In the absence of lung function, death occurs from hypoxemia, hypercarbia, and cardiac arrest in 5–10 min. When gas exchange is supported by ECMO normal metabolism and hemodynamic physiology continues in the absence of any lung function. Total loss of lung function often occurs during ECMO support. There is no tidal volume despite mechanical ventilation and chest x-ray shows total “white out” with no gas in the consolidated lungs. Lung recruitment maneuvers cause barotrauma and pneumothorax. Usually (50–60% of the time) this is a transient condition (days) eventually lung function resolves, alveoli are inflated, the chest x-ray clears, and gas exchange recovers. The lung recovers totally to normal function over a short period of time. In this situation ECMO simply supports life (gas exchange, hemodynamics, metabolism) until recovery of the native lung occurs.
When there is no lung function after an arbitrary period of time, the lung injury is considered irreversible, ECMO is discontinued, and the patient dies. The time interval defining irreversible lung injury has gradually progressed from 7 to 14 days to 4 weeks of time as experience with ECMO has increased.
When there is no lung function, multiple organ failure, uncontrolled sepsis, and ECMO complications often occur as the time on ECMO increases. After a few weeks of time, pulmonary hypertension occurs by occlusion of the pulmonary vasculature from thrombus or edema. Right ventricular strain occurs, and patients eventually die of arrhythmias secondary to right ventricular failure. The combination of no lung function with these other morbidities leads to a diagnosis of futility and ECMO is discontinued. At postmortem examination, the lungs are airless, edematous, and consolidated. Histologically there is obliteration of alveolar airspace, extensive inflammation, often capillary occlusion, and areas of necrosis. The pathologist and the clinician agree that the lung injury was indeed irreversible.
In the last ten years, the equipment and management techniques for ECMO have minimized circuit complications so ECMO can be continued without significant side effects for weeks or even months of time. ECMO is now managed with the patient awake and interacting with family and the care team even though there is no native lung function. This combination of events has resulted in continuing ECMO support beyond the time when the lung injury appears to be “irreversible.” Surprisingly, lung function has returned in some of these patients after weeks or months of no lung function. After recovery from no lung function, the lung function is essentially normal. At first, there were rare anecdotal cases but there are now hundreds of patients who have recovered to normal lung function and normal health after months of what appeared to be irreversible lung injury.
1. Documentation of late lung recovery during ECMO support
Many cases of late lung recovery have occurred and many of these are documented in the literature. Some of these cases are in children, others are in adults. Kon demonstrated lung recovery in 8 of 11 adult patients after more than 21 days of ECMO support.1 Camboni reported 52% survival in 19 patients supported more than 21 days.2 Review of the ELSO Registry indicates 974 patients on ECMO support beyond 14 days with 45% ultimate survival.3
The pattern of late recovery begins with appearance of some tidal volume with alveolar inflation which progresses to significant lung ventilation in a period of a few days. Oxygenation improves as inflated alveoli are matched to blood flow. The pulmonary vascular occlusion which led to right ventricular failure resolves and pulmonary blood flow gradually returns to normal. ECMO flow is decreased as native lung function increases to the point where oxygenation by the native lung is carried out at low ventilator pressure and low inspired oxygen. At this point, oxygenation and pulmonary blood flow has recovered but frequently the patient has significant CO2 retention and has to remain on ECMO support primarily for CO2 removal. The physiologic pattern is that of very large alveolar level dead space. Extreme hyperventilation can clear CO2 but this is deleterious and leads to exhaustion if the patient is breathing spontaneously. This appears to be irreversible chronic lung disease with profound hypercarbia. However, in almost all cases, this resolves to normal CO2 clearance in a matter of days or weeks. The reason for this phenomenon is probably due to the fact that areas of necrotic lung become inflated with no blood flow. CO2 clearance is compromised because of the very large alveolar level dead space. These areas of necrosis are lined with collagen as in any healing tissue and as the collagen contracts, the pneumatoceles and alveolar spaces decrease leading to more normal ventilation/perfusion relationships and CO2 clearance.
