Seediscussions,stats,andauthorprofilesforthispublicationat:https://www.researchgate.net/publication/23145873

AnApproachtoUsingRecombinantErythropoietinforNeuroprotectioninVeryPretermInfants

ARTICLEinPEDIATRICS·AUGUST2008

ImpactFactor:5.47·DOI:10.1542/peds.2007-2591·Source:PubMed

CITATIONS

88

READS

63

7AUTHORS,INCLUDING:

Jean-ClaudeFauchere

UniversityofZurich

68PUBLICATIONS950CITATIONS

SEEPROFILE

ReinhardVonthein

UniversitätzuLübeck

172PUBLICATIONS3,659CITATIONS

SEEPROFILE

HansUlrichBucher

UniversityofZurich

268PUBLICATIONS2,810CITATIONS

SEEPROFILE

Availablefrom:HansUlrichBucher

Retrievedon:15January2016

DOI: 10.1542/peds.2007-2591 2008;122;375Pediatrics

Arri, Martin Wolf and Hans Ulrich BucherJean-Claude Fauchère, Christof Dame, Reinhard Vonthein, Brigitte Koller, Sandra

Preterm InfantsAn Approach to Using Recombinant Erythropoietin for Neuroprotection in Very

  

  http://pediatrics.aappublications.org/content/122/2/375.full.html

located on the World Wide Web at: The online version of this article, along with updated information and services, is

 

of Pediatrics. All rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.Boulevard, Elk Grove Village, Illinois, 60007. Copyright © 2008 by the American Academy published, and trademarked by the American Academy of Pediatrics, 141 Northwest Pointpublication, it has been published continuously since 1948. PEDIATRICS is owned, PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly

by guest on June 9, 2013pediatrics.aappublications.orgDownloaded from

ARTICLE

An Approach to Using Recombinant Erythropoietinfor Neuroprotection in Very Preterm InfantsJean-Claude Fauchere, MDa, Christof Dame, Prof, MDb, Reinhard Vonthein, Drc, Brigitte Koller, RNa, Sandra Arri, MDa, Martin Wolf, PhDa,

Hans Ulrich Bucher, Prof, MDa

aClinic of Neonatology, University Hospital Zurich, Zurich, Switzerland; bDepartment of Neonatology, Campus Virchow-Klinikum, Charite-Universitatsmedizin Berlin,Berlin, Germany; cDepartment of Medical Biometry, University of Tubingen, Tubingen, Germany

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

What’s Known on This Subject

Epo has been shown to be neuroprotective against hypoxic-ischemic and inflammatoryinjuries in a broad range of tissues and organs. Recent studies using various models forneonatal or adult brain injury highly favor rhEpo as a novel, very effective pharmacologicagent for neuroprotection.

What This Study Adds

Our study provides important insights into the safety and short-term outcome of pre-term infants who received early high-dose rhEpo for its neuroprotective properties at atime point when their immature brain is at highest risk for damage.

ABSTRACT

OBJECTIVE.Erythropoietin has been shown to be protective against hypoxic-ischemic andinflammatory injuries in cell culture, animal models of brain injury, and clinical trials ofadult humans. The rationale for our study was that early administration of high-doserecombinant human erythropoietin may reduce perinatal brain injury (intraventricularhemorrhage and periventricular leukomalacia) in very preterm infants and improveneurodevelopmental outcome. We investigated whether administration of high-doserecombinant human erythropoietin to very preterm infants shortly after birth andsubsequently during the first 2 days is safe in terms of short-term outcome.

METHODS. This was a randomized, double-masked, single-center trial with a 2:1 allo-cation in favor of recombinant human erythropoietin. Preterm infants (gestationalage: 24

0⁄7 to 316⁄7 weeks) were given recombinant human erythropoietin or NaCl

0.9% intravenously 3, 12 to 18, and 36 to 42 hours after birth.

RESULTS. The percentage of infants who survived without brain injury or retinopathywas 53% in the recombinant human erythropoietin group and 60% in the placebogroup. There were no relevant differences regarding short-term outcomes such asintraventricular hemorrhage, retinopathy, sepsis, necrotizing enterocolitis, and bron-chopulmonary dysplasia. For 5 infants who were in the recombinant human eryth-ropoietin group and had a gestational age of �26

0⁄7 weeks, withdrawal of intensivecare was decided (3 of 5 with severe bilateral intraventricular hemorrhage, 2 of 5with pulmonary insufficiency); no infant of the control group died. Recombinanthuman erythropoietin treatment did not result in significant differences in bloodpressure, cerebral oxygenation, hemoglobin, leukocyte, and platelet count.

CONCLUSIONS.No significant adverse effects of early high-dose recombinant humanerythropoietin treatment in very preterm infants were identified. These resultsenable us to embark on a large multicenter trial with the aim of determining whetherearly high-dose administration of recombinant human erythropoietin to very pre-term infants improves neurodevelopmental outcome at 24 months’ and 5 years’corrected age. Pediatrics 2008;122:375–382

NOVEL STRATEGIES TO protect developing organ systems, in particular the centralnervous system, are of greatest interest in neonatal intensive care medicine,

because long-term disability remains a major problem in very preterm infants.1 Theseinfants have a significantly increased risk for a delay in psychomotor development and for cognitive deficits.2–4 Themost critical period centers around birth, when oxygenation, especially of the brain, may be impaired as a result ofrespiratory, circulatory, and nutritional disorders.

Erythropoietin (Epo), the primary regulator of red blood cell production, has been shown to be protective againsthypoxic-ischemic and inflammatory injuries in a broad range of tissues and organs.5–8 The neuroprotective effects ofrecombinant human Epo (rhEpo) have been investigated most intensively in neuronal cell cultures; in experimental

www.pediatrics.org/cgi/doi/10.1542/peds.2007-2591

doi:10.1542/peds.2007-2591

This trial has been registered atwww.clinicaltrials.gov (identifierNCT00413946).

