Resuscitation of Preterm Infants with Different Inspired Oxygen Fractions

Denise Rook, PhD1, Henk Schierbeek, PhD2, Maximo Vento, PhD3,4, Hester Vlaardingerbroek, MD, PhD1,

Anne C. van der Eijk, PhD1,5, Mariangela Longini, PhD6, Giuseppe Buonocore, PhD6, Javier Escobar, PhD4,

Johannes B. van Goudoever, PhD2,7, and Marijn J. Vermeulen, PhD1

Objective To test the hypothesis that an initial fraction of inspired oxygen (FiO2) of 30% during resuscitation ofpreterm infants results in less oxidative stress and is associated with improved clinical outcomes compared withan FiO2 of 65%.Study design Preterm infants of gestational age <32 weeks (n = 193) were randomized to start resuscitation witheither 30% oxygen (low-oxygen group) or 65% oxygen (high-oxygen group), after which the FiO2 was adjustedbased on oxygen saturation values. The primary outcome was bronchopulmonary dysplasia (BPD) assessed at36 weeks postmenstrual age. Secondary outcomes included major neonatal illnesses and markers of oxidativestress.Results The median gestational age of included infants was 286/7 weeks (IQR, 265/7-30

3/7 weeks). The incidenceof BPD was not significantly different between the low-oxygen and high-oxygen groups (24% vs 17%; P = .15). TheFiO2 in both groups was adjusted to a mean of 40% by 7 minutes in the low-oxygen group and by 11 minutes in thehigh-oxygen group. No differences in markers of oxidative stress were noted between groups.Conclusion Initial supplementation of preterm infants with 30% oxygen during the fetal-to-neonatal transition isas safe as 65% oxygen, with no differences in oxidative stress markers or BPD. (J Pediatr 2014;-:---).

Resuscitation of infants at birth with 100% oxygen is associated with increased morbidity and mortality1,2; thus, currentresuscitation guidelines advise starting the resuscitation of term infants with air or blended oxygen.3,4 The optimal frac-tion of inspired oxygen (FiO2) at which to start resuscitation of preterm infants remains unknown. The latest Interna-

tional Liaison Committee on Resuscitation guidelines state that both hyperoxemia and hypoxemia should be avoided in theabsence of information on the initial FiO2.

4

Small studies on FiO2 for the resuscitation of preterm infants have compared low and high initial FiO2.5-7 In those studies,

FiO2 was increased in the low-oxygen groups according to oxygen saturation (SpO2) targets, suggesting that starting resusci-tation of preterm infants with room air might provide insufficient oxygen. FiO2 was decreased in the high-oxygen groups basedon the SpO2 targets. Furthermore, Vento et al7 reported that resuscitation of preterm infants of gestational age (GA)#28 weekswith an initial FiO2 of 30% (n = 37) resulted in decreased oxidative stress markers and a decreased risk of bronchopulmonarydysplasia (BPD) compared with such infants with an initial FiO2 of 90% (n = 41).

Based on these data, it seems that starting the resuscitation of preterm infants with room air is too low and starting at 90%FiO2 is too high, and that the optimum level is most likely between these extremes. Thus, we aimed to assess the safety and

From the 1Department of Pediatrics, Division ofNeonatology, Erasmus Medical Center–Sophia

efficacy of starting the resuscitation of preterm infants with an intermediateFiO2. We hypothesized that starting the resuscitation of preterm infants (GA<32 weeks) with an initial FiO2 of 30% would decrease the incidence of BPDcompared with an initial FiO2 of 65%.

