antenatal magnesium sulfate for the prevention of cerebral
TRANSCRIPT
ApaA
Cmausbdimtprspbnapayqs
FCKHIaDMS
Ra
R
SRSHHS
0©d
www.AJOG.org Obstetrics Reviews
ntenatal magnesium sulfate for the prevention of cerebralalsy in preterm infants less than 34 weeks’ gestation:systematic review and metaanalysis
gustín Conde-Agudelo, MD, MPH; Roberto Romero, MD
lirt
cma32Mibtdtdc
pmtdbNcuwe
clppc9ssttrhtap
fdsnAtnwbdscb
erebral palsy describes a group ofdisorders affecting the develop-
ent of movement and posture, causingctivity limitation, and which are attrib-ted to nonprogressive disturbances. In-ults responsible for cerebral palsy areelieved to have occurred during fetalevelopment or infancy.1 Cerebral palsy
s the most prevalent chronic childhoodotor disability with an estimated life-
ime cost in 2003 of nearly $1 million pererson.2 The prevalence of cerebral palsyeported in recent population-basedtudies ranges between 1.5 and 3.6 caseser 1000 live births.3-5 The United Cere-ral Palsy Foundation estimates thatearly 800,000 children and adults of allges in the United States have cerebralalsy.6 Secular trends in the overall prev-lence of cerebral palsy over the last 40ears show a modest increase in the fre-uency that has been attributed to a sub-tantial increase in cerebral palsy in very
rom the Perinatology Research Branch (Drsonde-Agudelo and Romero), Euniceennedy Shriver National Institute of Childealth and Human Development/National
nstitutes of Health/Department of Healthnd Human Services, Bethesda, MD, andetroit, MI; and the Center for Molecularedicine and Genetics (Dr Romero), Wayne
tate University, Detroit, MI.
eceived March 5, 2009; revised April 2, 2009;ccepted April 15, 2009.
eprints not available from the authors.
upported was provided by the Intramuralesearch Program of the Eunice Kennedyhriver National Institute of Child Health anduman Development, National Institutes ofealth, Department of Health and Humanervices.
002-9378/$36.002009 Mosby, Inc. All rights reserved.
oi: 10.1016/j.ajog.2009.04.005
wSee related editorial, page 589
ow-birthweight infants, which, in turn,s attributable to their increased survivalesulting from improvements in neona-al intensive care.3,5
Preterm birth is a major risk factor forerebral palsy, and the risk increasesarkedly with decreasing gestational
ge.7 Currently, infants born at less than4 weeks of gestation constitute about5% of all new cases of cerebral palsy.5,8
ultiple pregnancy is associated with anncreased risk of cerebral palsy. This haseen attributed, at least in part, to pre-erm birth. However, fetal or neonataleath of a member of a multiple gesta-ion, twin-to-twin transfusion syn-rome, or intrapartum problems alsoontribute.9,10
Several observational studies have re-orted an association of antenatal treat-ent with magnesium sulfate for pre-
erm labor or preeclampsia with aecreased risk of cerebral palsy in low-irthweight or preterm infants. In 1995,elson and Grether11 reported a case-
ontrol study investigating whether intero exposure to magnesium sulfateas used to prevent convulsions in pre-
clampsia or whether a tocolytic agent
We conducted a systematic review and medetermine whether magnesium sulfate admbefore 34 weeks of gestation may reduce ttrials involving 4796 women and 5357 isulfate was associated with a significant rerisk [RR], 0.69; 95% confidence interval [palsy (RR, 0.64; 95% CI, 0.44-0.92), and s95% CI, 0.43-0.83). There was no overall d(RR, 1.01; 95% CI, 0.89-1.14). Minor sidreceiving magnesium sulfate. In conclusionat risk of delivery before 34 weeks of ges
Key words: cerebral palsy, infant developmneurologic handicap, pediatric mortality, pr
as associated with a lower prevalence of a
JUNE 2009 Am
erebral palsy in infants born weighingess than 1500 g. Children with cerebralalsy were less likely to have been ex-osed to magnesium sulfate than wereontrol subjects (odds ratio [OR], 0.14;5% confidence interval [CI], 0.05-0.51)uggesting a protective effect of magne-ium sulfate against cerebral palsy inhese very low-birthweight infants. Al-hough some observational studies haveeported similar findings,12-15 othersave reported no association betweenhe administration of magnesium sulfatend the subsequent risk of cerebralalsy.16-21