KeyWordspremature infant, erythropoietin,neurodevelopment, outcome, brain injury,very low birth weight infant,intraventricular hemorrhage, retinopathyof prematurity

AbbreviationsEpo—erythropoietinrhEpo—recombinant humanerythropoietinEpoR—erythropoietin receptorROP—retinopathy of prematurityIVH—intraventricular hemorrhagePVL—periventricular leukomalaciaGA—gestational ageTHI—total hemoglobin indexTOI—tissue oxygenation indexNIRS—near infrared spectroscopyOR—odds ratioCI—confidence intervalIQR—interquartile rangeCSF—cerebrospinal fluidBBB—blood-brain barrier

Accepted for publication Dec 28, 2007

Address correspondence to Jean-ClaudeFauchere, MD, University Hospital, Clinic ofNeonatology, Frauenklinikstrasse 10, CH-8091Zurich, Switzerland. E-mail: jean-claude.fauchere@usz.ch

PEDIATRICS (ISSN Numbers: Print, 0031-4005;Online, 1098-4275). Copyright © 2008 by theAmerican Academy of Pediatrics

PEDIATRICS Volume 122, Number 2, August 2008 375 by guest on June 9, 2013pediatrics.aappublications.orgDownloaded from

animal models; and in 3 clinical trials of adult humanswith stroke, schizophrenia, or chronic progressive mul-tiple sclerosis.9–11 These mechanisms by which rhEpoexerts its neuroprotective and neurotrophic effects in-clude the inhibition of glutamate release, modulation ofintracellular calcium metabolism, induction of antiapop-totic factors, reduction of inflammation, inhibition ofnitric oxide–mediated injury, and direct antioxidant ef-fects (for review see refs 5 and 12). Epo achieves itsneuroprotective effects by homodimerization of 2 Eporeceptors (EpoR) or heterodimerization of the EpoR withthe �-subunit common receptor (CD131).13 Recent ex-perimental studies that used a variety of models forneonatal or adult brain injury, and the 3 clinical trials ofadult patients highly favor rhEpo as a novel, very effec-tive pharmacologic agent for neuroprotection14,15 (forreview, see ref 12).

No clinical study on the use of rhEpo for neuropro-tection in human preterm and term neonates has beenreported; however, during the past decade, rhEpo hasbeen widely used in preterm infants to prevent or treatthe anemia of prematurity.16–18 Although there is stillcontroversy on the efficiency of this treatment and onadverse effects on retinopathy of prematurity (ROP), ingeneral, rhEpo has been considered to be safe and welltolerated in preterm infants (for review see refs 17 and 18).The long-term neurologic outcome of preterm infantswho received rhEpo for prevention of blood transfusionfor anemia of prematurity has so far been reported inonly 3 studies. Newton et al19 did not find relevantdifferences in the long-term neurodevelopmental out-come of 20 infants who weighed �1250 g at birth andwere treated with rhEpo (100 or 200 U/kg body weightgiven intravenously or subcutaneously, 5 times weeklyfor 6 weeks maximum) in comparison with control sub-jects. Ohls et al20 showed that extremely low birthweight infants (�1000 g) who were treated with rhEpo(400 U/kg body weight 3 times weekly given intrave-nously or subcutaneously from 96 hours after birth untilthe 35th postmenstrual week) did not benefit in neuro-developmental outcome. In a posthoc analysis, however,the same group reported higher developmental indexscores at 18 to 22 months’ corrected age for 6 infantswith Epo serum concentrations of �500 mU/mL com-pared with those (n � 6) with Epo serum concentrationsof �500 mU/mL.21

Conceptually, for proposing rhEpo as a neuroprotec-tive agent, it is important that the EpoR messenger RNAexpression in the brain is developmentally downregu-lated, but upregulated in response to hypoxia.22–24 Therationale for using rhEpo as a neuroprotective agent alsoincludes the consideration that exogenous Epo maycompensate for the delayed endogenous Epo synthesis.25

Evidence from animal experiments reveals that rhEpomust be given in high doses at the beginning or within ashort (up to 6 hours), critical time period after the onsetof brain injury to achieve a significant neuroprotectiveeffect (for review, see refs 5, 9, 26, and 27).

The aim of this study was to investigate the short-term safety of high-dose rhEpo given to very preterminfants immediately after birth and subsequently during

the first 2 days, before studying the long-term effect onneurodevelopmental outcome. The primary short-termsafety outcome measures were brain injury (intraven-tricular hemorrhage [IVH] and periventricular leukoma-lacia [PVL]) and ROP. The secondary safety outcomeswere sepsis, necrotizing enterocolitis, persistent ductusarteriosus, apnea of prematurity, and chronic lung dis-ease.

METHODSThis single-center phase II clinical trial was designed as arandomized, double-masked, placebo-controlled trial.Very preterm infants who were born between 24

0⁄7 and31

6⁄7 weeks of gestation from September 2005 throughNovember 2006 and admitted to our NICU were eligiblefor enrollment. The exclusion criteria were geneticallydefined syndromes, congenital malformations that ad-versely affect neurodevelopment, lack of adequate pa-rental information as a result of emergency cesareansection, and language barriers. The patients were ran-domly assigned within the first 3 hours after birth in a2:1 allocation in favor of rhEpo. The sequence of patientnumbers was assigned by the hospital pharmacy. Thestudy medication (rhEpo or NaCl 0.9%) was randomlyassigned to each patient number in advance, using acomputer-based random-number generator. Verum(rhEpo) and placebo drug solutions were indistinguish-able. Epoietin Beta (3000 U rhEpo/kg body weight atbirth, equal to 1 mL solution/kg birth weight; Roche,Basel, Switzerland) or an equivalent volume of normalsaline placebo was given intravenously 3 to 6, 12 to 18,and 36 to 42 hours after birth during a period of 10minutes. No infant was treated later with rhEpo foranemia of prematurity.

Neonatal adaptation was documented using the Ap-gar score at 1, 5, and 10 minutes and by the Clinical RiskIndex for Babies, the latter score being a simple tool toassess neonatal risk on the basis of the variables gesta-tional age (GA), birth weight, congenital malformation,lowest and highest appropriate fraction of inspired oxy-gen, and worst base deficit within the first 12 hours afterbirth.28 Moreover, placental histology was performedand Epo concentration in cord blood was determined induplicate by using the Quantikine human Epo Immu-noassay (R&D Systems, Wiesbaden, Germany) followingthe manufacturer’s protocol. The minimal detectionlimit of the assay was 2.5 mU/mL, and intra-assay vari-ability was �2%.

