Extracorporeal membrane oxygenation in the 21st century: A decade of change*

Although it is uncommon for case reports to have accompanying editorials, the presentation of a patient on extracorporeal membrane oxygenation (ECMO) with leukemia, infection, prolonged respiratory failure, persistent bacteremia, and an ECMO duration of 38 days in this issue of Pediatric Critical Care Medicine raises a number of issues that seem worthy of mention (1). It also highlights the recent increase in attention on extracorporeal life support from experience with pH1N1 (2), the publication of the randomized adult ECMO trial (3), and the wealth of new equipment that is making ECMO safer and easier for a wider variety of patients than ever before (4). The first issue to comment on is the use of ECMO in a child with underlying leukemia—a condition that was a contraindication to ECMO in years past and still a source of reticence in many centers. However, several recent reports challenge this old dogma and demonstrate that ECMO can be applied to patients with malignancy with good long-term survival obtained. A review of 107 children reported to the Extracorporeal Life Support Organization registry with malignancy who were placed on extracorporeal cardiopulmonary life support between 1994 and 2007 found that 42% were able to be weaned off ECMO support and that 35% survived to hospital discharge (5). Similar results have been reported for adults (6). These reviews should, at a minimum, provide ground for discussion among critical care providers and oncologists who are faced with a patient with cancer who has refractory pulmonary or cardiac dysfunction and failing conventional care rather than dismissing any consideration for extracorporeal support. Given changing cancer treatment regimens that may affect outcome and advances in ECMO technology and management, ongoing analysis is required to maintain recommendations that are based on current-day information. Other issues of note are the septic state of the patient coupled with neutropenia. These characteristics have been considered contraindications to ECMO in the past. Similar to cardiopulmonary bypass, ECMO has been associated with the systemic inflammatory response and an increase in cytokine production (7–9). Thus, exposing an infected patient to yet another stimulus for cytokine acceleration seems illogical and perhaps ill advised, especially in the face of a compromised immune system. There has also been concern that exposure to ECMO will alter the ability of white blood cells to fight infection, making the successful resolution of sepsis even more unlikely and the use of ECMO even more futile. However, comparison of neutrophil activity in neonates on ECMO vs. normal newborns noted that patients on ECMO had improved phagocytic ability and intracellular killing ability, even up to 5 days of ECMO exposure (10). The persistence of bacterial infection in the current patient may be related to adherence of bacteria to extracorporeal tubing and cannulas. Successful eradication of bacteria and fungal elements have been noted, however, in other ECMO reports (11, 12). A precise description of neutropenia apart from other factors that may be related to survival has not been well described in the ECMO population. The extent and duration of neutropenia in intensive care unit patients, even those not treated with ECMO, has been associated with risk of infection and outcome. Finally, the duration of mechanical ventilation before ECMO has been reported to correlate with survival (13). The patient in this report had repetitive need for mechanical ventilation and was outside the recommended 7-day window suggested by the Extracorporeal Life Support Organization guidelines (14). This recommendation is based on previous work that found a decrement in survival in patients on ECMO who received ECMO after 7 days of mechanical ventilation. In contrast, colleagues in Sweden and other sites have reported good survival with patients exposed to much longer durations of ventilation before ECMO (15, 16). Further exploration of this area to determine more precisely levels of ventilator support, duration, or other factors that indicate the extent of damage to the lung is an area of needed refinement. The authors also comment on the prolonged duration of ECMO. Just like in many other areas of ECMO support, there are little concrete data on how to predict futility or when support should be withdrawn. Signs of massive intracranial bleeding or intractable multiple organ failure often make the need to withdraw obvious. Patients who are “stable” but not improving, however, are more difficult. Factors that predict irreversible organ damage and can thus indicate when support is futile are not well defined. Clinicians must assess the overall milieu of the patient, ongoing resource use issues, and potential benefit to continuation of support to help guide the course of treatment. Prior work from the 1990s has shown that patients have been discontinued from ECMO with similar characteristics as those who survived with prolonged support; this work bears repeating in the current era (17). Taken together, the characteristics of the patient presented in this simple case report demonstrate how far the field of ECMO has advanced over the past decade (18). It also serves to point out how much further we have to go to more fully understand what patients may benefit or be harmed by expanding the horizons of extracorporeal support. Ongoing data collection, validation, and analysis are paramount to guide appropriate use of this technology in the future. Heidi J.