Available online at www.sciencedirect.com

www.elsevier.com/locate/semperi

S E M I N A R S I N P E R I N A T O L O G Y 3 8 ( 2 0 1 4 ) 9 7 – 1 0 3

0146-0005/14/$ - seehttp://dx.doi.org/10.

nCorresponding au98105-0371.

E-mail address:

Extracorporeal life support for the neonatal cardiacpatient: Outcomes and new directions

Andrew L. Mesher, MDb, and David Michael McMullan, MDa,b,n

aDivision of Cardiac Surgery, Seattle Children's Hospital, Seattle, WAbDepartment of Surgery, University of Washington School of Medicine, Seattle, WA

a r t i c l e i n f o

Keywords:

Neonate

Extracorporeal life support

Cardiac

Heart

Extracorporeal membrane

Oxygenation

front matter & 2014 Else1053/j.semperi.2013.11.00

thor at: Division of Cardi

michael.mcmullan@seat

a b s t r a c t

Extracorporeal life support is an important therapy for neonates with life-threatening

cardiopulmonary failure. Utilization of extracorporeal life support in neonates with

congenital heart disease has increased dramatically during the past three decades. Despite

increased usage, overall survival in these patients has changed very little and extracorpor-

eal life support-related morbidity, including bleeding, neurologic injury, and renal failure,

remains a major problem. Although survival is lower and neurologic complications are

higher in premature infants than term infants, cardiac extracorporeal life support

including extracorporeal cardiopulmonary resuscitation is effective in preventing death

in many of these high-risk patients. Miniaturized ventricular assist devices and compact

integrated extracorporeal life support systems are being developed to provide additional

therapeutic options for neonates.

& 2014 Elsevier Inc. All rights reserved.

Extracorporeal life support (ECLS) has become a criticallyimportant supportive therapy for neonates with life-threatening heart failure. Defined broadly as the use of anextracorporeal system to provide partial or full cardiopul-monary support, ECLS may be used to support neonateswith refractory cardiac and/or pulmonary failure as a bridgeto recovery, surgical intervention, or transplantation.Although recent advances in biotechnology have led to thedevelopment of miniaturized paracorporeal and implantableventricular assist devices (VAD) that may be useful inproviding medium- and long-term ECLS for young patients,1

extracorporeal membrane oxygenation (ECMO) remains, byfar, the most widely employed form of ECLS during theneonatal period.2 The clinical versatility of ECMO and thebreadth of clinical experience with this form of supportivetherapy have led to expanded indications and increasedutilization.

vier Inc. All rights reserv6

ac Surgery, Seattle Childr

tlechildrens.org (D.M. Mc

Trends in neonatal cardiac ECLS

Over 5000 cases of neonatal cardiac ECMO have been reportedto the Extracorporeal Life Support Organization (ELSO) inter-national registry. Categorically, neonatal cardiac ECMO rep-resents 9.2% of the total cases. The use of ECMO for primarilypulmonary indications has declined dramatically during thelast 20 years, whereas the use of ECMO to support neonateswith predominately cardiac failure has steadily climbed.3

Only 16 cases of neonatal cardiac ECMO were reported tothe ELSO registry in 1987, representing 2.4% of non-ECPRneonatal ECMO runs. The number of reported neonatalcardiac cases was 370 and represented 31.2% of all non-ECPR neonatal runs in 2012.2,4 This greater than 20-foldincrease in the utilization of ECMO is due to an increase inthe number of centers performing ECMO, greater clinicalfamiliarity, and expanded indications for mechanical

ed.

en's Hospital, 4800 Sand Point Way NE, M/S G-0035, Seattle, WA

Mullan).

S E M I N A R S I N P E R I N A T O L O G Y 3 8 ( 2 0 1 4 ) 9 7 – 1 0 398

circulatory support in neonates with life-threatening heartfailure. The proportional shift away from non-cardiac tocardiac ECMO is primarily related to the development ofimproved adjuvant medical therapies for respiratory failure,such as exogenous surfactant, high-frequency oscillatoryventilation, and inhaled nitric oxide,5 which enable criticallyill newborns to be supported without incurring the additionalrisks associated with ECMO.Neonatal cardiac ECMO survival has changed over time with

a 40% cumulative rate of survival to hospital discharge ortransfer.4 The reported survival rate was 69% prior to 1987. By1994, survival had decreased to 34% despite significantlyincreased utilization. The survival rate has been steadilyincreasing since 1994, with 48% of neonatal cardiac ECMOpatients surviving to hospital discharge in 2012 (Fig) Theremarkably high rate of survival observed during the late1980s was likely due to more conservative patient selectioncriteria during the initial clinical experience with ECMO. Thegradual trend of improved neonatal cardiac ECMO survivalobserved during the past 20 years likely reflects improvementsin ECLS equipment and patient management strategies thathave improved overall patient safety despite increased utiliza-tion in higher-risk patients. Additional factors, such as time toinitiation of support, patient selection, and the availability ofadditional forms of mechanical cardiac support, have undoubt-edly contributed to improved survival in the contemporary era.

