Journal of Pediatric Surgery (2005) 40, 1051 – 1057
Extracorporeal membrane oxygenation in pediatric cardiac transplantation Jae-O Baea, Jason S. Frischer a, Margarita Waicha, Linda J. Addoniziob, Eric L. Lazar a, Charles J.H. Stolar a,* a
Division of Pediatric Surgery, Columbia University, College of Physicians and Surgeons, Children’s Hospital of New York, New York, NY 10032-3784, USA b Division of Pediatric Cardiology, Columbia University, College of Physicians and Surgeons, Children’s Hospital of New York, New York, NY 10032-3784, USA Index words: Extracorporeal membrane oxygenation; Pediatric cardiac transplantation; Bridge to transplant; Congenital heart disease; Cardiac arrest
Abstract Background: We reviewed a single institution experience with extracorporeal membrane oxygenation (ECMO) in the perioperative management of cardiac transplantation. Methods: Of all pediatric cardiac transplant candidates (1984-2003), patients requiring ECMO pretransplantation/posttransplantation were identified, with particular attention to use of ECMO as a bridge to transplantation. Parameters reviewed included proportionate survival, incidence of pre-ECMO cardiac arrest, ECMO duration, and United Network for Organ Sharing list time. Results: Three hundred patients were listed for transplantation. Twenty-nine required ECMO: 18 pretransplant, 3 pretransplant and posttransplant, 6 posttransplant, and 2 for delayed acute rejection. There were 21 bridge-to-transplant candidates, of which 10 eventually transplanted with 60% survival; 11 not transplanted had no survivors ( P = .004). Thirteen of 21 had cardiac arrest pre-ECMO with 1 (8%) survivor; 8 of 21 had no arrest with 5 (63%) survivors ( P = .014). Mean ECMO duration and United Network for Organ Sharing list times between transplanted and not transplanted were not significant. Nine received ECMO posttransplantation for cardiopulmonary support; 5 (56%) of 9 survived. Two patients supported with ECMO for rejection-related cardiovascular collapse survived. Conclusion: ECMO can bridge children to cardiac transplantation. Survival is significantly impaired in bridge-to-transplant candidates stratified by pre-ECMO cardiac arrest. ECMO can also help transition from cardiopulmonary bypass after transplantation and provide effective support during acute rejection. D 2005 Elsevier Inc. All rights reserved.
Presented at the 56th Annual Meeting of the Section on Surgery of the American Academy of Pediatrics, San Francisco, California, October 8-10, 2004. This study was conducted with approval from the Columbia University Institutional Review Board; protocol IRB-AAAA6300 (Y1M00). T Corresponding author. Division of Pediatric Surgery, Children’s Hospital of New York, New York, NY 10032-3784, USA. Tel.: +1 212 305 2305; fax: +1 212 305 5971. E-mail addresses: [email protected]
(J.-O. Bae)8 [email protected]
(C.J.H. Stolar). 0022-3468/$ – see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.jpedsurg.2005.03.026
J.-O. Bae et al.
The role of extracorporeal membrane oxygenation (ECMO) in providing cardiopulmonary support in the setting of potentially reversible cardiac failure is well documented [1-3]. Over the past decade, the specific utility of ECMO in cardiac transplantation has been evolving. In addition to providing transitional support immediately after transplantation, ECMO has been increasingly used in the setting of irreversible cardiac failure, namely, as a bridge to heart transplantation. According to the January 2004 International Summary provided by the Extracorporeal Life Support Organization Registry, the overall survival outcomes surrounding ECMO use in transplantation are less than 50% . In an effort to delineate factors that may improve outcome and patient selection, we reviewed our institution’s experience with ECMO in the perioperative management of pediatric cardiac transplantation, with specific focus on the use of ECMO as a bridge to transplantation.
1. Materials and methods 1.1. Study population All pediatric cardiac transplant candidates listed for transplantation from January 1984 to December 2003 who required ECMO either before or after cardiac transplantation were retrospectively reviewed after obtaining approval from the Columbia University Institutional Review Board.
Records were evaluated for pre-ECMO clinical data: diagnosis of cardiac failure, incidence of pre-ECMO cardiac arrest (defined as requiring cardiopulmonary resuscitation with chest compressions), and duration on United Network for Organ Sharing (UNOS) list. Details of ECMO course were also evaluated: mode of cannulation, need for balloon atrial septoplasty/septostomy (BAS), duration of ECMO, and outcome. Survival was defined as survival to discharge from the hospital.
