Secondary Logo

Journal Logo


Extracorporeal Life Support in Pediatric Patients with Bronchopulmonary Dysplasia and Associated Pulmonary Hypertension

Pena Hernandez, Alejandra*; Carr, Nicholas R.; McCurnin, Donald*; Armijo-Garcia, Veronica

Author Information
doi: 10.1097/MAT.0000000000001134


Bronchopulmonary dysplasia (BPD) develops in 12.8% of all infants born ≤34 weeks gestation in the United States as reported by a recent population study with data from the National Inpatient Sample.1 The Neonatal Research network has reported stable rates of BPD (approximately 40%) in infants born ≤28 weeks gestation.2 Patients with BPD have structural and functional lung abnormalities that persist into young adulthood and predispose them to frequent serious respiratory infections.3–5 Secondary pulmonary hypertension (PH) occurs in approximately 17% of patients with BPD, and these patients have substantially higher morbidity and mortality.6–8 Extracorporeal life support in pediatric patients is offered as a bridge for recovery from not only cardiovascular etiologies but also acute respiratory processes.

A paucity of evidence exists regarding the use of extracorporeal life support (ECLS) in patients with a history of prematurity and BPD. Hibbs et al.9 reported 76 infants less than 1 year old with BPD that were treated with ECLS and had a survival to discharge of 78%, comparable to or better than other respiratory ECLS outcomes at that time.10 However, no data exist regarding outcomes for patients with BPD outside the early infancy period requiring ECLS.

The purpose of this study is to determine whether a diagnosis of BPD before ECLS management influences survival and other outcomes in pediatric patients. In addition, we wanted to examine a group known to have higher morbidity and mortality, patients with BPD and associated PH, to further establish their outcomes.

Materials and Methods

This is a retrospective cohort study using data from the ELSO registry. In 2017, more than 300 centers worldwide contributed to the ELSO registry data by voluntary standardized submission. Data include patient demographics, primary indication for ECLS, other diagnoses, type of therapy, complications, and survival at hospital discharge.

Data from 1982 through 2018 were queried for pediatric patients aged 60 days to 18 years with a diagnosis of BPD managed with ECLS. BPD was identified using the International Classification of Diseases Clinical Modification, Ninth Revision (ICD-9) code 770.7 (Chronic respiratory disease arising in the perinatal period) or ICD-10 codes P27.1 (Bronchopulmonary dysplasia originating in the perinatal period) or P27.8 (Other chronic respiratory diseases originating in the perinatal period). Children with BPD + PH were identified using the ICD-9 code 416.8 (Other chronic pulmonary heart diseases) and ICD-10 code I27.2 (Other secondary pulmonary hypertension). For patients undergoing ECLS more than once, only data from the first run were analyzed. Patients requiring ECLS for cardiac support, with a diagnosis of congenital heart disease, congenital diaphragmatic hernia, congenital anomalies of the airway or lungs (excluding tracheoesophageal fistula), or without a valid primary diagnosis available were excluded.

This study used only deidentified data and was compliant with the exempt categories of research as defined by the Institutional Board Review at The University of Texas Health Science Center at San Antonio.

The primary outcome was survival to hospital discharge. Survival to discharge was stratified by age groups and compared with a Fisher’s exact test. Descriptive statistics were used to compare patient characteristics. All continuous variables were assessed for distribution and based on the Shapiro–Wilk test for normality, none of the variables were normally distributed. Therefore, continuous variables were reported as median and range. Fisher’s exact test was used for categorical data, and Wilcoxon rank sum test was used for nonparametric data. Statistical analyses were performed using Microsoft Excel (Version 16.6.7) and JMP (Version 14. SAS Institute Inc., Cary, NC, 1989–2019). A p value of ≤0.05 was considered significant.

In the ELSO registry, the modality of ECLS is recorded as veno-arterial (VA), veno-venous (VV), VV double lumen, VV double lumen plus additional venous cannula, VA plus additional venous cannula, and VV converted to VA and vice versa. For this study, data were grouped as any VA modality, any VV modality, and conversions were noted.


Between January 1982 and July 2018, there were 281 pediatric patients (aged 60 days to 18 years) with BPD who were managed with ECLS as reported in the ELSO registry. The first reported case with BPD + PH in the ELSO registry was in 2004. No patients meeting our criteria were identified in the year 2018. A total of 143 patients managed with ECLS from 2004 to 2017 (128 patients with BPD and 15 patients with BPD + PH) were included in the analysis.