During the 24 hours after the injection of the studymedication, heart rate, transcutaneous arterial oxygensaturation (pulse oximetry arterial oxygen saturation,S9O2), transcutaneous PO2, and arterial blood pressurewere monitored. Standardized evaluations including ce-rebral sonography (at the latest before the second dose ofstudy medication) were performed at day 1, day 7, and36

0⁄7 weeks’ postmenstrual age (or earlier if discharged).IVH was graded according to Ment et al29 and cerebralwhite matter disease according to de Vries et al30 (per-sisting periventricular echodensity at 7 days and PVL at36 postmenstrual weeks). Hematologic examinationswere performed at day 1 and day 8 (range: days 7–10).

376 FAUCHERE et al by guest on June 9, 2013pediatrics.aappublications.orgDownloaded from

Both eyes were examined by an experienced ophthal-mologist to detect ROP. The severity of ROP was gradedaccording to the international classification of ROP.31

Growth (weight, length, and head circumference) wasdocumented before discharge from the hospital.

To assess whether the relatively high dose of rhEpohas any effect on brain oxygenation, we measured oxy-hemoglobin, deoxyhemoglobin, total hemoglobin index(THI), and tissue oxygenation index (TOI) using near-infrared spectroscopy (NIRS) (NIRO 300 Hamamatsu,Photonics; Hamamatsu, Japan) in a subset of infantsduring the second dose. The NIRS sensor, which wasplaced on the head of each infant, contains 1 light source(with 775-, 810-, 850-, and 910-nm wavelengths) and 1detector with 3 segments (SI photodiodes). The differ-ences between the baseline before and the five 10-minute intervals after injection (10, 20, 30, 40, and 50minutes) were calculated in each sample for the NIRSparameters.

Serious adverse events and adverse events were re-ported continuously to the safety monitoring board. Co-ordination and data management was provided by theSwiss Neonatal Network and by a study nurse (MsKoller) who was trained in data management. Dataquality was checked, and statistical calculations wereperformed by an independent person (Vonthein). Thetrial was approved by the ethical committee of the Uni-versity Children’s Hospital Zurich, by the Ethical Com-mittee of the Canton Zurich (KEK), and by SwissMedicsBerne. Written informed consent was obtained from theparents of eligible infants, ideally before birth.

The primary hypothesis of this pilot study was thatthe rate of survivors without brain injury (IVH and PVL)including retinopathy are not affected by administrationof 3 high doses of rhEpo early after birth. This chancewas 2:1 in our unit before starting the study; thereforeand because of a hope to double this chance, patientswere randomly assigned to receive rhEpo in 30 cases andplacebo in 15 cases. These numbers were calculated toprovide 95% probability of at least 1 event in each armso that odds ratios (ORs) could be determined. For cal-

culation of the posterior probability of effects of rhEpoon the occurrence of IVH or ROP, a historical controlbased on consecutive 620 preterm infants who were�32 gestational weeks from the Swiss Neonatal Networkwas included, yielding incidences for IVH of 23% and forROP of 19.5%.

Bayes factor (likelihood ratio) was used in lieu of theP value as a measure of the evidential strength.32 Fre-quencies of traits were compared by OR and exact mid-probability 95% confidence interval (CI) (StatsDirect2.4.4 [Ltd Altrincham, Cheshire, United Kingdom]).Continuous measurements that seemed to follow a nor-mal distribution were summarized by means, SD, and95% CIs for the difference of means. Median and inter-quartile ranges (IQRs) were used to summarize the othervariables (JMP IN 5.1 statistical software [SAS Institute,Cary, NC]).

RESULTSA total of 122 very preterm infants were assessed foreligibility. Of these, 77 were excluded: 56 for not meet-ing the inclusion criteria and 21 because of parentalrefusal, leaving a study group of 45 infants who wereallocated to the rhEpo group (nt � 30) or to the placebogroup (nc � 15). Table 1 summarizes the demographicdata. Importantly, the intervention and the controlgroups were comparable with regard to GA, birth weightand head circumference, gender, pregnancy-relatedcomplications, antenatal steroids, mode of delivery, um-bilical artery pH, Apgar score, Clinical Risk Index forBabies score, and Epo concentration in cord blood.

There were 16 (53%) of 30 survivors without IVH,PVL, or ROP in the rhEpo group and 9 (60%) of 15 in theplacebo group (Table 2). For 5 infants who had a GA of�26 weeks and belonged to the rhEpo group, with-drawal of intensive care and redirection of care weredecided because of severe bilateral IVH (3 infants, in 1 ofthem IVH was already present before dose 1) or severepulmonary interstitial emphysema with irreversibleglobal pulmonary insufficiency (2 infants). The charts ofthese 5 infants were carefully reviewed by the study

TABLE 1 Demographic Data

Parameter rhEpo (nt � 30) Placebo (nc � 15) Difference, OR (95% CI)

Girls, n (%) 15 (50) 8 (53) 0.875 (0.240 to 3.100)Gestational age, mean (SD), wk � d 28

3⁄7 (20⁄7) 28

4⁄7 (20⁄7) 0

3⁄7 (�12⁄7 to 1

2⁄7)Birth weight, mean (SD), g 1112 (347) 1081 (354) 31 (�198 to 259)z score, mean (SD) �0.03 (0.86) �0.23 (0.93) 0.20 (�0.39 to 0.79)

Birth head circumference, mean (SD), cm 26.1 (2.4) 26.2 (2.6) �0.1 (�1.8 to 1.5)z score, mean (SD) �0.16 (0.63) �0.21 (0.90) 0.05 (�0.49 to 0.58)

Maternal age, mean (SD), y 33.7 (4.5) 31.4 (5.5) 2.2 (�1.2 to 5.6)Chorioamnionitis, n (%) 6 (20) 1 (7) 3.50 (0.44 to 86.00)Antenatal steroids, n (%) 22 (73) 12 (80) 0.69 (0.13 to 3.10)Cesarean section, n (%) 24 (80) 15 (100) 0.0 (0.0 to 2.0)pH umbilical artery, mean (SD) 7.33 (0.08) 7.33 (0.07) 0.00 (�0.05 to 0.05)Maximum negative base excess, mean (SD) �5.0 (3.0) �5.1 (2.3) 0.1 (�1.6 to 1.7)Apgar score at 5 min, median (IQR) 7.5 (9 to 6) 8 (9 to 5) �0.5CRIB score, median (IQR) 2 (6 to 1) 2 (5 to 1) 0Epo concentration in cord blood �50 mU/mL, n (%)a 7/25 (23) 4/11 (36) 0.68 (0.15 to 3.40)a Cord blood specimens were available for only 25 of 30 and 11 of 15 patients, respectively.

