The Right Ventricle During Veno-Venous Extracorporeal Membrane Oxygenation in Acute Respiratory Distress Syndrome: Can We Protect the Injured Ventricle?

Abstract

Right ventricular (RV) and pulmonary vascular injury is common in acute respiratory distress syndrome (ARDS) and is a major determinant of adverse outcomes.1,2 The reported prevalence of RV injury (defined as RV dysfunction with hemodynamic compromise or RV failure or acute cor pulmonale) in ARDS is 21%.1 In a recent pooled meta-analysis of nine ARDS studies, the presence of RV injury was associated with significantly higher overall and short-term mortality.1 Recent data suggests that RV injury in the context of severe lung injury caused by the coronavirus disease 2019 (COVID-19), occurs in approximately 20% of patients and it is associated with a threefold increase in all-cause mortality.3 Veno-venous extracorporeal membrane oxygenation (VV ECMO) is increasingly being used as part of the algorithm to treat ARDS patients in whom conventional measures fail to preserve adequate gas exchange and mitigate ventilator-induced lung injury.4 By reversing hypoxemia, hypercapnia, and acidemia, VV ECMO theoretically reduces RV afterload leading to an improvement in pulmonary hemodynamics.5 However, RV injury can persist despite initiation of VV ECMO or even develop late in the course of VV ECMO support.6 The purpose of this editorial is to briefly discuss the pathophysiology and mechanisms of RV injury during VV ECMO support for ARDS, and potential strategies to mitigate injury and support the RV. Right Ventricular Injury in Acute Respiratory Distress Syndrome Dissociation and uncoupling between RV and pulmonary artery (PA) biomechanics is key in understanding the mechanisms of different RV injury phenotypes leading to RV failure and death. The end-systolic and pulmonary arterial elastance (Ees and Ea, representing RV contractility and afterload, respectively) are the main determinants of RV-PA coupling.7,8 Pulmonary vasoconstriction caused by hypoxemia, hypercapnia, and acidemia in patients with ARDS, as well as high driving pressure and high intensity of invasive ventilation (i.e. mechanical power) lead to the development of acute pulmonary arterial hypertension (PAH).7–9 As a consequence, the RV dilates (heterometric adaptation), together with a reduction in Ees:Ea ratio (<1) resulting in RV-PA uncoupling. This renders the RV unable to meet the flow demands without excessive use of Frank-Starling mechanism and systemic congestion ensues.7,8 There is currently lack of a consensual RV injury definition. In a recent systematic review of ARDS studies, the most frequently used definition was the composite of size ratio and elevated RV pressure or septal dyskinesia.10 In this article, we use RV injury as an umbrella term that encompasses one or more different RV and pulmonary vasculature pathologies (echocardiography phenotypes) which affect RV-PA coupling such as: acute PAH, RV dilatation, RV dysfunction, acute cor pulmonale, and RV failure. Does Veno-Venous Extracorporeal Membrane Oxygenation Unload the Injured Right Ventricle? Veno-venous ECMO reverses hypoxemia, hypercapnia, and acidemia when mechanical ventilatory support alone becomes insufficient in managing severe acute respiratory failure. This results in a reduction in pulmonary vasoconstriction caused by the aforementioned factors and a decrease in PAH and RV afterload.8,9 In a prospective observational study of ARDS patients requiring VV ECMO in whom a pulmonary artery catheter was inserted before ECMO cannulation, initiation of VV ECMO was associated with an immediate substantial decline in PA pressure and an increase in cardiac index.11 The effect of carbon dioxide (CO2) on pulmonary vasculature in the context of VV ECMO support for ARDS has been previously highlighted in a study by Schmidt et al12 whereby reduction in sweep gas flow resulted in a significant increase in PAP. Use of low flow extracorporeal CO2 removal in moderate to severe ARDS has been recently shown to improve RV systolic function parameters implying that adequate CO2 clearance potentially improves pulmonary hemodynamics.13 Retrospective data (including data from the extracorporeal life support organization registry) has shown that initiation of VV ECMO in ARDS patients requiring vasoactive hemodynamic support resulted in a significant reduction in vasoactive drug doses with a low