1.2. ECMO ECMO was initiated in patients with end-stage progressive heart failure refractory to all medical and surgical treatments, who had no clear contraindications to anticoagulation and ECMO (such as significant intracranial hemorrhage). Venoarterial ECMO was carried out via a Jostra HL-20 roller occlusion pump and standard membrane oxygenator. Cannulation was performed either centrally (transthoracic) or peripherally (via the neck, groin, or combination of both). Central cannulations were performed during cardiac surgery to correct congenital heart disease or after transplantation for postcardiotomy rescue. Peripheral cannulations were performed through the right neck in infants. In larger children, dual venous drainage (neck plus groin) with single arterial perfusion (neck or groin) was required to achieve adequate flow. One child was cannulated only through the right groin during an unsuccessful attempt
Listed for Orthotopic Heart Transplant n = 300
No ECMO n = 271
ECMO n = 29
Bridge to TXP n = 21
Non–Bridge to TXP n=8
(18 Pre-OHT; 3 Pre/Post OHT)
(6 Post -OHT; 2 Rejection)
Transplanted n = 10
Survived n=6 Fig. 1
Not Transplanted n = 11
Outcome of bridge to heart transplant patients.
Died n = 11
Extracorporeal membrane oxygenation in pediatric cardiac transplantation at extracorporeal cardiopulmonary resuscitation. Anticoagulation and blood product transfusions were managed according to standardized ECMO protocols.
1.3. Left heart decompression Most patients placed on venoarterial ECMO as a bridge to transplantation required left heart decompression to prevent severe left atrial hypertension in the setting of extremely poor left ventricular function. Unless there was a preexisting intracardiac shunt (patent foramen ovale, atrial septal defect [ASD], or ventricular septal defect [VSD]), left heart decompression was achieved by a blade/balloon atrial septostomy. BAS was performed in the cardiac catheterization suite either before or on ECMO. Rarely, in the setting of an inadequate preexisting atrial septal defect, a bedside septostomy was performed under echocardiography. In the last 3 years, we instituted a multidisciplinary (surgery, cardiology, and anesthesiology) approach to stabi-
lization with ECMO and BAS for the patient who presented with progressively deteriorating cardiac function. The patient was brought to the cardiac catheterization suite with a primed ECMO circuit. After intubation, cannulation was accomplished by percutaneous or open techniques. ECMO was not initiated at that point unless cardiac arrest occurred. BAS was then accomplished followed by ECMO initiation. If the patient was on ECMO before BAS, all catheter and wire exchanges for the septostomy were performed submerged in a saline-filled basin to prevent air embolism.
1.4. Transplantation Criteria for UNOS listing included diagnosis of cardiac failure deemed irreversible or uncorrectable in patients who had no contraindications precluding transplantation. All ECMO patients were listed as either UNOS status 1 (pre-1999) or status 1A (after 1999), which is the category for children with severe heart failure requiring multi-
Characteristics of bridge-to-transplant patients
Patient Age (mo)
UNOS ECMO Transplanted Cannulation Pre-ECMO Left heart Outcome Cause of (d) (h) site arrest decompression death
1 2 3 4 5 6
14 11 6 4 24 72
CHD Cardiomyopathy Cardiomyopathy CHD CHD Cardiomyopathy
5 44 24 98 6 38
168 349 533 185 262 586
Yes Yes Yes Yes Yes Yes
0.13 CHD 35 CHD
No RCA/RIJ/ RFV Cervical Yes Right iliacs Yes
CHD Cardiomyopathy CHD
Thoracic Cervical Cervical Thoracic Thoracic RIJ/RFV/ LFA Thoracic Thoracic
No No Yes No No No No Yes
A A A A A A
CHD 51 Cardiomyopathy 21
0.23 CHD 4 72 Cardiomyopathy 46
Thoracic RCA/RIJ/ RFV
D ASD BAS
D D D D
MSOF Neurological injury Hemorrhage— other MSOF Hemorrhage— other Withdrawal of care Neurological injury MSOF Neurological injury MSOF Withdrawal of care Neurological injury Neurological injury MSOF Withdrawal of care
A indicates alive; D, dead; CHD, congenital heart disease; RIJ, right internal jugular vein; RFV, right femoral vein; RCA, right carotid artery.
1054 ple inotropes and/or mechanical ventilation or mechanical assist devices.
