Does Erythropoietin monotherapy reduce mortality or moderate/severe disability in neonates with hypoxic ischemic encephalopathy?

December 30, 2017


Malla RR, Asimi R, Teli MA, Shaheen F, Bhat MA. Erythropoietin monotherapy in perinatal asphyxia with moderate to severe encephalopathy: a randomized placebo-controlled trial. Journal of Perinatology (2017): 37, 596-601. PMID:28277490


Joseph S Bircher, MD
Pediatrics Resident
Oregon Health & Science University

Dmitry Dukhovny, MD MPH
Associate Professor of Pediatrics
Oregon Health & Science University




In neonates with moderate or severe hypoxic ischemic encephalopathy (HIE) does erythropoietin monotherapy compared to placebo decrease death or moderate or severe disability at 19 months of age?


  • Design: Randomized Clinical Trial
  • Allocation: Concealed
  • Blinding: Triple blinded (investigators, caregivers, and families)
  • Follow-up period: Death or a mean age of 19 months
  • Setting: Single center study at the Srinagar Kashmir Institute of Medical Sciences (SKIMS ) – a neonatal intensive care unit of a tertiary care center in northern India. All infants enrolled in the study were born at SKIMS or from two local hospitals.
  • Patients: 100 term neonates; randomized December 2012 to November 2015.
  • Inclusion criteria (required all of the following):
    • ≥ 37 weeks gestational age
    • less than 6 hours of age
    • determined to have perinatal asphyxia:
      • Perinatal asphyxia was defined by a 10 minute Apgar score of <5 and two of the three following criteria: 1) History of fetal distress 2) Need for immediate neonatal ventilation for ≥ 10 minutes after delivery 3) Acidosis with base deficit ≥ 16 or pH ≤ 7.0 from cord blood or arterial blood obtained within 1 hour of life.
    • moderate or severe HIE
      • Moderate or severe HIE was diagnosed if there were abnormalities in > three1 of the following categories on physical exam: level of consciousness, spontaneous activity, tone, posture, primitive reflexes, and autonomic nervous system. Moderate and severe HIE were quantified based on the clinically determined level of dysfunction within these categories. The determination of encephalopathy was made by the Neonatologist, who had specific training for this determination.
  • Exclusion criteria (any of the following):
    • congenital abnormalities
    • chromosomal abnormalities
    • congenital infections
    • severe intrauterine growth retardation (defined as < 2 S.D. below the mean)
    • inborn errors of metabolism
  • The baseline characteristics of the two groups were similar including gestational age, birth weight, mode of delivery, mean HIE score (Thompson et al, 1997), and age at first dose.
  • Intervention: Patients were randomized to intravenous recombinant human erythropoietin (EPO) at a dose of 500 U/kg within 6 hours of birth and on alternative days for a total of 5 doses (intervention group) or to 2 ml of normal saline at same dosing regimen (placebo group). Randomization was done by permuted block algorithm with block size of two to four. The patients did not receive adjunct therapeutic hypothermia.
  • Primary Outcome: Death or moderate or severe disability at a mean age of 19 months. Level of disability was defined as moderate or severe based on neuromotor assessment using the Gross Motor Function Classification system (GMFCS), developmental assessment using the Bayley Scale of Infant Development II Mental Developmental Index (MDI) Score (mean score 100 ±15), vision, and hearing evaluation. Moderate disability was defined as an MDI score between 70-84 and any of the following: GMFCS grade 2 hearing impairment without need for amplification or persistent seizure disorder. Severe disability was defined by an MDI score < 70 or GMFCS grade 3-5, hearing impairment requiring hearing aid, and/or bilateral cortical visual impairment with no useful vision.
  • Secondary Outcomes: Disability, MDI, cerebral palsy, cortical visual impairment, hearing loss, and seizures requiring treatment. In addition, monitoring for adverse events included typical morbidities (e.g. intracranial hemorrhage, pulmonary hemorrhage, etc) and duration of hospitalization.
  • Analysis and Sample Size: A sample size of 50 infants in each group was determined using a two-tailed type 1 error rate of 0.05%, power of 80%, an assumed loss to follow up of 10%, and a presumed incidence of death or disability in the control group of 70% with 50% reduction in the intervention group. The analyses were unadjusted for co-variates. Although it is not explicitly stated in the manuscript, it does not appear that there was cross over between the two study groups and an intention to treat protocol was followed.
  • Patient Follow-up: 100% follow up.