Children’s Hospital, Rotterdam, The Netherlands;2Department of Pediatrics, Academic Medical Center–Emma Children’s Hospital, Amsterdam, TheNetherlands; 3Division of Neonatology, La Fe University

4

BPD

FiO2

GA

NICU

SpO2

Methods

and Polytechnic Hospital; Neonatal Research Unit, LaFe Health Research Institute, Valencia, Spain;5Department of Biomechanical Engineering, Faculty ofMechanical, Maritime, and Materials Engineering, DelftUniversity of Technology, Delft, The Netherlands;6Department of Pediatrics, Obstetrics, and ReproductiveMedicine, University of Siena, Siena, Italy; and7Department of Pediatrics, VUUniversityMedical Centre,Amsterdam, The Netherlands

M.V. and J.E. were supported by Instituto Carlos III(Spanish Ministry of Economy and Innovation; FIS PI011/0313). M.L. and G.B were supported by an unrestricted

This double-blind, randomized controlled trial was performed on the neonatalintensive care unit (NICU) of the Erasmus Medical Center–Sophia Children’sHospital, Rotterdam, The Netherlands, a level III NICU. The study protocolwas approved by the Erasmus Medical Center’s Medical Ethics Committee.

Inborn preterm infants of GA <32 weeks were eligible for this study. Determina-tion of GA was based on early fetal ultrasonography or on the date of the last men-

grant from the non-profit research foundation EURAIBI(Europe Against Infant Brain Injury) The Instituto Carlos IIIand EURAIBI foundation had no role in study design, datacollection, data analysis, interpretation of data, or in thewriting of the manuscript. The authors declare no con-flicts of interest.

Registered with TrialRegister.nl: NTR243.

0022-3476/$ - see front matter. Copyright ª 2014 Elsevier Inc.

All rights reserved. http://dx.doi.org/10.1016/j.jpeds.2014.02.019

Bronchopulmonary dysplasia

Fraction of inspired oxygen

Gestational age

Neonatal intensive care unit

Oxygen saturation

1

THE JOURNAL OF PEDIATRICS � www.jpeds.com Vol. -, No. -

strual period. Written informed consent was obtained from allparents antenatally, because the intervention was performeddirectly after birth. Mothers at risk for preterm delivery before32 weeks of gestation were approached for participation.Exclusion criteria included major congenital malformations,chromosome defects, and metabolic or endocrine disorders.

The study design and resuscitation procedures have beendescribed in detail previously.8 In short, the resuscitation unitswith an oxygen blender (Bird Ultrablender; Cardinal Health,Dublin, Ohio) as standard equipment were modified for thisstudy by adding an additional oxygen blender (PM5200; Pre-cision Medical, Northampton, Pennsylvania). This studyblender was randomized to either 30% or 65% oxygen via acomputer-generated list by an individual not involved in theresuscitation. Before the start of each resuscitation, the studyblender was activated by a switch on the resuscitation unit.From that moment, either 30% oxygen (low-oxygen group)or 65% oxygen (high-oxygen group) was provided. The physi-cian remained blinded to group assignment.

ResuscitationAll resuscitations of preterm infants were performed by aneonatologist or fellow. Delayed cord clamping was not prac-ticed in our center during the study period. Immediately aftercord clamping, the infant was placed on the resuscitation unitand resuscitated according to standard guidelines, with stim-ulation and prevention of heat loss. A disposable SpO2 sensor(Nellcor Max-N; Covidien, Dublin, Ireland) was applied tothe infant’s right hand or wrist before the pulse oximeter(Nellcor OxiMax N-600x; Covidien) was activated. Infantswere resuscitated with either a flow inflating bag (JacksonRees modification T-piece breathing system; Intersurgical,Wokingham, United Kingdom) or a T-piece resuscitator(Neopuff; Fisher & Paykel Healthcare, Auckland, New Zea-land), according to the physician’s preference.

The objective of the resuscitation was to keep the pulse ratestable and >100 bpm and to achieve a target SpO2 of 88%-94% at 10 minutes after birth. FiO2 was adjusted before 10minutes of age when SpO2 was >94% (FiO2 was reduced),heart rate was <100 bpm (FiO2 was increased), or SpO2

and/or heart rate dropped. To adjust the FiO2, the physiciandeactivated the research switch, which deactivated the studyblender. Subsequently, FiO2 was supplied via the regular ox-ygen blender, which by default was set to 21% oxygen andcould be manually adjusted to the desired FiO2 withoutknowledge of the initial FiO2 to which the infant had beenrandomized earlier.

