Abstract
We performed a multi-center retrospective cohort study of children aged 14 days to 18 years in the United States from 2011 to 2016 with cancer or hematopoietic cell transplant (HCT) who were supported with veno-venous extracorporeal membrane oxygenation (V-V ECMO). We compared the outcomes of children with oncological diagnoses or HCT supported with V-V ECMO to other children who have received V-V ECMO support. In this cohort of 204 patients supported with V-V ECMO, 30 (15%) had a diagnosis of cancer or a history of HCT. There were 21 patients who had oncological diagnoses without HCT and 9 children were post-HCT. The oncology/HCT group had a higher overall ICU mortality (67% vs. 28%, p<0.001), mortality on ECMO (43% vs 21%, p<0.01), and ICU mortality among ECMO survivors (35% vs 8%, p<0.01). The oncology/HCT group had a higher rate of conversion to veno-arterial (V-A) ECMO (23% vs. 9%, p=0.02) (RR 2.5, 95% CI 1.1–5.6). Children with cancer or HCT were older (6.6 years vs 2.9 years, p=0.02) and had higher creatinine levels (0.65 mg/dL vs 0.4 mg/dL, p=0.04) but were similar to the rest of the cohort for other pre-ECMO variables. For post-HCT patients, survival was significantly worse for those whose indication for HCT was cancer or immunodeficiency (0/6) as compared to other nonmalignant indications (3/3) (p=0.01).
Keywords: extracorporeal membrane oxygenation, pediatrics, acute respiratory distress syndrome, cancer, hematopoietic cell transplant, bone marrow transplant
Introduction
Survival for most childhood cancers and for children post HCT has improved over recent decades.1–3 This improved survival has resulted from both improvements in oncologic therapies and supportive critical care.1,4 However, the mortality for this cohort of patients in the setting of pediatric acute respiratory distress syndrome (PARDS) is significantly worse compared to other critically ill children,5,6 with one multi-center study showing a survival of only 24.6% for HCT patients with severe PARDS.7 ECMO is a rescue therapy for severe PARDS, but children with cancer or post-HCT are known to have poor outcomes when requiring ECMO support. Hospital survival for pediatric oncology subjects requiring ECMO has been previously reported at 27–35%.8,9 Survival to hospital discharge for HCT recipients supported with ECMO is worse, with a reported survival of 0–10%.9–13 Despite these low survival rates, 95% of physicians do not consider the presence of a cancer diagnosis as an absolute contraindication to ECMO support in children.8
In this study, our aim was to describe children with cancer or HCT supported with V-V ECMO and their outcomes in the current era. Secondly, we compared the oncology/HCT subgroup to other children who received V-V ECMO support. Lastly, we attempted to identify pre-ECMO risk factors for mortality within the oncology/HCT group to aid in determining appropriate candidacy for ECMO support.
Methods and Materials
This is a secondary analysis of the database produced from a previous study, thus the methods for this study have previously been described.14 A retrospective multi-center cohort study was conducted at 10 quaternary care children’s hospitals in the United States with established ECMO programs. All centers are associated with an established HCT program, and all centers have decades of experience in providing ECMO support. Each is a member of the Pediatric ECMO (PediECMO) research subgroup of the Pediatric Acute Lung Injury and Sepsis Investigators (PALISI) Network and the Extracorporeal Life Support Organization (ELSO). Subjects were managed according to local protocols and clinician preferences. Institutional review board authorization was obtained for all sites either centrally at the lead institution (Indiana University) or locally. Informed consent was waived.
The electronic medical records for children aged 14 days to 18 years who were cannulated for V-V ECMO from 2011 to 2016 were reviewed. Exclusion criteria were: 1) ECMO as a bridge to lung transplant, 2) asthma as the primary cause of acute respiratory failure, 3) unrepaired cyanotic congenital heart disease or single ventricle physiology, or 4) pre-existing chronic respiratory failure (defined as ventilator dependence, positive pressure ventilation, or home O2 not for obstructive sleep apnea).
