Neonatal enterovirus myocarditis outcomes following extracorporeal membrane oxygenation (NEMO-ECMO): An ELSO registry analysis Authors: Jack Luxford, Ashfaque Quadir, Philip Roberts, Claire Irving, Julian Ayer, Rugambwa M. Muhame, Nitesh Singhal Published: 2026 ABSTRACT BackgroundNeonates with enterovirus myocarditis may require extracorporeal membrane oxygenation (ECMO). The contemporary prognosis and risk factors for mortality of neonates with enterovirus myocarditis supported on ECMO is unknown.MethodsRetrospective cohort study of Extracorporeal Life Support Organization (ELSO) registry patients < 28 days old supported on V-A ECMO with enterovirus myocarditis. Primary outcome was survival to discharge. Secondary outcome was transplant-free survival to discharge. Demographic and ECMO-related variables were assessed as predictors of survival and transplant-free survival to discharge.Results163 neonates from 2000 to 2023 required ECMO for enterovirus myocarditis. The indication for ECMO was cardiac in 136 (83.4%) and eCPR in 16 (9.8%). Median duration of ECMO was 9.3 days (IQR 5.9-15.7). Hospital survival was 50.4% overall; 64.6% in the 2018 to 2023 era. ECMO bridge to VAD resulted in hospital mortality in 80%. Transplant-free survival at discharge was 44.2%. Variables negatively associated with hospital survival in multivariable regression analysis were age ≤ 8 days (OR 0.10, 95% CI 0.04-0.28), ECMO duration > 15.7 days (OR 0.29, 95% CI 0.11-0.78), central cannulation (OR 0.24, 95% CI 0.07-0.84), hyperbilirubinaemia (OR 0.15, 95% CI 0.03-0.89) and cerebral infarction (OR 0.07, 95% CI 0.01-0.72).ConclusionsECMO is an effective bridge to recovery in neonates with enterovirus myocarditis, with contemporary survival comparable to other neonatal indications for extracorporeal support. Risk factors including age ≤ 8 days and ECMO duration > 16 days define a high-risk cohort. Neonates bridged directly to VAD from ECMO have a dismal outcome. There is a critical need to improve outcomes in neonatal enterovirus myocarditis. Neonatal enterovirus myocarditis is a severe and potentially life-threatening manifestation of enteroviral infection occurring in the first weeks of life.1 Direct myocardial destruction and the secondary inflammatory cascade in a immune immature newborn without serotype-specific maternal antibodies contribute to disease lethality.2, 3, 4 One-third of those hospitalized with severe neonatal enterovirus develop myocarditis, of whom half require extra-corporeal membrane oxygenation (ECMO) for cardiogenic shock and malignant arrhythmia.5 Whilst the mortality of neonates hospitalized with severe enterovirus myocarditis approaches 39%, hospital mortality for the sub-group who require ECMO is reported as high as 67%.6, 7 There is limited evidence for disease-modifying therapy including immunoglobulin, steroids, and anti-virals and thus there remains a role for ECMO as a bridge to recovery in the setting of fulminant neonatal enteroviral myocarditis.8, 9, 10, 11 In a previous Extracorporeal Life Support Organization (ELSO) study of neonates requiring V-A ECMO for cardiac support, risk factors for mortality included lower birth weight, earlier era, single ventricle physiology, degree of acidosis at cannulation and increased intubation to cannulation time, with an overall survival of 41% and a median ECMO run duration of 4.6 days.12 The most robust survival estimates in neonatal enterovirus myocarditis are from a 2001 to 2007 ELSO cohort study of 24 neonates supported on ECMO with 33% survival to discharge.7 Within the limits of this small cohort, trends towards associations with hospital mortality were seen for renal dysfunction, disseminated intravascular coagulation and multiorgan dysfunction. Infrequent epidemics related to increased circulating pathogenic coxsackievirus serotypes necessitate accurate survival estimates to appropriately utilize high-value resources such as ECMO.2, 13, 14, 15, 16, 17 The contemporary outcomes of neonates supported with ECMO for neonatal enterovirus myocarditis, including hospital and transplant-free survival, are unknown, and there is an attendant need to update survival estimates to counsel families and inform clinical decisions about appropriate deployment and withdrawal of extracorporeal support. The NEMO-ECMO study aimed to define the contemporary practice patterns relating to use of ECMO for neonatal enterovirus myocarditis, to assess the overall and transplant-free survival following ECMO and pathways off ECMO, and to identify risk factors related both to overall and transplant-free hospital survival. Materials and methods Study design This was a retrospective, international, multi-centre cohort study using the ELSO registry. The ELSO registry is an international database including more than 600 centres that collects data on patients receiving ECMO. The ELSO registry was queried for neonates ≤ 28 days old at time of V-A ECMO cannulation, with a documented enterovirus infection (Infectious code for organism ID 74 (enterovirus), 108 (enterovirus D68) or 236 (coxsackievirus B) and a diagnosis of myocarditis. A diagnosis of myocarditis was considered to be met with ICD-9 (0.74.23 [coxsackievirus myocarditis], 422 [acute myocarditis], 422.0 [acute myocarditis in disease classified elsewhere], 422.91 [idiopathic myocarditis], and 429.0 [myocarditis, unspecified]) or ICD-10 codes (I40 [acute myocarditis], I40.0 [infective myocarditis], I40.1 [isolated myocarditis], I40.8 [other acute myocarditis] and I40.9 [acute myocarditis, unspecified]). To ensure maximum capture of potential cases, all neonates < 28 days coded for enterovirus infection were interrogated, and 16 patients with diagnoses and combinations thereof consistent with neonatal enterovirus myocarditis were included (see Table S1). Enterovirus sepsis, without myocarditis, was not included. Patients with a diagnosis of major congenital heart disease were excluded. Only the first ECMO run was included if patients had multiple ECMO runs. The primary outcome was survival status at hospital discharge, and patients who were transferred to another centre on ECMO at discharge were excluded (see Figure 1).Figure 1CONSORT flow diagram. CHD, Congenital heart disease; ECMO, Extracorporeal membrane oxygenation; ICD, International Statistical Classification of Diseases and Related Health Problems; TF, Transferred on discharge CONSORT flow diagram. CHD, Congenital heart disease; ECMO, Extracorporeal membrane oxygenation; ICD, International Statistical Classification of Diseases and Related Health Problems; TF, Transferred on discharge Variables extracted from the ELSO database included demographics, year, pre-ECMO clinical and blood