*EBNEO Write Author (SB & DD) Disclaimer*: Please note that the results below are described directly out of the publication by Malla et al. Journal of Perinatlogy (2017): 37, 596-601. There were some inconsistencies in the reporting of the some of the statistical results, including a 95% confidence interval that overlaps 1, yet a p-value of <0.05. This has been addressed by letter to the editor (by Bircher et al) and statistical response (by Bamat et al), as well as a response from the authors – all of which are currently in press in the Journal of Perinatology. The results below for the confidence interval and p-values where this applies are bolded and were interpreted by the authors of this write up as not-statistically significant if upper bound of the 95% confidence interval is than 1. Additionally, there is likely one typo on the upper confidence limit noted below for abnormalities on EEG.

For the primary outcome, the relative risk of death or moderate/severe disability at a mean 19 months of age was 0.57 (95% CI: 0.38-0.85, p=0.003) in the EPO group (40%) compared to placebo (70%).

There were also several secondary outcomes in the EPO group compared to placebo that were statistically improved, including (1) death or disability in neonates with moderate encephalopathy (21% vs. 61%; RR 0.33; 95% CI, 0.12-0.79, P=0.004); (2) neurodevelopmental abnormalities at follow up (29% vs 64%; RR 0.44; 95% CI, 0.25-0.76, P = 0.001); and (3) survival without neurologic abnormalities (71% vs 35%; RR 0.65; 95% CI, 0.45-0.94, P= 0.01).

There was no statistically significant difference between the EPO and placebo groups in several of the other secondary outcomes, including: cerebral palsy (23% vs. 45% RR 0.52; 95% CI 0.25-1.03, P = 0.04), hearing impairment (4.7% vs. 4.7% RR 1; 95% CI 0.33-2.9, P = 1),  seizures requiring anticonvulsants at the time of final study evaluation (19% vs 43% RR 0.47; 95% CI 0.20-1.01, P = 0.03), Bayley MDI and Psychomotor Development Index Scores (although the mean score for either component is not presented).  Finally, there was no significant difference in death or disability among neonates with severe encephalopathy between the treatment and placebo group (57% vs 79% RR 0.72; 95% CI 0.51-1.3, P = 0.10). The authors do not report on visual impairment component separately (it is a component of the primary outcome and a pre-specified secondary outcome).

Selected adverse outcomes were followed during these infant’s hospitalizations. There was no significant differences noted in oliguria, renal failure, hypotension requiring vasopressors, necrotizing enterocolitis, and disseminated intravascular coagulation between the EPO and placebo groups. Additionally, death in the neonatal period was the same between the EPO (12%) and placebo (12%) groups. However, the mean duration of hospital stay was significantly shorter in the EPO compared with the placebo group (9.7±6.9 days vs 13.5±8.1, P = 0.04) and there were fewer brain MRI abnormalities detected between day of life 10-14 in the treatment group (40% vs 60%; RR 0.66; 95% CI, 0.42 -1.03, P = 0.04) compared with the placebo group. At time of discharge there were fewer patients in the EPO group that had abnormalities on EEG (34% vs 66%; RR 0.81; CI, 0.51-0.32 p = 0.001) and fewer that had seizures requiring treatment (26% vs 57%; RR 0.46; CI, 0.25-0.80 p = 0.002) compared with the placebo group.

The only side-effect of EPO detected during this study was a mild elevation in hemoglobin levels (17.4±1.9 vs 15.7±2.2) and reticulocyte count (4.1±0.5 vs 3.2±0.5) compared with the placebo group on day 10 of life.


The use of EPO compared to placebo as monotherapy (i.e. without adjunct therapeutic hypothermia) for moderate or severe HIE within six hours of birth clinically and statistically reduces the risk of death or moderate/severe disability in infants greater than or equal to 37 weeks.

1 Some of the previous trials for management of moderate to severe HIE have included greater than or equal to three (instead of greater than 3) as the number of physical exam findings necessary to meet the definition whereas others did not specify. The definition by the authors could include babies that are more affected as compared to previous trials.