Outcome VariablesThe primary study outcome was the incidence of BPD at 36weeks postmenstrual age, diagnosed according to the physi-ological criteria of Walsh et al,9 including a room air chal-lenge test at 36 weeks postmenstrual age. Secondaryoutcomes included resuscitation variables, mortality, sur-vival without BPD, incidence of major neonatal diseases,and markers of oxidative stress. Clinical definitions havebeen reported previously.8

2

Oxidative StressGlutathione concentration and synthesis rates were deter-mined on postnatal day 2 using a stable isotope study asdescribed previously.10,11 Non–protein-bound iron concen-tration was assessed in blood samples collected within 24hours after birth and on day 6 of life. Samples were collectedin heparinized microtainers and immediately placed onmelting ice. After centrifugation at 3500 � g for 10 minutes,the plasma fraction was removed from the lower layer andstored at �80�C until further analysis. Plasma samples wereshipped on dry ice to the University of Siena, Siena, Italy,where the non–protein-bound iron was determined asdescribed previously.12

Urinary oxidative stress markers were determined within24 hours after birth and on day 6 of life. Urine was collectedby placing gauze in the infant’s diaper. After centrifugation at2800 � g for 5 minutes, the urine was stored at �20�C untilanalysis. Urine samples were shipped on dry ice to La Fe Uni-versity & Polytechnic Hospital, Valencia, Spain, where theurinary 8-hydroxy-20-deoxyguanosine/2-deoxyguanosine ra-tio (a marker of oxidative damage to DNA), ortho-tyrosine/phenylalanine ratio (a marker of oxidative damage to pro-tein), 3-nitro-tyrosine (a marker of nitrosative damage toprotein), and 3-chlor-tyrosine (a marker of inflammationcaused by free radicals) were determined by high-performance liquid chromatography coupled to mass spec-trometry,13 which was modified for 3-nitro-tyrosine and3-chlor-tyrosine based on the study of Orhan et al.14

Statistical AnalysesCohort studies reported an incidence of BPDof 30% inpreterminfants atGA<32weeks.9,15 For a reductionof 15%, 100 infantsper group were required to provide a power level of 80%with a1-sided type I error of 0.05. Differences between groups wereassessed using theMann-Whitney test for continuousmeasure-ments and the c2 test for categorical measurements (P < .05).All analyses were performed according to an intention-to-treat principle. Data are presented as mean � SD or median(IQR) unless stated otherwise. All analyses were performed us-ing SPSS version 20 (IBM, Armonk, New York).

Results

Out of 781 eligible preterm infants born between August2008 and February 2012, 193 were analyzed in this study(Figure 1; available at www.jpeds.com). The main reasonfor missed inclusion was imminent preterm delivery, whichleft no time to obtain informed consent before birth.Owing to the acute nature of the resuscitation of preterminfants, 67 infants were missed despite consent because theattending neonatologist neglected to include them. The 583infants not included in the study had a median GA of 292/7weeks (IQR, 272/7-30

6/7 weeks) and a median birth weightof 1150 g (IQR, 900-1425 g). After randomization, 5infants were excluded owing to a postnatally diagnosedmajor congenital abnormality. General demographic andobstetric characteristics are presented in Table I.

Rook et al

Table I. Clinical and obstetric baseline characteristics

Low oxygen, initialFiO2 30% (n = 99)

High oxygen, initialFiO2 65% (n = 94)

Female sex, n (%) 55 (56) 51 (54)GA, wk, median (IQR) 285/7 (27

1/7-303/7) 292/7 (26

3/7-304/7)

Birth weight, g,median (IQR)

1013 (820-1280) 1123 (790-1368)

Umbilical cord pH,median (IQR)

7.31 (7.27-7.34) 7.30 (7.25-7.34)

CRIB score, median (IQR)* 2 (1-7) 1 (1-5)Chorioamnionitis, n (%) 13 (13) 15 (16)Antenatal corticosteroiduse, n (%)