Indications for neonatal cardiac ECLS

The primary indication for mechanical circulatory support inneonates is inadequate cardiac output and end-organ injury

Fig – Neonatal Cardiac ECLS Trends in

that is refractory to optimal medical therapy. ECMO may beused for preoperative stabilization, failure to separate fromcardiopulmonary bypass, post-cardiotomy low cardiac outputstates, and as a bridge to transplantation. Single-ventricleheart disease, once considered a relative contraindication toECMO, now represents the most common underlying diag-nosis in neonates who require cardiac ECMO. Additionalindications for mechanical cardiopulmonary support includecardiomyopathy, myocarditis, pulmonary hypertension,intractable dysrhythmia, cardiac arrest, and low cardiac out-put states unrelated to structural heart disease (Table).

Preoperative stabilization

In 1970, Baffes et al.6 published what has been credited as thefirst series describing prolonged extracorporeal circulatorysupport for infants with congenital heart disease. ECMO maybe used to provide initial stabilization in neonates who presentwith refractory hypoxemia and/or severe cardiogenic shock,who might otherwise not survive to operative repair ofstructural heart disease. When used to stabilize patients withevidence of shock and end-organ damage, ECMO enhancesend-organ perfusion, reverses acidosis, and may facilitaterecovery of cardiac function. Most neonates with ductal-dependent cyanotic heart disease (e.g., Tetralogy of Fallot withpulmonary atresia) or isolated parallel circulations respondwell to prostaglandin infusion to establish or maintain ductalpatency or a balloon atrial septostomy to increase mixing atthe atrial level. However, ECMO may be necessary to stabilizepatients who present in a delayed manner with hypoxic shockthat is refractory to pharmacological restoration of pulmonary

Utilization and Survival 1987-2012.

Table – Indications for neonatal cardiac ECLS.

Preoperative stabilization of uncorrected congenital heart diseasePost-cardiotomyFailure to wean from cardiopulmonary bypassLow cardiac output statePulmonary hypertensionIntractable arrhythmiasCardiac arrestMyocarditisCardiomyopathyBridge to transplantationHemodynamic stabilization for proceduresRespiratory failure/sepsis

S E M I N A R S I N P E R I N A T O L O G Y 3 8 ( 2 0 1 4 ) 9 7 – 1 0 3 99

blood flow. Structural cardiac defects such as D-transpositionof the great arteries with refractory pulmonary hyperten-sion,7,8 pulmonary atresia with intact ventricular septum,tricuspid atresia, hypoplastic left heart syndrome,9 and severeEbstein's malformation may require ECMO for preoperativestabilization. The primary goal of ECMO is to optimize thehemodynamic state and improve end-organ oxygen deliveryin these patients to make them better surgical candidates.Although ECMO may also be used to stabilize patients withobstructed total anomalous pulmonary venous drainage, sur-gical repair should be undertaken as expeditiously as possiblein these patients. In an early report of preoperative cardiacECMO, Hunkeler et al.10 described eight neonates with cyanoticcongenital heart disease. The most common diagnosis wasTetralogy of Fallot with refractory hypoxemia and metabolicacidosis. Most patients were successfully supported to surgicalrepair and long-term survival was 63%. Renal failure, seizures,and significant surgical site bleeding were the most commoncomplications.In a more recent report of 26 patients, including 17 neo-

nates, with congenital heart disease who required preoper-ative ECMO,9 the most common indication for ECMO washypoxemia despite balloon atrial septostomy in patients withd-transposition of the great arteries. Overall survival tohospital discharge was 53% and duration of ECMO supportand failure to separate from cardiopulmonary bypass follow-ing surgical correction were associated with increased mor-tality. Central nervous system injury was the most commoncomplication, occurring in 81% of the patients. Neonates whorequire ECMO for preoperative stabilization are at anincreased risk for perioperative morbidity and death.