Erythropoietin (EPO) works by a number of mechanisms that gives it neuroprotective properties including decreasing neuronal apoptosis in the acute phase and enhancing angiogenesis and neurogenesis for long term brain healing (1). Malla et al (2) demonstrate improved neurologic outcomes in patients treated with EPO monotherapy vs placebo when treating moderate or severe hypoxic ischemic encephalopathy (HIE). A study done in China by Zhu et al (3) comparing EPO vs placebo, also without adjunctive therapeutic hypothermia (TH), showed similar results. Importantly, these studies showed a significant improvement in neurologic outcome in patients with moderate but not severe encephalopathy. There was also no difference in mortality between the placebo and EPO groups. This differs from trials with TH, the current standard of care in most resource rich areas, which show a similar reduction in combined mortality and morbidity, but also demonstrate a reduction in mortality (RR – 0.75 [95% CI 0.64, 0.88] (Table 1.2, reference (4)) and disability (RR – 0.77 [95% CI 0.63, 0.94] (Table 1.3, reference (4)), when stratified by moderate or severe encephalopathy (data not shown) (4). These differences could be attributed to the larger sample size of patients in the Cochrane review (4) compared to the single center study (2) and is certainly reflected in the forest plots when looking at individual trials (4).

In settings where TH is widely available, there is significant interest in the potential benefits of EPO as an adjunct therapy to TH. A phase I trial by Wu et al. showed that a presumably effective dosing regimen of EPO could be safely given to neonates undergoing TH for HIE (5). In a phase II trial, Wu et al. showed that infants treated with EPO as an adjunct to TH had less MRI brain injury and improved motor outcomes at 1 year of age (6). Currently, there are three phase III clinical trials underway (HEAL, PAEAN, and Neurepo) comparing mortality and neurodevelopmental disability in term infants with HIE treated with EPO + TH vs TH alone (7). These may reaffirm the safety of erythropoietin as adjunct therapy to TH and substantiate evidence that it improves neurologic outcome.

The number needed to treat (NNT) to prevent the composite measure of death or moderate or severe disability in this EPO monotherapy study was four (2), which is lower than most trials using TH alone, with an NNT of 7-9 (1), however the number of patients studied to date is far fewer. There has never been a head to head trial comparing EPO monotherapy vs TH. Given the findings of Malla and colleagues (2), is it an investigation worth considering given the resource intensity associated with TH compared to EPO monotherapy? Perhaps more importantly, would this ever be ethically feasible given the lack of clinical equipoise for the proven effectiveness of TH – although TH as compared to placebo has not been studied in the developing world either.  Therefore, if the addition of EPO as an adjunct to TH in the current phase three trials fails to show a benefit, there may be an opportunity to ask this question again. The more challenging issue will be if adjunct EPO therapy does show incremental benefit compared to placebo, does that close the door on a head to head EPO vs. TH? Use of simulation models may help to tease out some of the incremental cost-benefit challenges and potentially lay ground work for a head to head trial. Careful consideration, however, must be given to the benefits of TH beyond neuroprotection that EPO may not offer (i.e. cardiovascular, gastrointestinal system, etc.).


  1. Pet GJ, S.E. The potential of erythropoietin to treat asphyxia in newborns. Research and Reports in Neonatology. 2014;4:195-207.
  2. Malla RR, Asimi R, Teli MA, Shaheen F, Bhat MA. Erythropoietin monotherapy in perinatal asphyxia with moderate to severe encephalopathy: a randomized placebo-controlled trial. J Perinatol. 2017;37(5):596-601.
  3. Zhu C, Kang W, Xu F, Cheng X, Zhang Z, Jia L, et al. Erythropoietin improved neurologic outcomes in newborns with hypoxic-ischemic encephalopathy. Pediatrics. 2009;124(2):e218-26.
  4. Jacobs SE, Berg M, Hunt R, Tarnow-Mordi WO, Inder TE, Davis PG. Cooling for newborns with hypoxic ischaemic encephalopathy. Cochrane Database Syst Rev. 2013(1):CD003311.
  5. Wu YW, Bauer LA, Ballard RA, Ferriero DM, Glidden DV, Mayock DE, et al. Erythropoietin for neuroprotection in neonatal encephalopathy: safety and pharmacokinetics. Pediatrics. 2012;130(4):683-91.
  6. Wu YW, Mathur AM, Chang T, McKinstry RC, Mulkey SB, Mayock DE, et al. High-Dose Erythropoietin and Hypothermia for Hypoxic-Ischemic Encephalopathy: A Phase II Trial. Pediatrics. 2016;137(6).
  7. ClinicalTrials.Gov. Clinical Trials: NCT02811263, NCT03079167, NCT01732146 2017 [Date Accessed 07-15-2017]. Available from: