Extracorporeal Membrane Oxygenation and Cardiac Devices

Introduction

Extracorporeal membrane oxygenation or ECMO refers to a number of configurations of extracorporeal circulatory and/or respiratory support. As the technology evolved, other acronyms such as MCS (mechanical circulatory support) and ECLS (extracorporeal life support) have been used, but in North America ECMO continues to be the most commonly used term to describe a circuit, pump, and oxygenator that can perform the work of the heart and lungs by adding oxygen and removing carbon dioxide from circulating blood for prolonged intervals. Venovenous (VV) ECMO withdraws venous blood and returns oxygenated blood to the right side of the heart supporting only respiration; venoarterial (VA) ECMO withdraws venous blood and returns oxygenated blood to the arterial system, thereby supporting both respiration and circulation. Another configuration is venous-pulmonary artery (VPA) ECMO used to support the right heart and the lungs when there is right heart failure and respiratory failure but the left heart is not supported.

An ECMO circuit comprises cannulae inserted into large vessels to remove and return the blood from the patient, tubing to connect to a nonpulsatile centrifugal pump that generates blood flow, and an oxygenator, where an oxygen-air mixture flows through the blood, referred to as the “sweep.” This is a closed system without a reservoir, with some or all of the circuit components surface bonded with heparin, and is designed for extended use, such as days or weeks. This is in contrast to the extracorporeal circuit used for cardiopulmonary bypass (CPB) in the operating room that uses larger-diameter and longer tubing (larger “prime” volume), and is an open system with a reservoir designed to receive input not only from the venous cannula but also from the operative field. Such CPB circuits may or may not be heparin bonded, and are intended for surgical use of short duration (e.g., hours). There is some evidence in the lung transplant literature that use of ECMO during the surgical procedure may be associated with a smaller systemic inflammatory response than the use of the traditional CPB.

Pumpless extracorporeal lung assist (pECLA), or the Novalung, uses the patient’s arterial pressure rather than a pump to drive blood through an extracorporeal oxygenator. It therefore supports only respiration. This device requires less anticoagulation than traditional ECMO but due to lower flows and the membrane area also has a more limited capacity especially for oxygenation. It has been used to support patients with pulmonary hypertension awaiting lung transplantation, implanted in a pulmonary artery to left atrium configuration. It is available in North America but is less widely used than ECMO and will not be discussed further.

History of Extracorporeal Cardiorespiratory Support

The history of ECMO is inextricably bound to the development of CPB for cardiac surgery. The first successful use of CPB in a human was by Gibbon in 1953, to repair an atrial septal defect in an 18-year-old patient. The following year, Warden and colleagues reported cardiac surgery using extracorporeal circulation, after which there was an ever increasing number of reports from many centers. Among the major limitations in these early reports was the use of “bubble” oxygenators where oxygen is bubbled through a reservoir of blood to achieve gas exchange. These oxygenators are associated with trauma to formed elements of the blood and coagulopathy with prolonged use. The development of oxygenators with membranes separating flow of the respiratory gases from the blood reduced these effects, making the oxygenators suitable for longer-term use. Membrane oxygenators have replaced bubble oxygenators in cardiac surgery and their development led to the first successful reports of prolonged support outside of the operating room, in 1972 by Hill and associates in a 24-year-old trauma patient, and by Bartlett after neonatal cardiac surgery. In 1985 Bartlett and associates reported the successful use of ECMO for neonatal respiratory failure in 11 patients. In the following two decades a number of additional trials showed the benefit of ECMO in neonatal respiratory failure, the most definitive of which was published in 1996 by the UK Collaborative ECMO Trial Group in 185 infants. In the same year Green and colleagues published a trial demonstrating similar benefit in older pediatric patients with respiratory failure. Note the majority of use in the first decades was for respiratory failure, using VV ECMO. During this period the Extracorporeal Life Support Organization or ELSO was founded, first at the University of Michigan in 1989, then a European ELSO group formed in 1991. This organization has played a key role in documenting worldwide ECMO use, and pioneering education, research, and development in all types of ECMO support in all populations. Fig. 85.1 and Table 85.1 illustrate the increase in use of ECMO and the survival data since 1990 ( www.ELSO.org ).