Failure to separate from cardiopulmonary bypass

Despite advances in myocardial protection during cardiopul-monary bypass, myocardial dysfunction following neonatalcardiac surgery is common. Mild-to-moderate myocardialdysfunction due to the metabolic stress of cardiopulmonarybypass and ischemia/reperfusion injury is generally respon-sive to inotropic support, afterload reduction, and pulmonaryvasodilators. However, patients who experience significantmyocardial dysfunction after prolonged cardiopulmonarybypass and extended periods of myocardial arrest and thosewho have preexisting ventricular dysfunction are at anincreased risk for severe postoperative myocardial

dysfunction that may prevent successful separation fromcardiopulmonary bypass. Once residual anatomic lesionshave been excluded, ECMO may be used to improve systemicperfusion and facilitate myocardial recovery by reducingventricular wall stress and increasing myocardial oxygendelivery. Ventricular distention must be avoided during therecovery period. Interventions such as surgical or transcath-eter atrial septostomy, left atrial venting, or left ventricularventing should be undertaken if left ventricular distention isobserved on echocardiographic examination. Although veno-venous ECMO may be used to provide respiratory support forpatients who experience refractory hypoxemia due to cardi-opulmonary bypass-related lung injury or intraoperativepulmonary hemorrhage, venoarterial ECMO is the mostcommonly used modality in neonatal cardiac patients. Inmost cases, patients are transitioned from cardiopulmonarybypass to ECMO without changing the central arterial andvenous cannulae.

Low cardiac output state

Postoperative myocardial dysfunction may lead to a refractorylow cardiac output state immediately after separation fromcardiopulmonary bypass or several hours after surgery. In thesepatients, ECMO may be used to improve end-organ oxygendelivery and prevent hemodynamic collapse while reducingpharmacologic inotropic and vasopressor support. Additionally,ECMO may decrease clinically significant dysrhythmias byreducing myocardial workload and the need for potentiallyarrhythmogenic inotropic agents. The decision to place apatient on ECMO during the early postoperative period is oftenchallenging due to the evolving clinical state of the patientsagainst the backdrop of multiple simultaneous therapeuticinterventions. Several studies have examined the relationshipbetween interval to postoperative ECMO and outcome. Allanet al.11 found no difference in the survival rate of shunted-single-ventricle infants who were placed on ECMO due to failureto separate from cardiopulmonary bypass and of those whoreceived early ECMO support in the intensive care unit. Overallsurvival was 48%, but infants who required support for hypo-xemia or shunt obstruction had better survival than those whowere placed on ECMO for hypotension. Aharon et al.12 alsofound no association between survival and timing of cannula-tion following bypass in a group of infants and neonates whounderwent cardiac surgery. However, others have reportedimproved survival in cardiac surgery patients who receive earlyECMO support, including those who transition to ECMO in theoperating room,13,14 compared to those who receive supportlater in the postoperative period.15 Irrespective of the location ofcannulation, earlier initiation of ECLS facilitates unloading ofvulnerable myocardium, prevention or reduction of acidosis,and reduced risk of cardiovascular collapse.

Non-surgical indications

While the vast majority of neonatal cardiac ECLS occurs inpatients with structural congenital heart disease who areundergoing surgical repair or palliation, a number of patients

S E M I N A R S I N P E R I N A T O L O G Y 3 8 ( 2 0 1 4 ) 9 7 – 1 0 3100

without structural heart disease receive ECMO to managesevere heart failure. Cardiomyopathy (2.2%) and myocarditis(1.2%) are the most common non-structural etiologies ofheart failure in neonates reported to the ELSO registry.4

Overall survival is 60% for neonates with cardiomyopathyand 50% for those with myocarditis. Other indications forneonatal cardiac ECLS include unstable or refractory dys-rhythmia,16 myocardial infarction,17 and infection.