The reported use of ECLS in the BPD population has remained limited throughout the years; however, the reporting of patients with BPD + PH has increased since 2004, with more than 50% of the patients (8/15) reported after 2015 (Figure 1).

Figure 1.
Figure 1.:
Annual number of patients with BPD and BPD + PH managed with ECLS and survival. BPD, bronchopulmonary dysplasia; PH, pulmonary hypertension; ECLS, extracorporeal life support.

The median age of patients with BPD and BPD + PH was similar (217 [range: 60–4,924] days vs. 240 [range: 112–540] days, p = 0.62), with similar distribution of gender (BPD 53.2% males vs. BPD + PH 46.7% males, p = 0.79). Patients with BPD + PH had more comorbidities reported than patients with BPD to include intraventricular hemorrhage (IVH; 26.7% vs. 2.3%, p = 0.0023) and retinopathy of prematurity (ROP; 20.0% vs. 4.7%, p = 0.0537). Only one patient with a tracheostomy was identified in the group of BPD + PH and none in the BPD group, as well as one patient with ventilator dependency before ECLS in the same group (Table 1).

Table 1. - Characteristics 2004–2017
Characteristic BPD (n = 128) BPD + PH (n = 15) p Value
 Male, n (%)* 67 (53.2) 7 (46.7) 0.79
 Weight (kg)† 5 (1.9–50.2) 5.7 (2.5–9.2) 0.81
 Age (days)† 217 (60–4,924) 240 (112–544) 0.62
 Hours on ECLS† 272 (1–1,686) 258.5 (77–1,263) 0.88
ECLS modality, n (%)‡
 VA 66 (51.6) 5 (33.3) 0.17
 VV 54 (42.2) 10 (66.7)
 VA to VV 1 (0.8) 0 (0.0)
 VV to VA 6 (4.7) 0 (0.0)
 Unknown 1 (0.8) 0 (0.0)
Ventilatory support, n (%)§
 Conventional 43 (33.6) 6 (40.0) 0.76
 HFV 56 (43.8) 6 (40.0)
 Other HFV 4 (3.1) 2 (13.3)
 No report 25 (19.5) 1 (6.7)
Comorbidities, n (%)
 IVH 3 (2.3) 4 (26.7) 0.002
 ROP 6 (4.7) 3 (20.0) 0.05
 Cerebral palsy 2 (1.6) 0 (0.0) >0.99
 Developmental disorder 2 (1.6) 0 (0.0) >0.99
 NEC 0 (0.0) 0 (0.0) >0.99
 Short bowel syndrome 1 (0.8) 0 (0.0) >0.99
 Tracheostomy 0 (0.0) 1 (6.7) >0.99
 Ventilator dependency 0 (0.0) 1 (6.7) >0.99
 Oxygen dependency 1 (0.8) 0 (0.0) >0.99
 Gastrostomy 5 (3.9) 1 (6.7) 0.49
BPD, bronchopulmonary dysplasia; PH, pulmonary hypertension; ECLS, extracorporeal life support; VA, veno-arterial; VV, veno-venous; HFV, high-frequency ventilation; IVH, intraventricular hemorrhage; ROP, retinopathy of prematurity; NEC, necrotizing enterocolitis.
*Percentages accounted for missing data.
†Results presented in median (range) for nonparametric data.
‡Only VA vs. VV were compared.
§Only conventional vs. HFV were compared.
p Values indicate Fisher’s exact test or Wilcoxon rank sum test as appropriate.

Median duration of ECLS was not significantly different in patients with BPD and BPD + PH (272 [range: 1–1,686] hours vs. 258.5 [range 77–1,263] hours, p = 0.88). Ventilatory support strategy and modality of ECLS were not different in patients with BPD compared with BPD + PH. Pre-ECLS ventilatory support was similar in both groups, with oxygenation index (OI) of 49.4 in patients with BPD and 53.2 in patients with BPD + PH (p = 0.21). Patients with BPD + PH had more frequent use of sildenafil (40.0% vs. 1.6%, p = 0.0001) and milrinone (53.3% vs. 13.3%, p = 0.008) pre-ECLS compared with patients with BPD. The use of epoprostenol and nitric oxide was also higher in patients with BPD + PH but not significantly different (Table 2).