ECMO for Respiratory Failure (VV ECMO)

In contrast to the ECMO experience in neonatal and pediatric respiratory failure as previously described, demonstration of benefit in adults took much longer, delayed in part due to the publication of a trial in 1979 by Zapol and associates in 90 adult patients with respiratory failure. This trial had many limitations, including the use of VA rather than VV ECMO, patient selection, anticoagulation technique and bleeding complications, and the use of standard ventilation at the time—relatively high-tidal volume and low positive end-expiratory pressure (PEEP). Poor outcomes deterred adult use of ECMO for more than 20 years. From 2001 to 2006 a large, ambitious British trial, the CESAR trial, was performed to evaluate VV ECMO for respiratory failure in adults. This study was performed during the H1N1 influenza pandemic, and involved transferring patients with severe respiratory failure to a central expert ECMO center, where they were randomly assigned to ECMO or standard therapy. Despite some methodological and statistical limitations, the results supported the use of ECMO performed in a specialized center to improve survival: 63% versus 43% survival with ECMO versus standard treatment. At the same time, another report (case series) of VV ECMO in adults with severe acute respiratory distress syndrome (ARDS) due to H1N1 was reported from Australia and New Zealand (ANZ ECMO). This study found a 79% survival at 30 days in patients who received ECMO. Recently a large multicenter trial of VV ECMO in adults with severe ARDS, the EOLIA trial, was published in 2018, with the authors concluding no difference in mortality between ECMO and conventional therapy at 60 days: 35% versus 46% ( P = .09), with 28% of the control group crossing over to ECMO after randomization and a 57% mortality in this crossover group. Editorial comments on this study have challenged this conclusion, maintaining that it supports the use of early ECMO in adults with severe ARDS. The ELSO database reports survival to discharge with ECMO for adult respiratory failure as 60% with this percentage being relatively stable over 15 years.

Extracorporeal Membrane Oxygenation for Patients Awaiting and Undergoing Lung Transplantation

The history of ECMO for severe respiratory failure as described previously refers mostly to patients with acute or acute on chronic disease where recovery could be anticipated. Another population is patients with end-stage chronic lung disease awaiting lung transplantation. These patients often have a slow (in years) deterioration in function and an increased need for oxygen support, with the final stage being an acute deterioration where standard therapy with mechanical ventilation ultimately fails. Institution of ECMO to prolong survival until transplant, use of ECMO during transplant surgery, and extension or initiation of ECMO postoperatively for primary graft dysfunction (PGD) or other indications have all become common applications for this advanced therapy. As recently as 2010 there was concern that the use of ECMO resulted in a reduced long-term survival from lung transplantation, but this is no longer the case in 2018. Case reports, single center reports, and surveys have documented that the use of pretransplant ECMO, sometimes for months, was followed by successful transplant and good long-term outcomes. Raleigh and associates compared 10 studies on the use of preoperative ECMO for lung transplant patients and found similar outcomes to the patients who did not need ECMO and Loor and associates described factors that contributed to posttransplant survival in this population. Intraoperative use of ECMO during the procedure has also been shown to reduce the inflammatory response and leads to less PGD when compared to CPB and leads to improved short-term and longer-term outcomes even when compared to no use of either CPB or ECMO intraoperatively. While preoperative ECMO may be VV, VA, or VPA, intraoperatively VA ECMO is usually used due to surgical manipulation of the heart with its attendant hemodynamic compromise as well as the need for one-lung ventilation in patients with end-stage pulmonary disease and elevated pulmonary vascular pressures. Postoperatively, ECMO support may be needed to support the new lungs in the face of PGD, the right heart, the left heart, or any combination of these. Postoperative ECMO support is also associated with excellent outcomes.

ECMO for Circulatory Failure (VA ECMO)

VA ECMO can be used to support the heart and lungs temporarily in a patient with poor cardiac function who undergoes an invasive cardiology procedure, to continue postoperative cardiopulmonary support in a cardiac surgery patient who fails to separate from CPB, and in a patient with refractory cardiac failure with or without associated respiratory failure. These conditions may occur due to an acute recoverable illness (e.g., myocarditis) or may be in the setting of acute on chronic heart failure in patients being evaluated for long-term advanced therapies such as a durable ventricular assist device or cardiac transplant. Urgent use of ECMO in the acute setting of in-hospital cardiac arrest (ECPR) is also practiced in some centers. Finally, as VA ECMO can be instituted at the bedside without imaging, it may be the preferred technique for emergent cannulation in any form of respiratory or cardiac failure, with elective conversion to another form (e.g., VV ECMO) once the patient has stabilized.

