REVIEW ARTICLEPEDIATRICS Volume 139, number 5, May 2017:e20164317

Prophylactic Early Erythropoietin for Neuroprotection in Preterm Infants: A Meta-analysisHendrik S. Fischer, MD, Nora J. Reibel, MSc, Christoph Bührer, MD, Christof Dame, MD

abstractCONTEXT: Recombinant human erythropoietin (rhEPO) is a promising pharmacological agent for neuroprotection in neonates.OBJECTIVE: To investigate whether prophylactic rhEPO administration in very preterm infants improves neurodevelopmental outcomes in a meta-analysis of randomized controlled trials (RCTs).DATA SOURCES: Medline, Embase, and the Cochrane Central Register of Controlled Trials were searched in December 2016 and complemented by other sources.STUDY SELECTION: RCTs investigating the use of rhEPO in preterm infants versus a control group were selected if they were published in a peer-reviewed journal and reported neurodevelopmental outcomes at 18 to 24 months’ corrected age.DATA EXTRACTION: Data extraction and analysis followed the standard methods of the Cochrane Neonatal Review Group. The primary outcome was the number of infants with a Mental Developmental Index (MDI) <70 on the Bayley Scales of Infant Development. Secondary outcomes included a Psychomotor Development Index <70, cerebral palsy, visual impairment, and hearing impairment.RESULTS: Four RCTs, comprising 1133 infants, were included in the meta-analysis. Prophylactic rhEPO administration reduced the incidence of children with an MDI <70, with an odds ratio (95% confidence interval) of 0.51 (0.31–0.81), P < .005. The number needed to treat was 14. There was no statistically significant effect on any secondary outcome.CONCLUSIONS: Prophylactic rhEPO improved the cognitive development of very preterm infants, as assessed by the MDI at a corrected age of 18 to 24 months, without affecting other neurodevelopmental outcomes. Current and future RCTs should investigate optimal dosing and timing of prophylactic rhEPO and plan for long-term neurodevelopmental follow-up.

Department of Neonatology, Charité University Medical Center, Berlin, Germany

Dr Fischer designed the meta-analysis, searched for relevant studies, extracted, assessed, and analyzed the data, and drafted the manuscript; Ms Reibel and Dr Bührer contributed to data extraction and assessment; Dr Dame conceptualized the meta-analysis, searched for relevant studies, extracted and assessed the data, and drafted the manuscript; and all authors approved the final manuscript as submitted.

DOI: 10.1542/peds.2016-4317

Accepted for publication Feb 9, 2017

Address correspondence to Christof Dame, MD, Department of Neonatology, Charité - Universitätsmedizin Berlin, Augustenburger Platz 1, 13353 Berlin, Germany. E-mail: christof.dame@charite.de

To cite: Fischer HS, Reibel NJ, Bührer C, et al. Prophylactic Early Erythropoietin for Neuroprotection in Preterm Infants: A Meta-analysis. Pediatrics. 2017;139(5):e20164317

by guest on March 12, 2020www.aappublications.org/newsDownloaded from

FISCHER et al

Improving neurodevelopmental outcomes is a major goal in neonatology, especially with regard to the increasing survival rate of the most immature infants.1 – 3 In particular, extremely low birth weight (ELBW) infants may develop cognitive delay or cerebral palsy later in life, even if the cranial ultrasound examination at discharge is normal.4 Recombinant human erythropoietin (rhEPO) is 1 of the most promising pharmacological substances for neuroprotection in newborn infants and may be beneficial for regeneration and repair after neonatal brain injury.5, 6

Interestingly, the endogenous erythropoietin receptor system activates in response to hypoxic–ischemic injury. This means that rhEPO may be particularly potent for neuroprotection and neuronal regeneration if administered timely after an acute brain injury.7 – 9 Conversely, if rhEPO is administered prophylactically, higher doses may be needed to cross the blood–brain barrier.10, 11 Numerous in vitro and in vivo experiments (including models of the fetal or immature brain) have shown that rhEPO has antioxidant, antiexcitotoxic, and anti-inflammatory properties, prevents neuronal apoptosis, 12 and promotes neurogenesis and angiogenesis.7, 13 –15

Debate continues over whether premature infants at high risk of neurodevelopmental impairment may benefit from prophylactic rhEPO.16 Most clinical follow-up studies on preterm infants primarily treated with rhEPO (or its higher-glycosylated derivate darbepoetin) for anemia of prematurity indicated improved neurodevelopmental outcomes, 17 – 21 whereas only a few studies have shown no effect.22, 23 In 2005, the first phase I and II trials were initiated to establish the safety of early high-dose rhEPO for neuroprotection.24, 25 So far, only a limited number of randomized controlled trials (RCTs) have

reported data on the neuroprotective effects of rhEPO in preterm infants. This limitation was reflected in 2 previous meta-analyses26, 27 that included several quasi-RCTs25, 28– 30 and a cohort study.19 Recently, neurodevelopmental outcomes at a corrected age of 18 to 24 months from 2 prospective RCTs, intended primarily to evaluate the neuroprotective effects of rhEPO in a total of 978 very preterm infants, were published almost simultaneously.31, 32 Therefore, the present meta-analysis aimed to study whether prophylactic rhEPO improves the neurodevelopmental outcomes of very preterm infants at 18 to 24 months’ corrected age, solely including data from previous and recent RCTs.

MeThODs

Criteria for Considering studies for This Review

All RCTs that investigated the effect of prophylactic rhEPO in preterm infants were eligible for the meta-analysis. Included trials had to compare an rhEPO-treated group with a control group that received no treatment or placebo, and they had to report neurodevelopmental outcomes. Only studies published in full in peer-reviewed journals were accepted. No language restrictions were applied.