99 (100) 94 (100)

Incomplete course, n (%) 7 (7) 10 (11)Vaginal/cesareandelivery, n

31/68 33/61

CRIB, Clinical Risk Index for Babies.*CRIB score indicate the degree of illness.26

Figure 2. FiO2 and SpO2 during the first 15 minutes of resus-citation in preterm infants resuscitated with low (30%) or high(65%) initial FiO2 after birth. A, FiO2 was significantly differentbetween the groups during the first 5 minutes of resuscitation.*P < .01. B, SpO2 was significantly higher in the high-oxygengroup at 3, 4, and 5 minutes of life. *P < .01; #P < .05.

- 2014 ORIGINAL ARTICLES

ResuscitationIn 188 cases (97%), preterm infants required some kind ofrespiratory support immediately after birth. The variationin FiO2 at the start of the study is explained by the fact thatwe also included infants who did not require respiratory sup-port and thus were breathing room air in this analysis. Inaddition, we measured actual FiO2 instead of set FiO2,contributing to the variation. FiO2 was significantly higherin the high-oxygen group during the first 5 minutes of resus-citation (Figure 2, A). In the low-oxygen group, FiO2 wasincreased to a mean of 40% (IQR, 30%-42%) by 7 minutesof age, and in the high-oxygen group, FiO2 was decreasedto a mean of 40% (IQR, 21%-53%) by 11 minutes of age.Median time on the initial FiO2 was 5:29 minutes (IQR,3:18-10:05 minutes) in the low-oxygen group and 3:53minutes (IQR, 2:50-8:36 minutes) in the high-oxygengroup, which was not statistically different between groups.In both groups, approximately 26% of the infants receivedroom air at transport to the NICU (Table II).

At 3, 4, and 5minutes, SpO2 was higher in the high-oxygengroup (P < .05; Figure 2, B), and subsequently the time toachieving an SpO2 of 88% was shorter in this group (P <.001; Table II). Other general resuscitation variables, suchas Apgar scores, were not significantly different between thegroups (Table II).

Clinical OutcomesThe incidence of BPD was not significantly different in thelow-oxygen and high-oxygen groups (Table II). Moreover,differences in initial FiO2 were not associated with any ofthe secondary clinical outcomes. A subgroup analysis withinfants of GA #28 weeks revealed no differences in theincidence of BPD (P = .50), mortality (P = .39), or survivalwithout BPD (P = .65) between the 2 groups.

Oxidative StressGlutathione concentration and synthesis were not signifi-cantly different in the 2 groups (Figure 3; available at www.jpeds.com). In addition, there were no significantdifferences in non–protein-bound iron or urinary oxidative

Resuscitation of Preterm Infants with Different Inspired Oxygen F

stress markers on day 1 or day 6 of life (Figure 4; availableat www.jpeds.com).

Discussion

For years, there has been debate about the optimal concentra-tion of oxygen for the resuscitation of preterm infants afterbirth, because both hyperoxia and hypoxia can have detri-mental effects. The present study demonstrates that resuscita-tion with an initial FiO2 of 30% is safe, but is not associatedwith a reduced risk of BPD or decreased oxidative stresscompared with an initial FiO2 of 65%. The most immaturepreterm infants are more likely to need supplemental oxygenbecause of their immature lungs,16 butmight also bemore sus-ceptible to oxidative damage; thus, we performed a subgroupanalysis of preterm infants of #28 weeks GA, and found nodifference in the incidence of BPD in this subgroup as well.

ractions 3

Table II. Resuscitation variables and clinical outcomes

Low oxygen, initialFiO2 30% (n = 99)

High oxygen, initialFiO2 65% (n = 94)

ResuscitationApgar score at 5 min, median

(IQR)8 (7-9) 9 (8-9)

Time to SpO2 88%, min,median (IQR)

5:45 (3:49-7:51) 3:14 (2:08-5:24)*

Intubation in delivery room,n (%)