The longest period of ECMO support resulting in lung recovery was in a 7 year old girl who recovered after 605 days of extracorporeal support.4 This case is particularly illustrative of the sequence of events which can occur during recovery from total lung failure. Pulmonary hypertension leading to right ventricular failure was managed by conversion to right atrium to pulmonary artery access and pumped ECMO for a year. Episodes of sepsis and bleeding were managed. The patient was alert and keeping up with her schoolwork by internet. After recovery of oxygenation, ECCO2 removal was required for two months. Two years later lung function is essentially normal.
This phenomenon of lung recovery after weeks of no function is unique to ECMO. These patients would have died if not for lung replacement with mechanical devices. The regenerative microbiology is the same as in less severe cases of acute lung injury, but we are learning that lung recovery can occur in some patients months, even years after total loss of native lung function.
How does late lung recovery occur? How can we predict it? Most importantly, how can we enhance it in time and intensity? The cause of lung recovery must be endogenous cells or mechanisms that create new pulmonary epithelium and endothelium. In lung tissue these primordial cells are probably alveolar type 2 cells with unique regenerative capability.5 Exogenous stem cells have been studied in mice and humans and do not cause significant regeneration.6 The most promising of these exogenous stem cell interventions has been to perfuse the supernate or “secretome” from cultures of stem cells rather than the cells themselves.7 Those responses are minimal; however, it may be that the products of exogenous stem cell culture might provide exosomes or other components which could stimulate the endogenous stem cells in the lung.
At some time after severe lung injury, endogenous cells must be responsible for generating new, functional lung tissue. What factors combine to stimulate these cells to generate new lung endothelium and epithelium? How can the potential for lung recovery be identified in patients on prolonged ECMO support? Is basic lung skeletal structure necessary? It would seem to be, otherwise regenerating stem cells would likely form organoids or balls of lung tissue which are not related to the airway or the heart. If that is the case, what component of the lung skeletal structure is necessary for lung regrowth? All of these questions are being investigated.
In summary, the newly recognized phenomenon of total lung recovery after apparently irreversible injury is possible only because of prolonged ECMO support. The impact of this phenomenon will increase not only the use of ECMO but the duration of ECMO with particular attention to maintaining patients awake and rehabilitated in the ICU. This creates a problem of chronic ICU utilization by these patients. This stimulates research efforts to build wearable artificial lungs which will allow prolonged extracorporeal gas exchange support with patients out of the ICU or even out of the hospital.
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
 Kon ZN, Dahi S, Evans CF, Byrnes KA, Bittle GJ, Wehman B, et al. Long-term venovenous extracorporeal membrane oxygenation support for acute respiratory distress syndrome. Ann Thorac Surg 2015;100(6):2059-2063. Epub 2015 Aug 18 PMID: 2629626.
 Camboni D, Philipp A, Rottenkolber V, Zerdzitzki M, Holzamer A, Floerchinger B, et al. Long-term survival and quality of life after extracorporeal life support: a 10-year report. Eur J Cardiothorac Surg 2017 Aug 1;52(2):241-247.
 Posluszny J, Rycus PT, Bartlett RH, Engoren M, Haft JW, Lynch WR, et al. LM; ELSO member centers. outcome of adult respiratory failure patients receiving prolonged (≥14 Days) ECMO
. Ann Surg 2016;263(3):573-581.
 Nelson K. Lung recovery after 20 months of extracorporeal support. JASAIO 2018. In press.
 Zacharias WJ, Frank DB, Zepp JA, Morley MP, Alkhaleel FA, Kong J, et al. Regeneration
of the lung alveolus by an evolutionarily conserved epithelial progenitor. Nature 2018;555:251-255.
 Akram KM, Patel N, Spiteri MA, Forsyth NR. Lung regeneration
: endogenous and exogenous stem cell mediated therapeutic approaches. Int J Mol Sci 2016;17(1).
 Laffley JG, Kavanagh BP. Fifth years of research in ARDS
. Insight into acute respiratory distress syndrome. From models to patients. Am J Respir Crit Care Med 2017;196:18-28.