gas indices, ECMO indication, pre-ECMO therapies, ECMO run and cannulation details, procedure codes, and ELSO-defined complications. In-hospital outcomes including decannulation reason, time to and survival status at hospital discharge were recorded. Decannulation reasons within the ELSO registry are expected recovery, poor prognosis, ECLS complication, resource limitation, ventricular assist device (VAD) implantation or transplantation. Patients were identified as undergoing transplantation or VAD implantation by procedure code, decannulation reason, or transplantation status. The primary outcome of interest was survival to discharge. The secondary outcome was transplant-free survival to discharge. This study received approval from the Sydney Children’s Hospital Network Human Research Ethics Committee (2024/ETH01222) on June 26, 2024.The requirement for patient consent was waived because of the use of de-identified registry data. This study adhered to the STROBE guidelines for reporting cohort studies.18 Statistical analysis Baseline characteristics were summarized using frequency count and percentages, or median and interquartile range. Between-group differences in survivors and non-survivors at discharge were assessed using χ2 analysis or Fisher’s exact test for categorical variables, and the Mann Whitney U test for continuous variables. The primary and secondary outcomes were assessed at discharge. To identify factors related to the occurrence of the primary and secondary outcomes, 2 separate multivariable logistic regression models were built to assess the potential contribution of candidate demographic, pre-ECMO and ECMO-related risk factors and the incidence of complications. Missing variables were excluded listwise. All variables with p <0.1 on univariate analysis were included in the initial multivariable model, and variables excluded stepwise by backward elimination (likelihood ratio) until only factors with p <0.05 were retained. Age, weight, and duration of ECMO continuous variables were significantly associated with the outcome of interest on univariate analysis and were divided into quartiles, with the highest risk quartile compared to the reference lower risk 3 quartiles as a categorical variable in the multivariable analysis. Analysis was performed with SPSS Statistics v25.0 (SPSS Statistics, IBM Corp, Armonk, NY). Statistical significance was set at p <0.05 Results From January 1, 2000 to December 31, 2023, 163 neonates with enterovirus myocarditis on ECMO were reported to ELSO and included in this cohort (see Figure 1). Survival to hospital discharge occurred in 83 (50.9%) patients. Transplant-free survival to discharge occurred in 72 (44.2%). Annual case numbers and rates of hospital survival over time are displayed in Figure 2.Figure 2Case numbers and survival to hospital discharge over time for neonates supported on ECMO for enterovirus myocarditis. ECMO, extra-corporeal membrane oxygenation Case numbers and survival to hospital discharge over time for neonates supported on ECMO for enterovirus myocarditis. ECMO, extra-corporeal membrane oxygenation Demographic & ECMO-related characteristics The median age and weight at cannulation were 10.0 days (IQR 8.0-13.0) and 3.1 kg (IQR 2.8-3.5). Baseline characteristics at time of cannulation, and between those who did and did not survive to hospital discharge are displayed in Table 1 and S2. Those who did not survive to hospital discharge were younger, lower in weight, with lower systolic, diastolic and mean blood pressures prior to cannulation, and achieved a lower systolic blood pressure following 24 hours of ECMO support. The indication for ECMO was for cardiac support in the majority (136, 83.4%). Sixteen (9.8%) were cannulated onto ECMO during extracorporeal cardiopulmonary resuscitation (eCPR). The median duration of ECMO support was 9.3 days (IQR 5.9-15.7), and 142 (88.2%) of neonates were peripherally cannulated. The most common complications whilst on ECMO were a requirement for renal replacement therapy (67, 41.1%), arrythmia (58, 35.6%), a need for persistent inotropy (63, 38.7%), and the development of circuit-related thrombus (50, 30.7%). ECMO-related characteristics and the type and frequency of complications are displayed in Table 2.Table 1Demographic and Clinical Variables Overall and by Survival StatusVariableOverall (N = 163)Survivors (n = 83)Non-survivors (n = 80)p-valueAge (days)≤ 89-1011-13≥ 1410.0 (8.0-13.0)42 (25.8%)46 (28.2%)41 (25.2%)34 (20.9%)11.0 (9.0-14.0)9 (10.8%)23 (27.7%)29 (34.9%)22 (26.5%)9.0 (8.0-12.0)33 (41.3%)23 (28.7%)12 (15.0%)12 (15.0%)<0.001<0.001Weight (kg, n = 162)≤ 2.82.81-3.13.11-3.5≥ 3.53.1 (2.8-3.5)47 (29.0%)38 (23.5%)37 (22.8%)40 (24.7%)3.2 (2.9-3.6)15 (18.3%)18 (22.0%)21 (25.6%)28 (34.1%)3.0 (2.7-3.4)32 (40.0%)20 (25.0%)16 (20.0%)12 (15.0%)0.030.004Sex (male) (n = 161)95 (59.0%)47 (56.6%)48 (61.5%)0.53Gestational age (weeks, n = 127)38.0 (36.5-39.0)38.0 (37.0-39.0)37.0 (36.0-39.0)0.31Vaginal delivery (n = 123)66 (53.7%)33 (51.6%)24 (40.1%)0.23Era0.902000-20052006-20112012-20172018-202310 (6.1%)49 (30.1%)56 (34.4%)48 (29.4%)3 (3.6%)24 (28.9%)25 (30.1%)31 (37.3%)7 (8.8%)25 (31.3%)31 (38.8%)17 (21.3%)Meningitis57 (35%)33 (39.8%)24 (30.0%)0.19Arrhythmia78 (47.9%)39 (47.0%)39 (48.8%)0.82eCPR / arrest in 24 h prior to ECMO50 (30.7%)21 (25.3%)39 (36.3%)0.13HaemodynamicssBP (mmHg, n = 136)dBP (mmHg, n = 136)mBP (mmHg, n = 136)50.5 (44.0-60.0)36.5 (30.0-43.0)42.0 (35.0-50.0)53.0 (47.5-62.0)39.0 (32.0-44.0)44.0 (37.5-52.0)49.0 (41.0-58.0)33.0 (29.0-40.0)40.0 (32.7-48.0)0.0090.0130.011TherapiesNeuromuscular blockadeVasoactive(s)AdrenalineInodilatorInhaled nitric oxideProstaglandin E1Steroids91 (55.8%)145 (89.0%)95 (58.3%)56 (34.4%)29 (17.8%)9 (5.5%)20 (12.3%)47 (56.6%)74 (89.2%)52 (62.7%)33 (39.8%)15 (18.1%)4 (4.8%)13 (15.7%44 (55.0%)71 (88.8%)43 (53.8%)23 (28.7%)14 (17.5%)5 (6.3%)7 (8.8%)0.830.930.250.140.920.740.18Variables expressed as median (IQR) / n (%). ECMO, extracorporeal membrane oxygenation; eCPR, extracorporeal cardiopulmonary resuscitation; HFOV, high-frequency oscillatory ventilation; MAP, mean airway pressure; sBP / dBP / mBP, systolic / diastolic / mean blood pressure; PEEP, positive end-expiratory pressure; PIP, peak inspiratory pressureTable 2ECMO and Complication-Related Variables Overall and by Survival StatusVariableOverall (N = 163)Survivors (n = 83)Non-survivors (n = 80)p-valueIndication0.25CardiacPulmonaryeCPR136 (83.4%)11 (6.7%)16 (9.8%)72 (86.7%)6 (7.2%)5 (3.1%)64 (80.0%)5 (6.3%)11 (13.8%)Intubation-to-cannulation time (h) (n = 150)32.0 (9.0-63.0)28.0 (5.0-65.5)34 (10.8-58.0)0.52Central cannulation (n = 161)19 (11.8%)6 (7.3%)13 (16.5%)0.07Pump flowL/min @ 4hrs (n = 159)L/min @ 24hrs (n = 