Extracorporeal Membrane Oxygenation for Pulmonary Support

Origin and Technical Evolution of Membrane Oxygenators

Frey and Gruber performed the first experience with extracorporeal oxygenation of isolated organs at the end of 19th century. Fifty years later, John Gibbon started his laboratory investigations that led, in the mid-1950s, to the first use of a cardiopulmonary bypass in the operating room. Unfortunately the bubble oxygenators used at that time directly exposed blood to the oxygen flow, thus leading to an elevated risk of hemolysis and hemorrhage resulting from the high demand for systemic anticoagulation. Clowes tried to tackle this problem, introducing a cellophane membrane with the attempt to separate gas from blood phase. Over the years, with the purpose of optimizing oxygenators, different materials (silicone, polypropylene, polyethylene, polymethylpentene) and types of surfaces (microporous or continuous) were developed and tested. In the 1970s, Kolobow developed his membrane lung consisting of a silicone membrane long envelope containing a spacer net, wound around a central plastic spool. Oxygen flowed inside the envelope through the space created by the spacer net, while blood flowed in the outside. Kolobow's membrane lung earned great success and was used extensively, even for prolonged extracorporeal supports, for the following 30 to 40 years.

The latest generation of artificial lungs is represented by hollow-fiber devices with different designs (shell/tube and cross-flow). These membranes are included in circuits built of new materials such as polymethylpentene and heparin-coated surfaces, with great improvement of biocompatibility of the whole extracorporeal system.

Clinical Pioneering of Extracorporeal Respiratory Support

The first successful use of long-term ECMO was reported by Hill in 1972 in an adult patient with posttraumatic respiratory failure. Four years later, Robert Bartlett applied the first successful treatment in newborns. Several encouraging case series and two prospective randomized controlled studies with positive results made ECMO the standard treatment for neonatal respiratory failure.

Following the enthusiasm for this new technique, the National Institutes of Health (NIH, Bethesda, MD) sponsored a randomized trial comparing venous-arterial ECMO to conventional mechanical respiratory support in adult patients suffering from severe acute respiratory distress syndrome (ARDS). The trial was stopped for futility after the enrollment of 90 patients, because the mortality in both groups was around 90%. At that time the main concern for mechanically ventilated patients was identified with the high inspired oxygen fraction and not with the harm of ventilation. Therefore in this trial, the only difference concerning mechanical ventilation between treatment and control patients was a lower FiO ₂ in the ECMO group, whereas all the other parameters, today known as determinants of lung injury, were similar between the two arms. In addition, ECMO was applied with a venoarterial configuration, leading to severe increases in the ventilation-perfusion ratios of the native lungs. Moreover, the risks and complications were high, particularly bleeding, with a reported transfusion approximating 5 L of blood per day. The discouraging results of the NIH study led to the abandonment of the ECMO technique worldwide.

Gattinoni and Kolobow nevertheless proposed to exploit the ECMO support to “rest” the injured lungs by reducing respiratory rate, tidal volume, and airway pressure (low frequency positive pressure ventilation, LFPPV), thus fostering lung healing. The concept, when pushed to the extremes, involved the application of continuous positive airway pressure, maintained by a continuous oxygen supply to compensate for the natural lung oxygen consumption while carbon dioxide removal would be granted by the artificial lung (ECCO2R), with no need for any ventilation (apneic oxygenation).

This strategy first was applied in spontaneously breathing healthy sheep and, as a rescue treatment, in premature lambs. In 1980 it was applied successfully on three patients with refractory acute respiratory failure in whom conventional treatment had failed. Those patients were treated with an extremely low respiratory rate (3 bpm), thus avoiding possible pulmonary and extrapulmonary complications related to mechanical ventilation, while CO ₂ was removed through a venovenous extracorporeal bypass with low blood flow. The results of a clinical study designed to evaluate the effects of LFPPV-ECCO2R in 43 patients with severe acute respiratory failure were published in 1986. Lung function improved in 73% of the cases, and survival rate was 49% ; no major technical accidents were reported in more than 8000 hours of perfusion. Zwischenberger et al. refined the LFPPV-ECCO2R technique, developing a simplified arteriovenous extracorporeal CO ₂ removal technology, AVCO2-R, featuring a low-resistance membrane gas exchanger.