conversion rate to veno-arterial (VA) ECMO, possibly indicating RV unloading after commencement of VV ECMO.14,15 The effect of invasive ventilation on pulmonary vascular resistance (PVR) and RV afterload in ARDS patients depends on its net effect on alveolar and extraalveolar vessels.16 The increased lung volume and transpulmonary pressure (alveolar pressure–pleural pressure) during positive pressure ventilation results in collapse of alveolar vessels and dilatation (outward tethering) of extraalveolar vessels with the increased resistance of the alveolar vessels outweighing the falling resistance of the extraalveolar vessels.16 In ARDS with nonrecruitable lungs the overall effect is an acute rise in PVR and RV afterload16 (Figure 1).Figure 1.: Pre-extracorporeal membrane oxygenation (ECMO) acute cor pulmonale in a patient with coronavirus disease 2019 (COVID-19) related acute respiratory distress syndrome (ARDS). A: Transthoracic apical four chamber view; (B) transthoracic echocardiography parasternal short axis view; and (C) pulsed-wave doppler showing pulmonary hypertension.Veno-venous ECMO allows for a reduction in alveolar, plateau, and driving pressures (plateau pressure–total positive end-expiratory pressure [PEEP]) known to be risk factors for RV injury in severe ARDS.2,17 The resultant drop in transpulmonary pressure off-loads the RV by reducing PVR and PA pressure.11,17 In a prospective observational study, Lazzeri et al18 examined the prevalence and prognostic value of echocardiographic abnormalities in patients with moderate to severe ARDS requiring VV ECMO, before initiating extracorporeal support. Right ventricular injury was common (PAH with or without RV dilatation in 66.2% and impaired RV function in 28.4% of cases), and importantly, RV dilatation pre-ECMO implantation was an independent predictor of intensive care unit mortality despite ongoing VV ECMO support.18 In a more recent retrospective cohort study that included patients with severe ARDS supported with VV ECMO, Ortiz et al19 found that postcannulation RV injury (defined as RV dilatation and abnormal septal motion) was associated with decreased survival to liberation from ECMO and hospital discharge. Late-onset RV injury (defined as acute cor pulmonale), four to six weeks after VV ECMO support initiation has also been described (Figure 2).6,20Figure 2.: Right ventricular injury despite respiratory extracorporeal support. Transthoracic echocardiography-apical four chamber view on (A) extracorporeal membrane oxygenation (ECMO) day-1; and (B) ECMO day-35. The figure shows late-onset acute cor pulmonale, 35 days after veno-venous (VV) ECMO support initiation.Mechanisms of Right Ventricular Injury on Veno-Venous Extracorporeal Membrane Oxygenation Veno-venous ECMO provides a degree of RV protection by correcting reversible physiologic factors which affect pulmonary hemodynamics (hypoxemia, hypercapnia, and acidemia) and mitigates biophysical lung injury to a degree by facilitating a reduction in driving pressure shown to adversely affect RV function when ≥18 cm H2O.2,11 The presence or persistence of RV injury during VV ECMO support and despite correction of gas exchange and implementation of a least damaging ventilatory approach may be linked to different mechanisms of RV-PA uncoupling such as: (1) partial reversal of metabolic abnormalities19; (2) ongoing pulmonary vascular dysregulation19; (3) presence of untreated pulmonary embolism or microvascular thrombosis or immunothrombosis21; (4) reduction in PEEP immediately after initiation of ECMO which may result in atelectasis whereby extraalveolar vessels resistance increases overcoming the decreased alveolar vessels resistance causing high PVR16; and (5) continuous non-pulsatile flow through the tricuspid valve into the RV, especially in cases of prolonged support, which has been postulated by different authors to be a potential causative mechanism for RV injury on VV ECMO.11,20 The aforementioned proposed mechanistic links, however, are notions which must be confirmed in well-designed prospective studies. Right Ventricular Support and Mitigation of Further Injury on Veno-Venous Extracorporeal Membrane Oxygenation Right ventricular monitoring using echocardiography or pulmonary artery catheter with RV port (Paceport, Edwards Lifescience, Irvine, CA), for real-time RV and PA pressure monitoring pre- and post-ECMO implantation is paramount and