The historical perspective for use of VA ECMO for cardiac or combined cardiorespiratory support in adults is illustrated in the 2016 report from the ELSO registry. Adult cardiac ECMO was in its infancy in 1990 with little increase in use until 2006 when its use began to increase exponentially; there were more than 2000 adult cardiac ECMO runs reported to ELSO in 2015, comprising an ever-increasing proportion of all ECMO runs ( Fig. 85.2 ). This exponential increase was likely fueled by consistent success in the neonatal and pediatric populations; improvements in the ECMO circuits, pumps, and oxygenators; and the success and experience with VV ECMO for adult respiratory failure. Overall survival in adults who receive ECMO for cardiac indications is approximately 40% with a slight increasing trend over the last 10 years.

The Ethics of Extracorporeal Membrane Oxygenation

The initiation of any form of ECMO is lifesaving when the heart, or lungs, or both are failing despite maximum medical therapies. It is a very invasive and labor-intensive therapy, with associated severe complications, and may confine a patient to an ICU for days, weeks, or even months. As the practice has evolved over time, many groups have tried to address issues of patient appropriateness, possible exclusion criteria, and prognosis before initiating the therapy. An essential consideration is that ECMO is a “bridge” to something else and cannot be viewed as a long-term solution; Box 85.3 lists the uses of ECMO as a “bridging” therapy to a variety of possible scenarios. The following discussion addresses only the individual patient ethical dilemmas, and not the overriding issue of the use of a limited availability, expensive, and labor-intensive therapy with its cost:benefit implications, and overall implications for health care systems.

When an otherwise relatively young and healthy patient develops an acute severe illness resulting in acute cardiac failure and shock (e.g., viral cardiomyopathy) or acute refractory lung failure (e.g., viral pneumonia), the decision to initiate lifesaving extracorporeal support as a “bridge to recovery” seems relatively straightforward. Similarly, a patient with end-stage disease of any kind who already has a “do not resuscitate” status would likely not be a candidate for ECMO if the heart suddenly failed. Unfortunately, the clinical spectrum of potential candidates runs as a continuum between these two examples. From the patient and family, through all levels of caregivers and decision makers, tools to help assess the likelihood of successful outcome with this advanced therapy are needed.

The ELSO registry database has been used to study the likelihood of survival prior to ECMO initiation, both in respiratory failure ( Table 85.2 ) and cardiac failure ( Table 85.3 ). There are many common elements in these publications, including the duration and degree of respiratory support, age, other organ function, and acidosis. While these publications can be used as a general guide and can be quoted to referring physicians and families, they have to be placed in clinical context; decision making for life-sustaining treatment must be patient-specific. Courtwright and associates make a point of the need to emphasize to the patient’s family the “bridging” nature of ECMO therapy, and that a destination must be formulated at the outset or early on. Because of the need for anticoagulation and the risk for bleeding or thrombosis, patients who are not candidates for anticoagulation (e.g., intracranial hemorrhage) are generally not candidates for ECMO, even though more experience is accumulating with reduced or even no anticoagulation with the use of anticoagulant bonded cannulae, tubing, pump heads, and oxygenators. Patients with underlying severe disease who are not expected to survive more than some predetermined period (i.e., 6 months or a year) independent of the need for ECMO are unlikely to be considered as candidates. Rather than have a single physician or surgeon determine whether ECMO should be initiated in a given patient, especially when a request comes from an outside hospital, many institutions make this a shared decision by a small committee (i.e., 3 individuals) who are all familiar with and participate in ECMO management.

Overall survival for adult respiratory and cardiac ECMO is approximately 60% and 40%, respectively. Although this is certainly a major advance for diseases that were previously not survivable, the other side of this coin is that mortality remains at 40% and 60%. It makes good sense to engage palliative care and/or an ethics committee and other counselling services, where available and appropriate, at the outset for this therapy. When ECMO is being considered but there is uncertainty about the likelihood of recovery or candidacy for advanced durable therapy, discussions with family and caregivers should be similar to those made regarding “do not resuscitate,” assessing the values and goals of the patient. Counselling for ICU caregivers as well as family, including postmortem “debriefings,” can be very valuable to help staff deal with end-of-life issues. There are few settings where withdrawal of therapy can so immediately result in death, where patients have intact neurologic function but are on the “bridge to nowhere”; it can be extremely troubling to all involved when ECMO is stopped.

The Mechanics of Extracorporeal Membrane Oxygenation