Data collection was completed via the HIPAA-compliant online REDCap™ database (Vanderbilt University, Nashville, TN).15 Pre-ECMO data collected included demographics and the variables for the Pediatric Logistic Organ Dysfunction (PELOD),16 Pediatric Pulmonary Rescue with Extracorporeal Membrane Oxygenation Prediction (PPREP),17 and Ped-RESCUERS scores.18
Definitions
Cardiac arrest was defined as cessation of a perfusing rhythm for greater than 2 minutes. Acute kidney injury (AKI) was diagnosed if the subject met criteria for stage 3 disease based on Kidney Disease: Improving Global Outcomes (KIDGO)19 or Pediatric Risk, Injury, Failure, Loss, End Stage Renal Disease (pRIFLE) criteria.20 Hematologic malignancies were defined as any lymphoma or leukemia, and other tumors were defined as solid tumor malignancies. If the cancer diagnosis was within a month of ECMO initiation, it was considered a recent diagnosis. Hypotension was defined using PALS guidelines.21
Statistical Analysis
Analyses were performed to determine if there were significant associations between immunocompromised groups, as well as ICU survival for those who were immunocompromised. Due to data skewness, Wilcoxon rank-sum non-parametric tests were used for continuous variables and Chi-Square tests were performed for categorical variables, with Fisher’s Exact tests being used when cell counts were small. All analyses were performed using SAS v9.4 (SAS Institute, Cary, NC).
Results
Baseline
There were 204 subjects on V-V ECMO in the entire cohort, of which 30 (15%) had cancer or HCT. Description of the subject population can be found in a previous publication on this cohort.14 Table 1 shows selected pre-ECMO variables for the oncology/HCT group and the rest of the cohort. Oncology/HCT patients were older, had higher creatinine levels, lower platelet counts, and were more likely to have leukopenia compared to the general population. There were no differences between the groups in the other variables measured including incidence of AKI, aspartate aminotransferase (AST) level, International Normalized Ratio (INR), Oxygenation Index (OI), and days of intubation before ECMO. There were no significant differences in the pre-ECMO mechanical ventilation modes or settings used for the oncology/HCT patients compared to the rest of the cohort (Table 2).
Table 1:
Pre-ECMO Characteristics Comparing Children with Malignancy or Hematopoietic Cell Transplant to Rest of Cohort
| Oncology/HCT (n=30) | Rest of Cohort (n=174) | p-value | |
|---|---|---|---|
| Demographics | |||
| Age, years | 6.6 (2.3 – 13.4) | 2.9 (0.6 – 10.6) | 0.02 |
| Weight, kilograms | 19.7 (11.0 – 58.7) | 13.5 (8.0 – 41.5) | 0.05 |
| Genetic disorder | 13% | 9% | 0.49 |
| Days of intubation to ECMO | 2 (1 – 7) | 2 (1 – 6) | 0.47 |
| Cardiac arrest | 7% | 10% | 0.74 |
| Organ dysfunction variables | |||
| Acute kidney injury | 23% | 11% | 0.06 |
| Highest creatinine prior to ECMO, mg/dL | 0.7 (0.4 – 0.9) | 0.4 (0.3 – 0.7) | 0.04 |
| Hypotension prior to ECMO | 46% | 37% | 0.35 |
| Highest heart rate 24 hours before ECMO | 158 (147 – 174) | 161 (143 – 178) | 0.73 |
| White blood cell count 2000/per mm3 or less | 60% | 8% | <0.01 |
| Lowest platelet count, per mm3 | 52 (27 – 66) | 209 (110 – 324) | <0.01 |
| Highest AST in 24 hours before ECMO | 60 (34 – 122) | 43 (29 – 120) | 0.30 |
| Highest INR in 24 hours before ECMO | 1.2 (1.1 – 1.4) | 1.2 (1.1 – 1.4) | 0.59 |
| Respiratory Parameters | |||
| Oxygenation index | 40 (32 – 62) | 47 (36 – 63) | 0.50 |
| Lowest pH 6 hours prior to ECMO | 7.2 (7.1 – 7.3) | 7.2 (7.1 – 7.3) | 0.97 |
| Highest pCO2 6 hours prior to ECMO | 74 (62 – 98) | 77 (60 – 95) | 0.98 |
Continuous variables are expressed as median (interquartile range), categorical variables are expressed as a percentage.
Variables displayed are all previously established risk factors for mortality on ECMO.