Types of Outcome Measures

Outcome measures were specified beforehand in a review protocol, considering the outcomes reported by recently published RCTs.31, 32 The primary outcome of the meta-analysis was the number of infants with a Mental Developmental Index (MDI) <70 on the Bayley Scales of Infant Development, Second Edition (BSID-II), at 18 to 24 months’ corrected age.33 If infants were assessed according to the Third Edition (BSID-III), 34 we used cognitive scores <85 as the

primary outcome, because a recent study showed that a cognitive score <85 (BSID-III) predicts an MDI <70 (BSID-II), with an overall agreement of 97.3%.35 A planned subgroup analysis assessed the incidence of an MDI <70 (BSID-II) or cognitive score <85 (BSID-III) in infants <28 weeks’ gestational age. Other predetermined secondary outcomes were the number of infants with a Psychomotor Development Index (PDI) <70 (BSID-II), cerebral palsy, severe visual impairment, and severe hearing impairment, respectively. After the literature search, we added the secondary outcome “any neurodevelopmental impairment at 18 to 24 months’ corrected age.”

search Methods for Identification of studies

For the literature search, we followed the standard search methods of the Cochrane Collaboration.36 Two authors independently searched the databases Medline, Embase, and the Cochrane Central Register of Controlled Trials (CENTRAL). The results were combined with cross-referencing of previous reviews and trials and with the use of expert information. Additional information about ongoing trials was sought through the search portal of the International Clinical Trials Registry Platform.37 Medline, Embase, and CENTRAL were searched on May 31, 2016. The search was updated on December 18, 2016. A highly sensitive search strategy was applied, as outlined in the Cochrane Handbook for Systematic Reviews of Interventions, and complied with specific recommendations for identifying RCTs in Medline and Embase.36, 38 A variety of search terms were applied to identify RCTs concerned with rhEPO use in premature infants (search concept: premature infant AND erythropoietin AND randomized controlled trial) or with follow-up studies on children of all ages that reported neurodevelopmental outcomes

2 by guest on March 12, 2020www.aappublications.org/newsDownloaded from

PEDIATRICS Volume 139, number 5, May 2017

after rhEPO treatment (search concept: child AND erythropoietin AND [RCT OR follow-up study] AND [neurodevelopmental testing OR neurodevelopmental outcomes OR neuroprotection]). Altogether, we used >60 free-text and indexed search terms, including many from the Medical Subject Headings (MeSH) and Excerpta Medica Tree indexing systems. The detailed search strategy is available in the Supplemental Information. We followed the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) Statement.39, 40

Assessment of Methodological Quality

The methodological quality of the selected studies was assessed according to Cochrane’s guidelines.36 The risks of selection bias (quality of randomization, allocation concealment), performance bias (blinding of intervention), detection bias (blinding of outcome assessment), attrition bias (completeness of follow-up), selective reporting (reporting bias), and other potential biases were independently examined by 2 study authors (H.S.F., N.J.R.) and categorized as low risk, high risk, or unclear risk. Any discrepancies were resolved by a third author (C.D.).

Data Collection and Analysis

Trial searching, study selection, and data extraction were performed independently by 2 study authors (H.S.F., C.D.). At each stage, differences were discussed and resolved by consensus or by a third author (C.B.). We planned to contact the corresponding authors of the included RCTs up to 3 times via e-mail to request stratified data or additional information about their trials, allowing 2 months for their response. RevMan version 5.3 (Cochrane Collaboration, Nordic Cochrane Centre, Copenhagen) was used for data analysis. The impacts of rhEPO on primary and

secondary outcomes were calculated via the Mantel–Haenszel method for dichotomous outcomes in a random effects model. Results for all studies and for the pooled data were reported as odds ratios (ORs) and 95% confidence intervals (CIs). Statistical significance was defined as P < .05. For pooled outcomes with significant effects, a number needed to treat (NNT) was calculated. To estimate heterogeneity, we applied the χ2 test and I2 value. To assess for publication bias, the treatment effects of rhEPO were plotted against SE (log[OR]) in funnel plots.36, 41

ResulTs

A total of 576 records were identified, 564 through database searches and 12 through other sources (Fig 1). From them, 114 potentially met the inclusion criteria and were assessed as full-text articles. A subset of 21 articles were published in languages other than English: 4 in Bulgarian, 1 in Chinese, 2 in French, 3 in Italian, 2 in Japanese, 2 in Polish, 6 in Spanish, and 1 in Turkish.

excluded studies

After the full-text assessment, 110 articles had to be excluded. As shown in Fig 1, 19 articles did not report RCTs, 2 reported RCTs that compared different schemes of rhEPO administration but lacked a control group, 42, 43 and 86 RCTs had rhEPO and control groups but did not report neurodevelopmental outcomes. Three studies were excluded for other reasons: Newton et al22 presented neurodevelopmental outcomes of 40 infants enrolled in 2 small pilot RCTs (n = 28) and 1 multicenter trial (n = 22) conducted between 1988 and 1993.44 – 46 However, these studies were excluded because the follow-up article presented only pooled follow-up data and did not allow any outcome analysis at the level of the original RCTs. Bierer et al18

correlated neonatal peak serum erythropoietin concentrations with neurodevelopmental outcomes at 18 to 22 months’ corrected age in a single-center follow-up study on an RCT of rhEPO versus placebo, but the study was excluded because the population was a subset of the Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network Trial included in our meta-analysis.17 A third RCT (NCT01207778), recently published by Ohls et al, 21 was excluded because it reported neurodevelopmental outcomes at 3.5 to 4 years.