31 (31) 28 (30)

Received room air at transferto NICU, n (%)

24 (24) 27 (28)

Time to transfer to NICU, min,median (IQR)

20 (17-28) 21 (16-27)

Clinical outcomeBPD, n (%)† 23 (24) 14 (17)Mortality, n (%) 6 (6) 10 (11)Survival without BPD, n (%) 71 (72) 70 (75)Patent ductus arteriosus,

n (%)Pharmacologically treated 28 (28) 23 (24)Surgically treated 7 (7.1) 5 (5.3)

Necrotizing enterocolitisgrade $2, n (%)

4 (4.0) 3 (3.2)

Retinopathy of prematuritygrade $2, n (%)

6 (6.1) 5 (5.3)

Intraventricular hemorrhagegrade $2, n (%)

8 (8.1) 10 (10.5)

Duration of oxygen treatment,d, median (IQR)

2 (0-31) 2 (0-23)

Duration of mechanicalventilation, d, median (IQR)

2 (0-12) 1 (0-9)

Surfactant treatment, n (%) 45 (45) 41 (43)

*P < .001.†BPD was diagnosed according to the physiological criteria described by Walsh et al.9

THE JOURNAL OF PEDIATRICS � www.jpeds.com Vol. -, No. -

Our results somewhat resemble those reported by Ventoet al7 which showed that resuscitation with an initial FiO2

of 30% decreased oxidative stress and reduced the incidenceof BPD compared with an FiO2 of 90% in infants at <28weeks GA. In that study, the FiO2 required individual adjust-ment in both groups to a mean FiO2 of 55% by 5 minutes ofage. In the present study, FiO2 was adjusted in both groupsbased on individual SpO2 recordings. This resulted in amean FiO2 of �40% after 6 minutes in both groups, whichremained relatively constant during the entire course ofresuscitation in most cases. Differences in the adjusted oxy-gen dose between the 2 studies cannot be explained by differ-ences in target SpO2, because our target was higher. A morelikely explanation is the higher GA, and hence more fullydeveloped lungs, of our infants. Considering these datatogether, we can conclude that the optimal dose lies between30% and 65% and should not be as high as 90%, and that in-dividual monitoring and adjustments are necessary.

The amount of oxygen administered differed significantlybetween our low-oxygen and high-oxygen groups during thefirst 5 minutes. Because the newborn lungs take some time tobecome aerated, the oxygen dose that actually enters thelungs is unknown; however, the difference in oxygen admin-istered did result in increased SpO2 levels in the high-oxygengroup. Although we found no differences in short-term clin-ical outcome ormarkers of oxidative stress, this may still havesubtle effects on the long-term outcome.

4

We acknowledge the following limitations of this study.First, the study was not powered to detect differences in mor-tality. In developed countries, mortality is relatively low inpreterm infants of GA <32 weeks, with reported mortalityrates varying between 10% and 13%.17,18 There was a trendtoward higher mortality in the high-oxygen group, butmore studies with larger numbers of infants are needed todraw a firm conclusion. An even more relevant outcome issurvival without BPD, because those infants who died arealso those who might have been at increased risk, but didnot have the opportunity to develop BPD.Second, the generalizability of our results for daily practice

is hampered by the selection of cases for which sufficient timewas available to obtain antenatal informed consent. Richet al19 reported that eligible but nonenrolled infants haddifferent demographics (eg, less exposure to antenatal ste-roids) and worse clinical outcomes compared with enrolledinfants. In our study, all enrolled infants received at least 1dose of antenatal steroids, which resulted in a selectionbias. Antenatal steroids are indicated for the prevention ofrespiratory distress and are associated with increased antiox-idant enzyme activity, which reduces susceptibility to oxida-tive damage as a result of hyperoxia.20 Thus, our cases mighthave had less morbidity (eg, less BPD) than the general pop-ulation of very preterm infants. On the other hand, the ma-jority of included infants were born prematurely because ofpreeclampsia (n = 73) and chorioamnionitis (n = 23), whichare associated with increased oxidative stress.21,22 The oxida-tive burden in these infants might already be so high that thedifferences in oxygen supplementation between our groupsmight not have had a significant influence on the oxidativeburden. Future studies should use a waiver of consent toinclude preterm infants after imminent delivery.Third, the blinded supply of oxygen was administered for a

median of only 5 minutes, until the caregivers decided tochange the oxygen dose. One could question whether thisrelatively short intervention could lead to any relevantchange in outcome. On the other hand, previous studieshave demonstrated a significant impact of interventions inFiO2 for only limited durations.1,7 Another limitation ofour study design is the inability to measure ventilation vari-ables, such as applied pressure, tidal volume, and respiratoryrate. Thus, we have no data on the total supplied oxygen loador on possible differences in vigor of ventilation.The main strengths of this study are the large number of

included infants, the study of initial FiO2 in a double-blindtrial, and the inclusion of very preterm infants in a trialwith an acute intervention. The blinded setting ensuredthat FiO2 adjustments during resuscitation were not influ-enced by knowledge of the initial FiO2 but instead were basedsolely on the infant’s condition. Moreover, we studied well-known and reliable biomarkers of oxidative stress, whichsupported our clinical results.Adequate ventilation is important for oxygen supplemen-

tation. Without appropriate aeration of the lung (ie, estab-lishment of functional residual capacity), oxygensupplementation will have no effect and may result in

Rook et al

- 2014 ORIGINAL ARTICLES

overexposure of the lung to oxygen.16 Ventilation of preterminfants is difficult in the immediate postnatal phase, however,owing to immaturity and low compliance of the lungs.Furthermore, preterm infants lack sufficient pulmonary anti-oxidant defenses to overcome the sudden exposure to oxy-gen,23,24 making them more susceptible to oxidative lungdamage. A recent meta-analysis comparing low and high ox-ygen concentrations for the resuscitation of preterm infantsshowed a possible reduction in mortality after resuscitationwith low oxygen concentrations25; however, this effect disap-peared when only studies with concealed allocation wereincluded, indicating the need for more well-conductedstudies. Adding our own data to this meta-analysis, andalmost doubling the number of infants studied, does not alterthe conclusion with regard to an immediate effect on mortal-ity. Despite the lack of effect on clinical outcomes, it seemspreferable to initiate oxygen supplementation after birthwith the lowest safe oxygen concentration.

In the present study, we have demonstrated that resuscita-tion of preterm infants with 30% oxygen is as safe as resusci-tation with 65% oxygen, but offers no benefits with regard tothe risk of BPD or biomarkers of oxidative stress. However,the amount of oxygen administered to the infants in thelow-oxygen group was significantly lower for several mi-nutes. Planned long-term follow-up of these infants willreveal whether subtle differences not discerned in theneonatal period might result in different long-term out-comes. Based on the present data, we advise starting theresuscitation of preterm infants of GA <32 weeks with aninitial FiO2 of 30%. FiO2 should be subsequently adjustedbased on SpO2 measurements according to the approachproposed by Dawson et al.16 n

Submitted for publication Oct 7, 2013; last revision received Dec 20, 2013;

accepted Feb 5, 2014.

Reprint requests: Johannes B. van Goudoever, PhD, AMC-Emma Children’s

Hospital, Department of Pediatrics, Room H7-276, PO Box 22660, 1100 DD

Amsterdam, The Netherlands. E-mail: h.vangoudoever@amc.uva.nl

References

1. Saugstad OD, Ramji S, Soll RF, Vento M. Resuscitation of newborn in-

fants with 21% or 100% oxygen: an updated systematic review andmeta-

analysis. Neonatology 2008;94:176-82.

2. Tan A, Schulze A, O’Donnell CP, Davis PG. Air versus oxygen for

resuscitation of infants at birth. Cochrane Database Syst Rev

2005;CD002273.

3. Kattwinkel J, Perlman JM, Aziz K, Colby C, Fairchild K, Gallagher J, et al.

Part 15: neonatal resuscitation. 2010 AmericanHeart Association Guide-

lines for Cardiopulmonary Resuscitation and Emergency Cardiovascular

Care. Circulation 2010;122:S909-19.

4. Perlman JM, Wyllie J, Kattwinkel J, Atkins DL, Chameides L,

Goldsmith JP, et al. Part 11: neonatal resuscitation. 2010 International

Consensus on Cardiopulmonary Resuscitation and Emergency Cardio-

vascular Care Science with Treatment Recommendations. Circulation

2010;122:S516-38.

5. Wang CL, Anderson C, Leone TA, Rich W, Govindaswami B, Finer NN.

Resuscitation of preterm neonates by using room air or 100% oxygen.

Pediatrics 2008;121:1083-9.

6. Escrig R, Arruza L, Izquierdo I, Villar G, Saenz P, Gimeno A, et al.

Achievement of targeted saturation values in extremely low gestational

Resuscitation of Preterm Infants with Different Inspired Oxygen F

age neonates resuscitated with low or high oxygen concentrations: a pro-

spective, randomized trial. Pediatrics 2008;121:875-81.

7. Vento M, Moro M, Escrig R, Arruza L, Villar G, Izquierdo I, et al. Pre-

term resuscitation with low oxygen causes less oxidative stress, inflam-

mation, and chronic lung disease. Pediatrics 2009;124:e439-49.

8. Rook D, Schierbeek H, van der Eijk A, Longini M, Buonocore G,

Vento M, et al. Resuscitation of very preterm infants with 30% vs 65%

oxygen at birth: study protocol for a randomized controlled trial. Trials

2012;13:65.

9. Walsh MC, Yao Q, Gettner P, Hale E, Collins M, Hensman A, et al.

Impact of a physiologic definition on bronchopulmonary dysplasia rates.

Pediatrics 2004;114:1305-11.

10. Schierbeek H, Te Braake F, Godin JP, Fay LB, van Goudoever JB. Novel

method formeasurement of glutathione kinetics in neonates using liquid

chromatography coupled to isotope ratio mass spectrometry. Rapid

Commun Mass Spectrom 2007;21:2805-12.

11. Schierbeek H, Rook D, te Braake FW, Dorst KY, Voortman G, Godin JP,

et al. Simultaneous analysis of (13)C-glutathione as its dimeric form

GSSG and its precursor [1-(13)C]glycine using liquid chromatog-

raphy/isotope ratio mass spectrometry. Rapid Commun Mass Spectrom

2009;23:2897-902.

12. Paffetti P, Perrone S, Longini M, Ferrari A, Tanganelli D, Marzocchi B,

et al. Non–protein-bound iron detection in small samples of biological

fluids and tissues. Biol Trace Elem Res 2006;112:221-32.

13. Solberg R, Andresen JH, Escrig R, Vento M, Saugstad OD. Resuscitation

of hypoxic newborn piglets with oxygen induces a dose-dependent in-

crease in markers of oxidation. Pediatr Res 2007;62:559-63.

14. Orhan H, Vermeulen NP, Tump C, Zappey H, Meerman JH. Simulta-

neous determination of tyrosine, phenylalanine and deoxyguanosine

oxidation products by liquid chromatography-tandem mass spectrom-

etry as non-invasive biomarkers for oxidative damage. J Chromatogr B

Analyt Technol Biomed Life Sci 2004;799:245-54.

15. Ehrenkranz RA,WalshMC, Vohr BR, Jobe AH,Wright LL, Fanaroff AA,

et al. Validation of the National Institutes of Health consensus definition

of bronchopulmonary dysplasia. Pediatrics 2005;116:1353-60.

16. Dawson JA, Vento M, Finer NN, Rich W, Saugstad OD, Morley CJ, et al.

Managing oxygen therapy during delivery room stabilization of preterm

infants. J Pediatr 2011;160:158-61.