Table 2:
Pre-ECMO Mechanical Ventilation for Children with Malignancy or Hematopoietic Cell Transplant Compared to Rest of Cohort
| Mode | Setting | Oncology/HCT (n=30) | Rest of Cohort (n=174) | p-value |
|---|---|---|---|---|
| Conventional Mechanical Ventilation | 8 (26.7) | 47 (27.0) | 0.90 | |
| High Frequency Oscillatory Ventilation | 18 (60.0) | 105 (60.3) | ||
| Airway Pressure Release Ventilation | 2 (6.7) | 14 (8.1) | ||
| High Frequency Percussive Ventilation | 0 (0) | 1 (0.6) | ||
| Extubated | 0 (0) | 0 (0) | ||
| Other | 2 (6.7) | 7 (4.0) | ||
| Mean Airway Pressure (all modes) | 29.5 (25, 32) | 27 (20, 30) | 0.11 | |
| FiO2 (all modes) | 1 (0.8, 1) | 1 (0.9, 1) | 0.66 | |
| Conventional Mechanical Ventilation | Rate, breaths/minute | 26.5 (16.0, 33.0) | 23.5 (20.0, 32.5) | 0.89 |
| FiO2 | 1 (0.88, 1) | 1 (0.70, 1) | 0.36 | |
| PEEP, cm H2O | 11 (9, 14) | 10 (7.5, 12) | 0.47 | |
| Mean Airway Pressure, cm H2O | 23 (17, 28) | 18 (15, 22) | 0.11 | |
| Peak Inspiratory Pressure, cm H2O | 43 (33, 50) | 37 (30, 44) | 0.25 | |
| Driving Pressure, cm H2O | 27 (22, 40) | 25 (19, 36) | 0.51 | |
| Compliance, mL/cm H2O | 3.5 (2.3, 5.1) | 4.3 (1.6, 8.8) | 0.82 | |
| Tidal Volume, mL/kg | 7.4 (5.8, 9.6) | 5.6 (4.3, 6.7) | 0.05 | |
| Inspiratory Time, seconds | 1.0 (1.0, 1.0) | 0.8 (0.6, 1.0) | 0.19 | |
| High Frequency Oscillatory Ventilation | Mean Airway Pressure, cm H2O | 30 (28, 32) | 29 (25, 32) | 0.12 |
| Frequency, Hertz | 7 (6, 8) | 7 (6, 8) | 0.67 | |
| FiO2 | 1 (0.90, 1) | 1 (0.92, 1) | 0.34 | |
| Amplitude | 62.5 (50, 88) | 60 (50, 75) | 0.59 |
Values are medians (IQRs) with p-values from Wilcoxon rank-sum test or frequencies (percentage) with Fisher’s exact test.
FiO2: Fraction of inspired oxygen
PEEP: Positive end-expiratory pressure
There were 21 subjects with an oncologic diagnosis without HCT, and characteristics of these patients are detailed in Table 3. Characteristics of the nine patients post-HCT are detailed in Table 4.
Table 3:
Oncology Patients Supported with ECMO
| Primary Disease | Disease Details | Age (Years) | Gender | Organism | Survival to ICU Discharge | Cause of Death |
|---|---|---|---|---|---|---|
| Neuroblastoma | Stage 4, new diagnosis | 0.4 | Female | RSV | Yes | Not applicable |
| Neuroblastoma | Stage 4 high risk, Diagnosed 10 months prior | 1.5 | Female | Rhinovirus/Enterovirus | Yes | Not applicable |
| Rhabdomyosarcoma | Grade 4, neutropenic, diagnosed 2 months prior | 2.8 | Male | RSV and pertussis | No | Bleeding complication |
| Medulloblastoma | None available | 1.5 | Female | RSV | No | Bleeding complication |
| ALL | Neutropenic, diagnosed about 4 weeks prior to ECMO | 12.6 | Female | RSV | No | Multi-organ failure |
| Relapsed T-Cell ALL | Diagnosed >1.5 years prior, ongoing chemo, neutropenic | 7.9 | Female | No organism recorded | No | Withdraw due to poor neurological prognosis |
| ALL | New diagnosis, about 2 weeks prior to ECMO, neutropenic | 6.5 | Male | HMPV | No | Refractory lung disease |
| Acute Promyelocytic Leukemia | High risk, new diagnosis | 13.4 | Male | No organism recorded | Yes | Not applicable |
| ALL | Standard risk, diagnosis 2 weeks prior to ECMO | 2.5 | Female | RSV | Yes | Not applicable |
| ALL | Standard risk, diagnosis 4 weeks prior to ECMO | 18.3 | Female | Rhinovirus/Enterovirus | No | Bleeding complication |
| ALL | History of hemophagocytic lymphoproliferative disease and aplastic anemia | 11.8 | Male | Rhinovirus/Enterovirus | No | Multi-organ failure |
| T-cell ALL | Diagnosed 6 weeks prior, on chemotherapy | 5.9 | Male | RSV | No | Bleeding complication |
| ALL | Diagnosed 7 months prior to admission, on chemotherapy, high risk | 12.2 | Female | HSV | No | Refractory lung disease |
| AML | New diagnosis after ECMO initiation | 16.4 | Female | RSV | Yes | Not applicable |
| T-cell ALL | Intermediate risk, neutropenic, diagnosed 4 months prior | 16.3 | Male | Streptococcus viridans bacteremia | Yes | Not applicable |
| ALL | High risk, neutropenic, diagnosed 4 weeks prior | 2.1 | Female | RSV and Streptococcus viridans bacteremia | Yes | Not applicable |
| ALL | Neutropenic on admission, recovered counts prior to ECMO, high risk, diagnosed >1 year prior | 5.8 | Male | Disseminated CMV | No | Accidental ECMO cannula displacement |