Table 2. - Pre-ECLS Support 2004–2017
BPD (n = 128) BPD + PH (n = 15) p Value
Ventilatory support*
 FiO2 (%) 100.0 (30–100) 100.0 (70–100) 0.65
 PIP (cm H2O) 43.0 (18–93) 40.0 (26–55) 0.57
 PEEP (cm H2O) 10.0 (2–23) 8.5 (5–11) 0.64
 MAP (cm H2O) 21.0 (9–40) 22.2 (10–33) 0.84
 OI 45.6 (4.8–184.6) 52.7 (29.4–80.9) 0.21
 pH 7.3 (6.8–7.8) 7.3 (6.8–7.6) 0.56
 PCO2 (mmHg) 63.5 (27.3–235) 66.0 (17.6–160) 0.87
 PO2 (mmHg) 46.5 (11–441) 44.0 (19–68) 0.60
 HCO3 (mmol/L) 29.6 (13.7–59) 30.7 (15–60) 0.59
 SaO2 (%) 82.0 (21.1–100) 82.5 (17–95) 0.64
Medications, n (%)
 Epoprostenol 3 (2.3) 1 (6.7) 0.36
 Milrinone 17 (13.3) 8 (53.3) 0.008
 Nitric oxide 84 (65.6) 13 (86.7) 0.14
 Sildenafil 2 (1.6) 6 (40.0) 0.0001
 Surfactant 7 (5.5) 0 (0.0) >0.99
 Systemic steroids 21 (16.4) 4 (26.7) 0.3
ECLS, extracorporeal life support; PIP, peak inspiratory pressure; PEEP, positive end expiratory pressure; MAP, mean airway pressure; OI, oxygenation index.
*Results presented in median (range) for nonparametric data.

The most common primary diagnosis in patients with BPD was viral respiratory infection (45.3%), and within this group, pneumonia caused by respiratory syncytial virus was coded in 50.4% of the cases. In patients with BPD + PH, the most common primary diagnosis was respiratory failure without mention of infection (40.0%) followed by secondary PH (26.7%; Table 3).

Table 3. - Indication for ECLS
Primary Diagnosis BPD (n = 128) BPD + PH (n = 15)
n % n %
Viral respiratory infection 58 45.3 2 13.3
Bacterial pneumonia 6 4.7 0 0.0
Unspecified pneumonia 2 1.6 1 6.7
Aspiration pneumonitis 1 0.8 1 6.7
Respiratory failure, no mention of infection 27 21.1 6 40.0
BPD 5 3.9 0 0.0
Secondary PH 2 1.6 4 26.7
Sepsis 5 3.9 0 0.0
Tracheoesophageal fistula 2 1.6 0 0.0
Circulatory failure 1 0.8 0 0.0
Cardiac arrest 7 5.5 1 6.7
Myocardial or pericardial disease 2 1.6 0 0.0
Other 10 7.8 0 0.0
BPD, bronchopulmonary dysplasia; PH, pulmonary hypertension.

Overall survival to discharge for patients with BPD was 68.0% and for patients with BPD + PH was 86.7% (p = 0.55). We stratified the survival by age groups because older patients were included in the BPD group. Survival was still not significantly different in infants or toddlers (66.3% vs. 83.3%, p = 0.35 and 65.7% vs. 100%, p = 0.54, respectively). No patients with BPD + PH were reported in preschool or older years (>3 years; Table 4). Lung recovery was the most common reason for discontinuation of ECLS in both groups (BPD 78.9%, BPD + PH 93.3%; Table 5).

Table 4. - Survival by Age Group
Age Group BPD (n = 128) BPD + PH (n = 15) p Value
Cases Survival, n (%) Cases Survival, n (%)
Infant (2 months to 1 year) 83 55 (66.3) 12 10 (83.3) 0.33
Toddler (>1 year to 3 years) 35 23 (65.7) 3 3 (100.0) 0.54
Preschool (>3 years to 5 years) 5 5 (100.0) 0 NA NA NA
School age and above (>5 years) 5 4 (80.0) 0 NA NA NA
Grand total 128 87 (68.0) 15 13 (86.7) 0.23
BPD, bronchopulmonary dysplasia; PH, pulmonary hypertension.

Table 5. - Reason for Discontinuation of ECLS Support
ECLS Discontinuation BPD (n = 128) BPD + PH (n = 15)
n % n %
Lung recovery 101 78.9 14 93.3
Family request 4 3.1 0 0.0
Hemorrhage 4 3.1 0 0.0
Diagnosis incompatible with life 3 2.3 0 0.0
Organ failure 9 7.0 0 0.0
Not specified 7 5.5 1 6.7
ECLS, extracorporeal life support; BPD, bronchopulmonary dysplasia; PH, pulmonary hypertension.