could aid risk stratification and an individualized approach to RV injury management.22 Pharmacological treatment options include vasoactive drugs which improve RV perfusion pressure and RV-PA coupling such as: vasopressors (e.g. norepinephrine and vasopressin), inodilators (milrinone, enoximone, and levosimendan), inopressors (epinephrine), and drugs which reduce PVR and RV stroke work such as pulmonary vasodilators (prostaglandins or nitric oxide).23–27 Nonpharmacological therapies include (1) ultraprotective ventilation following the ECMO to Rescue Lung Injury in Severe ARDS trial ventilatory strategy ensuring optimal PEEP, especially in the first three days of ECMO support and low driving pressure12,17,28; (2) prone ventilation, which is feasible and safe during VV ECMO, and promotes RV unloading but its effect on persistent RV injury during VV ECMO has not been investigated to date29–31; and (3) mechanical RV support including: intraaortic balloon pump,32 conversion of VV ECMO to veno-arterial-venous (VAV) ECMO,19 and veno-pulmonary arterial (V-PA) ECMO.33–36 Veno-Pulmonary Arterial Extracorporeal Membrane Oxygenation in Acute Respiratory Distress Syndrome The role of mechanical RV support in severe respiratory failure refractory to conventional management has been examined recently in patient cohorts with COVID-19 and pre-ECMO evidence of RV injury.33–35 Mustafa et al34 used percutaneous right ventricular assist device (RVAD) with oxygenator to support 40 patients with COVID-19 lung and RV injury.34 Veno-pulmonary (V-PA) ECMO was initiated using a 29 or 31 Fr Protek-Duo right atrium to pulmonary artery TandemHeart cannula (CardiacAssist Inc., Pittsburgh, PA) under fluoroscopic and echocardiographic guidance providing flows up to 4.5 L/min (Figure 3).34 In addition to mechanical RV support and extracorporeal oxygenation all patients received a bundle of care which included: early extubation, early corticosteroid therapy, high dose anticoagulation, inhaled pulmonary vasodilators, and diuretics.34 Survival to hospital discharge was 73% but it is unclear whether the outcome benefit can be attributed to RV mechanical support from the outset or the combination of ECMO, early rehabilitation and pharmacological bundle of care.34 More recently, the same group used propensity match analysis to compare COVID-19 patients on V-PA ECMO (dual-stage right atrium to pulmonary artery cannula) to ECMO-eligible COVID-19 patients on maximum ventilation alone and the effect of those interventions on outcomes.33 Veno-pulmonary arterial ECMO led to a threefold survival benefit compared with maximum ventilation support alone and that was observed across all age and body mass index groups.33 In a small retrospective analysis, Cain et al35 found that early use of percutaneous RVAD (at the time of ECMO initiation) may improve mortality in patients with severe COVID-19 lung injury. It is possible that early mechanical RV unloading/support in the context of severe acute respiratory failure combined with extracorporeal oxygenation, which allows application of lung- and RV-protective ventilation, accounts for the observed outcome.Figure 3.: Veno-pulmonary extracorporeal membrane oxygenation (ECMO) using a 29 Fr Protek-Duo from the right atrium to pulmonary artery. A: Protek-Duo (white arrows) in a chest –X Ray; (B) mid-esophageal right ventricular (RV) inflow-outflow view in transesophageal echocardiography showing the course of the Protek-Duo through the left ventricle (LV) and across the pulmonary valve (white arrows).In conclusion, RV injury in ARDS increases mortality despite evidence-based practices and it is often overlooked by intensive care clinicians who focus primarily on ARDS management. Understanding the mechanisms of RV-PA uncoupling would potentially aid early risk stratification and implementation of personalized approaches to monitor and mitigate RV injury. Such approaches may include application of V-PA ECMO in ECMO-eligible ARDS patients with echocardiographic evidence of RV injury who fail to improve with conventional management, combined with RV-unloading adjuncts such as proning and pulmonary vasodilators. Although the signal is a beneficial effect of the proposed interventions, the feasibility, outcome benefit and importantly cost-benefit must be examined in large multicentre randomized trials comparing standard VV ECMO with RV extracorporeal support technologies.36