| ALL | Neutropenic | 6.8 | Female | No organism recorded | Yes | Not applicable |
| AML | Neutropenic | 16.9 | Female | No organism recorded | No | Bleeding complication |
| AML | Not available | 0.6 | Male | Parainfluenza | No | Multi-organ failure |
| ALL | High risk | 0.3 | Female | Rhinovirus/Enterovirus | No | Refractory lung disease |
ALL: Acute lymphoblastic leukemia
AML: Acute myeloid leukemia
ARF: Acute respiratory failure
RSV: Respiratory syncytial virus
HMPV: Human metapneumovirus
HSV: Herpes simplex virus
CMV: Cytomegalovirus
Table 4:
Hematopoietic Cell Transplant Patients Supported with ECMO
| Primary Disease | Cancer Details | HCT Details | Age (years) | Gender | Organism(s) | Survival to ICU Discharge | Cause of Death |
|---|---|---|---|---|---|---|---|
| AML | Diagnosed 2 years prior | Matched unrelated donor transplant | 17.4 | Male | Mycobacterium | No | Multi-organ failure |
| Severe combined immunodeficiency and bronchiolitis obliterans | Not applicable | None available | 2.3 | Male | Parainfluenza | No | Refractory lung disease |
| Sickle Cell Disease | Not applicable | Engrafted at time of ECMO cannulation, but with loss of donor cells after ECMO | 3.6 | Male | CMV | Yes | Not applicable |
| ALL | High risk | Matched sibling donor, VOD, pulmonary hypertension | 14.2 | Female | No organism recorded | No | Refractory lung disease |
| Osteopetrosis | Not applicable | Matched unrelated donor transplant, engrafted | 0.6 | Female | No organism recorded | Yes | Not applicable |
| Common variable immunodeficiency | Not applicable | Day +20 from HCT at time of ECMO cannulation | 9.4 | Female | No organism recorded | No | Multi-organ failure |
| Common variable Immunodeficiency | Not applicable | Haploidentical donor, father | 7.7 | Female | Aspergillus, CMV, and adenovirus | No | Multi-organ failure |
| Burkitt lymphoma | None available | Autologous bone marrow transplant, engrafted | 5.9 | Male | HMPV | No | Bleeding complication |
| Diamond Blackfan anemia | Not applicable | None available | 18.3 | Female | No organism recorded | Yes | Not applicable |
ALL: Acute lymphoblastic leukemia
AML: Acute myeloid leukemia
ARF: Acute respiratory failure
CMV: Cytomegalovirus
HCT: Hematopoietic Cell Transplant
HMPV: Human metapneumovirus
VOD: Veno-occlusive disease
Outcomes
The oncology/HCT subjects had a higher rate of conversion to V-A ECMO compared to non-oncology/HCT subjects (23% vs. 9%, p=0.02) (RR 2.54, 95% CI 1.1–5.6). Upon analysis of the entire cohort of V-V ECMO patients, the only identified risk factor for the need for conversion to V-A ECMO was the presence of oncologic disease or HCT. Table 5 compares the mechanical ventilation modes, settings, and use of bronchoscopy during ECMO for the oncology/HCT subgroup compared to the rest of the cohort.
Table 5:
On ECMO Mechanical Ventilation for Children with Malignancy or Hematopoietic Cell Transplant Compared to Rest of Cohort
| Oncology/HCT (n=30) | Rest of Cohort (n=174) | p-value | |
|---|---|---|---|
| Mode | |||
| Conventional Mechanical Ventilation | 22 (73.3) | 133 (76.4) | 0.30 |
| High Frequency Oscillatory Ventilation | 1 (3.3) | 8 (4.6) | |
| Airway Pressure Release Ventilation | 4 (13.3) | 24 (13.8) | |
| High Frequency Percussive Ventilation | 1 (3.3) | 7 (4.0) | |
| Extubated | 2 (6.7) | 1 (0.6) | |
| Other | 0 (0) | 1 (0.6) | |
| Mechanical Ventilation Settings and Support | |||
| Number of Bronchoscopies | 2 (1, 3) | 1 (1, 3) | 0.30 |
| PEEP, cm H2O, Three Day Average | 10 (10, 12) | 10 (10, 10) | 0.02 |
| Delta Pressure, cm H2O, Three Day Average | 12 (10, 16) | 11 (10, 16) | 0.43 |
| Mean Airway Pressure, cm H2O, Three Day Average | 15.0 (13.7, 17.3) | 14.2 (12.7, 16.7) | 0.09 |
| Tidal Volume, ml/kg, Three Day Average | 1.6 (0.8, 3.7) | 2.1 (1.0, 3.9) | 0.36 |
| Compliance, mL/cm H2O, Three Day Average | 2.7 (0.8, 5.9) | 2.0 (1.1, 5.6) | 0.87 |
| FiO2, Three Day Average | 0.4 (0.3, 0.6) | 0.4 (0.3, 0.5) | 0.20 |
| SpO2, Three Day Average | 90.0 (84.0, 92.3) | 91.3 (87.8, 94.7) | 0.02 |
Values are median (IQRs) with p-values from Wilcoxon rank-sum test or frequencies (percentage) with Fisher’s Exact test.