In response to the conflicting evidencerom observational studies, several ran-omized controlled trials of magnesiumulfate administered to mothers for fetaleuroprotection have been performed.recent systematic review that assessed
he effects of magnesium sulfate as a fetaleuroprotective agent when given toomen considered at risk of pretermirth concluded that this therapy re-uced the risk of cerebral palsy and sub-tantial gross motor dysfunction in earlyhildhood.22 This review, however,ased its main conclusions on meta-
nalysis of randomized controlled trials totered to women at risk of preterm deliveryisk of cerebral palsy in their children. Six
nts were included. Antenatal magnesiumtion in the risk of cerebral palsy (relative
, 0.55-0.88), moderate or severe cerebraltantial gross motor dysfunction (RR, 0.60;rence in the risk of total pediatric mortalityffects were more frequent among womenagnesium sulfate administered to womenn reduces the risk of cerebral palsy.
t, magnesium sulfate, metaanalysis,aturity, preterm birth, systematic review
taainishe rnfaducCI]ubsiffee e, m
tatio
enem
nalysis that included preterm infants
erican Journal of Obstetrics & Gynecology 595
ldtoc
mcamptb
MTtapgi
SWa3w3oic(RctauknttmlvaNAdmtpwtp
SWapwwwniwpsaw““
wdct
OTcibccippfcophcnd3mtfam8idcadses
vta
SWup3fbpnsnitstdepamrrdotdscpdoqqvpiTemwvas
DOsttrd
Reviews Obstetrics www.AJOG.org
5
ess than 37 weeks of gestational age andid not perform any economic evalua-ion for estimating the cost-effectivenessf this intervention for the prevention oferebral palsy.
We carried out a systematic review andetaanalysis of all available randomized
ontrolled trials to determine the efficacynd safety of antenatal administration ofagnesium sulfate to women at risk of
reterm delivery before 34 weeks of ges-ational age for the prevention of cere-ral palsy in their children.
aterials and methodshe systematic review was conducted af-
er a prospectively prepared protocolnd reported using the Quality of Re-orting of Metaanalysis (QUOROM)uidelines for metaanalysis of random-zed controlled trials.23
earche searched PubMed, Embase, Cinahl,
nd Lilacs (all from inception to March1, 2009), ISI Web of Science (http://ww.isiknowledge.com) (1960 to March1, 2009), the Cochrane Central Registerf Controlled Trials (http://www.mrw.nterscience.wiley.com/cochrane/ochrane_clcentral_articles_fs.html)1960 to March 31, 2009), and Researchegisters of ongoing trials (www.linicaltrials.gov, www.controlled-rials.com, www.centerwatch.com, www.ctr.org.au, www.nrr.nhs.uk, and www.min.ac.jp/ctr) using a combination ofey words and text words related to mag-esium, cerebral palsy, and neuroprotec-ion. Proceedings of the Society for Ma-ernal-Fetal Medicine and international
eetings on cerebral palsy, referenceists of identified studies, textbooks, pre-iously published systematic reviews,nd review articles were also searched.o language restrictions were applied.ll searches were carried out indepen-ently by the 2 authors and results wereerged. For studies that resulted in mul-
iple publications, the data from theublication with the largest sample sizeere used and supplemented if addi-
ional information appeared in the other
ublications. s96 American Journal of Obstetrics & Gynecology
tudy selectione included randomized controlled tri-
ls comparing magnesium sulfate withlacebo or no magnesium sulfate foromen at risk of preterm birth before 34eeks of gestation, whose primary aimas to prevent cerebral palsy and othereurologic abnormalities in the unborn
nfant or if the primary aim was other-ise but data on cerebral palsy were re-orted for the infants. Quasirandomizedtudies were excluded. We classified tri-ls according to aim of the treatmentith magnesium sulfate into 2 groups:
neuroprotection of the fetus” andother aims.”