PEDIATRICS Volume 122, Number 2, August 2008 377 by guest on June 9, 2013pediatrics.aappublications.orgDownloaded from

safety board. In each case, a causal relationship withrhEpo treatment was unlikely. None of the neonates inthe control group died.

The posterior probability of the hypothesis that rhEpohas an effect on the occurrence of IVH was 0.20 whenthe previous probability was 0.5 (Bayes factor 4.11; Fig1). Including historical controls of combined weight, 15,to make the trial balanced, would alter that posteriorprobability to 0.27 and the Bayes factor to 2.76. Simi-larly, the posterior probability of rhEpo effect on theoccurrence of ROP was 0.36 without and 0.47 withprevious information.

Continuous monitoring of arterial blood pressure, ar-

terial oxygen saturation, heart rate, cerebral oxygen-ation (TOI), and cerebral perfusion (THI) did not showany relevant deviation from baseline during injection ofrhEpo or placebo, respectively (Fig 2). There was norelevant difference for these parameters between 16 in-vestigated patients who received rhEpo and 10 investi-gated patients who received placebo at any time beforeand after drug injection.

Hemoglobin, platelets, and white cell counts weresimilar at day 8 in both groups. For 39 of 45 patients, Epoconcentrations in the umbilical arterial cord blood weredetermined (Table 3). For 11 (28.2%) of 39 specimens,Epo concentrations were elevated (�50 mU/mL33).

DISCUSSIONWe report on the first randomized, double-blind, placebo-controlled clinical trial to use high-dose rhEpo as a neu-roprotective agent given shortly after birth to very pre-term infants. The first objective of this trial was to assessthe safety of early high-dose rhEpo administration andwhether this intervention alters the incidence of com-plications that typically are associated with preterm birth(IVH, PVL, ROP, septicemia, necrotizing enterocolitis,bronchopulmonary dysplasia) and, finally, whethercomplications that are associated with rhEpo treatmentin adult humans, such as arterial hypertension andthrombotic events, occur also in preterm infants (Table4).

The most important finding of this trial is that therewere no differences regarding the short-term outcomebetween both groups. Importantly, any relevant increasein typical adverse effects of rhEpo, in particular ROP,18

were not observed in our patients. The potential draw-back of Epo therapy, namely an increase in red cell mass,

TABLE 2 Neonatal Morbidity According to Treatment

Parameter rhEpo (nt � 30) Placebo (nc � 15) Difference, OR (9% CI)

Death, n/N (%) 5/30 (17) 0/15 (0) ∞ (0.640 to ∞)IVH grades I–IV, n/N (%) 11/30 (37) 5/15 (33) 1.200 (0.310 to 4.600)Severe IVH grades III–IV, n/N (%) 4/30 (13) 0/15 (0) ∞ (0.460 to ∞)Persisting PVE (�7 d), n/N (%) 20/28 (71) 14/15 (93) 0.180 (0.008 to 1.300)ROP stages 1–4, n/N (%) 2/24 (8) 2/15 (13) 0.640 (0.061 to 6.800)Severe ROP stages 3 and 4, n/N (%) 1/24 (4) 0/15 (0) ∞ (0.033 to ∞)Sepsis, n/N (%) 7/30 (23) 4/15 (27) 0.840 (0.200 to 3.900)NEC, n/N (%) 9/29 (31) 2/15 (13) 0.000 (0.000 to 1.700)PDA (treatment needed), n/N (%) 14/29 (48) 4/15 (27) 2.600 (0.650 to 11.000)Apnea of prematurity (treatment needed), n/N (%) 29/30 (97) 15/15 (100) 0.000 (0.000 to 38.000)Chronic lung disease (oxygen dependence at 36

0⁄7

gestational wk PMA), n/N (%)13/30 (43) 7/15 (47) 0.870 (0.240 to 3.200)

Survivors without IVH, PVL, or ROP, n/N (%) 16/30 (53) 9/15 (60) 0.760 (0.210 to 2.700)Hematologic parametersHemoglobin at day 8 (g/dL)Mean (SD, % of mean) 15.8 (20) 14.6 (17) �8% (�7% to 26%)n 16 11

Platelet count at day 8 (/�L)Mean (SD, % of mean) 209 (61) 244 (31) �14% (�39% to 21%)n 16 11

White blood cell count at day 8 (/�L)Mean (SD, % of mean) 15 617 (73) 10 535 (244) �48% (�44% to 292%)n 16 11

PVE indicates periventricular echodensity; NEC, necrotizing enterocolitis; PDA, persistent ductus arteriosus; PMA, postmenstrual age.

0.05 0.25 0.5 0.75 0.95

0.05

0.25

0.5

0.75

0.95

Prior probability

Pos

terio

r pro

babi

lity

IVH

ROP

0.05 0.25 0.5 0.75 0.95

0.05

0.25

0.5

0.75

0.95

IVH

ROP

FIGURE 1If-then diagram. Shown is the posterior probability of the null hypothesis of exactly noeffect of rhEpo on the occurrence of IVH (fine lines) and ROP (heavy lines) by previousprobability, for historical controls who weighed as much as 15 patients in the controlgroup (dashed lines) or without historical controls (solid lines).