Single ventricle patients

Neonates with single-ventricle heart disease represent one ofthe most challenging groups of patients to manage. Approx-imately 30% of patients die within 1 year of undergoing theNorwood procedure and over half of these patients die priorto hospital discharge.18 Although the clinical benefit of ECMOto provide perioperative support for Norwood patients whoexperience hemodynamic instability was initially ques-tioned,13 ECMO has been used with increasing frequency tosupport Norwood patients during the past decade. Currently,patients with hypoplastic left heart syndrome who undergothe Norwood procedure make up the largest group of neo-nates who require ECMO for cardiac support. Overall hospitalsurvival in this group of patients has remained around 30%for the past decade.4 Patients who require ECMO are at agreater risk of dying in the years following Norwood pallia-tion than those who do not require ECMO.19 Risk factors forrequiring ECMO after the Norwood procedure include birthweight o2.5 kg and increased duration of cardiopulmonarybypass,20 whereas risk factors for death once ECMO has beeninitiated include duration of support and ECMO-related com-plications such as neurologic injury and renal failure.15 FewNorwood patients survive more than nine days of ECMOsupport.Surgical palliation of single-ventricle heart disease has

evolved significantly during the past decade. Improved earlysurvival has led many centers to preferentially perform themodified Norwood procedure (aortic arch reconstruction þright ventricle-to-pulmonary artery shunt) rather than theclassic Norwood procedure (aortic arch reconstruction þsystemic-to-pulmonary artery shunt). Additionally, somecenters have abandoned the Norwood procedure altogetherin favor of a hybrid approach to palliation that combinesplacement of bilateral pulmonary artery bands with insertionof a stent to maintain patency of the ductus arteriosus. ECMOcannulation strategies must be thoughtfully tailored tounderlying anatomy and the palliative procedure performed.ECMO cannulation following the Norwood procedure typi-cally occurs in the operating room due to failure to separatefrom cardiopulmonary bypass or in the intensive care unitduring the early postoperative period before the sternum hasbeen closed. The need for ECMO support later during thehospitalization is less common. Consequentially, most Nor-wood patients receive VA-ECMO support utilizing central(aorta/right atrium) cannulation. The theoretical concernabout the potential risk of pulmonary over-circulation inthe setting of a systemic-to-pulmonary artery shunt in tradi-tional Norwood patients prompted some to advocate partialor total shunt occlusion during ECMO support. However,

there is evidence of improved survival when shunt patencyis maintained during the course of ECMO.21 Total ECMO flowmay need to be increased to 200 cc/kg/min to compensate forshunt run-off into the pulmonary circulation. Respiratorysupport with lower tidal volumes and FiO2 may reduce shuntrun-off and improve systemic oxygen delivery when on fullECMO support. The risk of pulmonary over-circulation is verylow when ECMO is used to support patients who haveundergone the modified Norwood procedure. Perhaps, thegreatest shunt-related risk to these patients is shunt throm-bosis. Maintaining some degree of ejection through the aortaand the right ventricle-to-pulmonary artery shunt mayreduce the likelihood of shunt thrombosis during ECMO.Patients who have undergone a hybrid palliation are uniquelychallenging. Options for arterial cannulation site include themain pulmonary artery/ductus arteriosus and the carotidartery. Stent-related compression of the distal transverseaortic arch is a recognized complication of the hybrid proce-dure that can lead to inadequate myocardial and centralperfusion. In this situation, ductal cannulation may beinadequate to achieve cerebral and myocardial oxygen deliv-ery. Conversely, carotid cannulation may result in inadequatesystemic end-organ perfusion. A clear understanding of thepatient's aortic arch anatomy and factors contributing tohemodynamic instability is necessary for selecting the mostappropriate cannulation strategy. Outcome data is sparse forpatients who require ECMO after hybrid palliation.

Extracorporeal cardiopulmonary resuscitation

Extracorporeal cardiopulmonary resuscitation (ECPR) is beingused with increasing frequency as rescue therapy for neo-nates who experience refractory cardiac arrest. Data from theELSO registry indicates that the use of ECPR is associated with42% overall hospital survival in infants and children withheart disease.22 When compared to cardiopulmonary resus-citation alone, ECPR appears to improve survival by approx-imately 12–23% in pediatric patients.23 Overall neonatal ECPRsurvival is better (39%) than adult ECPR survival (27%).4 Age-related differences in survival are likely related to differencesin age-associated etiology of cardiac arrest and comorbidconditions. Because most neonates (83%) in the ELSO registrywho receive ECPR have an underlying cardiac diagnosis,23

neonatal ECPR data provides useful clinical outcome and riskstratification information about neonatal cardiac patients atthe worst extreme of the clinical spectrum. Lower birthweight and lower pre-ECPR arterial oxygenation, as well asseveral complications occurring on ECLS including CNS orpulmonary hemorrhage, acidosis, the receipt of renal replace-ment therapy, and mechanical complications, are associatedwith reduced survival. Survival following ECPR is only 21% inthe small number of neonates born before 34 weeks (n ¼ 12)and 27% for all preterm neonates but rises to approximately40% in early-term and full-term neonates.24 The rate of ECPR-related neurologic injury, which is an independent risk factorfor death, falls significantly with increasing gestational age.Despite increased risks associated with low gestational age,approximately 20% of neonates under 2 kg survive to hospital

S E M I N A R S I N P E R I N A T O L O G Y 3 8 ( 2 0 1 4 ) 9 7 – 1 0 3 101

discharge, underscoring the potential benefit of ECPR in thesevery young patients.