The most common complications reported in all patients with BPD from 1982 to 2017 were inotropes use (36.7% with 54.4% survival), hypertension requiring vasodilators (31.0% with 72.4% survival), and cannula problems (19.6% with 67.3% survival). In our analyzed group from 2004 to 2017, the most common complications for patients with BPD were still hypertension requiring vasodilators (37.5% with 79.2% survival), inotropes use (35.9% with 52.2% survival), and cannulation site bleeding (28.1% with 75% survival). In patients with BPD + PH, inotropes use (33.3% with 80.0% survival), cannulation site bleeding (26.7% with 75.0% survival), and requirement of renal replacement therapy (26.7% with 75.0% survival) were the most commonly reported complications. Severe neurologic complications such as central nervous system (CNS) or IVH hemorrhage or infarction were not reported in the BPD + PH group, and for the patients with BPD they were reported in less than 8% of the runs (see Table 1, Supplemental Digital Content 1,


Survival to hospital discharge for patients with BPD and BPD + PH (68.0% and 86.7%, respectively) is encouraging and favorable compared with other pediatric patients who received respiratory ECLS as reported by ELSO. In the ELSO International Report from January 2019,10 survival to discharge for pediatric respiratory ECLS cases was 58%. In addition, potentially longer ECLS runs expected in patients with BPD were not seen; our findings suggest that the length of ECLS runs is similar in patients with BPD (272 hours), BPD + PH (258.5 hours), and all pediatric patients reported by ELSO (285 hours).

Despite PH being historically associated with an expected increase in morbidity and mortality, there was no difference in survival outcomes of pediatric patients managed on ECLS. The lack of a reported diagnosis of associated PH in patients with BPD older than 3 years of age is an interesting finding. One of the possible explanations for this observation is the natural history of PH which includes the possibility of improvement of symptoms or resolution overtime.11 We should also acknowledge that this retrospective review of self-reported data can only speculate on the accuracy of the workup for PH and its reporting to the ELSO database.

ECLS has been used in patients with BPD for more than 3 decades, but patients with BPD + PH were not reported in the ELSO registry until 2004. As it is noted in our results, the number of patients with BPD + PH supported with ECLS is increasing. The coding system used for the ELSO registry since the early years allows for the report of PH. We are uncertain regarding the reason for the absence of patients with BPD + PH before 2004. We hypothesize that the increasing use of ECLS as a rescue therapy in patients with BPD + PH may be because of increased experience and comfort of providers with patients with BPD over the years, or there has been an increase in investigation and reporting of patients with BPD + PH because of the advancing awareness that PH worsens the clinical course of patients with BPD.12

We described the use of common pulmonary vasodilators before ECLS to help characterize our cohorts and found patients with BPD + PH to be more likely to have therapy with sildenafil and milrinone than patients without PH. Surprisingly, the use of nitric oxide and epoprostenol, even though more likely in the BPD + PH group, was not statistically different. Possible reasons for this observation include the limited number of patients in our BPD + PH group, the lack of reporting of PH in some patients with BPD, the empirical use of pulmonary vasodilators as rescue medications without a diagnosis of PH, or the real-time assessment and management of PH by bedside clinicians that does not necessarily reflect a prior diagnosis of PH. Prospective collection of data would help elucidate the reason for this finding, which is outside the scope of this review.

The most common diagnosis leading to ECLS was viral respiratory infection, which is in concordance with other studies that reported ex-premature infants with BPD.9,13 In the group with BPD + PH, the most common primary diagnosis was respiratory failure with no mention of infection, followed by secondary PH. Given the retrospective nature of this review, it is difficult to elucidate whether the reason for ECLS in patients with BPD + PH was a pulmonary hypertensive crisis, or a respiratory infection exacerbating their baseline disease.

Within the complications reported in the ELSO International Report10 for pediatric respiratory ECLS, inotrope use while on ECLS was similar in occurrence between our groups (ELSO 27.6%, BPD 35.9%, BPD + PH 33.3%) and associated with reasonable survival (ELSO 52%, BPD 52.2%, BPD + PH 80.0%).