All published studies deemed suitableere retrieved and reviewed indepen-ently by the 2 authors to determine in-lusion. Disagreements were resolvedhrough consensus.
utcome measureshe primary outcomes of interest wereerebral palsy and total pediatric mortal-ty (fetal death � any mortality of liveirths that occurred during the 2 years oforrected age after the birth). We in-luded total pediatric mortality becauset is a competing outcome that wouldreclude the assessment of cerebralalsy. Prespecified secondary outcomes
or the neonate and infant included milderebral palsy, any intraventricular hem-rrhage, moderate or severe cerebralalsy, grade III or IV intraventricularemorrhage, periventricular leukomala-ia, Apgar score � 7 at 5 minutes, neo-atal seizures, respiratory distress syn-rome, need for supplemental oxygen at6 weeks, bronchopulmonary dysplasia,echanical ventilation, necrotizing en-
erocolitis, substantial gross motor dys-unction, major neurologic disability,ny neurologic impairment, Bayleyental development index � 70 and �
5, Bayley psychomotor developmentndex � 70 and � 85, blindness, andeafness. For the mother, secondary out-omes were death, cardiac or respiratoryrrest, pulmonary edema, respiratoryepression, hypotension, tachycardia,evere postpartum hemorrhage, cesar-an delivery, and clinical and self-as-essed maternal side effects of the infu-
ion such as flushing, nausea or bJUNE 2009
omiting, sweating, problems at injec-ion site, stopping of infusion because ofdverse effects, and any side effect.
tudy quality assessmente assessed study methodologic quality
sing a modified scoring system pro-osed by Jadad et al,24 which is based onitems: randomization, blinding, and
ollow-up. Points were awarded on theasis of the quality of randomization (2oints: computer-generated randomumbers or similar; 1 point: not de-cribed; 0 points: quasirandomized orot randomized [we excluded such stud-
es]); double blinding (2 points: neitherhe person doing the assessments nor thetudy participant could identify the in-ervention being assessed; 1 point: notescribed; 0 points: no blinding or inad-quate method), and follow-up (2oints: number or reasons for dropoutsnd withdrawals described, and assess-ent of primary outcomes in � 95% of
andomized fetuses; 1 point: number oreasons for dropouts and withdrawalsescribed but assessment of primaryutcomes in � 95% of randomized fe-uses; 0 points: number or reasons forropouts and withdrawals not de-cribed). In addition, we assessed con-ealment of allocation as follows: 2oints: adequate method (central ran-omization; or drug containers orpaque, sealed envelopes that were se-uentially numbered and opened se-uentially only after they had been irre-ersibly assigned to the participant); 0oints: no concealment of allocation or
nadequate method or not described.hus, the total score ranged from 0 (low-st quality) to 8 (highest quality). Theethodologic quality of included trialsas assessed individually by the 2 re-iewers who were not associated withny of the trials. When differences incoring existed, a consensus was reached.
ata abstractionne reviewer (A.C.-A.) scanned ab-
tracts and titles. Potentially relevant ar-icles were acquired and data were ex-racted in duplicate from all reports andecorded on a piloted form indepen-ently by the 2 reviewers. There was no
linding by authorship. All outcomednopbuaindtmltaeitpacptbabas
SSccotcpfamrCswveoidissptw
epcfvo9wcwtmgataeng
btialmmmca
tctCe9dpR
N
Ibw3ttctw3
twe
oAdadnaagtpScgofaa9ecin
ecinroppsSvrmmaNbaf(K
RT1tl
www.AJOG.org Obstetrics Reviews
ata were further verified with the origi-al articles. Information was extractedn study characteristics (randomizationrocedure, allocation concealment,lind assessment at baseline and follow-p, follow-up period, intention-to-treatnalysis, and losses to follow-up), partic-pants (inclusion and exclusion criteria,umbers of mothers and infants in ran-omized groups, baseline characteris-ics, and country and date of recruit-
ent), details of intervention (aim,oading dose, maintenance dose, medianotal dose, duration, and retreatment),nd outcomes (number of outcomevents including mortality and morbid-ty). The number of infants was used ashe denominator for primary (cerebralalsy and pediatric mortality), neonatal,nd infant neurodevelopmental out-omes. When maternal outcomes wereresented, numerators and denomina-ors were calculated based on the num-er of mothers. In an attempt to obtaindditional data, we contacted 3 authorsy e-mail of whom 1 responded. Dis-greements in extracted data were re-olved by discussion among reviewers.
tatistical analysistatistical analysis was performed ac-ording to the guidelines of the Co-hrane Collaboration.25 We analyzedutcomes on an intend-to-treat basis. Ifhis was not clear from the original arti-le, then we carried out reanalysis whereossible. If data for similar outcomes
rom 2 or more separate studies werevailable, we combined the data in aetaanalysis and calculated a summary
elative risk (RR) with associated 95%I. Heterogeneity of the results among
tudies was tested with the quantity I2,hich describes the percentage of totalariation across studies that is due to het-rogeneity rather than chance.26 A valuef 0% indicates no observed heterogene-
ty, whereas I2 values of 50% or more in-icate a substantial level of heterogene-
ty.26 We planned to pool data acrosstudies using a fixed-effects model if sub-tantial statistical heterogeneity was notresent. We used random-effects modelso pool data across studies if the I2 values
ere � 50%. tWe conducted sensitivity analyses toxplore the robustness of findings for therimary outcomes according to statisti-al model (fixed effects vs random ef-ects), modified Jadad quality score (� 4s � 4), and completeness of follow-upf randomized fetuses (� 95% vs �5%). Additional subgroup analysesere planned to assess the primary out-
omes by primary aim of the treatmentith magnesium sulfate (neuroprotec-
ive vs other aims), median total dose ofagnesium sulfate used (� 4 g vs � 4 g),
estational age at trial entry (� 34, � 32,nd � 30 weeks), and plurality (single-on and multiple pregnancy). The meta-nalysis by plurality of pregnancy, how-ver, was not possible because there wereot sufficient data available from thereat majority of studies.We assessed publication and related
iases visually by examining the symme-ry of funnel plots and statistically by us-ng the Egger test.27 The larger the devi-tion of the intercept of the regressionine from zero, the greater was the asym-
etry and the more likely it was that theetaanalysis would yield biased esti-ates of effect. As suggested by Egger, we
onsidered P � .1 to indicate significantsymmetry.