1.5. Data analysis Proportionate survival and pre-ECMO cardiac arrest data were evaluated by Fisher’s Exact test using Epi Info 3.2.2. ECMO duration and UNOS list times were compared using Student t test.
2. Results From January 1984 to December 2003, 300 patients were listed for cardiac transplantation with 185 transplanted (including 9 retransplants). Of the 300 listed patients, 29 required ECMO: 18 pretransplant, 3 both pretransplant and posttransplant, 6 immediately posttransplant only, and 2 for adjuvant therapy in the treatment of delayed acute rejection. The median age of the patients who required ECMO was 17 months (range 4 days to 13.7 years).
2.1. Bridge-to-transplant group There were 21 bridge-to–heart transplant ECMO patients: 18 pretransplant (bridge to transplant only) and 3 patients requiring pretransplant and posttransplant ECMO (bridge to transplant and posttransplant cardiac rescue for failure to wean from bypass) (Fig. 1). Of these 21 children, 10 were eventually transplanted with 60% (6/10) survival. Eleven were not transplanted and were removed from support. Withdrawal of support was based on complications from ECMO precluding transplantation or family desires and was not based on organ availability or any arbitrary time limit. The indications for discontinuation of ECMO were neurological devastation in 4, multisystem organ failure (MSOF) in 3, withdrawal of care by family request in 3, and ongoing pulmonary hemorrhage in 1. There were no survivors among these 11 patients compared with the 60% survival in patients transplanted ( P = .004). No patient placed on ECMO for cardiac failure as a bridge to transplantation was weaned from ECMO without transplantation. Of the 21 bridge-to-transplant patients, 13 had left heart decompression (3 with preexisting ASD, 10 required BAS). Duration of ECMO, UNOS list time, mode of cannulation, and outcome data are detailed on Table 1. Of those eventually transplanted, the mean and median duration of ECMO were 285 and 224 hours, respectively (range 24 -586 hours). In contrast, the mean and median duration of support for those who were not successfully bridged to a transplant were 172 and 144 hours, respectively (range 20 -597 hours). Although the mean and median ECMO duration were longer for those transplanted than not transplanted, this was not significant ( P = .15 by Student t test). Likewise, mean UNOS list time for those transplanted (27 days) versus not transplanted (21 days) was not significant ( P = .62).
J.-O. Bae et al. Primary cardiac diagnosis requiring ECMO included congenital heart disease in 12, cardiomyopathy in 8, and myocarditis in 1. Congenital heart disease patients had 25% (3/12) survival, whereas cardiomyopathy patients had 38% (3/8) survival. Survival between these 2 groups was not significant ( P = .46). The 1 patient with myocarditis did not survive. There were 15 deaths in the bridge-to-transplant group (Fig. 2). Primary causes included neurological injury (intracranial hemorrhage or infarction) in 5, MSOF/sepsis in 5, hemorrhage (nonneurological) in 2, and withdrawal of care by family request in 3. Of the bridge-to-transplant patients, 13 had cardiac arrest before initiation of ECMO, with 23% (3/13) successfully bridged and transplanted. Eight patients did not have a pre-ECMO arrest, with 88% (7/8) successfully bridged to transplantation. An arrest before ECMO was associated with significantly decreased rate of transplantation ( P = .007). Examining the variable of pre-ECMO cardiac arrest independent of whether the patient was successfully bridged, 13 had a cardiac arrest pre-ECMO with only 8% (1/13) survival; 8 had no arrest with 63% (5/8) survival ( P = .014) (Table 2). The majority of patients with cardiac arrest pre-ECMO (9/13) had severe progressive cardiac failure on multiple inotropic medications and/or mechanical ventilation and were already listed for transplantation for an average of 28 days. Three of 8 patients who did not have cardiac arrest pre-ECMO were likewise on inotropes and/or mechanical ventilation and were already listed for transplantation for an average of 24 days before the placement of ECMO. The remaining 5 patients were placed on ECMO in the operating room when they failed to be weaned from bypass despite maximal pressor therapy and were listed for transplantation either at that time (n = 3) or 6 days later (n = 2). The difference in mean UNOS list time of pre-ECMO cardiac
Hemorrhage - Other 2 MSOF 5
Withdrawal of Care 3
Neurologic Injury 5 Hemorrhage - Other
Withdrawal of Care
Cause of mortality in bridge-to-transplant candidates.
