.
11. Lee S, Jung J, Chung J, et al. Right heart failure
during veno-venous extracorporeal membrane oxygenation
for H1N1 induced acute respiratory distress syndrome
. Korean J Thorac Cardiovasc Surg 2015; 48(4): 289293.