Contraindications/risk factors

Neonatal cardiac ECLS is generally reserved for patients witha refractory low cardiac output state, impending circulatorycollapse, or cardiac arrest, who are unlikely to survive with-out ECLS. Consequentially, it is difficult to establish consen-sus regarding absolute contraindications to ECLS in thesehigh-risk patients. There is a general agreement that neo-nates with lethal congenital malformations or severe irrever-sible brain injury should not receive ECLS. However, manyfactors that have historically been considered absolute con-traindications to ECLS, such as gestational age o34 weeks,weight o2.5 kg, end-organ damage, and intracranial or solidorgan hemorrhage, are now considered relative contraindi-cations. In 1982, Bartlett et al.25 reported 25% overall survivaland a high rate of neurological morbidity in infants o35weeks' gestation who received ECLS for respiratory failure.Contemporary data from the ELSO registry indicates that theoverall survival rate for premature infants (o37 weeks'gestation) who receive cardiac ECMO is 31%, compared to41% survival in term infants.4 Survival is only 19% for neo-nates o33 weeks' gestational age.When stratified by age group, a stepwise improvement in

survival is observed with increasing gestational age. Sim-ilarly, the risk of central nervous system hemorrhage isinversely related to increasing gestational age. Weighto2 kg has historically also been a relative contraindicationto ECMO. Gelehrter et al.26 recently reported 10% survival inneonates with hypoplastic left heart syndrome who weighedo2.5 kg when placed on ECMO. In another review of neonatesrequiring postoperative cardiac ECLS, Bhat et al.27 reported asurvival rate of 33% in infants o3 kg. Because the ELSOregistry contains relatively few patients with extremely lowweight or gestational age, it is not possible to establish anabsolute gestational age or weight below which ECMO shouldnot be used. However, any potential survival benefit must beweighed against increased risk of neurologic injury inyounger and smaller infants.

Venovenous ECLS for cardiac patients

Although venoarterial ECMO is most commonly used tosupport neonatal cardiac patients, this mode of ECLS hasseveral disadvantages when compared to venovenous ECMO.The risk of coronary, cerebral, and systemic embolism ishigher in patients without intracardiac shunts supportedwith venoarterial ECMO than those supported with venove-nous ECMO. However, the risk of systemic embolization islikely similar in patients with single-ventricle heart diseaseand unrestrictive intra-atrial shunting. Increased systemicafterload related to venoarterial ECMO may lead to dilatationof the failing systemic ventricle, increased end-diastolicpressures, and decreased myocardial perfusion. VenoarterialECMO can result in coronary ischemia in patients withseverely impaired native pulmonary gas exchange with

ejection through the systemic semilunar valve. The effect ofimpaired lung function on the coronary circulation must beconsidered in patients who are unable to wean from ECMOdue to myocardial ischemia.Although venovenous ECMO does not involve arterial

access, there is evidence that venovenous ECMO mayimprove systemic perfusion in neonates with impaired car-diac function. In small observation studies of infants whorequired inotropic support,28,29 systemic blood pressure, leftventricular shortening fraction, and aortic peak velocitieswere maintained, whereas inotrope score decreased inpatients who received venovenous ECMO. A retrospectivereview of the ELSO registry indicates that venovenous ECMOis used in only 1.6% of pediatric patients who receive ECMOfor primary cardiac indications.30 Overall, 42% of the patientsin the study survived to hospital discharge. However, approx-imately one-quarter of the patients had to be converted tovenoarterial ECMO.