Extracorporeal membrane oxygenation in pediatric cardiac transplantation Table 2 survival
Impact of cardiac arrest on bridge-to-transplant
Total Survived Died % Survival
No cardiac arrest
13 1 12 8%
8 5 3 63%
P = .014.
arrest patients (20 days) and patients without pre-ECMO arrest (29 days) was not significant ( P = .50).
2.2. ECMO posttransplant group Of the 300 patients listed for heart transplantation, 185 were transplanted with 9 patients requiring ECMO posttransplantation for cardiopulmonary support: 6 had purely posttransplantation, and 3 required support both as a bridge to transplantation and posttransplantation. In these 9 patients, 5 (56%) survived to hospital discharge. Cause of mortality included 3 patients with MSOF and/or sepsis and 1 patient with uncontrollable postoperative mediastinal hemorrhage.
2.3. ECMO for delayed acute rejection Two patients were supported with ECMO 5 and 11 weeks after transplantation for rejection-related cardiovascular collapse. Duration of ECMO support was 5 and 6 days, respectively, during which time the patients received antirejection therapy. Myocardial improvement was documented by echocardiography, and both patients were successfully weaned from ECMO and survived to hospital discharge. However, the second patient had recurrent rejection 3 months after her discharge and died. ECMO was not offered during that recurrent rejection.
3. Discussion 3.1. Bridge to transplantation The role of ECMO in pediatric end-stage cardiac failure is evolving to encompass supporting the patient before and after transplantation. The lack of appropriate ventricular assist devices in children remains a challenge. ECMO can function to bridge patients to transplantation but is neither implantable nor usable for more than a few weeks. In an earlier multicenter review on the role of ECMO in the perioperative support in pediatric heart transplantation, 20 patients were placed on ECMO as an adjunct to cardiac transplantation, with 4 being placed as a bridge to cardiac transplantation . Only 1 of the 4 survived. Other series have demonstrated increased use of ECMO as a bridge to cardiac transplantation with improved survival of 43% to 48% [6-8]. In this single institution review of all transplant-related use of ECMO (n = 29), 21 (72%) patients were placed on extracorporeal life support as a bridge to cardiac transplan-
tation. Although the aggregate survival of these patients was 29% (6/21), those who were successfully bridged to a cardiac transplant (ie, survived on ECMO until transplanted) had 60% (6/10) survival. Conversely, the group that was not successfully bridged (ie, never received a transplant) had no survivors. Survival in the bridge-to-transplant ECMO group was dependent on whether the patient was transplanted ( P = .004). Stated another way, those who were maintained on ECMO until an organ was available for transplantation, without incurring a fatal complication, had statistically increased survival compared with those who were unable to be supported until organ availability. Although successful wean from ECMO without transplantation is reported, we had no such patients in the bridge-to-transplant group [6,7]. Although mean and median duration of ECMO support were longer for those bridge-to-transplant patients who were eventually transplanted than those who did not survive until transplantation, this was not statistically significant ( P = .15). The shorter duration in the nontransplanted group may reflect the increased premorbidity of this subgroup: 91% (10/11) patients in the nontransplanted group had a pre-ECMO arrest, compared with 30% (3/10) patients in the transplanted subgroup. Neurological complications and MSOF/sepsis accounted for most (10 of 15) of the mortality in the bridge-to-transplant group. They are the main factors limiting prolonged extracorporeal life support. Decreasing the incidence of these main complications may increase the matriculation of those being transplanted and improve overall survival.