In 1994 Morris published the results of a second randomized clinical trial. Pressure-controlled inverse ratio ventilation followed by LFPPV-ECCO2R (21 patients) was compared with conventional positive pressure ventilation (19 patients) in ARDS patients. The trial did not show an improved survival in the patients treated with the extracorporeal support (42% in the control group vs. 33% in the ECMO group). However, the survival rate was improved significantly compared with the previous trial. However, a lot of criticism has been raised, mainly regarding the inhomogeneous ventilatory settings applied in the ECMO group, the high peak pressure used, and the elevated number of complications related to systemic blood anticoagulation. After this trial, only few centers around Europe continued to provide venovenous extracorporeal support in selected series of patients, usually as a last resource. In the United States, Bartlett et al. continued to provide extracorporeal support as an alternative to mechanical ventilation with very encouraging results. Since 1989 The Extracorporeal Life Support Organization (ELSO) maintains the largest registry of data on patients receiving ECMO and provides yearly data about the worldwide use of this technique ( http://www.elso.med.umich.edu ).

Cesar Trial and H1N1 Flu Pandemics

A renewed interest on ECMO rose after the publication of the conventional ventilatory support versus extracorporeal membrane oxygenation for severe adult respiratory failure (CESAR) trial, a prospective randomized trial conducted in the United Kingdom. The trial clearly showed outcome advantages when patients with severe respiratory failure were centralized in an experienced center and eventually treated with ECMO if required, compared with conventional ventilatory support performed in peripheral centers. The study included 180 patients from 68 centers; 90 patients were enrolled in the ECMO group (68 effectively treated with ECMO) and 90 in the conventional treatment group. In the control group the intensivists could use any type of management they felt appropriate, but the NIH ARDS strategy was recommended. The authors found that the primary end point, survival at 6 months free of disabilities, was 63% in the ECMO group versus 47% in the control group, and they concluded that referral of severely hypoxemic ARDS patients to a specialized center able to provide ECMO may increase survival.

The study is characterized by two major limitations. First, not all patients allocated to the ECMO group received ECMO, because they either died before or during transportation (5 patients) or improved with conventional treatment after transportation to the ECMO center (17 patients). Second, the nonstandardized protocol for mechanical ventilation in the control group resulted in few patients receiving a protective ventilatory strategy. However, despite these limitations, the results of this study have certainly contributed to strengthen the motivation of “ECMO believers” and to increase the interest in ECMO around the world.

However, the widespread use of extracorporeal support was due to its use as a rescue therapy in Australia and New Zealand during the 2009 H1N1 flu pandemics. Between June and August 2009, 68 patients with severe H1N1 influenza-associated ARDS were treated with ECMO with a survival rate of 78%. Before ECMO institution, these patients, characterized by a median age of 34 years, had severe respiratory failure despite advanced mechanical ventilatory support. After the Australian experience, several countries worldwide instituted a national ECMO network and numerous case series, reporting 70% to 80% survival rates, were published.

In the United Kingdom, a cohort study compared ECMO-referred patients with matched patients who were not referred for ECMO. ECMO-referred patients were defined as patients with H1N1-related ARDS who were referred, accepted, and transferred to one of the four adult ECMO centers in the United Kingdom during the H1N1 pandemic in winter 2009 to 2010. The study clearly demonstrated an impressive advantage of this strategy compared with conventional mechanical ventilation. In Italy, as in other countries, a national network was established, reporting a survival rate of about 70%.

However, a similar case-matching study conducted in France did not show the same results. Although the French study was conducted with a different matching approach, causing some limitations, its results should be taken as well into account.

The H1N1 flu pandemic coupled with several technical improvements, promoted, far beyond any previous randomized trial, the use of ECMO in patients with severe ARDS.

The Extracorporeal Membrane Oxygenation for Severe Acute Respiratory Distress Syndrome (EOLIA) randomized control trial is currently ongoing in France (NCT01470703). This trial is testing the usefulness of ECMO as an adjunct to mechanical ventilation in patients with ARDS.