PEEP: Positive end-expiratory pressure
FiO2: Fraction of inspired oxygen
SpO2: Peripheral oxygen saturation
Mortality on ECMO was higher in the oncology/HCT subgroup (43%) when compared to the rest of the cohort (21%) (p<0.01) (OR, 3.0, 95% Confidence Interval [CI], 1.3–6.7). Isolating just the ECMO survivors, the oncology/HCT group had a higher ICU mortality post-ECMO (35% vs 8%, p<0.01). Overall ICU mortality was higher in the oncology/HCT subgroup (67%) compared to the rest of the cohort (28%) (p<0.001) (OR, 4.6, 95% CI, 2.1–10.5). The cause of death did not differ between the oncology/HCT subgroup and the rest of the cohort, with bleeding complications being the most common cause (Figure 1). However, it is notable that when examining the entire oncology/HCT group of 30 patients, 20% (6) died due to bleeding complications, 20% (6) died due to multiple organ failure, and 17% (5) died of refractory lung disease. This compares to the cohort of 174 non-oncology/HCT patients, in which only 9% (16) died due to bleeding complications, 7% (13) died due to multiple organ failure, and 5% (9) died due to refractory lung disease.
Figure 1.
Causes of death in general cohort and oncology/hematopoietic cell transplant cohort.
Subgroup of oncology/HCT subjects
In the oncology/HCT group, there were no differences between survivors and non-survivors in any of the pre-ECMO variables investigated, including days intubated before ECMO, OI, AKI, platelet count, and incidence of leukopenia. Table 6 shows selected pre-ECMO variables for just the oncology/HCT group by survival.
Table 6:
Pre-ECMO Variables in Children with Malignancy or Hematopoietic Cell Transplant
| Demographics | ICU Survivors (n=11) | ICU Non-Survivors (n=19) | p-value |
|---|---|---|---|
| Age (years) | 3.6 (1.5 – 16.3) | 7.7 (2.8 – 12.6) | 0.54 |
| Cardiac Arrest | 0% | 11% | 0.52 |
| Genetic Disorder | 18% | 11% | 0.61 |
| Days of Intubation to ECMO | 2 (1 – 5) | 2 (1 – 8) | 0.61 |
| Organ Dysfunction Variables | |||
| Acute Kidney Injury | 36% | 16% | 0.37 |
| Hypotensive Prior to ECMO | 27% | 42% | 0.42 |
| Highest Creatinine Prior to ECMO, mg/dL | 0.6 (0.3 – 1.3) | 0.7 (0.5 – 0.8) | 0.75 |
| White Blood Cell Count 2000/per mm3 or Less | 73% | 53% | 0.44 |
| Lowest Platelet Count, per mm3 | 49 (46 – 70) | 53 (24 – 66) | 0.85 |
| Respiratory Measures | |||
| Oxygenation Index | 54.0 (39.0 – 67.6) | 37.1 (30.8 – 41.5) | 0.15 |
| Lowest pH 6 Hours Prior to ECMO | 7.2 (7.1 – 7.3) | 7.1 (7.1 – 7.3) | 0.12 |
| Highest PCO2 6 Hours Prior to ECMO | 69 (53 – 74) | 84 (64 – 107) | 0.10 |
Continuous variables are expressed as median (interquartile range), categorical variables are expressed as a percentage.
In the 21 subjects with a cancer diagnosis but no HCT, survival was 38%. Type of cancer (hematologic versus solid tumor) was not associated with survival. In the same group of 21 subjects, survival was not different for those placed on ECMO within 1 month of diagnosis as compared to being more remote from diagnosis.
In all subjects with HCT, survival was 33% (3/9). This survival was significantly worse for those whose indication for HCT was cancer or immunodeficiency (0/6) as compared to other indications (sickle cell disease, osteopetrosis, Diamond-Blackfan anemia) (3/3) (p=0.01).