Management of bleeding during ECMO

Postoperative surgical site bleeding is an important cause ofmorbidity and mortality in neonates who undergo surgicalcorrection of complex congenital heart diseases. Cardiopul-monary bypass results in platelet and clotting factor con-sumption as well as reduced platelet function related tocontact activation of the hemostatic system. In addition,the effect of hemodilution on coagulation factors, antithrom-bin III levels, and number of platelets is most pronounced inneonates, further increasing the risk of clinically significantpostoperative hemorrhage. The coagulopathic impact of car-diopulmonary bypass on neonates and infants is often morepronounced due to the immaturity of the hemostatic systemin these patients. Vitamin-K-dependent factors, prothrombin,antithrombin III, protein C, and protein S levels are depressedat birth and many factors do not achieve adult levels forseveral months.31 Although the risk of life-threatening post-operative hemorrhage is generally low in infants whoundergo cardiac surgery, the incidence of significant surgicalsite bleeding is as high as 32% in neonates who require ECMOafter heart surgery. Furthermore, the risk of death in ECMOpatients with postoperative surgical site hemorrhage is sub-stantially greater in neonates (71%) than in older patients(60%).4 Recognizing these risks, some centers are reluctant touse postoperative ECMO in all but the most extremecircumstances.Advances in ECMO circuitry, including circuit coatings,

polymethylpentene oxygenators, highly efficient centrifugalpumps, and shorter blood pathways, have reduced the like-lihood of circuit thrombosis by decreasing blood transit timeand sites of clot activation. Although roller-pumps have beenroutinely used for neonatal ECMO for decades, a growingnumber of centers are using centrifugal pumps to supportneonates. Early centrifugal pumps that frequently causedclinically significant pump-related hemolysis have largelybeen replaced by newer, low-friction centrifugal pumps thatappear to have the same hemolysis profile as roller-pumps.4

Over 90% of North American centers now use polymethyl-pentene oxygenators, which, like contemporary centrifugal

S E M I N A R S I N P E R I N A T O L O G Y 3 8 ( 2 0 1 4 ) 9 7 – 1 0 3102

pumps, are highly efficient and less prone to thrombosis.Taking advantage of the relatively low thrombotic risk ofnewer circuits, many centers significantly reduce or delayheparin anticoagulation during the early postoperative periodto reduce the risk of bleeding. When a specialized ECMOcircuit and a low-anticoagulation (ACT o180 s) or no-anticoagulation strategy is used during the early postoper-ative period, patients experience less postoperative bleedingand require fewer blood transfusions.32

New directions

Unlike roller-pumps and older membrane oxygenators thatwere specifically designed for ECMO, none of the ECMOcircuitry that is currently available in the United States hasbeen approved for use beyond 6 h by the U.S. Food and DrugAdministration. Despite this, the perceived clinical benefit ofmany contemporary circuit components has resulted inincreased use and, in many cases, the off-label use of someequipment has become the standard of care. A growing trendin mechanical circulatory support is the development of fullyintegrated extracorporeal life support systems, in which theblood pump and oxygenator are contained in a single-usemodular component of a multi-use platform. The utility ofthese specialized systems in the neonatal patient populationhas yet to be established. Lessons learned from the manage-ment of end-stage heart failure patients who requiremechanical cardiac support have underscored the challengesof applying adult therapies to children and neonates. Ven-tricular assist devices (VADs) are widely used to support adultpatients with acute or chronic heart failure, with over 1700adults per year receiving mechanical cardiac support beyondthe perioperative period. In contrast, VADs are used in fewerthan 100 pediatric patients annually in the United States.Currently, the Berlin Heart EXCORs VAD is the only approvedneonatal ventricular assist device system that is commer-cially available in the United States. In contrast to the lowrate of device-related complications observed in adult VADpatients, significant VAD-related neurologic injury occurs inover 30% of neonates supported with the Berlin HeartEXCORs system, prompting many to question the advantageof this mode of support over ECMO in small neonates.33 Somecenters have elected to utilize short-term extracorporealpumps in neonates in an off-label manner as a bridge totransplantation to avoid the high risk of neurologic injuryassociated with the Berlin Heart EXCORs system.In recognition of the limited support options available for

neonates and young children, the National Heart, Lung, andBlood Institute (NHLBI) established the Pediatric CirculatorySupport Program in 2004 to provide $22.5 million in researchfunding to develop five novel circulatory assist devices for thesepatients. The NHLBI subsequently provided an additional $23.6million in funding in 2010 to support further development ofsome of these devices as part of the Pumps for Kids, Infants,and Neonates (PumpKIN) program. The initial goal of thePumpKIN program was to fund the development of four novelpediatric and neonatal ECMO and VAD systems and to com-plete preclinical testing and begin a clinical trial by 2013.However, funding for two of the devices was discontinued in