3.2. Impact of cardiac arrest on survival Cardiac arrest before or at the time of ECMO cannulation has been shown to be a negative prognostic indicator in children with severe cardiac failure . Recently, however, Kolovos et al  reported no difference in survival in patients with treated cardiac arrest before cannulation; although those having cardiopulmonary resuscitation during cannulation did have significantly worse survival. Del Nido  likewise reported improved survival in children placed on ECMO for resuscitation of witnessed cardiac arrests with 64% (7/11) weaning from ECMO and surviving. However, no series has delineated an association of pre-ECMO cardiac arrest in children listed for cardiac transplantation with successful bridging and survival. In our experience, cardiac arrest before initiating ECMO in the end-stage cardiac failure child significantly impairs survival. An arrest before initiating extracorporeal life support was associated with significantly decreased rate of transplantation than those without an arrest (23% vs 88%; P = .007). Because survival was dependent on successful bridging to transplantation, patients with arrests were less likely to be transplanted and therefore survive. Survival of those with pre-ECMO arrest was statistically worse than those without an arrest (8% vs 63%; P = .014). The national average of UNOS time to pediatric heart transplant is 2.1 to
1056 2.3 months . In this study, the mean UNOS list time for patients who had an arrest was only 20 days and reflects that the high mortality in these patients is a result of the premorbidity of this cohort and not necessarily caused by prolonged waiting for organ availability. Given the significantly impaired survival in bridge-to-transplant patients with cardiac arrests before ECMO, the data suggest that earlier ECMO intervention performed in a bmore electiveQ manner, before arrest, may increase survival. In an effort to improve outcome and decrease the incidence of pre-ECMO cardiac arrests, we have recently instituted a multidisciplinary approach to stabilization with ECMO and BAS in the patient with deteriorating cardiac function. In anticipation of requiring ECMO, the patient is brought to the cardiac catheterization suite with a primed ECMO circuit. If not already intubated, an anesthesiologist intubates the patient in a controlled environment with the pediatric surgeon, and operating team prepared for immediate cannulation. Interventional cardiology then performs the BAS followed by initiation of ECMO. ECMO is not initiated before the septostomy unless cardiac arrest occurs. Our proactive approach of anticipatory ECMO is an evolving experience, and the impact on survival has yet to be determined.
3.3. ECMO posttransplantation ECMO after cardiac transplantation usually results from failure to wean from operative cardiopulmonary bypass. The transplanted heart may require prolonged support, beyond the scope of normal operative bypass circuits, for myocardial recovery. In patients with secondary pulmonary hypertension from end-stage cardiac disease, a period of support may be required for the pulmonary vascular resistance to normalize. In this series of 185 transplanted patients, 9 required ECMO after transplantation. Five (56%) of 9 survived to discharge.
3.4. ECMO for delayed acute rejection There are several case reports of the use of extracorporeal life support in the setting of graft failure [12,13]. Transplanted patients with acute rejection can have rapidly deteriorating cardiac function which may require ECMO support, for a defined period, as an adjunct to antirejection therapy. Our experience (n = 2) is anecdotal, but serves only to report that ECMO has been used to successfully provide full cardiopulmonary support in the setting of acute rejection until recovery of myocardial function. The utility of extracorporeal life support in this setting has yet to be definitively determined.
J.-O. Bae et al. and complications inherent to prolonged extracorporeal life support, specifically, neurological accidents, MSOF, and infection. In bridge-to-transplant ECMO patients, survival depends on successful transplantation and is significantly impaired by pre-ECMO cardiac arrest. Because cardiac arrest before ECMO in bridge-to-transplant candidates is associated with impaired survival, earlier ECMO intervention, before arrest, may increase survival. ECMO may also help transition from cardiopulmonary bypass after transplantation and provide effective support during acute rejection.
References  Bartlett RH, Gazzaniga AB, Fong SW, et al. Extracorporeal membrane oxygenator support for cardiopulmonary failure. Experience in 28 cases. J Thorac Cardiovasc Surg 1977;73(3):375 - 86.  Meliones JN, Custer JR, Snedecor S, et al. Extracorporeal life support for cardiac assist in pediatric patients: review of ELSO registry data. Circulation 1991;84(5 Suppl):III168 - 72.  Dalton HJ, Siewers RD, Fuhrman BP, et al. Extracorporeal membrane oxygenation for cardiac rescue in children with severe myocardial dysfunction. Crit Care Med 1993;21(7):1020 - 8.  Extracorporeal Life Support Organization. International summary. 2004 [Ann Arbor (Mich)].  Galantowicz ME, Stolar CJ, King TC. Extracorporeal membrane oxygenation support as a bridge to pediatric heart transplantation. J Thorac Cardiovasc Surg 1991;102(1):148 - 52.  Gajarski RJ, Mosca RS, Ohye RG, et al. Use of extracorporeal life support as a bridge to pediatric cardiac transplantation. J Heart Lung Transplant 2003;22(1):28 - 34.  Fiser WP, Yetman AT, Gunselman RJ, et al. Pediatric arteriovenous extracorporeal membrane oxygenation (ECMO) as a bridge to cardiac transplantation. J Heart Lung Transplant 2003;22(7):770 - 7.  del Nido PJ, Armitage JM, Fricker FJ, et al. Extracorporeal membrane oxygenation support as a bridge to pediatric heart transplantation. Circulation 1994;90(5 Part 2):II-66 - 9.  Kolovos NS, Bratton SL, Moler FW, et al. Outcome of pediatric patients treated with extracorporeal life support after cardiac surgery. Ann Thorac Surg 2003;76(5):1435 - 41.  del Nido PJ. Extracorporeal membrane oxygenation for cardiac support in children. Ann Thorac Surg 1996;61(1):336 - 9.  Addonizio LJ, Zangwill SD, Rosenthal DN, et al. Have changes in UNOS status system improved allocation in pediatric heart recipients? J Heart Lung Transplant 2005 [Print in progress].  Ko WJ, Chen YS, Chou NK, et al. Extracorporeal membrane oxygenation rescue after heart transplantation. Transplant Proc 2000;32(7):2388 - 91.  Hoffman TM, Spray TL, Gaynor JW, et al. Survival after acute graft failure in pediatric thoracic organ transplant recipients. Pediatr Transplant 2000;4(2):112 - 7.