Discussion
The use of ECMO for children with cancer and HCT has grown.9 Previous studies of this population have been either single center reports or data from the ELSO registry, and all have shown poor outcomes.8–13 This multicenter study demonstrated a similarly high mortality in 10 pediatric centers in the current era. Children with cancer or HCT were 3 times more likely to die on ECMO and 4.6 times more likely to die prior to ICU discharge compared to children supported on V-V ECMO without oncologic disease or HCT.
The increased risk of mortality for this vulnerable population is not a new finding. However, this study provides new insights into the outcomes for children with cancer or HCT supported with V-V ECMO. First, almost one quarter of children in the oncology/HCT group were converted to V-A ECMO. The incidence of conversion to V-A was significantly higher than the rest of the cohort despite having similar hemodynamic measures before ECMO cannulation. We suspect the need for V-A conversion was higher due to the increased likelihood for subsequent organ failure, including cardiovascular failure, in the oncology/HCT group. When initiating V-V ECMO for a subject with a cancer diagnosis or HCT, providers and families should be aware of the possible need for conversion to V-A ECMO. Secondly, more than one third of children with cancer or HCT who survived ECMO died before ICU discharge. This post-ECMO mortality was significantly higher than other children on V-V ECMO. This increased risk of post ECMO mortality in children with cancer or HCT has not previously been described, but this finding is similar to the high mortality HCT subjects suffer post ICU discharge for PARDS.7
In this contemporary cohort, children with oncologic diagnosis/HCT had worse outcomes despite being similar to other children on ECMO in many observed variables that have been previously associated with mortality.17,18,22 As expected, subjects in the oncology/HCT group had more bone marrow failure defined by a higher incidence of leukopenia and lower platelet counts. Poor bone marrow function resulting in leukopenia and an increased susceptibility to infection, as well as an increased risk of bleeding due to thrombocytopenia have been concerns when utilizing ECMO in children with cancer or HCT.12 However, neither the presence of leukopenia nor platelet count was associated with mortality within the oncology/HCT group. However, it is notable that odds of death due to bleeding complications were twice as high in the oncology/HCT group compared to the non-oncology/HCT group. Bone marrow failure may be a contributing factor in the outcome differences seen between the oncology/HCT group and the non-oncology/HCT group, but within the oncology/HCT group, the degree of bone marrow failure was not associated with a higher risk of mortality.
Risk factors that have been established in the general respiratory ECMO population may not apply to the oncology/HCT subgroup. When the oncology/HCT subjects were isolated, no pre-ECMO variables were associated with mortality. This included many previously identified risk factors for mortality in children on ECMO such as oxygenation index (OI), pre-ECMO cardiac arrest, pH, pCO2, days of mechanical ventilation prior to cannulation, and AKI. This may be confounded due to the relatively small size of the oncology/HCT cohort. In a previous report on this subgroup, higher OI was associated with mortality. However, the median OI was 52 in the previous study8 compared to a median OI of 40 in the present study. Longer time from intubation to ECMO initiation has previously been associated with mortality in HCT subjects.11 Historically, duration of invasive mechanical ventilation before ECMO was longer than the current cohort (median of 5 days vs 2 days).11 These likely indicate changes in practice due to the previously identified risk factors of duration of mechanical ventilation before ECMO and OI. While not statistically significant, it is notable that the median OI among the oncology/HCT patients in this study was higher in survivors than non-survivors (54.0 versus 37.1, p-value 0.15). It is possible that those patients with an elevated OI were placed on ECMO sooner, and the timing of ECMO initiation could have influenced the outcome. It would appear that children in the current study were placed on ECMO earlier and with less severe lung disease than in previous studies. However, survival was similarly poor in the current cohort despite changes in practice that would be thought to improve outcomes.
The survival for the subgroup of children with HCT supported with V-V ECMO was 33%, which is slightly better than previous reports (0–10%).10,11,13 The higher survival rate may be due to expanding indications for HCT. When HCT subjects were assessed by indication, none of the subjects with an impaired immune system prior to HCT (cancer or immunodeficiency) survived. However, all 3 subjects with an intact immune system prior to HCT (sickle cell disease, osteopetrosis, and Diamond-Blackfan anemia) did survive. In a previous ELSO database report of 29 HCT subjects on ECMO, 1 of 23 (4%) children whose indication for HCT was cancer or immunodeficiency survived, while 2 out of 6 (33%) with other indications survived.13 Therefore, when evaluating children for ECMO who have had a HCT, the indication, immune status (pre-HCT and current), and toxicities of therapies prior to HCT should be considered.