2013. One of the remaining devices in the PumpKIN program isthe Jarvik 2000 Infant VAD, which is a fully implantable axialflow pump that is intended to provide long-term support. TheJarvik VAD is the size of a AAA battery and is supposed to becapable of supporting infants as small as 3.5 kg. The otherPumpKIN device, the Levitronix PediPL, is an integrated ECMOsystem that is intended to provide full cardiopulmonary sup-port for patients as small as 2.5 kg for up to 30 days. Clinicaltrials involving these devices may begin as early as 2014.While the overall volume of neonatal non-cardiac ECMO

has declined over the past two decades, neonatal cardiacECMO volume continues to increase as centers recognize thesurvival benefit in these high-risk patients. ECMO circuitry,like the neonatal patient population, has changed over time.Contemporary ECMO circuits have shorter blood pathways,more efficient oxygenators, and are increasingly incorporat-ing blood pump technology borrowed from ventricular assistdevices. Government-sponsored initiatives, such as thePumpKIN program, and industry spin-off programs shouldfurther increase the efficiency and, hopefully, the safety ofECMO systems for our most vulnerable patients.

r e f e r e n c e s

1. Fynn-Thompson F, Almond C. Pediatric ventricular assistdevices. Pediatr Cardiol. 2007;28(2):149–155.

2. Paden ML, Conrad SA, Rycus PT, Thiagarajan RR. ELSORegistry. Extracorporeal Life Support Organization RegistryReport 2012. ASAIO J. 2013;59(3):202–210.

3. Hintz SR, Benitz WE, Colby CE, et al. Utilization and outcomesof neonatal cardiac extracorporeal life support: 1996–2000n.Pediatr Crit Care Med. 2005;6(1):33.

4. 2013 ELSO International Summary. Extracorporeal Life Sup-port Organization Registry, Ann Arbor, MI 2013 Jun 18;1–26.

5. Roy BJ, Rycus P, Conrad SA, Clark RH. The changing demo-graphics of neonatal extracorporeal membrane oxygenationpatients reported to the Extracorporeal Life Support Organ-ization (ELSO) Registry. Pediatrics. 2000;106(6):1334–1338.

6. Baffes TG, Fridman JL, Bicoff JP, Whitehill JL. Extracorporealcirculation for support of palliative cardiac surgery in infants.Ann Thorac Surg. 1970;10(4):354–363.

7. Jaillard S, Belli E, Rakza T, et al. Preoperative ECMO in trans-position of the great arteries with persistent pulmonaryhypertension. Ann Thorac Surg. 2005;79(6):2155–2158.

8. Chang AC, Wernovsky G, Kulik TJ, Jonas RA, Wessel DL.Management of the neonate with transposition of the greatarteries and persistent pulmonary hypertension. Am J Cardiol.1991;68(11):1253–1255.

9. Bautista-Hernandez V, Thiagarajan RR, Fynn-Thompson F,et al. Preoperative extracorporeal membrane oxygenation as abridge to cardiac surgery in children with congenital heartdisease. Ann Thorac Surg. 2009;88(4):1306–1311.

10. Hunkeler NM, Canter CE, Donze A, Spray TL. Extracorporeallife support in cyanotic congenital heart disease beforecardiovascular operation. Am J Cardiol. 1992;69(8):790–793.

11. Allan CK, Thiagarajan RR, del Nido PJ, Roth SJ, Almodovar MC,Laussen PC. Indication for initiation of mechanical circulatorysupport impacts survival of infants with shunted single-ventricle circulation supported with extracorporeal membraneoxygenation. J Thorac Cardiovasc Surg. 2007;133(3):660–667.

12. Aharon AS, Drinkwater DC, Churchwell KB, et al. Extracorpor-eal membrane oxygenation in children after repair of con-genital cardiac lesions. Ann Thorac Surg. 2001;72(6):2095–2101([discussion2101–2]).

S E M I N A R S I N P E R I N A T O L O G Y 3 8 ( 2 0 1 4 ) 9 7 – 1 0 3 103

13. Pizarro C, Davis DA, Kerins PJ, Raphaely RC, Spurrier EA,Norwood WI. Extracorporeal membrane oxygenation for neo-nates with single ventricle and parallel circulations. J HeartLung Transplant. 2001;20(2):239–240.

14. Chaturvedi RR, Macrae D, Brown KL, et al. Cardiac ECMO forbiventricular hearts after paediatric open heart surgery. Heart.2004;90(5):545–551.

