Extracorporeal membrane oxygenation

VV ECMO support is a potential management option for patients with severe acute respiratory failure, in whom mechanical ventilation and adjunctive therapies such as inhaled nitric oxide cannot provide adequate gas exchange (Box 5). ECMO facilitates “lung rest” by reducing ventilation demands, allowing lung-protective strategies to be employed, and minimising ventilator-induced lung injury. It is appropriate for reversible, acute lung disease, but not chronic progressive conditions. Patients who have already received prolonged mechanical ventilation are unlikely to be appropriate candidates.

Until the early 1980s, VA ECMO was predominantly used in respiratory failure. However, VV ECMO is now the preferred modality for isolated respiratory disease because it provides oxygenated pulmonary blood flow, maintains systemic pulsatile flow, and avoids the risks of arterial cannulation (such as systemic arterial embolisation).

VA ECMO support is appropriate for managing patients with severe cardiac failure, whose condition is refractory to other management strategies (such as intra-aortic balloon pumps), and patients with intractable, in-hospital cardiac arrest (Box 5). In these settings, ECMO can prevent end-organ damage, prevent further cardiac damage due to shock, enable timely decisions regarding the diagnosis and treatment of underlying conditions, and facilitate selection of the most appropriate clinical end point. ECMO may be used as a rescue therapy in patients with reversible disease aetiologies, or as a bridge to heart transplantation or ventricular assist device insertion. In patients who cannot be weaned from bypass after cardiac surgery, ECMO can provide support until the recovery of heart function. Use of ECMO for cardiac support after heart transplantation is well established, and relied upon in cases of primary graft failure — which are more likely with recent trends such as the use of donor tissue with long ischaemic times.5 It should not be offered to patients who are unlikely to recover cardiac function and who are not suitable for ongoing mechanical support or heart transplantation.

Data from four randomised controlled trials support the use of ECMO in neonatal respiratory failure.8-12 The first adopted an unorthodox “play-the-winner” design: neonates with severe respiratory failure had a probability of being randomly assigned to the intervention or control arms that was influenced by the outcome of the preceding patients in each arm. With the survival of the first patient, who was assigned to the ECMO group, and the death of the second patient, who was assigned to the control group, 10 subsequent patients were randomly assigned to receive ECMO therapy and survived, leading to the conclusion that ECMO provided a survival benefit (P = 10-8).7,8

Despite the theoretical validity of this trial design, the uneven distribution of patients invited considerable scepticism, and more conventionally randomised controlled trials followed.9-12 The largest of these demonstrated that, in 185 neonates with severe respiratory failure, referral for ECMO (VA or VV) conferred a relative risk of death before 1 year of age of 0.55 (95% CI, 0.39–0.77),12 and data from follow-up studies suggested that an advantage in terms of death or disability persisted at 7 years.13

A subsequent meta-analysis reported an increased survival for neonates with severe respiratory failure who are supported with ECMO. Relative risk of death before discharge was 0.44 (95% CI, 0.31–0.61) for neonates receiving ECMO compared with those receiving conventional mechanical ventilation, or 0.33 (95% CI, 0.21–0.53) when neonates with congenital diaphragmatic hernia were excluded.14

No randomised studies of ECMO for respiratory support in children have been published, although a retrospective multicentre analysis of 29 ECMO patients with one or two matched controls revealed a significant reduction in mortality for children supported with ECMO compared with those who received conventional support (26% v 47%, P < 0.01).15

The history of research into ECMO use in adult respiratory failure reflects the challenges of studying new invasive technologies. A prospective multicentre randomised controlled trial conducted between 1974 and 1977 showed no survival benefit above conventional management,16 “effectively stopping research on ECMO for adult respiratory failure for the next 15 years”.7 Owing to the use of VA ECMO, the overly inclusive selection criteria, and inexperience of the researchers with the technology, it is now widely acknowledged that this study no longer has clinical applicability.17 Widespread interest in ECMO in adult respiratory failure was revived in the late 1980s, when the significance of ventilator-induced lung injury was identified.18

A study of a low-flow form of VV ECMO for extracorporeal carbon dioxide removal (ECCO2R) in combination with lung-protective ventilation showed promising but uncontrolled results in patients with respiratory failure. In patients with a 90% mortality risk before the commencement of extracorporeal life support, treatment with ECCO2R and “rest ventilation” was associated with 48.8% survival.19 However, a subsequent randomised controlled trial evaluating this technique in 40 patients with acute respiratory distress syndrome revealed no 30-day survival benefit for patients receiving ECCO2R compared with those receiving conventional mechanical ventilation.20 Although a high rate of haemorrhagic complications and inexperience of the researchers with extracorporeal support were blamed for the trial’s negative results, ECCO2R was largely abandoned.21

Centres with staff experienced with ECMO continued to offer high-flow ECMO (providing full oxygenation) as respiratory support for adults, and soon began to demonstrate survival rates exceeding 50% for patients likely to die if given conventional treatment.22-24 These results have been attributed to strict selection criteria (including restrictions regarding prior mechanical ventilation), the use of VV ECMO, and minimal concurrent ventilation.

Success with the use of ECMO at various institutions has culminated in the recent conclusion of a randomised controlled trial of ECMO for adult respiratory failure. The Conventional ventilation or ECMO for Severe Adult Respiratory failure (CESAR) trial enrolled 180 patients aged 18–65 years from 2001 to 2006.21,25 Patients were randomly assigned to be referred for ECMO at Glenfield Hospital (Leicester, United Kingdom) or to receive conventional treatment (typically at the referring hospital). Promising results from this trial have been presented and are pending publication (Peek GJ, presentation at Extracorporeal Life Support Symposium, Melbourne, 29 May 2008).

As no controlled studies of ECMO use in cardiac failure have been conducted, an analysis of registry data provides the best indication of the performance of ECMO for cardiac support. Aggregate data from the ELSO registry to July 2009 suggest that for neonatal (age 0 to 30 days), paediatric (age 31 days to < 16 years) and adult (age 16 years or over) patients receiving ECMO, survival to discharge or transfer was 38%, 46% and 35%, respectively, per period of support.2 In 2008, survival in these groups was 43%, 54% and 35%, respectively, despite a perceived increase in the severity of illness in supported patients over time.2 However, these data are difficult to interpret in isolation, the adult dataset is limited, and ongoing evolution of the indications for ECMO restricts the use of historical controls. In addition, for various reasons, ELSO aggregate data are skewed to reflect the support of patients with congenital defects, who exhibit higher mortality rates than most other patients in whom ECMO is indicated. The highest survival rates were observed in neonates with cardiomyopathy (63%) and older patients with myocarditis (66%).2