158)0.40 (0.33-0.47)0.40 (0.35-0.46)0.40 (0.34-0.46)0.40 (0.35-0.48)0.38 (0.32-0.48)0.41 (0.35-0.46)0.630.38Haemodynamics @ 24hrssBP (mmHg, n = 78)dBP (mmHg, n = 78)mBP (mmHg, n = 78)SvO2 (%, n = 71)58.0 (50.0-64.0)47.0 (42.8-52.0)50.0 (44.0-56.0)84.0 (79-90.0)60.5 (54.0-65.3)48.0 (42.8-53.3)51.5 (46.3-57.5)84.0 (79.0-87.0)52.5 (46.0-62.5)45.0 (40.0-50.8)47.0 (43.0-56.0)85.0 (78.0-90.0)0.0090.130.050.59Left heart decompression0.21NoneBASLeft atrial ventOther113 (69.3%)38 (23.3%)10 (6.1%)2 (1.2%)61 (73.5%)19 (22.9%)3 (3.6%)052 (65.0%)19 (23.8%)7 (8.8%)2 (2.5%)Run duration (days)< 5.95.9-9.39.31-15.7> 15.79.3 (5.9-15.7)41 (25.2%)41 (25.5%)41 (25.2%)40 (24.5%)8.5 (5.9-13.0)21 (25.3%)27 (32.5%)21 (25.3%)14 (16.9%)10.9 (5.9-17.2)20 (25%)14 (34.1%)20 (25.0%)26 (32.5%)0.100.05ComplicationsCardiacArrythmiaCPR on ECMOHypertensionInotropesMyocardial stunPatent ductus arteriosusTamponade58 (35.6%)2 (1.2%)26 (16.0%)63 (38.7%)10 (6.1%)10 (6.1%)7 (4.3%)26 (31.3%)1 (1.2%)14 (16.9%)27 (32.5%)5 (6.0%)3 (3.6%)3 (3.6%)32 (40.0%)1 (1.3%)12 (15.0%)36 (45.0%)5 (6.3%)7 (8.8%)4 (5.0%)0.261.00.750.101.00.210.72MechanicalCircuit thrombusHaemofilter clotsCircuit clotsAir in circuitCannula problemsCircuit changeOxygenator failurePump failure50 (30.7%)11 (6.7%)48 (29.4%)8 (4.9%)25 (15.3%)12 (7.4%)8 (4.9%)5 (3.1%)26 (31.3%)2 (2.4%)26 (31.3%)3 (3.6%)12 (14.5%)6 (7.2%)3 (3.6%)1 (1.2%)24 (30.0%)9 (11.3%)22 (27.5%)5 (6.3%)13 (16.3%)6 (7.5%)5 (6.3%)4 (5.0%)0.850.0240.590.490.750.950.490.21HemorrhagicCannulation site bleeding25 (15.3%)11 (13.3%)14 (17.5%)0.45DICGI hemorrhageSurgical site bleeding8 (4.9%)5 (3.1%)10 (6.1%)1 (1.2%)1 (1.2%)4 (4.8%)7 (8.8%)4 (5.0%)6 (7.5%)0.0320.210.53InfectiousNon-enteroviral virusBacterialFungal22 (13.5%)29 (17.8%)4 (2.5%)15 (18.1%)10 (12.0%)07 (8.8%)19 (23.8%)4 (5.0%)0.080.050.06MetabolicHemolysisHypoglycemicHyperglycemiaHyperbilirubinaemia36 (22.1%)3 (1.8%)13 (8.0%)16 (9.8%)16 (19.3%)1 (1.2%)7 (8.4%)4 (4.8%)20 (25.0%)2 (2.5%)6 (7.5%)12 (15.0%)0.380.620.830.029RenalCreatinine > 1.5 mg/dLRenal replacement therapy20 (12.3%)67 (41.1%)8 (9.6%)27 (32.5%)12 (15.0%)40 (50.0%)0.300.023PulmonaryPulmonary hemorrhage15 (9.2%)5 (6.0%)10 (12.5%)0.15NeurologicalSeizuresBrain deathDiffuse ischemiaCerebral hemorrhageCerebral infarction12 (7.4%)1 (0.6%)2 (1.2%)17 (10.4%)9 (5.5%)5 (6.0%)01 (1.2%)4 (2.8%)2 (2.4%)7 (8.8%)1 (1.3%)1 (1.3%)13 (16.3%)7 (8.8%)0.510.491.00.170.10Variables expressed as median (IQR / n (%). BAS, balloon atrial septostomy; DIC, disseminated intravascular coagulopathy; e/CPR, extracorporeal / cardiopulmonary resuscitation; GI, gastrointestinal ECMO, extracorporeal membrane oxygenation; HFOV, high-frequency oscillatory ventilation; MAP, mean airway pressure; PEEP, positive end-expiratory pressure; PIP, peak inspiratory pressure; sBP / dBP / mBP, systolic / diastolic / mean blood pressure. Table 1Demographic and Clinical Variables Overall and by Survival StatusVariableOverall (N = 163)Survivors (n = 83)Non-survivors (n = 80)p-valueAge (days)≤ 89-1011-13≥ 1410.0 (8.0-13.0)42 (25.8%)46 (28.2%)41 (25.2%)34 (20.9%)11.0 (9.0-14.0)9 (10.8%)23 (27.7%)29 (34.9%)22 (26.5%)9.0 (8.0-12.0)33 (41.3%)23 (28.7%)12 (15.0%)12 (15.0%)<0.001<0.001Weight (kg, n = 162)≤ 2.82.81-3.13.11-3.5≥ 3.53.1 (2.8-3.5)47 (29.0%)38 (23.5%)37 (22.8%)40 (24.7%)3.2 (2.9-3.6)15 (18.3%)18 (22.0%)21 (25.6%)28 (34.1%)3.0 (2.7-3.4)32 (40.0%)20 (25.0%)16 (20.0%)12 (15.0%)0.030.004Sex (male) (n = 161)95 (59.0%)47 (56.6%)48 (61.5%)0.53Gestational age (weeks, n = 127)38.0 (36.5-39.0)38.0 (37.0-39.0)37.0 (36.0-39.0)0.31Vaginal delivery (n = 123)66 (53.7%)33 (51.6%)24 (40.1%)0.23Era0.902000-20052006-20112012-20172018-202310 (6.1%)49 (30.1%)56 (34.4%)48 (29.4%)3 (3.6%)24 (28.9%)25 (30.1%)31 (37.3%)7 (8.8%)25 (31.3%)31 (38.8%)17 (21.3%)Meningitis57 (35%)33 (39.8%)24 (30.0%)0.19Arrhythmia78 (47.9%)39 (47.0%)39 (48.8%)0.82eCPR / arrest in 24 h prior to ECMO50 (30.7%)21 (25.3%)39 (36.3%)0.13HaemodynamicssBP (mmHg, n = 136)dBP (mmHg, n = 136)mBP (mmHg, n = 136)50.5 (44.0-60.0)36.5 (30.0-43.0)42.0 (35.0-50.0)53.0 (47.5-62.0)39.0 (32.0-44.0)44.0 (37.5-52.0)49.0 (41.0-58.0)33.0 (29.0-40.0)40.0 (32.7-48.0)0.0090.0130.011TherapiesNeuromuscular blockadeVasoactive(s)AdrenalineInodilatorInhaled nitric oxideProstaglandin E1Steroids91 (55.8%)145 (89.0%)95 (58.3%)56 (34.4%)29 (17.8%)9 (5.5%)20 (12.3%)47 (56.6%)74 (89.2%)52 (62.7%)33 (39.8%)15 (18.1%)4 (4.8%)13 (15.7%44 (55.0%)71 (88.8%)43 (53.8%)23 (28.7%)14 (17.5%)5 (6.3%)7 (8.8%)0.830.930.250.140.920.740.18Variables expressed as median (IQR) / n (%). ECMO, extracorporeal membrane oxygenation; eCPR, extracorporeal cardiopulmonary resuscitation; HFOV, high-frequency oscillatory ventilation; MAP, mean airway pressure; sBP / dBP / mBP, systolic / diastolic / mean blood pressure; PEEP, positive end-expiratory pressure; PIP, peak inspiratory pressure Demographic and Clinical Variables Overall and by Survival Status Variables expressed as median (IQR) / n (%). ECMO, extracorporeal membrane oxygenation; eCPR, extracorporeal cardiopulmonary resuscitation; HFOV, high-frequency oscillatory ventilation; MAP, mean airway pressure; sBP / dBP / mBP, systolic / diastolic / mean blood pressure; PEEP, positive end-expiratory pressure; PIP, peak inspiratory pressure Table 2ECMO and Complication-Related Variables Overall and by Survival StatusVariableOverall (N = 163)Survivors (n = 83)Non-survivors (n = 80)p-valueIndication0.25CardiacPulmonaryeCPR136 (83.4%)11 (6.7%)16 (9.8%)72 (86.7%)6 (7.2%)5 (3.1%)64 (80.0%)5 (6.3%)11 (13.8%)Intubation-to-cannulation time (h) (n = 150)32.0 (9.0-63.0)28.0 (5.0-65.5)34 (10.8-58.0)0.52Central cannulation (n = 161)19 (11.8%)6 (7.3%)13 (16.5%)0.07Pump flowL/min @ 4hrs (n = 159)L/min @ 24hrs (n = 158)0.40 (0.33-0.47)0.40 (0.35-0.46)0.40 (0.34-0.46)0.40 (0.35-0.48)0.38 (0.32-0.48)0.41 (0.35-0.46)0.630.38Haemodynamics @ 24hrssBP (mmHg, n = 78)dBP (mmHg, n = 78)mBP (mmHg, n = 78)SvO2 (%, n = 71)58.0 (50.0-64.0)47.0 (42.8-52.0)50.0 (44.0-56.0)84.0 (79-90.0)60.5 (54.0-65.3)48.0 (42.8-53.3)51.5 (46.3-57.5)84.0 (79.0-87.0)52.5 (46.0-62.5)45.0 (40.0-50.8)47.0 (43.0-56.0)85.0 (78.0-90.0)0.0090.130.050.59Left heart decompression0.21NoneBASLeft atrial ventOther113 (69.3%)38 (23.3%)10 (6.1%)2 (1.2%)61 (73.5%)19 (22.9%)3 (3.6%)052 (65.0%)19 (23.8%)7 (8.8%)2 (2.5%)Run duration (days)< 5.95.9-9.39.31-15.7> 15.79.3 (5.9-15.7)41 (25.2%)41 (25.5%)41 (25.2%)40 (24.5%)8.5 (5.9-13.0)21 (25.3%)27 (32.5%)21 (25.3%)14 (16.9%)10.9 (5.9-17.2)20 (25%)14 (34.1%)20 (25.0%)26 (32.5%)0.100.05ComplicationsCardiacArrythmiaCPR on ECMOHypertensionInotropesMyocardial stunPatent ductus arteriosusTamponade58 (35.6%)2 (1.2%)26 (16.0%)63 (38.7%)10 (6.1%)10 (6.1%)7 (4.3%)26 (31.3%)1 (1.2%)14 (16.9%)27 (32.5%)5 (6.0%)3 (3.6%)3 (3.6%)32 (40.0%)1 (1.3%)12 (15.0%)36 (45.0%)5 (6.3%)7 (8.8%)4 (5.0%)0.261.00.750.101.00.210.72MechanicalCircuit thrombusHaemofilter clotsCircuit clotsAir in circuitCannula problemsCircuit changeOxygenator failurePump failure50 (30.7%)11 (6.7%)48 (29.4%)8 (4.9%)25 (15.3%)12 (7.4%)8 (4.9%)5 (3.1%)26 (31.3%)2 (2.4%)26 (31.3%)3 (3.6%)12 (14.5%)6 (7.2%)3 (3.6%)1 (1.2%)24 (30.0%)9 (11.3%)22 (27.5%)5 (6.3%)13 (16.3%)6 (7.5%)5 (6.3%)4 (5.0%)0.850.0240.590.490.750.950.490.21HemorrhagicCannulation site bleeding25 (15.3%)11 (13.3%)14 (17.5%)0.45DICGI hemorrhageSurgical site bleeding8 (4.9%)5 (3.1%)10 (6.1%)1 (1.2%)1 (1.2%)4 (4.8%)7 (8.8%)4 (5.0%)6 (7.5%)0.0320.210.53InfectiousNon-enteroviral virusBacterialFungal22 (13.5%)29 (17.8%)4 (2.5%)15 (18.1%)10 (12.0%)07 (8.8%)19 (23.8%)4 (5.0%)0.080.050.06MetabolicHemolysisHypoglycemicHyperglycemiaHyperbilirubinaemia36 (22.1%)3 (1.8%)13 (8.0%)16 (9.8%)16 (19.3%)1 (1.2%)7 (8.4%)4 (4.8%)20 (25.0%)2 (2.5%)6 (7.5%)12 (15.0%)0.380.620.830.029RenalCreatinine > 1.5 mg/dLRenal replacement therapy20 (12.3%)67 (41.1%)8 (9.6%)27 (32.5%)12 (15.0%)40 (50.0%)0.300.023PulmonaryPulmonary hemorrhage15 (9.2%)5 (6.0%)10 (12.5%)0.15NeurologicalSeizuresBrain deathDiffuse ischemiaCerebral hemorrhageCerebral infarction12 (7.4%)1 (0.6%)2 (1.2%)17 (10.4%)9 (5.5%)5 (6.0%)01 (1.2%)4 (2.8%)2 (2.4%)7 (8.8%)1 (1.3%)1 (1.3%)13 (16.3%)7 (8.8%)0.510.491.00.170.10Variables expressed as median (IQR / n (%). BAS, balloon atrial septostomy; DIC, disseminated intravascular coagulopathy; e/CPR, extracorporeal / cardiopulmonary resuscitation; GI, gastrointestinal ECMO, extracorporeal membrane oxygenation; HFOV, high-frequency oscillatory ventilation; MAP, mean airway pressure; PEEP, positive end-expiratory pressure; PIP, peak inspiratory pressure; sBP / dBP / mBP, systolic / diastolic / mean blood pressure. ECMO and Complication-Related Variables Overall and by Survival Status Variables expressed as median (IQR / n (%). BAS, balloon atrial septostomy; DIC, disseminated intravascular coagulopathy; e/CPR, extracorporeal / cardiopulmonary resuscitation; GI, gastrointestinal ECMO, extracorporeal membrane oxygenation; HFOV, high-frequency oscillatory ventilation; MAP, mean airway pressure; PEEP, positive end-expiratory pressure; PIP, peak inspiratory pressure; sBP / dBP / mBP, systolic / diastolic / mean blood pressure. Predictors of survival & transplant-free survival Univariate logistic regression was performed to identify candidate variables predictive of hospital survival. Variables with p < 0.1 were included in the multivariable model (Table 3), comprised of 134 (82.2% of total cohort) patients with all available variables. Variables negatively associated with hospital survival included age ≤ 8 days, central cannulation, ECMO run duration > 15.7 days, hyperbilirubinaemia and cerebral infarction. Fungal infection exclusively occurred in patients experiencing mortality and was not included in the logistic regression. Post hoc analysis demonstrated central cannulation was associated with a higher incidence of tamponade (3, 15.8% vs 4, 2.8%, p = 0.036), all-cause bleeding (9, 47.4% vs 34, 23.9%, p = 0.03) cannulation-site related bleeding (6, 31.6% vs 17, 12.0%, p = 0.034), and circuit thrombus (10, 52.6% vs 40, 28.2%, p = 0.03) compared to peripheral cannulation. Central cannulation occurred following eCPR in only 2 cases.Table 3Univariate and Multivariable Logistic Regression for Hospital SurvivalVariableUnivariate Odds Ratio (95% CI)p-valueMultivariate Odds Ratio (95% CI)p-valueClinical and demographicWeight (kg)Weight ≤ 2.8 kg2.43 (1.24-4.75)0.33 (0.16-0.68)0.0090.002**Sex (male)0.82 (0.44-1.53)0.53Age (days)1.11 (1.03-1.21)0.009Age ≤ 8 days0.17 (0.08-0.39)<0.001**0.10 (0.04-0.28)<0.001Run duration (days)0.96 (0.92-0.99)0.039Run > 15.7 days0.42 (0.20-0.88)0.022**0.29 (0.11-0.78)0.014Recent era (2018-2023)2.21 (1.10-4.43)0.026**Cardiac indication (ref)Pulmonary eCPR1.07 (0.31-3.66)0.40 (0.13-1.23)0.920.11eCPR / arrest < 24 h0.60 (0.30-1.17)0.13Meningitis1.54 (0.80-2.95)0.19Arrhythmia0.93 (0.50-1.72)0.82Blood gaspHpCO2 (mmHg)pO2 (mmHg)HCO3 (mmol/L)SaO2 (%)Lactate (mmol/L)1.11 (0.20-6.02)1.00 (0.98-1.02)1.00 (0.99-1.00)1.02 (0.97-1.07)0.99 (0.98-1.01)1.02 (0.94-1.11)0.910.920.120.530.590.64VentilatorHFOVFiO2 (%)PIP (cm H20)PEEP (cm H20)MAP (cm H20)(0.28-3.62)0.99 (0.98-1.00)0.98 (0.94-1.03)0.96 (0.80-1.15)0.95 (0.86-1.03)1.000.190.460.670.22HemodynamicsBP (mmHg)dBP (mmHg)mBP (mmHg)1.03 (1.0-1.06)1.04 (1.01-1.08)1.04 (1.01-1.08)0.026**0.025**0.023**Pump flow @ 4 (L/min)Pump flow @ 24 (L/min)1.02 (0.87-1.20)1.02 (0.92-1.13)0.800.73TimingIntubation-to-cannulation time (hrs)1.0 (0.99-1.00)0.87PerinatalGestational AgePremature1.12 (0.91-1.37)0.57 (0.26-1.27)0.300.17Vaginal delivery0.64 (0.32-1.32)0.23Pre-ECMO therapiesNeuromuscular blockadeVasoactive(s)AdrenalineInodilatorInhaled nitric oxideProstaglandin E1Steroids1.07 (0.58-2.00)1.04 (0.39-2.78)1.44 (0.77-2.70)1.64 (0.85-3.15)1.04 (0.47-2.32)0.76 (0.20-2.94)1.94 (0.73-5.14)0.830.930.250.140.920.690.18TechnicalCentral cannulation0.40 (0.14-1.11)0.080**0.24 (0.07-0.84)0.026Left heart decompression0.67 (0.34-1.31)0.24ECMO-related complicationsCardiacArrythmiaCPR on ECMOHypertensionInotropesMyocardial stunPatent ductus arteriosusTamponade0.25 (0.36-1.30)0.96 (0.06-15.67)1.15 (0.50-2.67)0.59 (0.31-1.11)0.96 (0.27-3.46)0.39 (0.10-1.57)0.71 (0.15-3.29)0.250.980.750.100.950.190.66MechanicalCircuit thrombus1.06 (0.55-2.07)0.85Haemofilter clotsCircuit clots Air in circuitCannula problemsCircuit changeOxygenator failurePump failure0.20 (0.04-0.93)1.20 (0.61-2.36)0.56 (0.13-2.44)0.87 (0.37-2.04)0.96 (0.30-3.11)0.56 (0.13-2.44)0.23 (0.03-2.12)0.040**0.590.440.750.950.440.200.18 (0.03-1.10)0.063HemorrhagicCannulation site bleeding0.72 (0.31-1.70)0.45DICGI hemorrhageSurgical site bleeding0.13 (0.02-1.06)0.23 (0.03-2.12)0.62 (0.17-2.30)0.056**0.200.480.10 (0.01-1.14)0.064InfectiousNon-enteroviral virusBacterialFungal2.30 (0.88-5.98)0.44 (0.19-1.02)00.088**0.055**NAMetabolicHemolysis0.72 (0.34-1.51)0.38Hypoglycemic0.48 (0.04-5.35)0.55HyperglycemiaHyperbilirubinaemia1.14 (0.37-3.54)0.29 (0.09-0.93)0.830.038**0.15 (0.03-0.89)0.036RenalCreatinine > 1.5 mg/dLRenal replacement therapy0.60 (0.23-1.57)0.48 (0.26-0.91)0.300.024**PulmonaryPulmonary hemorrhage0.45 (0.15-1.38)0.16NeurologicalSeizures0.67 (0.20-2.20)0.51Cerebral hemorrhage0.26 (0.08-0.84)0.024**Cerebral infarction0.26 (0.05-1.28)0.097**0.07 (0.01-0.72)0.025**Included in multivariable analysis. Odds ratios reflect 1-unit increase in continuous variables. DIC, disseminated intravascular coagulopathy; GI, gastrointestinal; ECMO, extracorporeal membrane oxygenation; eCPR, extracorporeal cardiopulmonary resuscitation; HFOV, high-frequency oscillatory ventilation; MAP, mean