To our knowledge, this is the only report looking at outcomes of patients with BPD managed on ECLS to include older patients outside the neonatal and early infancy period as well as to describe outcomes in patients with associated secondary PH. As former extremely premature patients now approach adulthood, the long-term manifestations of their prematurity and related BPD will require additional evaluation and management strategies. Although the median age of our patients was 7 months for BPD and 8 months for BPD + PH, our study also included school-aged and adolescent patients. A previous study by Hibbs et al.9 reported small numbers from a single center of ex-premature infants with BPD managed on ELSO for noncardiac reasons before 1 year of uncorrected age with a survival of 78%. An additional retrospective study by Brown et al.13 reporting data from the United Kingdom of ex-premature infants with BPD published survival rates of 80%. Our study supports that older patients beyond infancy diagnosed with PH and BPD before the need of ECLS may still have positive outcomes of survival.

Interpretation of our retrospective data is limited by the subjective and variable nature of the ELSO registry that relies on coding. Other noted limitations include the small number of patients with BPD + PH reported to the ELSO database, the absence of documentation on gestational age at birth, documentation on severity of BPD, baseline oxygen requirement, and further characterization of PH.


ECLS support for pediatric patients with a history of BPD and with BPD with secondary PH should be considered as a bridging therapy to lung recovery given encouraging survival outcomes. A prospective multicenter study would help describe better long-term outcomes of pediatric patients with a history of BPD when managed on ECLS.


The authors would like to acknowledge and thank Dr. Curtis Froehlich for his collaboration and support during the development of this project.


1. Donda K, Vijayakanthi N, Dapaah-Siakwan F, Bhatt P, Rastogi D, Rastogi S. Trends in epidemiology and outcomes of respiratory distress syndrome in the United States. Pediatr Pulmonol 2019.54: 405–414.
2. Stoll BJ, Hansen NI, Bell EF, et al. Trends in care practices, morbidity, and mortality of extremely preterm neonates, 1993–2012. JAMA. 2015314:1039–1051.
3. Filippone M, Sartor M, Zacchello F, Baraldi E. Flow limitation in infants with bronchopulmonary dysplasia and respiratory function at school age. Lancet 2003.361: 753–754.
4. O’Reilly M, Sozo F, Harding R. Impact of preterm birth and bronchopulmonary dysplasia on the developing lung: long-term consequences for respiratory health. Clin Exp Pharmacol Physiol 2013.40: 765–773.
5. Doyle LW, Faber B, Callanan C, Freezer N, Ford GW, Davis NM. Bronchopulmonary dysplasia in very low birth weight subjects and lung function in late adolescence. Pediatrics 2006.118: 108–113.
6. Lagatta JM, Hysinger EB, Zaniletti I, et al.; Children’s Hospital Neonatal Consortium Severe BPD Focus Group: The impact of pulmonary hypertension in preterm infants with severe bronchopulmonary dysplasia through 1 year. J Pediatr 2018.203: 218–224.e3.
7. Mourani PM, Abman SH. Pulmonary hypertension and vascular abnormalities in bronchopulmonary dysplasia. Clin Perinatol 2015.42: 839–855.
8. An HS, Bae EJ, Kim GB, et al. Pulmonary hypertension in preterm infants with bronchopulmonary dysplasia. Korean Circ J 2010.40: 131–136.
9. Hibbs A, Evans JR, Gerdes M, Hunter JV, Cullen JA. Outcome of infants with bronchopulmonary dysplasia who receive extracorporeal membrane oxygenation therapy. J Pediatr Surg 2001.36: 1479–1484.
10. ECMO Registry of the Extracorporeal Life Support Organization (ELSO). January, 2019. Ann Arbor, Michigan, Available at:
11. Altit G, Bhombal S, Hopper RK, Tacy TA, Feinstein J. Death or resolution: the “natural history” of pulmonary hypertension in bronchopulmonary dysplasia. J Perinatol 2019.39: 415–425.
12. Berkelhamer SK, Mestan KK, Steinhorn R. An update on the diagnosis and management of bronchopulmonary dysplasia (BPD)-associated pulmonary hypertension. Semin Perinatol 2018.42: 432–443.
13. Brown KL, Walker G, Grant DJ, et al. Predicting outcome in ex-premature infants supported with extracorporeal membrane oxygenation for acute hypoxic respiratory failure. Arch Dis Child Fetal Neonatal Ed 2004.89: F423–F427.

extracorporeal life support; pediatric; extracorporeal membrane oxygenation; bronchopulmonary dysplasia; pulmonary hypertension

Supplemental Digital Content

Copyright © 2020 by the ASAIO