378 FAUCHERE et al by guest on June 9, 2013pediatrics.aappublications.orgDownloaded from

was not seen in our study, an adverse effect that we didnot expect because of the short duration of therapy in apopulation at risk rather for anemia than for polycy-themia. This result is in accordance with the report byEhrenreich et al,14 who showed that short-term ad-ministration of Epo as a neuroprotective agent to pa-tients with stroke did not result in increased hemato-crit levels. Other complications of rhEpo treatment,such as arterial hypertension and thrombocytosis, alsodid not occur. Moreover, no perceptible median groupdifferences in brain perfusion (THI) and cerebral TOIwere found (Fig 2).34

A finding that warrants a closer analysis is the highermortality in rhEpo-treated infants (5 of 30) in compar-ison with control subjects (0 of 15; Table 2). All 5 deaths

occurred in extremely preterm infants with a GA of �26weeks (extremely low GA). For all of these infants,redirection of care was decided together with the par-ents. Mechanical ventilation was withdrawn between 6and 174 hours after birth for either severe bilateral IVH(3 of 5 infants, with 1 infant presenting with IVH beforereceiving the first dose of rhEpo) or global respiratoryinsufficiency (2 of 5 with severe bilateral pulmonaryinterstitial emphysema and air leaks).

Figure 1 shows how previous probabilities of the nullhypotheses change after our pilot study. Although it ismore probable than not that a high dose of rhEpo doesnot alter the incidence rates of IVH and ROP in very lowbirth weight infants, the inclusion of historical controlsshifts the assessment toward equipoise, because histori-cal incidences were higher than in the placebo group.

Both the incidence of severe IVH (III–IV; 13%) andthe incidence of death as a result of pulmonary insuffi-ciency (9%) in rhEpo-treated infants of extremely lowGA were within the range reported in recent interna-tional studies: 11% to 25% for severe IVH35–37 and 34%for death as a result of pulmonary insufficiency.38

The optimal dosage and treatment regimen in a neu-roprotection approach in neonates is not clear yet. In astudy that used high-dose rhEpo for neuroprotection inadult humans with stroke, rhEpo was administered at adosage of 33 333 U (�450 U/kg) during 30 minuteswithin 5 hours after the onset of symptoms and on thesubsequent 2 days.14 In patients with long-term schizo-phrenia, rhEpo was given as neuroprotectant in aweekly dosage of 40 000 U for 3 months.15 Most re-cently, the effects of 48 000 vs 8000 U of rhEpo (12weeks of weekly followed by 12 weeks of biweeklytreatment) were tested in an open-label study of patientswith long-term progressive multiple sclerosis.11 In these3 human trials that used rhEpo as neuroprotectant,treatment was found to be safe and beneficial.

Experimental data on the effects of rhEpo in animalmodels of neonatal brain injury have been summarizedrecently.12 On the one hand, escalating dosages from1000 to 5000 U/kg showed most efficient neuroprotec-tion at Epo concentrations in the cerebrospinal fluid(CSF) between 20 and 200 mU/mL.39–44 On the otherhand, recent in vitro and animal experiments of neona-tal brain injury that used single or multiple injections ofhighest rhEpo dosages (up to 30 000 U/kg or 40 U/mL inbrain slide cultures) showed no additional benefit on

TABLE 3 Association of Mortality and Morbidity With Fetal CordBlood Epo Levels

Parameter Elevated Cord BloodEpo Level (�50 mU/L)

(np � 11)

Normal Cord Blood EpoLevel (�50 mU/L )

(nn � 25)

RhEpo(n � 7)

Placebo(n � 4)

RhEpo(n � 18)

Placebo(n � 7)

Neonatal death 3 0 1 0IVH 2 0 4 0PVE 4 3 13 7ROP 0 1 1 1

FIGURE 2Mean arterial pressure (MAP), TOI, and THI before and after rhEpo/placebo injection. Boxplots show the median, quartiles, and extreme values.

PEDIATRICS Volume 122, Number 2, August 2008 379 by guest on June 9, 2013pediatrics.aappublications.orgDownloaded from

neuroprotection or even neurotoxic effects, particularlywhen combined with mild hypoxia.42,45 Evidence fromanimal experiments reveals that rhEpo must be given inhigh doses at the beginning or within a short (up to 6hours) critical time period after the onset of brain injuryto achieve a significant neuroprotective effect26,27 (forreview, see refs 5 and 9). Because the highly glycosy-lated Epo molecule needs to cross the blood-brain barrier(BBB), most likely by a saturable, active transport mech-anism,26 intravenous administration is favorable, as sug-gested by pharmacokinetic analysis. Recent studies inadult rats, fetal sheep, and juvenile or adult nonhumanprimates indicated that Epo concentrations in the CSFincrease between 1 and 2 hours after systemic (intra-peritoneal or intravenous) application of high-doserhEpo (5000 U/kg) to concentrations of �100 mU/mLand peak between 3 and 4 hours at concentrations of�200 mU/mL.26,40 On the basis of a model for predictedconcentration-time profiles of Epo in the CSF after singleand repetitive rhEpo administration,46 we decided togive 3 doses of 3000 U/kg rhEpo intravenously, yieldingEpo concentrations in CSF not higher than 300 mU/mL.

A limitation of our study is the small number ofinfants included. A larger number would be needed toassess the statistical significance of trends such as thoseobserved for mean arterial pressure and TOI towardlower values in the rhEpo group (Fig 2) and of higherpercentages in severe IVH and severe ROP (Table 3).Pooling our results with those of other studies that areunder way will help to narrow the CIs.

On the basis of our preliminary results, several aspectsmay be important for future clinical studies. (1) Potentialdanger of high-dose rhEpo in infants with IVH: In pre-term infants with IVH, the BBB is impaired; conse-quently, higher amounts of rhEpo may penetrate thebrain. An increased uptake of rhEpo around very vul-nerable brain capillaries may change their autoregula-tion if endothelial EpoRs are activated. Importantly, ex-tremely high concentrations of rhEpo may exert adverseeffects on neuronal cells and increase the risk for braindamage, in particular in combination with additionalfactors such as hypoxia and hyperoxia.47 (2) Upregula-tion of EpoR in extremely preterm infants: As shown inmice, EpoR expression is 10-fold higher in the embryonicbrain than in the adult brain and decreases immediatelyafter birth48; therefore, it is very likely that EpoR levelsare relatively high in newly born preterm infants. (3)Elevated cord blood Epo levels: In 11 (28.2%) of 39

umbilical cord blood specimens, Epo concentrationswere significantly elevated (�50 mU/mL), indicatingpossible fetal distress. In 9 of 11 infants, placental insuf-ficiency, pre/eclampsia, amnion infection, or fetal ane-mia could accounted for increased fetal Epo production,suggesting preconditioning that has been developedslowly over time. In the remaining 2 infants, increasedEpo concentrations could have resulted from acute pre-natal fetal distress. In our short-term analysis, no con-sistent pattern between neonatal complications and thepresence of increased fetal Epo concentrations or theabsolute Epo levels was obvious (Table 3). The observa-tion of possibly less periventricular echodensity in thepresence of increased fetal blood Epo concentrationsmight be of interest when it comes to the analysis ofneurodevelopmental outcome.

Upregulation of EpoR and an increase in circulatingEpo may also be important aspects for the design ofadditional studies to investigate the neuroprotective ef-fect of rhEpo in term infants with acute hypoxic-isch-emic brain injury, likely in combination with inducedbrain hypothermia; however, as recently discussed byBuhrer et al,47 caution is warranted in hypoxic condi-tions because various changes occur that may divert Epofrom its neuroprotective and restorative actions. Theseobserved changes include (1) a significantly increasedtransport of blood-borne proteins (eg, Epo) through theBBB,49 exposing the central nervous system to a higheramount of administered rhEpo, (2) an increase in EpoRtranscription in neurons,50 and (3) an induction of EpoRby Epo51 with a decreased ratio of heterodimeric CD131/EpoR, leading to an increased binding of Epo to thehigh-affinity homodimeric EpoR/EpoR, thereby losingits protective action.

CONCLUSIONSOur study provides important insights into the safetyand short-term outcome of preterm infants who re-ceived early high-dose rhEpo for neuroprotection. Asindependently analyzed by the external safety boardand by the Epo trial center, no signs of adverse effectsof early high-dose rhEpo treatment in very preterminfants were identified. These results enable us toembark on a larger multicenter study with the aim todetermine whether early high-dose administration ofrhEpo in very preterm infants finally improves neu-rodevelopmental outcome at 24 months’ and 5 years’corrected age.

TABLE 4 Clinical Status at Discharge According to Treatment

Parameter rhEpo (nt � 25) Placebo (nc � 15) Difference, OR (95% CI)

PMA, median (IQR), wk � d 365⁄7 (35

1⁄7 to 376⁄7) 37

3⁄7 (361⁄7 to 39

3⁄7) �5⁄7

Weight, mean (SD), g 2567 (467) 2653 (557) �86.00 (�438.00 to 267.00)z score, mean (SD) �0.90 (1.02) �1.16 (1.13) 0.27 (�0.46 to 1.00)

Head circumference, mean (SD), cm 33.3 (1.5) 33.7 (2.1) �0.40 (�1.70 to 0.90)z score, mean (SD) �0.41 (1.00) �0.52 (1.28) 0.12 (�0.68 to 0.91)

Weight growth, mean (SD) g 1375 (496) 1572 (569) �197.00 (�561.00 to 167.00)Head growth, mean (SD) cm 6.6 (2.3) 7.4 (2.9) �0.90 (�1.70 to 1.00)Length of hospital stay, mean (SD), d 49 (22) 51 (31) �9.50 (�28.00 to 9.00)

PMA indicates postmenstrual age.

380 FAUCHERE et al by guest on June 9, 2013pediatrics.aappublications.orgDownloaded from

ACKNOWLEDGMENTSThis pilot study was supported by the RoFAR Founda-tion (Roche Foundation for Anemia Research). Epoietin� was kindly provided by Roche (Basel, Switzerland).

We particularly thank the parents who have contrib-uted greatly to this project by consenting to the enroll-ment of their preterm infant and the NICU staff for theirsupport. We are indebted to the independent data andsafety monitoring board (president Theo Gasser, Zurich).We acknowledge the invaluable support and thoughtfulcomments of Christian Bauer (Zurich), Max Gassmann(Institute of Veterinary Physiology, University of Zu-rich), Hugo Marti (Max-Planck Institute, Molecular CellBiology, Bad Nauheim, Germany), Joachim Riethmuller(University of Tubingen, Tubingen, Germany), and nu-merous other colleagues. We also acknowledge the helpby Daniel Fetz of the Pharmacy of the Canton Zurich forpreparing the rhEpo and saline vials and for the randomassignment of our patients. In large part, the NIRS mea-surements were performed by Katrin Egli, Andrea Baus-chatz, Tanja Karen, and Ruben Barbaro.

REFERENCES1. Marlow N, Wolke D, Bracewell MA, Samara M, for the

EPICure Study Group. Neurologic and developmental disabilityat six years of age after extremely preterm birth. N Engl J Med.2005;352(1):9–19

2. Bucher HU, Killer C, Ochsner Y, Vaihinger S, Fauchere JC.Growth, developmental milestones and health problems in thefirst 2 years in very preterm infants compared with terminfants: a population based study. Eur J Pediatr. 2002;161(3):151–156

3. Winter S, Autry A, Boyle C, Yeargin-Allsopp M. Trends in theprevalence of cerebral palsy in a population-based study. Pedi-atrics. 2002;110(6):1220–1225

4. Meadow W, Lee G, Lin K, Lantos J. Changes in mortality forextremely low birth weight infants in the 1990s: implicationsfor treatment decisions and resource use. Pediatrics. 2004;113(5):1223–1229

5. Dame C, Juul S, Christensen RD. The biology of erythropoietinin the central nervous system and its neurotrophic and neuro-protective potential. Biol Neonate. 2001;79(3–4):228–235

6. Gassmann M, Heinicke K, Soliz J, Ogunshola OO. Non-erythroid functions of erythropoietin. Adv Exp Med Biol. 2003;543:323–330

7. Jelkmann W, Wagner K. Beneficial and ominous aspects of thepleiotropic action of erythropoietin. Ann Hematol. 2004;83(11):673–686