Included studies

Four RCTs, which assessed a total of 1133 patients at 18 to 24 months' corrected age, were included in the meta-analysis; their main characteristics are summarized in Table 1. All studies included preterm infants of ≤32 0/7 weeks’ gestational age, but only 2 used the infants’ birth weight to define the inclusion criteria.17, 20 Notably, the timing, dosing, and duration of rhEPO administration varied. Whereas 1 RCT applied early high-dose rhEPO only (first dose within 3 hours after birth, cumulative dose of 9000 IU/kg within the first 42 hours), 32 the other 3 RCTs started rhEPO later (within the first 48–96 hours), applied lower doses (1200–1750 IU/kg per week), and prescribed sustained treatment of several weeks.17, 20, 31 One RCT was designed primarily to prevent red blood cell transfusions by reducing the anemia of prematurity.17 Another RCT targeted both hematologic and neurodevelopmental outcomes, comparing rhEPO, placebo, and a third intervention group receiving darbepoetin.20 For methodological reasons, however, the latter was not included in the meta-analysis.

To obtain additional data or information, we contacted the first or senior authors of all included RCTs. We highly appreciate the cooperation

3 by guest on March 12, 2020www.aappublications.org/newsDownloaded from

FISCHER et al

of 1 author who provided us with previously unpublished data for infants <28 0/7 weeks’ gestational age.32

Risk of Bias in the Included studies

Selection Bias

Random sequence generation and allocation concealment were adequate in all studies (low risk).

Performance Bias

In 3 RCTs, considerable efforts were made to blind parents and personnel to the study intervention (low risk).17, 20, 32 By contrast, doctors and nurses responsible for the treatment were not blinded in the trial by Song et al31 (high risk).

Detection Bias

Investigators performing long-term outcome assessments were blinded to the patients’ group allocation in all studies (low risk).

Attrition Bias

The proportion of the surviving study patients who completed follow-up varied from 71% to 91%, according to the particular study and group, and the maximum difference in follow-up between intervention and control group was 91% versus 80% in 1 study20 (unclear risk).

Reporting Bias

Two RCTs were registered at ClinicalTrials.gov before or shortly after recruitment began (low risk), 20, 32 and 1 RCT was not registered (unclear risk).17 The study by Song et al31 was registered 6 months after enrollment was completed, and the registered inclusion criteria and primary outcomes differed from those reported in the publication (high risk).

Other Bias

Two RCTs were co-funded by pharmaceutical companies (low risk).17, 32

Visual inspection of the funnel plots did not reveal significant

asymmetries in the distribution of effect sizes around the mean, making publication bias less likely. The funnel plots, detailed characteristics of the included studies, and the complete risk of bias assessment are available as Supplemental Information (Supplemental Tables, Supplemental Fig 4).

effects of Intervention

The effects of rhEPO on MDI <70 and PDI <70 are shown in Fig 2.

Prophylactic rhEPO in very preterm infants significantly reduced the incidence of an MDI <70 at 18 to 24 months’ corrected age, with an OR (95% CI) of 0.51 (0.31–0.81), P = .005. Statistical measures did not indicate a heterogeneity problem (χ2 = 4.01, P = .26, I2 = 25%). The absolute risk of an MDI <70 was reduced from 15.7% to 8.4%, corresponding to an NNT of 14. In a sensitivity analysis excluding the data of the largest trial by Song et al31 the

4

FIGuRe 1PRISMA flow diagram.39

by guest on March 12, 2020www.aappublications.org/newsDownloaded from

PEDIATRICS Volume 139, number 5, May 2017

calculated effect size of rhEPO on MDI <70 remained in favor of rhEPO but was no longer statistically significant, with an OR (95% CI) of 0.68 (0.4–1.18), P = .17 (data not shown). RhEPO had no significant impact on MDI <70 in infants <28 weeks’ gestational age (Fig 2B) or on PDI <70 (Fig 2C). The definitions of the other secondary outcomes (cerebral palsy, severe visual impairment, severe hearing impairment, and any neurodevelopmental impairment) all showed some variation between the included studies (see Supplemental Tables for details). As shown in Fig 3, there was no significant effect of rhEPO on any of these outcomes in the meta-analysis. The combined OR (95% CI) of any neurodevelopmental impairment was 0.55 (0.28–1.08), P = .08, with substantial statistical heterogeneity between the included studies (χ2 = 10.38, P = .02, I2 = 71%).

DIsCussIOn

This is the largest meta-analysis to date to investigate the neuroprotective effects of prophylactic rhEPO in very preterm infants of 18 to 24 months’ corrected age and the first to include only RCTs. Our analysis is highly topical, because 2 of the 4 RCTs, which included 978 of the 1133 patients studied, had only been reported recently.31, 32 The meta-analyzed data showed that rhEPO reduced the number of patients with an MDI <70 at 18 to 24 months’ corrected age but had no significant effect on motor development, hearing, or vision.

Quality of evidence

The included RCTs were mostly of high methodological quality, with a low risk of bias (Supplemental Tables). The only major problems related to the RCT by Song et al31 were the lack of blinding of personnel and indications of selective reporting. Incomplete outcome data were considered as a minor point in all

studies. However, the reported follow-up rates at 18 to 24 months’ corrected age were satisfactory from a practical point of view. In accordance with Cochrane methods, we used funnel plots to assess for publication bias, 36 but we did not identify major asymmetries (Supplemental Fig 4). However, we were aware that this method had low power to detect bias, because only 4 RCTs were included in the meta-analysis. More importantly, there remains a possibility of unpublished data, in that 86 RCTs of rhEPO were excluded from the meta-analysis because they did not report neurodevelopmental outcomes (Fig 1).