Scott Engum, MD, FAAP (Indianapolis, IN): Does your facility have the availability of other types of assist devices that were used in your patient population, and was ECMO just one of those?
ECMO can successfully bridge children to cardiac transplantation but is limited by premorbid conditions
Jae-O Bae, MD (New York, NY): In our experience, ECMO is the preferred method of support in the pediatric
Extracorporeal membrane oxygenation in pediatric cardiac transplantation population. In larger children, left ventricular assist device or biventricular assist device have been used. Unfortunately, 1 of the limitations of current ventricular assist device technology in bridging to transplantation is the inability of implanting these devices in smaller children. Half of the patients in our bridge-to-transplant group were younger than 1 year. Ronald B. Hirschl, MD, FAAP (Ann Arbor, MI): Very nice paper. I’m wondering how many patients went on ECMO for cardiac disease who were not transplanted? So in other words, you probably had a number of them that went on support for cardiac disease, did well, came off, and weren’t transplanted. And also I’m just curious, did these postoperative patients have cardiomyopathy? What were the sort of circumstances by which you put them on ECMO? Jae-O Bae, MD (New York, NY): This paper specifically focused on UNOS-listed pediatric transplant candidates who required ECMO and did not address non –transplantrelated uses of cardiac ECMO for postcardiotomy rescue or myocarditis. Ronald B. Hirschl, MD, FAAP (Ann Arbor, MI): Well let me follow up with that. What is your protocol? Maybe you can help us out here, because this is a big problem. When you put patients with cardiac disease on ECMO, at what point do you list them for transplantation? When do you decide that they are going to be transplant candidates? Jae-O Bae, MD (New York, NY): Transplant cardiologists are familiar with cardiomyopathy patients and have generally listed them before requiring ECMO. The timing of listing patients for postcardiotomy ECMO rescue depends on the cardiac surgeon’s subjective assessment on the likelihood of recovery. Two thirds of the postcardiotomy patients were listed simultaneously with initiating ECMO; the other third were listed after
3 to 6 days of ECMO. We are fairly aggressive about listing patients, especially those with cardiomyopathy. The only exception is the patient with myocarditis. Richard R. Ricketts, MD, FAAP (Atlanta, GA): This may be a similar question, but you did not have a category for myocarditis. Are you considering these patients to be in the cardiomyopathy group, or are they in a separate category? Jae-O Bae, MD (New York, NY): Of the 21 patients placed on ECMO as a bridge to transplant, 1 patient had myocarditis and did not survive. In the literature, there is a significant population of patients with myocarditis who are placed on ECMO and recover. Because in this study we started with patients who were already listed for heart transplantation and required ECMO, we had excluded myocarditis patients placed on ECMO who recovered without ever being listed. Myocarditis patients are not listed as aggressively as children with cardiomyopathies because of the high likelihood that they will recover with extracorporeal support alone. Arlet Kurkchubasche, MD, FAAP (Providence, RI): One of the conclusions of your paper was that earlier ECMO might improve survival. Do you have specific suggestions for revisions to the indications for ECMO which would enable earlier intervention with consequent patient salvage? Is this even a tenable proposition? Jae-O Bae, MD (New York, NY): Preparedness to provide bemergency ECMOQ requires considerable logistic and personnel commitments. It may not be realistic to expect all institutions to provide this. We have adopted a strategy of bpreemptive,Q as opposed to bemergencyQ ECMO preparation for specific patients at imminent risk of needing ECMO. This kind of preparedness to provide preemptive ECMO is dependent on good communication among the intensivist, cardiology service, and the ECMO team.