15. Sherwin ED, Gauvreau K, Scheurer MA, et al. Extracorporealmembrane oxygenation after stage 1 palliation for hypo-plastic left heart syndrome. J Thorac Cardiovasc Surg. 2012;144(6):1337–1343.

16. Dyamenahalli U, Tuzcu V, Fontenot E, et al. Extracorporealmembrane oxygenation support for intractable primaryarrhythmias and complete congenital heart block in new-borns and infants: short-term and medium-term outcomes.Pediatr Crit Care Med. 2012(1):47–52.

17. Tometzki AJ, Pollock JC, Wilson N, Davis CF. Role of ECMO inneonatal myocardial infarction. Arch Dis Child Fetal NeonatalEd. 1996;74(2):F143–F144.

18. Ohye RG, Schonbeck JV, Eghtesady P, et al. Cause, timing, andlocation of death in the Single Ventricle Reconstruction trial.J Thorac. Cardiovasc Surg.. 2012;144(4):907–914.

19. Debrunner MG, Porayette P, Breinholt JP, Turrentine MW,Cordes TM. Midterm survival of infants requiring postoper-ative extracorporeal membrane oxygenation after Norwoodpalliation. Pediatr Cardiol. 2013;34(3):570–575.

20. Friedland-Little JM, Hirsch-Romano JC, Yu S, et al. Risk factorsfor requiring extracorporeal membrane oxygenation supportafter a Norwood operation. J Thorac Cardiovasc Surg. 2013 Oct5. pii: S0022-5223(13)00971-9. doi:10.1016/j.jtcvs.2013.08.051.[Epub ahead of print].

21. Jaggers JJ, Forbess JM, Shah AS, et al. Extracorporeal membraneoxygenation for infant postcardiotomy support: significance ofshunt management. Ann Thorac Surg. 2000;69(5):1476–1483.

22. Chan T, Thiagarajan RR, Frank D, Bratton SL. Survival afterextracorporeal cardiopulmonary resuscitation in infants andchildren with heart disease. J Thorac Cardiovasc Surg. 2008;136(4):984–992.

23. Tajik M, Cardarelli MG. Extracorporeal membrane oxygen-ation after cardiac arrest in children: what do we know? Eur JCardiothorac Surg. 2008;33(3):409–417.

24. McMullan DM, Thiagarajan RR, Smith KM, Rycus P, BroganTV. Extracorporeal cardiopulmonary resuscitation outcomesin term and premature neonates. Pediatr Crit Care Med.

25. Bartlett RH, Andrews AF, Toomasian JM, Haiduc NJ,Gazzaniga AB. Extracorporeal membrane oxygenation fornewborn respiratory failure: forty-five cases. Surgery. 1982;92(2):425–433.

26. Gelehrter S, Fifer CG, Armstrong A, Hirsch J, Gajarski R.Outcomes of hypoplastic left heart syndrome in low-birth-weight patients. Pediatr Cardiol. 2011;32(8):1175–1181.

27. Bhat P, Hirsch JC, Gelehrter S, et al. Outcomes of infantsweighing three kilograms or less requiring extracorporealmembrane oxygenation after cardiac surgery. Ann ThoracSurg. 2013;95(2):656–661.

28. Roberts N, Westrope C, Pooboni SK, et al. Venovenousextracorporeal membrane oxygenation for respiratory failurein inotrope dependent neonates. ASAIO J. 2003;49(5):568–571.

29. Strieper MJ, Sharma S, Dooley KJ, Cornish JD, Clark RH. Effectsof venovenous extracorporeal membrane oxygenation oncardiac performance as determined by echocardiographicmeasurements. J Pediatr. 1993;122(6):950–955.

30. Kim K, Mazor RL, Rycus PT, Brogan TV. Use of venovenousextracorporeal life support in pediatric patients for cardiacindications: a review of the Extracorporeal Life SupportOrganization registry. Pediatr Crit Care Med. 2012;13(3):285–289.

31. Andrew M, Paes B, Milner R, et al. Development of the humancoagulation system in the full-term infant. Blood. 1987;70(1):165–172.

32. McMullan DM, Emmert JA, Permut LC, et al. Minimizingbleeding associated with mechanical circulatory supportfollowing pediatric heart surgery. Eur J Cardiothorac Surg.2011;39(3):392–397.

33. Almond CS, Morales DL, Blackstone EH, et al. Berlin Heart EXCORpediatric ventricular assist device for bridge to heart trans-plantation in US children. Circulation. 2013;127(16):1702–1711.