effect on Cognitive Outcome

The beneficial effect of prophylactic rhEPO on the incidence of MDI <70 at 18 to 24 months’ corrected age was consistent across all included RCTs (Fig 2A), although timing and dosing of rhEPO administration varied between trials. With an absolute risk reduction from 15.7% to 8.4% and an NNT of 14 patients, the estimated effect is clinically relevant. The observed benefits of rhEPO are robust,

because they are in agreement with available evidence from longer-term follow-up studies, which also reported improved cognitive outcomes. In the second follow-up on the previously mentioned RCT of rhEPO, darbepoetin, or placebo, Ohls et al21, 47 assessed 53 children via the Wechsler Preschool and Primary Scale of Intelligence, Third Edition at the age of 3.5 to 4 years. The study compared the pooled data of the rhEPO group (n = 24) and darbepoetin group (n = 15) with the placebo group (n = 14) and showed that the children treated by erythropoiesis-stimulating agents scored significantly better for full-scale IQ, performance IQ, overall executive function, and working memory. When cognitive function was analyzed, results at 18 to 22 months and 4 years correlated significantly between BSID-II composite score and full-scale IQ.20, 21 Neubauer et al19 analyzed neurodevelopmental outcomes at the age of 10 to 13 years in a cohort study of 89 ELBW infants who received rhEPO from the first week of life to near-term and 57 untreated ELBW infants. rhEPO-treated infants attained higher verbal, nonverbal, and composite

5

TABle 1 Characteristics of Included Studies

Author Year N Gestational Age, Birth Wt

Time Point of Intervention

Intervention Recruitment

Ohls et al17 2004 102 ≤32 0/7, ≤1000 g

24–96 h of age rhEPO 400 IU/kg IV or SC, 3 times per wk until 35 0/7 wk gestation

1997–1998

Ohls et al20 2014 53a Any GA, b 500–1250 g

≤48 h of age rhEPO 400 IU/kg SC, 3 times per wk until 35 0/7 wk gestation

2006–2010

Natalucci et al32

2016 365 26 0/7–31 6/7, any BW

<3 h of age rhEPO 3000 IU/kg IV at <3 h, 12–18 h and 36–42 h of age

2005–2012

Song et al31 2016 613 ≤32 0/7, any BW

<72 h of age rhEPO 500 IU/kg IV every other day for 2 wk

2009–2013

BW, birth wt; GA, gestational age; IV, intravenously; SC, subcutaneously.a This study had 3 groups: rhEPO (n = 29) versus placebo (n = 24) versus darbepoetin (n = 27). The darbepoetin group was not included in the meta-analysis.b Median (interquartile range) GA of included infants 28 (26–29) wk; study entry criteria: preterm infants with a birth wt of 500–1250 g.

by guest on March 12, 2020www.aappublications.org/newsDownloaded from

FISCHER et al

IQ scores in the Hamburg–Wechsler Intelligence Test for Children III, and a subgroup analysis showed a particular benefit of rhEPO in infants with intraventricular hemorrhage.

Because infants <28 0/7 weeks’ gestational age are at particularly high risk of poor developmental outcomes, the effects of rhEPO in this subpopulation deserve closer examination. In our planned subgroup analysis, there was no

significant benefit of rhEPO on the outcome measure MDI <70 in such infants (Fig 2B). However, only stratified data from the recent Swiss and Chinese rhEPO neuroprotection RCTs were available for the subgroup analysis, and the resulting subset of 60 rhEPO-treated and 57 placebo-treated infants <28 0/7 weeks’ gestational age was less than the optimal information size. We are aware that a multicenter RCT in the United States is currently

recruiting 941 preterm infants <28 0/7 weeks’ gestational age to answer this question (PENUT Trial, NCT01378273). The PENUT Trial enrolls infants within 24 hours after birth to investigate an early high-dose strategy of 6 doses of intravenous Epo 1000 U/kg per dose at 48-hour intervals, followed by subcutaneous Epo 400 U/kg per dose 3 times per week up to 32 6/7 weeks’ postmenstrual age. Results are expected in 2019 and will be

6

FIGuRe 2Effects of rhEPO on neurodevelopment, as assessed by BSID-II or BSID-III at 18 to 24 months’ corrected age. Forest plots show the effects on the number of infants with an MDI <70 (BSID-II) or cognitive score <85 (BSID-III) in all infants (primary outcome, A), in infants <28 0/7 weeks’ gestational age (planned subgroup analysis, B), and on the number of infants with a PDI <70 as assessed by BSID-II (secondary outcome, C). M-H, Mantel–Haenszel.

by guest on March 12, 2020www.aappublications.org/newsDownloaded from

PEDIATRICS Volume 139, number 5, May 2017 7

FIGuRe 3Effects of rhEPO on secondary outcomes at 18 to 24 months’ corrected age. Forest plots show the effects on cerebral palsy (A), severe visual impairment (B), severe hearing impairment (C), and any neurodevelopmental impairment (D). M-H, Mantel–Haenszel.

by guest on March 12, 2020www.aappublications.org/newsDownloaded from

FISCHER et al

stratified by gestational age at birth (<26 or 26–27 6/7 weeks).48

effect on Motor Outcomes, Vision, and hearing

In the meta-analysis, prophylactic rhEPO had no significant effect on the incidence of PDI <70, cerebral palsy, severe visual impairment, or severe hearing impairment at 18 to 24 months’ corrected age (Figs 2C and 3 A–C). Admittedly, these adverse outcomes are rare, and the meta-analysis may not be sufficiently powered to detect minor effects. The latter also applies to longer-term follow-up studies that did not detect prophylactic rhEPO to have any impact on gross and fine motor activity, cerebral palsy, vision, or hearing.19, 21 Reassuringly, the lack of effect on visual outcome in the meta-analysis is consistent with previous meta-analyses showing that early rhEPO did not increase the risk of retinopathy of prematurity (ROP)49, 50 and with experimental data suggesting that rhEPO may even protect the neuronal cells of the retina.51 Most recently, a meta-analysis by Fang et al52 evaluated the influence of rhEPO (independent of early or late treatment initiation) on ROP and found no impact on the rates of any stage ROP, severe ROP, or ROP necessitating intervention.

effect on Any neurodevelopmental Impairment

Despite the robust beneficial effect of rhEPO on cognitive development, this meta-analysis showed only a trend toward lowering the combined outcome of any neurodevelopmental impairment (OR 0.55; 95% CI, 0.28–1.08; P = .08; Fig 3D). One possible explanation is that rhEPO targets only a subset of cells or certain areas of the brain responsible for cognitive development. Alternatively, the aforementioned discrepancy may result from the fact that data on the incidence of PDI <70 were not available for 2 RCTs.20, 31 In the meta-analysis, this limitation may

have biased the calculated effect of rhEPO on PDI <70 and on any neurodevelopmental impairment in favor of the placebo (Figs 2C and 3D). Notably, Song et al31 reported absolute PDIs, which were significantly higher in their rhEPO group compared with placebo (median PDI 99 vs 90; P < .001).