8. Fandrey J. A cordial affair: erythropoietin and cardioprotec-tion. Cardiovasc Res. 2006;72(1):1–2

9. Maiese K, Li F, Chong ZZ. New avenues of exploration forerythropoietin. JAMA. 2005;293(1):90–95

10. Noguchi CT, Asavaritikrai P, Teng R, Jia Y. Role of erythropoi-etin in the brain. Crit Rev Oncol Hematol. 2007;64(2):159–171

11. Ehrenreich H, Fischer B, Norra C, et al. Exploring recombinanthuman erythropoietin in chronic progressive multiple sclero-sis. Brain. 2007;130(Pt 10):2577–2588

12. Juul S, Felderhoff-Mueser U. Epo and other hematopoieticfactors. Semin Fetal Neonatal Med. 2007;12(4):250–258

13. Brines M, Grasso G, Fiordaliso F, et al. Erythropoietin mediatestissue protection through an erythropoietin and common beta-subunit heteroreceptor. Proc Natl Acad Sci U S A. 2004;101(41):14907–14912

14. Ehrenreich H, Hasselblatt M, Dembowski C, et al. Erythropoi-

etin therapy for acute stroke is both safe and beneficial. MolMed. 2002;8(8):495–505

15. Ehrenreich H, Hinze-Selch D, Stawicki S, et al. Improvement ofcognitive functions in chronic schizophrenic patients by re-combinant human erythropoietin. Mol Psychiatry. 2007;12(2):206–220

16. Maier RF, Obladen M, Muller-Hansen I, et al. Early treatmentwith erythropoietin beta ameliorates anemia and reducestransfusion requirements in infants with birth weights below1000 g. J Pediatr. 2002;141(1):8–15

17. Aher SM, Ohlsson A. Early versus late erythropoietin for pre-venting red blood cell transfusion in preterm and/or low birthweight infants. Cochrane Database Syst Rev. 2006;(3):CD004865

18. Ohlsson A, Aher S. Early erythropoietin for preventing redblood cell transfusion in preterm and/or low birth weightinfants. Cochrane Database Syst Rev. 2006;(3):CD004863. DOI:10.1002/14651858. CD004863.pub2.

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

20. Ohls RK, Ehrenkranz RA, Das A, et al. Neurodevelopmentaloutcome and growth at 18 to 22 months’ corrected age inextremely low birth weight infants treated with early erythro-poietin and iron. Pediatrics. 2004;114(5):1287–1291

21. Bierer R, Peceny MC, Hartenberger CH, Ohls RK. Erythropoi-etin concentrations and neurodevelopmental outcome inpreterm infants. Pediatrics. 2006;118(3). Available at: www.pediatrics.org/cgi/content/full/118/3/e635

22. Chin K, Yu X, Beleslin-Cokic B, et al. Production and process-ing of erythropoietin receptor transcripts in brain. Brain Res MolBrain Res. 2000;81(1–2):29–42

23. Spandou E, Papoutsopoulou S, Soubasi V, et al. Hypoxia-ischemia affects erythropoietin and erythropoietin receptor ex-pression pattern in the neonatal rat brain. Brain Res. 2004;1021(2):167–172

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

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

26. Brines ML, Ghezzi P, Keenan S, et al. Erythropoietin crossesthe blood-brain barrier to protect against experimental braininjury. Proc Natl Acad Sci U S A. 2000;97(19):10526–10531

27. Wang L, Zhang Z, Wang Y, Zhang R, Chopp M. Treatment ofstroke with erythropoietin enhances neurogenesis and angio-genesis and improves neurological function in rats. Stroke.2004;35(7):1732–1737

28. Parry G, Tucker J, Tarnow-Mordi W. CRIB II: an update of theclinical risk index for babies score. Lancet. 2003;361(9371):1789–1791

29. Ment LR, Bada HS, Barnes P, et al. Practice parameter: neuro-imaging of the neonate—report of the Quality Standards Sub-committee of the American Academy of Neurology and thePractice Committee of the Child Neurology Society. Neurology.2002;58(12):1726–1738

30. De Vries LS, Van Haastert IL, Rademaker KJ, Koopman C,Groenendaal F. Ultrasound abnormalities preceding cerebralpalsy in high-risk preterm infants. J Pediatr. 2004;144(6):815–820

31. International Committee for the Classification of the LateStages of Retinopathy of Prematurity. An international classi-fication of retinopathy of prematurity: II—the classification ofretinal detachment. Arch Ophthalmol. 1987;105(7):906–912

PEDIATRICS Volume 122, Number 2, August 2008 381 by guest on June 9, 2013pediatrics.aappublications.orgDownloaded from

32. Goodman SN. Toward evidence-based medical statistics:2—the Bayes factor. Ann Intern Med. 1999;130(12):1005–1013

33. Fahnenstich H, Dame C, Allera A, Kowalewski S. Biochemicalmonitoring of fetal distress with serum-immunoreactive eryth-ropoietin. J Perinat Med. 1996;24(1):85–91

34. Wolf M, Egli K, Karen T, Barbaro R, Bauschatz A, Bucher H.Effects of Erythropoietin on Cerebral Blood Circulation Measured byNear-infrared Spectroscopy in Preterm Neonates. Barcelona, Spain:European Academy of Paediatrics; 2006:125

35. Vanhaesebrouck P, Allegaert K, Bottu J, et al. The EPIBELStudy: outcomes to discharge from hospital for extremely pre-term infants in Belgium. Pediatrics. 2004;114(3):663–675

36. Markestad T, Kaaresen PI, Ronnestad A, et al. Early death,morbidity, and need of treatment among extremely prematureinfants. Pediatrics. 2005;115(5):1289–1298

37. Mikkola K, Ritari N, Tommiska V, et al. Neurodevelopmentaloutcome at 5 years of age of a national cohort of extremely lowbirth weight infants who were born in 1996–1997. Pediatrics.2005;116(6):1391–1400

38. Costeloe K, Hennessy E, Gibson AT, Marlow N, Wilkinson AR,for the EPICure Study Group. The EPICure Study: outcomes todischarge from hospital for infants born at the threshold ofviability. Pediatrics. 2000;106(4):659–671