There has not been significant progress in recent decades with respect to outcomes for children with cancer or HCT on ECMO. These children have worse outcomes than the general pediatric ECMO population despite having similar pre-ECMO risk factors. Additionally, no risk factors for mortality within the oncology/HCT group were identified in this study. The vast majority of pediatric centers do not consider an oncological diagnosis as a contraindication to ECMO. Therefore, there is a need to continue to work towards improving risk prognostication and subject selection.
The overall ECMO survival for this group is similar to other groups of patients needing ECMO, such as those with adenovirus infection (38%),23 pertussis infection (28%),24 and ECPR (43%).25 Despite the poor outcomes demonstrated in our cohort, having an oncologic diagnosis or being a HCT recipient should not exclude ECMO as a support option. Instead, we need to consider the individual patient and understand the factors that may impact survival. These complex patients have additional factors that must be considered when determining if an individual is an appropriate candidate for ECMO, including primary disease, existing organ failures, engraftment status, the presence of graft versus host disease, and bleeding risk.
There are limitations to this study. Primarily, this is a retrospective study that can only identify associations and cannot comment on causation. Secondly, this was a secondary analysis of a relatively small number of patients from an existing database. The primary goal was not to investigate subjects with cancer or HCT, therefore all variables specific to cancer and HCT were not collected. In particular, we lacked data on hemolysis and anticoagulation requirements in this population. Nonetheless, this represents the first multicenter report on children with cancer or HCT supported with V-V ECMO using data outside of the ELSO registry, and this contributes to the literature for this high-risk population.
Conclusions
Oncology and HCT subjects on V-V ECMO have significantly worse clinical outcomes by multiple measures, despite having similar pre-ECMO clinical characteristics to other children supported with V-V ECMO. Outcomes for Oncology and HCT subjects are similar to other high risk ECMO patient populations. No Pre-ECMO risk factors for mortality were identified within the oncology/HCT subgroup. Prognostication and subject selection in this unique group of subjects remains difficult.
Acknowledgments
This study was financially supported by the Extracorporeal Life Support Organization and the Pediatrics Department at Indiana University School of Medicine.
Disclosures:
Dr. Barbaro is the chair of the Extracorporeal Life Support Organization Registry. Dr. Bembea has received funding from the NIH NICHD and NINDS for work unrelated to this study. Dr. Cheifetz is a medical advisor for Philips, a contributing author for Up-to-Date, and has received grant support from NHLBI for work unrelated to this study. Dr. Barbaro has received funding from the Training to Advance Care Through Implementation Science in Cardiac and Lung illnesses (TACTICAL) NHLBI, NIH K12 HL138039 unrelated to this study. Dr. Rowan has received funding from the NIH K12 HD068371–01A1 unrelated to this study. The remaining authors do not have any potential conflicts of interest to disclose.
References
- 1.Tamburro RF, Barfield RC, Shaffer ML, et al. Changes in outcomes (1996–2004) for pediatric oncology and hematopoietic stem cell transplant patients requiring invasive mechanical ventilation. Pediatr Crit Care Med. 2008;9(3):270–277. [DOI] [PubMed] [Google Scholar]
- 2.Curtin SC, Minino AM, Anderson RN. Declines in Cancer Death Rates Among Children and Adolescents in the United States, 1999–2014. NCHS Data Brief. 2016(257):1–8. [PubMed] [Google Scholar]
- 3.Gooley TA, Chien JW, Pergam SA, et al. Reduced mortality after allogeneic hematopoietic-cell transplantation. N Engl J Med. 2010;363(22):2091–2101. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Chima RS, Daniels RC, Kim MO, et al. Improved outcomes for stem cell transplant recipients requiring pediatric intensive care. Pediatr Crit Care Med. 2012;13(6):e336–342. [DOI] [PubMed] [Google Scholar]
- 5.Erickson S, Schibler A, Numa A, et al. Acute lung injury in pediatric intensive care in Australia and New Zealand: a prospective, multicenter, observational study. Pediatr Crit Care Med. 2007;8(4):317–323. [DOI] [PubMed] [Google Scholar]
- 6.Ben Abraham R, Toren A, Ono N, et al. Predictors of outcome in the pediatric intensive care units of children with malignancies. J Pediatr Hematol Oncol. 2002;24(1):23–26. [DOI] [PubMed] [Google Scholar]
- 7.Rowan CM, Smith LS, Loomis A, et al. Pediatric Acute Respiratory Distress Syndrome in Pediatric Allogeneic Hematopoietic Stem Cell Transplants: A Multicenter Study. Pediatr Crit Care Med. 2017;18(4):304–309. [DOI] [PubMed] [Google Scholar]
- 8.Gow KW, Heiss KF, Wulkan ML, et al. Extracorporeal life support for support of children with malignancy and respiratory or cardiac failure: The extracorporeal life support experience. Crit Care Med. 2009;37(4):1308–1316. [DOI] [PubMed] [Google Scholar]
- 9.Gupta M, Shanley TP, Moler FW. Extracorporeal life support for severe respiratory failure in children with immune compromised conditions. Pediatr Crit Care Med. 2008;9(4):380–385. [DOI] [PubMed] [Google Scholar]
- 10.Zabrocki LA, Brogan TV, Statler KD, Poss WB, Rollins MD, Bratton SL. Extracorporeal membrane oxygenation for pediatric respiratory failure: Survival and predictors of mortality. Crit Care Med. 2011;39(2):364–370. [DOI] [PubMed] [Google Scholar]
- 11.Gow KW, Wulkan ML, Heiss KF, et al. Extracorporeal membrane oxygenation for support of children after hematopoietic stem cell transplantation: the Extracorporeal Life Support Organization experience. J Pediatr Surg. 2006;41(4):662–667. [DOI] [PubMed] [Google Scholar]
- 12.Maue DK, Hobson MJ, Friedman ML, Moser EA, Rowan CM. Outcomes of pediatric oncology and hematopoietic cell transplant patients receiving extracorporeal membrane oxygenation. Perfusion. 2019:267659119842471. [DOI] [PubMed] [Google Scholar]
- 13.Di Nardo M, Locatelli F, Palmer K, et al. Extracorporeal membrane oxygenation in pediatric recipients of hematopoietic stem cell transplantation: an updated analysis of the Extracorporeal Life Support Organization experience. Intensive Care Med. 2014;40(5):754–756. [DOI] [PubMed] [Google Scholar]
- 14.Friedman ML, Barbaro RP, Bembea MM, et al. Mechanical Ventilation in Children on Venovenous ECMO. Respir Care. 2020. [DOI] [PubMed] [Google Scholar]
- 15.Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42(2):377–381. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Leteurtre S, Martinot A, Duhamel A, et al. Validation of the paediatric logistic organ dysfunction (PELOD) score: prospective, observational, multicentre study. Lancet. 2003;362(9379):192–197. [DOI] [PubMed] [Google Scholar]
- 17.Bailly DK, Reeder RW, Zabrocki LA, et al. Development and Validation of a Score to Predict Mortality in Children Undergoing Extracorporeal Membrane Oxygenation for Respiratory Failure: Pediatric Pulmonary Rescue With Extracorporeal Membrane Oxygenation Prediction Score. Crit Care Med. 2017;45(1):e58–e66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Barbaro RP, Boonstra PS, Paden ML, et al. Development and validation of the pediatric risk estimate score for children using extracorporeal respiratory support (Ped-RESCUERS). Intensive Care Med. 2016;42(5):879–888. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Kidney Disease: Improving Global Outcomes (KDIGO) Acute Kidney Injury Work Group: KDIGO Clinical Practice Guideline for Acute Kidney Injury. Kidney Int Suppl. 2012;2:1–138. [Google Scholar]
- 20.Akcan-Arikan A, Zappitelli M, Loftis LL, Washburn KK, Jefferson LS, Goldstein SL. Modified RIFLE criteria in critically ill children with acute kidney injury. Kidney Int. 2007;71(10):1028–1035. [DOI] [PubMed] [Google Scholar]
- 21.American Academy of Pediatrics., American Heart Association. Pediatric advanced life support : Provider manual. Dallas, TX: American Heart Association; 2016. [Google Scholar]
- 22.Pollack MM, Holubkov R, Funai T, et al. The Pediatric Risk of Mortality Score: Update 2015. Pediatr Crit Care Med. 2016;17(1):2–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Prodhan P, Bhutta AT, Gossett JM, et al. Extracorporeal membrane oxygenation support among children with adenovirus infection: a review of the Extracorporeal Life Support Organization registry. ASAIO J. 2014;60(1):49–56. [DOI] [PubMed] [Google Scholar]
- 24.Domico M, Ridout D, MacLaren G, et al. Extracorporeal Membrane Oxygenation for Pertussis: Predictors of Outcome Including Pulmonary Hypertension and Leukodepletion. Pediatr Crit Care Med. 2018;19(3):254–261. [DOI] [PubMed] [Google Scholar]
- 25.Barbaro RP, Paden ML, Guner YS, et al. Pediatric Extracorporeal Life Support Organization Registry International Report 2016. ASAIO J. 2017;63(4):456–463. [DOI] [PMC free article] [PubMed] [Google Scholar]