Future Directions

The results of the present meta-analysis clearly indicate that more clinical trials are warranted to investigate open questions about prophylactic rhEPO administration in preterm infants. Most importantly, major variations in the study protocols of the 4 RCTs included in the meta-analysis indicate that the optimal timing and dosing of rhEPO for neuroprotection in preterm infants are still unknown. It is therefore interesting that the follow-up on a previous phase I and II dose escalation trial of early high-dose rhEPO showed that rhEPO improved cognitive and motor outcome measures at 4 to 36 months’ corrected age.53 The protocol of the Swiss RCT (first dose within 3 hours after birth, cumulative dose of 9000 IU/kg within the first 42 hours)32 was the only RCT that closely followed experimental data, arguing for early high-dose rhEPO to achieve therapeutic levels in the brain within the most vulnerable early postnatal period.54, 55 Because a subset of the study patients who were examined by cranial MRI scans at term-corrected age showed significantly lower white and gray matter injury scores after rhEPO treatment, 56 it was disappointing that BSID-II scores at 18 to 24 months were not improved, neither in the subset nor in the entire study population.32 However, benefits in neurodevelopmental outcome may not be ascertainable until school age or even later.19, 57, 58 Looking at the encouraging outcome data of preterm rhEPO-treated infants in

the retrospective study of Neubauer et al19 which extend the follow-up examinations from 2 to 13 years of age, the 2-year time point was the only one that showed no effects of rhEPO. Therefore, patients from the 4 analyzed RCTs need long-term neurodevelopmental follow-up examinations at 5 to 6 and 10 to 13 years of age, and additional funding should be provided for this purpose.

The fact that in the other 3 RCTs, repeated low-dose rhEPO treatment (started between 24 and 96 hours after birth, in doses of 1200–1750 IU/kg per week, sustained for 2 weeks31 or longer17, 20) significantly improved neurologic outcomes provides an indication that continued treatment may be essential. So far, it is unclear whether this effect is caused by a direct impact of rhEPO on the premature brain or by stimulated erythropoiesis and reduced red blood cell transfusions. However, the question deserves more research, particularly considering pharmacodynamically optimized rhEPO treatment to prevent transfusions.59 To date, clinical data from meta-analyses and studies of early high-dose rhEPO indicate no increased risk for typical disorders in very preterm infants such as necrotizing enterocolitis, bronchopulmonary dysplasia, and other adverse outcomes.24, 25, 27, 32, 49, 50, 53 Two cohort studies even found a lower incidence of bronchopulmonary dysplasia in infants treated with rhEPO to prevent red blood cell transfusions.60, 61 However, more skin hemangiomas were reported in 2 other cohort studies.25, 62

Future RCTs of prophylactic rhEPO in preterm infants should investigate treatment protocols that combine an early high-dose strategy with continued treatment. The aforementioned PENUT trial in the United States (NCT01378273)48

8 by guest on March 12, 2020www.aappublications.org/newsDownloaded from

PEDIATRICS Volume 139, number 5, May 2017

and another ongoing trial in China (NCT02601872, NCT02550054) apply such strategies. Moreover, the EPO REPAIR trial in Switzerland currently investigates a similar dosing protocol for the rescue treatment of preterm infants with intraventricular hemorrhage (NCT02076373).6 However, the results of these studies will not be available before the end of 2018.

limitations

The limitations of the available evidence are discussed above. Importantly, the RCT by Song et al31 had a substantial impact on the primary outcome but involved a high risk of bias in 2 domains. In addition, there were limitations at the meta-analysis level; as a primary outcome, we accepted a cognitive score <85 (BSID-III) as equivalent to an MDI <70 (BSID-II), as recommended by Johnson et al.35 However, authors of other studies found slightly different cutoff levels.63, 64 With regard to the secondary outcomes, visual impairment, hearing impairment, and any neurodevelopmental impairment, we deliberately accepted minor deviations of the respective outcome definitions between the included studies (Supplemental Tables).

COnClusIOns

The current meta-analysis of 4 RCTs, including 574 rhEPO-treated and 547 placebo-treated premature infants, indicates that prophylactic rhEPO improves neurocognition, as assessed by the MDI at 18 to 24 months’ corrected age. By contrast, rhEPO had no significant effects on other neurodevelopmental outcomes. These findings demonstrate the promising potential of rhEPO for neuroprotection in very preterm infants. New data from current and future RCTs are needed to identify the optimal timing and dosing of prophylactic rhEPO administration in this age group. Ongoing RCTs and those already completed should be funded for long-term neurodevelopmental follow-up.

ACknOwleDGMenTs

We thank Dr Giancarlo Natalucci for providing previously unpublished data and additional information on his study.32 We appreciate the expert neuropediatric advice of Drs Elisabeth Walch and Gitta Reuner regarding the comparison of the different editions of the Bayley Scales of Infant Development. In addition, we thank Lukas Bomke, Judith Hollnagel, Christina Dame-Paasch,

and Drs Makoto Kashiwabara, Paulina Aleksander, and Banu Orak for translating articles published in Bulgarian, Chinese, Italian, Japanese, Polish, and Turkish, respectively.

ReFeRenCes

1. Patel RM, Kandefer S, Walsh MC, et al; Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network. Causes and timing of death in extremely premature infants from 2000 through 2011. N Engl J Med. 2015;372(4):331–340

2. Jeschke E, Biermann A, Günster C, et al; Routine Data-Based Quality

Improvement Panel. Mortality and major morbidity of very-low-birth weight infants in Germany 2008–2012: a report based on administrative data. Front Pediatr. 2016;4:23

3. Serenius F, Ewald U, Farooqi A, et al; Extremely Preterm Infants in Sweden Study Group. Neurodevelopmental outcomes among extremely preterm infants 6.5 years after active perinatal

care in Sweden. JAMA Pediatr. 2016;170(10):954–963

4. Laptook AR, O’Shea TM, Shankaran S, Bhaskar B; NICHD Neonatal Network. Adverse neurodevelopmental outcomes among extremely low birth weight infants with a normal head ultrasound: prevalence and antecedents. Pediatrics. 2005;115(3):673–680

9

ABBReVIATIOns

BSID-II:  Bayley Scales of Infant Development, Second Edition

BSID-III:  Bayley Scales of Infant Development, Third Edition

CENTRAL:  Cochrane Central Register of Controlled Trials

CI:  confidence intervalELBW:  extremely low birth

weightMDI:  Mental Developmental

IndexMeSH:  Medical Subject HeadingsNNT:  number needed to treatOR:  odds ratioPDI:  Psychomotor Development

IndexPRISMA:  Preferred Reporting

Items for Systematic Reviews and Meta-analyses

RCT:  randomized controlled trialrhEPO:  recombinant human

erythropoietinROP:  retinopathy of prematurity

PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275).

Copyright © 2017 by the American Academy of Pediatrics

FInAnCIAl DIsClOsuRe: The authors have indicated they have no financial relationships relevant to this article to disclose.

FunDInG: No external funding.

POTenTIAl COnFlICT OF InTeResT: The authors have indicated they have no potential conflicts of interest to disclose.

by guest on March 12, 2020www.aappublications.org/newsDownloaded from

FISCHER et al

5. Robertson NJ, Tan S, Groenendaal F, et al. Which neuroprotective agents are ready for bench to bedside translation in the newborn infant? J Pediatr. 2012;160(4):544–552.e4

6. Rüegger CM, Hagmann CF, Bührer C, Held L, Bucher HU, Wellmann S; EpoRepair Investigators. Erythropoietin for the repair of cerebral injury in very preterm infants (EpoRepair). Neonatology. 2015;108(3):198–204

7. Tsai PT, Ohab JJ, Kertesz N, et al. A critical role of erythropoietin receptor in neurogenesis and post-stroke recovery. J Neurosci. 2006;26(4):1269–1274

8. Chikuma M, Masuda S, Kobayashi T, Nagao M, Sasaki R. Tissue-specific regulation of erythropoietin production in the murine kidney, brain, and uterus. Am J Physiol Endocrinol Metab. 2000;279(6):E1242–E1248

9. Wallach I, Zhang J, Hartmann A, et al. Erythropoietin-receptor gene regulation in neuronal cells. Pediatr Res. 2009;65(6):619–624

10. Brines ML, Ghezzi P, Keenan S, et al. Erythropoietin crosses the blood–brain barrier to protect against experimental brain injury. Proc Natl Acad Sci USA. 2000;97(19):10526–10531

11. Juul SE, McPherson RJ, Farrell FX, Jolliffe L, Ness DJ, Gleason CA. Erythropoietin concentrations in cerebrospinal fluid of nonhuman primates and fetal sheep following high-dose recombinant erythropoietin. Biol Neonate. 2004;85(2):138–144

12. Alnaeeli M, Wang L, Piknova B, Rogers H, Li X, Noguchi CT. Erythropoietin in brain development and beyond.Anat Res Int. 2012;2012:953264

13. Iwai M, Cao G, Yin W, Stetler RA, Liu J, Chen J. Erythropoietin promotes neuronal replacement through revascularization and neurogenesis after neonatal hypoxia/ischemia in rats. Stroke. 2007;38(10):2795–2803

14. Jantzie LL, Miller RH, Robinson S. Erythropoietin signaling promotes oligodendrocyte development following prenatal systemic hypoxic–ischemic brain injury. Pediatr Res. 2013;74(6):658–667

15. Gonzalez FF, Larpthaveesarp A, McQuillen P, et al. Erythropoietin

increases neurogenesis and oligodendrogliosis of subventricular zone precursor cells after neonatal stroke. Stroke. 2013;44(3):753–758

16. Ohls RK, Christensen RD, Widness JA, Juul SE. Erythropoiesis stimulating agents demonstrate safety and show promise as neuroprotective agents in neonates. J Pediatr. 2015;167(1):10–12

17. Ohls RK, Ehrenkranz RA, Das A, et al; National Institute of Child Health and Human Development Neonatal Research Network. Neurodevelopmental outcome and growth at 18 to 22 months’ corrected age in extremely low birth weight infants treated with early erythropoietin and iron. Pediatrics. 2004;114(5):1287–1291

18. Bierer R, Peceny MC, Hartenberger CH, Ohls RK. Erythropoietin concentrations and neurodevelopmental outcome in preterm infants. Pediatrics. 2006;118(3). Available at: www. pediatrics. org/ cgi/ content/ full/ 118/ 3/ e635

19. Neubauer AP, Voss W, Wachtendorf M, Jungmann T. Erythropoietin improves neurodevelopmental outcome of extremely preterm infants. Ann Neurol. 2010;67(5):657–666

20. Ohls RK, Kamath-Rayne BD, Christensen RD, et al. Cognitive outcomes of preterm infants randomized to darbepoetin, erythropoietin, or placebo. Pediatrics. 2014;133(6):1023–1030

21. Ohls RK, Cannon DC, Phillips J, et al. Preschool assessment of preterm infants treated with darbepoetin and erythropoietin. Pediatrics. 2016;137(3):e20153859

22. Newton NR, Leonard CH, Piecuch RE, Phibbs RH. Neurodevelopmental outcome of prematurely born children treated with recombinant human erythropoietin in infancy. J Perinatol. 1999;19(6 pt 1):403–406

23. Luciano R, Fracchiolla A, Ricci D, et al. Are high cumulative doses of erythropoietin neuroprotective in preterm infants? A two year follow-up report. Ital J Pediatr. 2015;41:64

24. Fauchère JC, Dame C, Vonthein R, et al. An approach to using recombinant erythropoietin for neuroprotection

in very preterm infants. Pediatrics. 2008;122(2):375–382

25. Juul SE, McPherson RJ, Bauer LA, Ledbetter KJ, Gleason CA, Mayock DE. A phase I/II trial of high-dose erythropoietin in extremely low birth weight infants: pharmacokinetics and safety. Pediatrics. 2008;122(2):383–391

26. Zhang J, Wang Q, Xiang H, Xin Y, Chang M, Lu H. Neuroprotection with erythropoietin in preterm and/or low birth weight infants. J Clin Neurosci. 2014;21(8):1283–1287

27. Wang H, Zhang L, Jin Y. A meta-analysis of the protective effect of recombinant human erythropoietin (rhEPO) for neurodevelopment in preterm infants. Cell Biochem Biophys. 2015;71(2):795–802

28. He JS, Huang ZL, Yang H, Weng KZ, Zhu SB. Early use of recombinant human erythropoietin promotes neurobehavioral development in preterm infants. Zhongguo Dang Dai Er Ke Za Zhi. 2008;10(5):586–588

29. Wang YH. Effects of erythropoietin on anemia and nervous system of premature infants. J Med Forum. 2005;26(7):49–52

30. Wang YH. Effects of recombinant human erythropoietin on neurobehavioral development in premature infants. Chin J Pract Pediatr. 2006;12(1):59–60

31. Song J, Sun H, Xu F, et al. Recombinant human erythropoietin improves neurological outcomes in very preterm infants. Ann Neurol. 2016;80(1):24–34

32. Natalucci G, Latal B, Koller B, et al; Swiss EPO Neuroprotection Trial Group. Effect of early prophylactic high-dose recombinant human erythropoietin in very preterm infants on neurodevelopmental outcome at 2 years: a randomized clinical trial. JAMA. 2016;315(19):2079–2085

33. Bayley N. Bayley Scales of Infant Development. 2nd ed. San Antonio, TX: Psychological Corporation; 1993

34. Bayley N. Bayley Scales of Infant and Toddler Development Screening Test. 3rd ed. San Antonio, TX: Harcourt Assessment; 2006

35. Johnson S, Moore T, Marlow N. Using the Bayley-III to assess neurodevelopmental delay: which

10 by guest on March 12, 2020www.aappublications.org/newsDownloaded from

PEDIATRICS Volume 139, number 5, May 2017

cut-off should be used? Pediatr Res. 2014;75(5):670–674

36. Higgins JPT, Green S, eds. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0. London, UK: The Cochrane Collaboration; 2011. Available at: www. cochrane- handbook. org. Accessed May 15, 2016

37. World Health Organization. International Clinical Trials Registry Platform (ICTRP). World Health Organization website. Available at: www. who. int/ ictrp/ en/ . Accessed December 18, 2016

38. Wong SS, Wilczynski NL, Haynes RB. Developing optimal search strategies for detecting clinically sound treatment studies in EMBASE. J Med Libr Assoc. 2006;94(1):41–47

39. Moher D, Liberati A, Tetzlaff J, Altman DG; PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6(7):e1000097

40. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. BMJ. 2009;339:b2700

41. Sterne JA, Sutton AJ, Ioannidis JP, et al. Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials. BMJ. 2011;343:d4002

42. Donato H, Vain N, Rendo P, et al. Effect of early versus late administration of human recombinant erythropoietin on transfusion requirements in premature infants: results of a randomized, placebo-controlled, multicenter trial. Pediatrics. 2000;105(5):1066–1072

43. Maier RF, Obladen M, Kattner E, et al. High-versus low-dose erythropoietin in extremely low birth weight infants. The European Multicenter rhEPO Study Group. J Pediatr. 1998;132(5):866–870

44. Shannon KM, Mentzer WC, Abels RI, et al. Recombinant human erythropoietin in the anemia of prematurity: results of a placebo-controlled pilot study. J Pediatr. 1991;118(6):949–955

45. Shannon KM, Mentzer WC, Abels RI, et al. Enhancement of erythropoiesis by recombinant human erythropoietin in low birth weight infants: a pilot

study. J Pediatr. 1992;120(4 pt 1): 586–592

46. Shannon KM, Keith JF III, Mentzer WC, et al. Recombinant human erythropoietin stimulates erythropoiesis and reduces erythrocyte transfusions in very low birth weight preterm infants. Pediatrics. 1995;95(1):1–8

47. Ohls RK, Christensen RD, Kamath-Rayne BD, et al. A randomized, masked, placebo-controlled study of darbepoetin alfa in preterm infants. Pediatrics. 2013;132(1). Available at: www. pediatrics. org/ cgi/ content/ full/ 132/ 1/ e119

48. Juul SE, Mayock DE, Comstock BA, Heagerty PJ. Neuroprotective potential of erythropoietin in neonates; design of a randomized trial. Matern Health Neonatol Perinatol. 2015;1:27

49. Ohlsson A, Aher SM. Early erythropoietin for preventing red blood cell transfusion in preterm and/or low birth weight infants. Cochrane Database Syst Rev. 2014;4:CD004863

50. Xu XJ, Huang HY, Chen HL. Erythropoietin and retinopathy of prematurity: a meta-analysis. Eur J Pediatr. 2014;173(10):1355–1364

51. Grimm C, Wenzel A, Groszer M, et al. HIF-1-induced erythropoietin in the hypoxic retina protects against light-induced retinal degeneration. Nat Med. 2002;8(7):718–724

52. Fang JL, Sorita A, Carey WA, Colby CE, Murad MH, Alahdab F. Interventions to prevent retinopathy of prematurity: a meta-analysis. Pediatrics. 2016;137(4):e20153387

53. McAdams RM, McPherson RJ, Mayock DE, Juul SE. Outcomes of extremely low birth weight infants given early high-dose erythropoietin. J Perinatol. 2013;33(3):226–230

54. Kersbergen KJ, Makropoulos A, Aljabar P, et al. Longitudinal regional brain development and clinical risk factors in extremely preterm infants. J Pediatr. 2016;178:93–100.e6

55. Raets MM, Dudink J, Govaert P. Neonatal disorders of germinal matrix. J Matern Fetal Neonatal Med. 2015;28(suppl 1):2286–2290

56. Leuchter RH, Gui L, Poncet A, et al. Association between early administration of high-dose erythropoietin in preterm infants and

brain MRI abnormality at term-equivalent age. JAMA. 2014;312(8):817–824

57. Marlow N. Is survival and neurodevelopmental impairment at 2 years of age the gold standard outcome for neonatal studies? Arch Dis Child Fetal Neonatal Ed. 2015;100(1):F82–F84

58. Marlow N, Morris T, Brocklehurst P, et al. A randomised trial of granulocyte–macrophage colony-stimulating factor for neonatal sepsis: childhood outcomes at 5 years. Arch Dis Child Fetal Neonatal Ed. 2015;100(4):F320–F326

59. Rosebraugh MR, Widness JA, Nalbant D, Cress G, Veng-Pedersen P. Pharmacodynamically optimized erythropoietin treatment combined with phlebotomy reduction predicted to eliminate blood transfusions in selected preterm infants. Pediatr Res. 2014;75(2):336–342

60. Rayjada N, Barton L, Chan LS, Plasencia S, Biniwale M, Bui KC. Decrease in incidence of bronchopulmonary dysplasia with erythropoietin administration in preterm infants: a retrospective study. Neonatology. 2012;102(4):287–292

61. Meyer MP. Bronchopulmonary dysplasia and erythropoietin. Concerning the article by N. Rayjada et al.: decrease in incidence of bronchopulmonary dysplasia with erythropoietin administration in preterm infants: a retrospective study [Neonatology 2012;102:287–292]. Neonatology. 2013;103(2):123

62. Doege C, Pritsch M, Frühwald MC, Bauer J. An association between infantile haemangiomas and erythropoietin treatment in preterm infants. Arch Dis Child Fetal Neonatal Ed. 2012;97(1):F45–F49

63. Lowe JR, Erickson SJ, Schrader R, Duncan AF. Comparison of the Bayley II Mental Development Index and the Bayley III Cognitive Scale: are we measuring the same thing? Acta Paediatr. 2012;101(2):e55–e58

64. Moore T, Johnson S, Haider S, Hennessy E, Marlow N. Relationship between test scores using the second and third editions of the Bayley Scales in extremely preterm children. J Pediatr. 2012;160(4):553–558

11 by guest on March 12, 2020www.aappublications.org/newsDownloaded from

DOI: 10.1542/peds.2016-4317 originally published online April 7, 2017; 2017;139;Pediatrics 

Hendrik S. Fischer, Nora J. Reibel, Christoph Bührer and Christof DameMeta-analysis

Prophylactic Early Erythropoietin for Neuroprotection in Preterm Infants: A

ServicesUpdated Information &

http://pediatrics.aappublications.org/content/139/5/e20164317including high resolution figures, can be found at:

Referenceshttp://pediatrics.aappublications.org/content/139/5/e20164317#BIBLThis article cites 60 articles, 20 of which you can access for free at:

Subspecialty Collections

http://www.aappublications.org/cgi/collection/neonatology_subNeonatologysubhttp://www.aappublications.org/cgi/collection/fetus:newborn_infant_Fetus/Newborn Infantrning_disorders_subhttp://www.aappublications.org/cgi/collection/cognition:language:leaCognition/Language/Learning Disordersal_issues_subhttp://www.aappublications.org/cgi/collection/development:behaviorDevelopmental/Behavioral Pediatricsfollowing collection(s): This article, along with others on similar topics, appears in the

Permissions & Licensing

http://www.aappublications.org/site/misc/Permissions.xhtmlin its entirety can be found online at: Information about reproducing this article in parts (figures, tables) or

Reprintshttp://www.aappublications.org/site/misc/reprints.xhtmlInformation about ordering reprints can be found online:

by guest on March 12, 2020www.aappublications.org/newsDownloaded from

DOI: 10.1542/peds.2016-4317 originally published online April 7, 2017; 2017;139;Pediatrics 

Hendrik S. Fischer, Nora J. Reibel, Christoph Bührer and Christof DameMeta-analysis

Prophylactic Early Erythropoietin for Neuroprotection in Preterm Infants: A

http://pediatrics.aappublications.org/content/139/5/e20164317located on the World Wide Web at:

The online version of this article, along with updated information and services, is

http://pediatrics.aappublications.org/content/suppl/2017/04/05/peds.2016-4317.DCSupplementalData Supplement at:

1073-0397. ISSN:60007. Copyright © 2017 by the American Academy of Pediatrics. All rights reserved. Print

the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elk Grove Village, Illinois,has been published continuously since 1948. Pediatrics is owned, published, and trademarked by Pediatrics is the official journal of the American Academy of Pediatrics. A monthly publication, it

by guest on March 12, 2020www.aappublications.org/newsDownloaded from