39. Aydin A, Genc K, Akhisaroglu M, Yorukoglu K, Gokmen N,Gonullu E. Erythropoietin exerts neuroprotective effect in neo-natal rat model of hypoxic-ischemic brain injury. Brain Dev.2003;25(7):494–498

40. Juul SE, McPherson RJ, Farrell FX, Jolliffe L, Ness DJ, GleasonCA. Erythropoietin concentrations in cerebrospinal fluid ofnonhuman primates and fetal sheep following high-dose re-combinant erythropoietin. Biol Neonate. 2004;85(2):138–144

41. Sola A, Wen TC, Hamrick SE, Ferriero DM. Potential for pro-tection and repair following injury to the developing brain: arole for erythropoietin? Pediatr Res. 2005;57(5 pt 2):110R–117R

42. Kellert BA, McPherson RJ, Juul SE. A comparison of high-doserecombinant erythropoietin treatment regimens in brain-injured neonatal rats. Pediatr Res. 2007;61(4):451–455

43. Statler PA, McPherson RJ, Bauer LA, Kellert BA, Juul SE.Pharmacokinetics of high-dose recombinant erythropoietin inplasma and brain of neonatal rats. Pediatr Res. 2007;61(6):671–675

44. Wang Y, Zhang ZG, Rhodes K, et al. Post-ischemic treatmentwith erythropoietin or carbamylated erythropoietin reducesinfarction and improves neurological outcome in a rat model offocal cerebral ischemia. Br J Pharmacol. 2007;151(8):1377–1384

45. Weber A, Dzietko M, Berns M, et al. Neuronal damage aftermoderate hypoxia and erythropoietin. Neurobiol Dis. 2005;20(2):594–600

46. Xenocostas A, Cheung WK, Farrell F, et al. The pharmacoki-netics of erythropoietin in the cerebrospinal fluid after intra-venous administration of recombinant human erythropoietin.Eur J Clin Pharmacol. 2005;61(3):189–195

47. Buhrer C, Felderhoff-Mueser U, Wellmann S. Erythropoietinand ischemic conditioning: why two good things may be bad.Acta Paediatr. 2007;96(6):787–789

48. Liu ZY, Chin K, Noguchi CT. Tissue specific expression ofhuman erythropoietin receptor in transgenic mice. Dev Biol.1994;166(1):159–169

49. Plateel M, Teissier E, Cecchelli R. Hypoxia dramatically in-creases the nonspecific transport of blood-borne proteins to thebrain. J Neurochem. 1997;68(2):874–877

50. Ehrenreich H, Hasselblatt M, Knerlich F, et al. A hematopoieticgrowth factor, thrombopoietin, has a proapoptotic role in thebrain. Proc Natl Acad Sci U S A. 2005;102(3):862–867

51. Beleslin-Cokic BB, Cokic VP, Yu X, Weksler BB, Schechter AN,Noguchi CT. Erythropoietin and hypoxia stimulate erythropoi-etin receptor and nitric oxide production by endothelial cells.Blood. 2004;104(7):2073–2080

ON THEIR TOES AND ASKING FOR TROUBLE, SELF-TAUGHT BALLERINAS GO ONLINE:YOUTUBE PEER REVIEW HORRIFIES DANCE SCHOOLS; EN POINTE ACCIDENTSWAITINGTO HAPPEN

“Since Louis XIV established the Royal Academy of Dance in 1661, formalballet has never been a self-taught art. Aspiring dancers have been trainedunder the supervision of professionals. And ever since the pointe shoe wasinvented in the early 1800s, only girls who had spent three years or so takinglessons ever got permission from their ballet teachers to buy toe shoes. Butnow Web videos are shaking loose the rigid hierarchy of the ballet world.Aspiring ballerinas are recording videos of themselves dancing and postingthe results for people to look at and critique on the Internet. Young hopefulsput video cameras on their kitchen or bathroom floor, then do simpleexercises in pointe shoes. The videos, which generally aren’t more than aminute long, attract viewer chat pointing out mistakes and offering tips.Dancing in pointe shoes can indeed be a danger for children who aren’ttrained to get up on their toes without hurting themselves. Possible injuriesinclude stress fractures, sprained ankles, tendonitis, damage to the growthplates in the feet, shin-splints and bunions. Shoes that don’t fit properly canpermanently deform young feet.”

Steel E.Wall Street Journal. June 8, 2008Noted by JFL, MD

382 FAUCHERE et al by guest on June 9, 2013pediatrics.aappublications.orgDownloaded from

DOI: 10.1542/peds.2007-2591 2008;122;375Pediatrics

Arri, Martin Wolf and Hans Ulrich BucherJean-Claude Fauchère, Christof Dame, Reinhard Vonthein, Brigitte Koller, Sandra

Preterm InfantsAn Approach to Using Recombinant Erythropoietin for Neuroprotection in Very

  

ServicesUpdated Information &

mlhttp://pediatrics.aappublications.org/content/122/2/375.full.htincluding high resolution figures, can be found at:

References

ml#ref-list-1http://pediatrics.aappublications.org/content/122/2/375.full.htat:This article cites 47 articles, 17 of which can be accessed free

Citations

ml#related-urlshttp://pediatrics.aappublications.org/content/122/2/375.full.htThis article has been cited by 13 HighWire-hosted articles:

Subspecialty Collections

_disorders_subhttp://pediatrics.aappublications.org/cgi/collection/neurologicNeurologic Disorders

_subhttp://pediatrics.aappublications.org/cgi/collection/neurologyNeurologythe following collection(s):This article, along with others on similar topics, appears in

Permissions & Licensing

mlhttp://pediatrics.aappublications.org/site/misc/Permissions.xhttables) or in its entirety can be found online at: Information about reproducing this article in parts (figures,

Reprints http://pediatrics.aappublications.org/site/misc/reprints.xhtml

Information about ordering reprints can be found online:

rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.Grove Village, Illinois, 60007. Copyright © 2008 by the American Academy of Pediatrics. All and trademarked by the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elkpublication, it has been published continuously since 1948. PEDIATRICS is owned, published, PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly

by guest on June 9, 2013pediatrics.aappublications.orgDownloaded from