17. Guerrot AM, Chadie A, Torre S, Rondeau S, Pinto Cardoso G, Abily-

Donval L, et al. Compared outcomes of very preterm infants born in

2000 and 2005. Acta Paediatr 2012;101:731-5.

18. Horbar JD, Carpenter JH, Badger GJ, Kenny MJ, Soll RF, Morrow KA,

et al. Mortality and neonatal morbidity among infants 501 to 1500 grams

from 2000 to 2009. Pediatrics 2012;129:1019-26.

19. Rich W, Finer NN, Gantz MG, Newman NS, Hensman AM, Hale EC,

et al. Enrollment of extremely low birth weight infants in a clinical

research study may not be representative. Pediatrics 2012;129:480-4.

20. Vento M, Aguar M, Escobar JJ, Arduini A, Escrig R, Brugada M, et al.

Antenatal steroids and antioxidant enzyme activity in preterm in-

fants: influence of gender and timing. Antioxid Redox Signal 2009;

11:2945-55.

21. Raijmakers MT, Roes EM, Poston L, Steegers EA, Peters WH. The tran-

sient increase of oxidative stress during normal pregnancy is higher and

persists after delivery in women with pre-eclampsia. Eur J Obstet Gyne-

col Reprod Biol 2008;138:39-44.

22. Cheah FC, Pillow JJ, Kramer BW, Polglase GR, Nitsos I, Newnham JP,

et al. Airway inflammatory cell responses to intra-amniotic lipopolysac-

charide in a sheep model of chorioamnionitis. Am J Physiol Lung Cell

Mol Physiol 2009;296:L384-93.

23. Frank L, Sosenko IR. Prenatal development of lung antioxidant enzymes

in four species. J Pediatr 1987;110:106-10.

24. Frank L, Sosenko IR. Failure of premature rabbits to increase antioxidant

enzymes during hyperoxic exposure: increased susceptibility to pulmo-

nary oxygen toxicity compared with term rabbits. Pediatr Res 1991;29:

292-6.

25. Brown JV, Moe-Byrne T, Harden M, McGuire W. Lower versus higher

oxygen concentration for delivery room stabilisation of preterm neo-

nates: systematic review. PloS ONE 2012;7:e52033.

ractions 5

Figure 1. Trial profile.

THE JOURNAL OF PEDIATRICS � www.jpeds.com Vol. -, No. -

5.e1 Rook et al

Figure 3. Glutathione concentration and synthesis rates on day 2 in preterm infants resuscitated with low (30%) or high (65%)initial FiO2 after birth. Data are expressed as boxplots, with the box representing the median and IQR and the whiskers repre-senting the range. Glutathione concentration and synthesis rates were not different between the low-oxygen group (n = 15) andhigh-oxygen group (n = 13). ASR, absolute synthesis rate; FSR, fractional synthesis rate.

- 2014 ORIGINAL ARTICLES

Resuscitation of Preterm Infants with Different Inspired Oxygen Fractions 5.e2

Figure 4. Oxidative stressmarkers on day 0 and day 6 in preterm infants resuscitatedwith a low (30%) or high (65%) initial FiO2

after birth. Data are expressed as boxplots, with the box representing the median and IQR and the whiskers representing therange.A,Plasma non–protein-bound iron did not differ between the 2 groups (low-oxygen group: n = 28 on day 0, n = 22 on day6; high-oxygen group: n = 19 on day 0, n = 19 on day 6). B, Urinary oxidative stress markers also did not differ between the 2groups (low-oxygen group: n = 41 on day 0, n = 46 on day 6; high-oxygen group: n = 34 on day 0, n = 34 on day 6). 8-OhdG/2dG,8-hydroxy-20-deoxyguanosine/2-deoxyguanosine ratio; O-Tyr/Phenyl, ortho-tyrosine/phenylalanine ratio; 3-Cl-Tyr, 3-chlor-tyrosine; 3-N-Tyr, 3-nitro-tyrosine.

THE JOURNAL OF PEDIATRICS � www.jpeds.com Vol. -, No. -

5.e3 Rook et al