airway pressure; sBP / dBP / mBP, systolic / diastolic / mean blood pressure; PEEP, positive end-expiratory pressure, PIP, peak inspiratory pressure Table 3Univariate and Multivariable Logistic Regression for Hospital SurvivalVariableUnivariate Odds Ratio (95% CI)p-valueMultivariate Odds Ratio (95% CI)p-valueClinical and demographicWeight (kg)Weight ≤ 2.8 kg2.43 (1.24-4.75)0.33 (0.16-0.68)0.0090.002**Sex (male)0.82 (0.44-1.53)0.53Age (days)1.11 (1.03-1.21)0.009Age ≤ 8 days0.17 (0.08-0.39)<0.001**0.10 (0.04-0.28)<0.001Run duration (days)0.96 (0.92-0.99)0.039Run > 15.7 days0.42 (0.20-0.88)0.022**0.29 (0.11-0.78)0.014Recent era (2018-2023)2.21 (1.10-4.43)0.026**Cardiac indication (ref)Pulmonary eCPR1.07 (0.31-3.66)0.40 (0.13-1.23)0.920.11eCPR / arrest < 24 h0.60 (0.30-1.17)0.13Meningitis1.54 (0.80-2.95)0.19Arrhythmia0.93 (0.50-1.72)0.82Blood gaspHpCO2 (mmHg)pO2 (mmHg)HCO3 (mmol/L)SaO2 (%)Lactate (mmol/L)1.11 (0.20-6.02)1.00 (0.98-1.02)1.00 (0.99-1.00)1.02 (0.97-1.07)0.99 (0.98-1.01)1.02 (0.94-1.11)0.910.920.120.530.590.64VentilatorHFOVFiO2 (%)PIP (cm H20)PEEP (cm H20)MAP (cm H20)(0.28-3.62)0.99 (0.98-1.00)0.98 (0.94-1.03)0.96 (0.80-1.15)0.95 (0.86-1.03)1.000.190.460.670.22HemodynamicsBP (mmHg)dBP (mmHg)mBP (mmHg)1.03 (1.0-1.06)1.04 (1.01-1.08)1.04 (1.01-1.08)0.026**0.025**0.023**Pump flow @ 4 (L/min)Pump flow @ 24 (L/min)1.02 (0.87-1.20)1.02 (0.92-1.13)0.800.73TimingIntubation-to-cannulation time (hrs)1.0 (0.99-1.00)0.87PerinatalGestational AgePremature1.12 (0.91-1.37)0.57 (0.26-1.27)0.300.17Vaginal delivery0.64 (0.32-1.32)0.23Pre-ECMO therapiesNeuromuscular blockadeVasoactive(s)AdrenalineInodilatorInhaled nitric oxideProstaglandin E1Steroids1.07 (0.58-2.00)1.04 (0.39-2.78)1.44 (0.77-2.70)1.64 (0.85-3.15)1.04 (0.47-2.32)0.76 (0.20-2.94)1.94 (0.73-5.14)0.830.930.250.140.920.690.18TechnicalCentral cannulation0.40 (0.14-1.11)0.080**0.24 (0.07-0.84)0.026Left heart decompression0.67 (0.34-1.31)0.24ECMO-related complicationsCardiacArrythmiaCPR on ECMOHypertensionInotropesMyocardial stunPatent ductus arteriosusTamponade0.25 (0.36-1.30)0.96 (0.06-15.67)1.15 (0.50-2.67)0.59 (0.31-1.11)0.96 (0.27-3.46)0.39 (0.10-1.57)0.71 (0.15-3.29)0.250.980.750.100.950.190.66MechanicalCircuit thrombus1.06 (0.55-2.07)0.85Haemofilter clotsCircuit clots Air in circuitCannula problemsCircuit changeOxygenator failurePump failure0.20 (0.04-0.93)1.20 (0.61-2.36)0.56 (0.13-2.44)0.87 (0.37-2.04)0.96 (0.30-3.11)0.56 (0.13-2.44)0.23 (0.03-2.12)0.040**0.590.440.750.950.440.200.18 (0.03-1.10)0.063HemorrhagicCannulation site bleeding0.72 (0.31-1.70)0.45DICGI hemorrhageSurgical site bleeding0.13 (0.02-1.06)0.23 (0.03-2.12)0.62 (0.17-2.30)0.056**0.200.480.10 (0.01-1.14)0.064InfectiousNon-enteroviral virusBacterialFungal2.30 (0.88-5.98)0.44 (0.19-1.02)00.088**0.055**NAMetabolicHemolysis0.72 (0.34-1.51)0.38Hypoglycemic0.48 (0.04-5.35)0.55HyperglycemiaHyperbilirubinaemia1.14 (0.37-3.54)0.29 (0.09-0.93)0.830.038**0.15 (0.03-0.89)0.036RenalCreatinine > 1.5 mg/dLRenal replacement therapy0.60 (0.23-1.57)0.48 (0.26-0.91)0.300.024**PulmonaryPulmonary hemorrhage0.45 (0.15-1.38)0.16NeurologicalSeizures0.67 (0.20-2.20)0.51Cerebral hemorrhage0.26 (0.08-0.84)0.024**Cerebral infarction0.26 (0.05-1.28)0.097**0.07 (0.01-0.72)0.025**Included in multivariable analysis. Odds ratios reflect 1-unit increase in continuous variables. DIC, disseminated intravascular coagulopathy; GI, gastrointestinal; ECMO, extracorporeal membrane oxygenation; eCPR, extracorporeal cardiopulmonary resuscitation; HFOV, high-frequency oscillatory ventilation; MAP, mean airway pressure; sBP / dBP / mBP, systolic / diastolic / mean blood pressure; PEEP, positive end-expiratory pressure, PIP, peak inspiratory pressure Univariate and Multivariable Logistic Regression for Hospital Survival **Included in multivariable analysis. Odds ratios reflect 1-unit increase in continuous variables. DIC, disseminated intravascular coagulopathy; GI, gastrointestinal; ECMO, extracorporeal membrane oxygenation; eCPR, extracorporeal cardiopulmonary resuscitation; HFOV, high-frequency oscillatory ventilation; MAP, mean airway pressure; sBP / dBP / mBP, systolic / diastolic / mean blood pressure; PEEP, positive end-expiratory pressure, PIP, peak inspiratory pressure Similar logistic regression was performed for candidate variables to predict transplant-free survival to discharge, displayed in Table 4. Age ≤ 8 days, ECMO run duration > 15.7 days and the composite occurrence of cardiac arrest in the preceding 24 hours prior to ECMO or eCPR were negatively associated with transplant-free survival on multivariable analysis. Fungal infection and disseminated intravascular coagulopathy exclusively occurred in patients experiencing transplantation or in-hospital mortality and were not included in the logistic regression.Table 4Univariate and Multivariable Logistic Regression for Transplant-Free Hospital SurvivalVariableUnivariate Odds Ratio (95% CI)p-valueMultivariate Odds Ratio (95% CI)p-valueClinical and demographicWeight (kg)Weight ≤ 2.8 kg1.77 (0.93-3.37)0.49 (0.24-0.99)0.080.047**Sex (male)0.70 (0.37-1.31)0.26Age (days)1.10 (1.02-1.18)0.02Age ≤ 8 days0.17 (0.07-0.42)<0.001**0.16 (0.06-0.45)<0.001Run duration (days)0.96 (0.92-0.99)0.03Run > 15.7 days0.33 (0.15-0.73)0.006**0.30 (0.12-0.76)0.011Recent era (2018-2023)2.0 (1.01-3.96)0.046**Cardiac indication (ref)PulmonaryeCPR1.43 (0.42-4.92)0.40 (0.12-1.30)0.570.13eCPR / arrest < 24hrs0.42 (0.21-0.86)0.017**0.32 (0.13-0.80)0.014Meningitis1.36 (0.71-2.60)0.35Arrhythmia0.96 (0.52-1.78)0.89Blood gaspHpCO2 (mmHg)pO2 (mmHg)HCO3 (mmol/L)SaO2 (%)Lactate (mmol/L)0.73 (0.13-4.00)1.01 (0.99-1.03)1.01 (0.99-1.01)1.01 (0.96-1.07)0.99 (0.98-1.01)1.05 (0.96-1.14)0.720.420.150.660.610.31VentilatorHFOVFiO2 (%)PIP (cm H20)PEEP (cm H20)MAP (cm H20)1.32 (0.36-4.76)0.99 (0.99-1.01)1.00 (0.96-1.05)0.99 (0.82-1.18)0.97 (0.89-1.06)0.680.560.900.880.55HemodynamicsBP (mmHg)dBP (mmHg)mBP (mmHg)1.04 (1.01-1.07)1.04 (1.00-1.08)1.04 (1.01-1.08)0.012**0.031**0.020**Pump flow @ 4 (L/min)Pump flow @ 24 (L/min)1.02 (0.88-1.19)1.02 (0.93-1.12)0.780.70TimingIntubation-to-cannulation time (hrs)0.99 (0.99-1.00)0.70PerinatalGestational AgePremature (< 37 weeks)1.06 (0.87-1.30)0.77 (0.35-1.71)0.550.52Vaginal delivery0.71 (0.35-1.45)0.35Pre-ECMO therapiesNeuromuscular blockadeVasoactive(s)AdrenalineInodilatorInhaled nitric oxideProstaglandin E1Steroids0.80 (0.43-1.49)0.77 (0.29-2.05)1.11 (0.59-2.08)1.43 (0.75-2.74)1.22 (0.55-2.73)(0.26-3.91)2.08 (0.80-5.39)0.490.600.740.280.621.00.13TechnicalCentral cannulation0.55 (0.20-1.52)0.25Left heart decompression0.62 (0.31-1.22)0.16ECMO-related complicationsCardiacArrythmiaCPR on ECMOHypertensionInotropesMyocardial stunPatent ductus arteriosusTamponade0.60 (0.31-1.16)1.27 (0.08-20.62)1.1 (0.47-2.55)0.67 (0.35-1.27)0.83 (0.23-3.07)0.52 (0.13-2.09)0.49 (0.09-2.61)0.130.870.820.220.780.360.40MechanicalCircuit thrombusHaemofilter clotsCircuit clots Cannula problemsCircuit changeOxygenator failurePump failure1.11 (0.57-2.17)0.26 (0.05-1.25)1.24 (0.63-2.44)0.67 (0.28-1.62)1.29 (0.39-4.18)0.75 (0.17-3.24)0.31 (0.03-2.80)0.760.920.530.370.670.700.30HemorrhagicCannulation site bleedingDICGI hemorrhageSurgical site bleeding0.82 (0.34-1.95)00.31 (0.03-2.80)0.83 (0.23-3.07)0.65NA0.300.78InfectiousNon-enteroviral virusBacterialFungal1.62 (0.66-4.00)0.61 (0.26-1.41)00.300.25NAMetabolicHemolysis0.65 (0.30-1.40)0.27Hypoglycemic0.63 (0.06-5.05)0.71Hyperglycemia1.09 (0.35-3.40)0.88Hyperbilirubinaemia0.26 (0.71-0.95)0.042**0.23 (0.04-1.25)0.09RenalCreatinine > 1.5mg/dLRenal replacement therapy0.65 (0.24-1.72)0.45 (0.24-0.86)0.380.016**PulmonaryPulmonary hemorrhage0.60 (0.20-1.90)0.38NeurologicalSeizures0.61 (0.18-2.11)0.44Cerebral hemorrhage0.35 (0.11-1.13)0.080**Cerebral infarction0.15 (0.02-1.20)0.073**0.14 (0.02-1.33)0.09**Included in multivariable analysis. Odds ratios reflect 1-unit increase in continuous variables. DIC, disseminated intravascular coagulopathy; GI, gastrointestinal; ECMO, extracorporeal membrane oxygenation; eCPR, extracorporeal cardiopulmonary resuscitation; HFOV, high-frequency oscillatory ventilation; MAP, mean airway pressure; sBP / dBP / mBP, systolic / diastolic / mean blood pressure; PEEP, positive end-expiratory pressure; PIP, peak inspiratory pressure Table 4Univariate and Multivariable Logistic Regression for Transplant-Free Hospital SurvivalVariableUnivariate Odds Ratio (95% CI)p-valueMultivariate Odds Ratio (95% CI)p-valueClinical and demographicWeight (kg)Weight ≤ 2.8 kg1.77 (0.93-3.37)0.49 (0.24-0.99)0.080.047**Sex (male)0.70 (0.37-1.31)0.26Age (days)1.10 (1.02-1.18)0.02Age ≤ 8 days0.17 (0.07-0.42)<0.001**0.16 (0.06-0.45)<0.001Run duration (days)0.96 (0.92-0.99)0.03Run > 15.7 days0.33 (0.15-0.73)0.006**0.30 (0.12-0.76)0.011Recent era (2018-2023)2.0 (1.01-3.96)0.046**Cardiac indication (ref)PulmonaryeCPR1.43 (0.42-4.92)0.40 (0.12-1.30)0.570.13eCPR / arrest < 24hrs0.42 (0.21-0.86)0.017**0.32 (0.13-0.80)0.014Meningitis1.36 (0.71-2.60)0.35Arrhythmia0.96 (0.52-1.78)0.89Blood gaspHpCO2 (mmHg)pO2 (mmHg)HCO3 (mmol/L)SaO2 (%)Lactate (mmol/L)0.73 (0.13-4.00)1.01 (0.99-1.03)1.01 (0.99-1.01)1.01 (0.96-1.07)0.99 (0.98-1.01)1.05 (0.96-1.14)0.720.420.150.660.610.31VentilatorHFOVFiO2 (%)PIP (cm H20)PEEP (cm H20)MAP (cm H20)1.32 (0.36-4.76)0.99 (0.99-1.01)1.00 (0.96-1.05)0.99 (0.82-1.18)0.97 (0.89-1.06)0.680.560.900.880.55HemodynamicsBP (mmHg)dBP (mmHg)mBP (mmHg)1.04 (1.01-1.07)1.04 (1.00-1.08)1.04 (1.01-1.08)0.012**0.031**0.020**Pump flow @ 4 (L/min)Pump flow @ 24 (L/min)1.02 (0.88-1.19)1.02 (0.93-1.12)0.780.70TimingIntubation-to-cannulation time (hrs)0.99 (0.99-1.00)0.70PerinatalGestational AgePremature (< 37 weeks)1.06 (0.87-1.30)0.77 (0.35-1.71)0.550.52Vaginal delivery0.71 (0.35-1.45)0.35Pre-ECMO therapiesNeuromuscular blockadeVasoactive(s)AdrenalineInodilatorInhaled nitric oxideProstaglandin E1Steroids0.80 (0.43-1.49)0.77 (0.29-2.05)1.11 (0.59-2.08)1.43 (0.75-2.74)1.22 (0.55-2.73)(0.26-3.91)2.08 (0.80-5.39)0.490.600.740.280.621.00.13TechnicalCentral cannulation0.55 (0.20-1.52)0.25Left heart decompression0.62 (0.31-1.22)0.16ECMO-related complicationsCardiacArrythmiaCPR on ECMOHypertensionInotropesMyocardial stunPatent ductus arteriosusTamponade0.60 (0.31-1.16)1.27 (0.08-20.62)1.1 (0.47-2.55)0.67 (0.35-1.27)0.83 (0.23-3.07)0.52 (0.13-2.09)0.49 (0.09-2.61)0.130.870.820.220.780.360.40MechanicalCircuit thrombusHaemofilter clotsCircuit clots Cannula problemsCircuit changeOxygenator failurePump failure1.11 (0.57-2.17)0.26 (0.05-1.25)1.24 (0.63-2.44)0.67 (0.28-1.62)1.29 (0.39-4.18)0.75 (0.17-3.24)0.31 (0.03-2.80)0.760.920.530.370.670.700.30HemorrhagicCannulation site bleedingDICGI hemorrhageSurgical site bleeding0.82 (0.34-1.95)00.31 (0.03-2.80)0.83 (0.23-3.07)0.65NA0.300.78InfectiousNon-enteroviral virusBacterialFungal1.62 (0.66-4.00)0.61 (0.26-1.41)00.300.25NAMetabolicHemolysis0.65 (0.30-1.40)0.27Hypoglycemic0.63 (0.06-5.05)0.71Hyperglycemia1.09 (0.35-3.40)0.88Hyperbilirubinaemia0.26 (0.71-0.95)0.042**0.23 (0.04-1.25)0.09RenalCreatinine > 1.5mg/dLRenal replacement therapy0.65 (0.24-1.72)0.45 (0.24-0.86)0.380.016**PulmonaryPulmonary hemorrhage0.60 (0.20-1.90)0.38NeurologicalSeizures0.61 (0.18-2.11)0.44Cerebral hemorrhage0.35 (0.11-1.13)0.080**Cerebral infarction0.15 (0.02-1.20)0.073**0.14 (0.02-1.33)0.09**Included in multivariable analysis. Odds ratios reflect 1-unit increase in continuous variables. DIC, disseminated intravascular coagulopathy; GI, gastrointestinal; ECMO, extracorporeal membrane oxygenation; eCPR, extracorporeal cardiopulmonary resuscitation; HFOV, high-frequency oscillatory ventilation; MAP, mean airway pressure; sBP / dBP / mBP, systolic / diastolic / mean blood pressure; PEEP, positive end-expiratory pressure; PIP, peak inspiratory pressure Univariate and Multivariable Logistic Regression for Transplant-Free Hospital Survival **Included in multivariable analysis. Odds ratios reflect 1-unit increase in continuous variables. DIC, disseminated intravascular coagulopathy; GI, gastrointestinal; ECMO, extracorporeal membrane oxygenation; eCPR, extracorporeal cardiopulmonary resuscitation; HFOV, high-frequency oscillatory ventilation; MAP, mean airway pressure; sBP / dBP / mBP, systolic / diastolic / mean blood pressure; PEEP, positive end-expiratory pressure; PIP, peak inspiratory pressure Hospital survival by quartile for age, weight, run duration and era is reported in Table 5, and for era is displayed graphically in Figure 4.Table 5Hospital Survival by Age, Weight, ECMO Duration and Era QuartilesVariable (N)Hospital Survival (n, %)p-valueAge (days)<0.001 ≤ 8 (42)9-10 (46)11-13 (41)≥ 14 (34)9 (21.4%)23 (50.0%)29 (70.7%)22 (64.7%)Weight (kg)0.004 ≤ 2.8 (47)2.81-3.1 (38)3.11-3.5 (37)≥ 3.5 (40)15 (31.9%)18 (47.4%)21 (56.8%)28 (58.3%)Era0.10 2000-2005 (10)2006-2011 (49)2012-2017 (56)2018-2023 (48)3 (30.0%)24 (48.9%)25 (44.6%)31 (64.6%)Run duration (days)0.05 < 5.9 (41)5.9-9.3 (41)9.31-15.7 (41)> 15.7 (40)21 (51.2%)27 (65.8%)21 (51.2%)14 (35.0%)ECMO, extracorporeal membrane oxygenation Table 5Hospital Survival by Age, Weight, ECMO Duration and Era QuartilesVariable (N)Hospital Survival (n, %)p-valueAge (days)<0.001 ≤ 8 (42)9-10 (46)11-13 (41)≥ 14 (34)9 (21.4%)23 (50.0%)29 (70.7%)22 (64.7%)Weight (kg)0.004 ≤ 2.8 (47)2.81-3.1 (38)3.11-3.5 (37)≥ 3.5 (40)15 (31.9%)18 (47.4%)21 (56.8%)28 (58.3%)Era0.10 2000-2005 (10)2006-2011 (49)2012-2017 (56)2018-2023 (48)3 (30.0%)24 (48.9%)25 (44.6%)31 (64.6%)Run duration (days)0.05 < 5.9 (41)5.9-9.3 (41)9.31-15.7 (41)> 15.7 (40)21 (51.2%)27 (65.8%)21 (51.2%)14 (35.0%)ECMO, extracorporeal membrane oxygenation Hospital Survival by Age, Weight, ECMO Duration and Era Quartiles ECMO, extracorporeal membrane oxygenation Pathways off ECMO The clinical trajectory of neonates following ECMO separation is displayed in Figure 3. Ninety-seven (59.5%) children were decannulated from ECMO support at a median of 9.0 days (IQR 6-12.8), anticipating successful separation. Of these, 7 received an in-hospital heart transplant and 1 underwent delayed VAD placement and then received a transplant. All 7 who received a primary heart transplant, and the 1 child who underwent VAD followed by transplantation survived to discharge at a median of 103 days (IQR 96-148.5). Seventeen (18%) died in-hospital following decannulation despite anticipated recovery, at a median of 37.5 days (IQR 22.5-56) post ECMO separation. The overall hospital survival, and transplant-free survival of the 97 who were separated from ECMO anticipating recovery was 80 (82.5%), and 72 (74.2%), respectively. The median time to hospital discharge in survivors was 46 days (IQR 26.5-80).Figure 3Hospital pathway following ECMO separation. ECMO, extra-corporeal membrane oxygenation; HTx, heart transplantation; VAD, ventricular assist deviceFigure 4Survival to hospital discharge for neonates supported on ECMO for enterovirus myocarditis, stratified by era. ECMO, extra-corporeal membrane oxygenation Hospital pathway following ECMO separation. ECMO, extra-corporeal membrane oxygenation; HTx, heart transplantation; VAD, ventricular assist device Survival to hospital discharge for neonates supported on ECMO for enterovirus myocarditis, stratified by era. ECMO, extra-corporeal membrane oxygenation Ten children were bridged off ECMO directly onto durable VAD at a median of 14.2 days (IQR 10.8-17) of ECMO support (see Table S3). Three of these children underwent transplantation from VAD. Eight of these 10 children who transitioned directly from ECMO to VAD died following a median of 35 days (IQR 23-63.5) in hospital including 2 of the 3 who underwent transplantation from VAD. One patient survived to discharge at 201 days of hospitalization following VAD implantation following 14.5 days of ECMO but was not coded for transplantation and so is assumed to have been weaned from VAD for myocardial recovery. One child received an urgent heart transplant directly off ECMO at 18.3 days and survived to hospital discharge. Fifty-five neonates (33.7%) were decannulated at a median of 11.0 days (IQR 121-397), anticipating demise or poor prognosis. All neonates decannulated anticipating a poor prognosis, including 4 who received a salvage transplant, died prior to hospital discharge. Discussion This study represents the largest cohort of neonates supported on ECMO for enterovirus myocarditis to date, and demonstrates a number of important novel insights. Overall 51% of neonates survived to discharge following ECMO, which is higher than previously reported, and has improved to 65% in the most recent era.7 Transplant-free survival to discharge was 44%. Younger age (< 8 days) and prolonged ECMO duration (> 16 days) were associated with substantially lower survival, as was central compared to peripheral cannulation. In those transitioned directly to VAD from ECMO, survival to hospital discharge was dismal. The improvement over time is likely multifactorial, and may relate to combined institutional and technical improvements in mechanical circulatory support for this cohort of high-risk neonates. Contemporary hospital survival of 65% is comparable to the large numbers of neonates supported on V-A ECMO for alternative cardiac indications, and supports the use of ECMO as an appropriate mechanical support strategy in the setting of cardiogenic shock, severe heart failure or following cardiac arrest.12, 19 Multiple high risk sub-groups with decreased odds of survival to discharge were identified on multivariable analysis, and these can inform institutional decisions to deploy and withdraw ECMO support. Neonates < 8 days of age at time of cannulation survived to discharge in only 21% of cases and these youngest patients may be most at risk of overwhelming viraemia with multiorgan dysfunction including significant rates of hepatic necrosis.20 Survival in neonates supported on ECMO for > 16 days was only 35%. The median ECMO duration of 9.3 days was twice that of neonatal cardiac ECMO indications previously reported, and a third longer than paediatric myocarditis ECMO runs.12, 21 This underscores a need for strategies that adequately unload the left ventricle to facilitate rapid myocardial recovery and shorten the period of exposure to ECMO and related complications. Neonates supported on ECMO beyond 16 days have diminished likelihood of successful survival to discharge, and discussions regarding alternative care pathways including comfort-focused and palliative care should be initiated.22 Neonates who were centrally cannulated had increased mortality risk. This may relate to accrual of complications associated with a prolonged open sternum, such as bleeding and repeated inflammatory and infectious insults. Central cannulation in paediatric V-A ECMO is associated with increased bleeding risk and the accrual of complications increases ECMO mortality.23, 24, 25, 26, 27 Importantly, central cannulation was not simply a proxy for eCPR in this study, but an independent and potentially modifiable risk factor for clinicians embarking on what is likely to be a prolonged mechanical circulatory support course. In this cohort, hepatic injury portended a poor outcome. Hyperbilirubinaemia on ECMO was associated with decreased hospital survival, and may reflect ischemic or congestive hepatitis, hemolysis or enterovirus-associated hepatic necrosis. Isolated hemolysis was not associated with increased mortality and so is unlikely to be the explanation for the prognostic significance of hyperbilirubinaemia. Other markers of liver injury are not captured within the ELSO registry, but univariate analysis suggested that lower blood pressure at 24 hours of ECMO support were also associated with decreased odds of survival. This supports the hypothesis that failure to achieve adequate end-organ perfusion may compromise recovery, in addition to the established impact of hepatic necrosis.20 Cerebral infarction was also associated with decreased survival to discharge, and is likely to reflect the increased thrombotic burden associated with longer run duration, in addition to cerebral hypoperfusion.28 Whilst the delivery of renal replacement therapy was very common (41%) it did not portend a more severe prognosis in multivariable analysis, and perhaps conceiving of renal replacement therapy as a complication does not accurately reflect many institutions routine use of slow continuous ultrafiltration (SCUF) in the armamentarium of neonatal fluid management strategies on ECMO. However, a persistent need for renal replacement therapy post ECMO is likely to have a different risk profile to SCUF. Transplant-free survival to discharge was 44%, with 10 successful heart transplantations enabling hospital discharge. An additional risk factor that predicted decreased odds of transplant-free survival in this cohort was the composite incidence of eCPR or cardiac arrest in the 24 hours preceding ECMO cannulation. Timely deployment of ECMO prior to cardiac arrest is key, although it can be difficult to reliably identify a child at imminent risk of arrest, and there is a confounding risk of peri-operative arrest in placing a stable but unwell child on ECMO. This variable predicted decreased odds of transplant-free survival but not overall survival, and so the occurrence of cardiac arrest or eCPR may identify a group at high risk of persistent myocardial dysfunction post ECMO most likely to benefit from consideration for transplantation. Transplantation from ECMO and interval rehabilitation following ECMO decannulation followed by delayed primary transplant or VAD achieved 100% survival to discharge in 9 children. Where donor supply can support neonatal transplantation, this is a valuable option in cases of refractory heart failure.29, 30 A key novel insight in this cohort was identification of the very high mortality (80%) in neonates directly bridged from ECMO to VAD, the majority of which occurred in the most recent decade. Currently available VAD technology inadequately supports these smallest of patients, and this data demonstrates outcomes even worse than previous reports of mortality of 71%-77% in children < 5 kg supported on pulsatile para-corporeal VAD when bridged from ECMO.31 A previous single-centre study reported persistent biventricular failure as the reason for withdrawal of life-sustaining therapies in 2 neonates with enterovirus myocarditis bridged from ECMO to centrifugal LVAD.6 Whilst the specific reasons for mortality for patients on VAD in this study cannot be identified, the high incidence of stroke, likely need for biventricular support in the setting of global myocardial dysfunction, and the risk of thrombosis in the pro-inflammatory environment of sustained myocarditis are likely to complicate the delivery of VAD support. Given the near-universally poor outcomes for this sub-group, and limited donor supply if a mechanical bridge to transplantation is planned, if there has been inadequate myocardial recovery to support independence from mechanical circulatory support, compassionate decannulation to support palliation and comfort care may be a more appropriate consideration than VAD implantation. Late in-hospital mortality occurred in 18% of patients decannulated successfully from ECMO. Given the late hazard of transplantation, VAD and death in this group, there is likely an ongoing risk of heart failure and arrhythmia, emphasizing a need to maintain vigilance for the development of late or persistent cardiac symptoms. Additional unmeasured factors, including the accumulation of non-cardiac comorbidity may additionally impact post ECMO in-hospital survival. Epidemiologically, the virtual absence of cases in 2020 to 2021 is striking. This likely reflects decreased circulating pathogenic enterovirus serotypes related to the SARS-CoV-19 pandemic, but was followed by reports of regional epidemics in 2022 and 2023 in England and Australia, respectively.17, 32 Limitations in the present study are those inherent to a retrospective multi-centre registry. This cohort does not include patients in whom ECMO was not required, was considered but discounted, or those who died prior to cannulation. Relevant data-points not captured within the registry include echocardiographic, electrocardiographic, and biochemical indices that might influence timing of ECMO deployment and recovery of end-organ function. Raw creatinine values are not reported, and accounting for variable postnatal renal maturity was not possible. This registry reflects the experience of ELSO centres, who may represent more mature ECMO programs. Centre specific variation in ECMO management and provision of potentially disease-modifying therapy could not be assessed. It was not possible to robustly define multiorgan dysfunction using the ad-hoc diagnostic codes within the ELSO registry, as this was a field with significant variability in completeness, and the timing of organ dysfunction could not be identified with respect to ECMO. A small number of the present cohort from 2000 to 2008 have previously been reported, although this prior publication captured only neonates less than 15 days old.7 Further study is needed to define the acute and long-term outcome for all neonates with enterovirus myocarditis, regardless of ECMO requirement, which will help to define the specific indications for ECMO in this cohort of children who may suffer the highest risk on ECMO. The rarity of neonatal enterovirus myocarditis will necessitate multi-centre and multi-national collaboration. Identification of undertreated heart failure and arrhythmia post ECMO, and improvements in the technology of durable mechanical support may lead to further improvements in survival with the deployment of advanced cardiac therapies.2, 5, 33 The role of disease-modifying therapy—immunoglobulin, corticosteroids and anti-viral agents including pocapavir and pleconaril—remains investigational, and given the high incidence of morbidity and mortality it is imperative we develop tools to identify and treat this highly lethal condition, before an ECMO requirement eventuates, or to attenuate ECMO-related morbidity. ECMO is an effective bridge to recovery in neonates with severe cardiac failure secondary to enterovirus myocarditis, with survival in the recent era comparable to other neonatal cardiac indications for extracorporeal support. Prolonged ECMO into the third week, and ECMO as a bridge to VAD were unsuccessful mechanical support strategies which should be considered carefully. Cautious ECMO use in neonates ≤ 8 days, timely initiation prior to cardiac arrest, with preferentially peripheral cannulation anticipating a prolonged ECMO run may lead to more strategic deployment of ECMO and continued improvement in transplant-free and overall survival. Financial support JL was supported by a Cardiac Society of Australia and New Zealand Heart Failure Council Travel Grant. All other authors have no relevant disclosures. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.