We also calculated the number neededo treat (NNT) for an additional benefi-ial outcome and the NNT for an addi-ional harmful outcome with their 95%Is for outcomes in which the treatment
ffect was significant at the 5% level (the5% CI for the absolute risk differenceid not include zero).28 NNT was com-uted from the results of metaanalysis ofRs as follows:
NT ��1 ⁄ Control group event rate
� (1 � relative risk)�n this review, NNT for an additionaleneficial outcome is the number ofomen at risk of preterm delivery before4 weeks of gestation who need to bereated with magnesium sulfate ratherhan with placebo to prevent 1 case oferebral palsy. The NNT for an addi-ional harmful outcome is the number ofomen at risk of preterm delivery before4 weeks of gestation who need to be
reated with magnesium sulfate rather cJUNE 2009 Am
han with placebo for 1 additionaloman to be harmed by an adverse
vent.We estimated the hypothetical impact
f universal use of magnesium sulfate inmerican women at high risk of pretermelivery before 34 weeks of gestationalge to prevent cerebral palsy in their chil-ren. Initially, we calculated the totalumber of new cases of cerebral palsy di-gnosed each year in the United Statesmong infants born before 34 weeks ofestational age using published data onhe total number of new cases of cerebralalsy recognized each year in the Unitedtates and the percentage of infants witherebral palsy born before 34 weeks ofestational age. Then, the total numberf new cases of cerebral palsy among in-ants born before 34 weeks of gestationalge was multiplied by the summary RRnd the upper and lower bounds of its5% CI obtained in our metaanalysis tostimate the hypothetical number of newases of cerebral palsy with correspond-ng 95% CI that could be prevented an-ually using this intervention.Finally, we performed an economic
valuation to calculate the incrementalost of preventing 1 case of cerebral palsyn the United States through use of ante-atal magnesium sulfate in women atisk of preterm delivery before 34 weeksf gestational age. First, we searched dataublished recently on the total cost peratient of administrating magnesiumulfate as a tocolytic agent in the Unitedtates. Then, the incremental cost of pre-enting 1 case of cerebral palsy (with cor-esponding 95% CI) by using antenatalagnesium sulfate was estimated byultiplying the total cost per patient of
dministering magnesium sulfate by theNT for benefit and the upper and lowerounds of its 95% CI obtained in ournalyses. All statistical analyses were per-ormed with the StatsDirect version 2.7.2StatsDirect Ltd, Cheshire, Unitedingdom).
esultshe flow of the search is shown in Figure. Of the 331 potentially relevant cita-ions identified, 5 studies (6 trials) pub-ished in 7 articles fulfilled the inclusion
riteria after a detailed review of 92 stud-erican Journal of Obstetrics & Gynecology 597
ierasarsfsbaifecTw
aw1Zctgfw(pmddtwaaeaiwbwastpchbri61at
(eMbriwaawescpMnapl
scnRintbMowadnlcwsh
hnfwatttnntoait
i4
stwmdtodubmawwcliamthqlawrwffnfacfa4rtmMl
PSpcpsrgsd
Reviews Obstetrics www.AJOG.org
5
es.29-35 Because the study by Mittendorft al29,30 had 2 arms (tocolytic and neu-oprotective), it was considered as 2 sep-rate trials in this review. Of the other 4tudies included, 3 studies31,33,35 evalu-ted magnesium sulfate as an infant neu-oprotective agent, and 1 study32 as-essed the efficacy of magnesium sulfateor preventing eclampsia. Eighty-seventudies were excluded, the main reasoneing the lack of data on cerebral palsynd/or pediatric mortality in random-zed controlled trials on magnesium sul-ate for preventing preterm birth orclampsia. Overall agreement on the in-lusion of studies was 100% (� � 1.00).he 6 trials included a total of 4796omen and 5357 infants.The characteristics of included studies
re presented in Table 1. Two studiesere performed in the United States,29,35
each in France,33 Australia, and Newealand,31 and the remaining study wasonducted in 19 countries across 5 con-inents.32 Trials included women withestational ages � 34 weeks29 (165 in-ants), � 33 weeks33 (688 infants), � 32eeks35 (2444 infants), and � 30 weeks31
1255 infants). The Magpie trial32 re-orted data for 3283 infants whoseothers were randomly assigned before
elivery. Of these, 805 infants were ran-omly assigned before 34 weeks of gesta-ional age and 788 between 34 and 36eeks. Data for the 805 infants randomly
ssigned before 34 weeks of gestationalge included in our metaanalyses werextracted from the review by Doyle etl.22 Multiple pregnancies were includedn all trials. Two trials31,33 includedomen at high risk of preterm deliveryecause of expected or planned birthithin 24 hours. The trial by Rouse et
l35 included women at high risk forpontaneous delivery because of prema-ure rupture of membranes or advancedreterm labor and women with indi-ated preterm delivery within 2-24ours. The main reasons for pretermirth were preterm labor and prematureupture of the membranes (63% and 9%n the study by Crowther et al,31 85% and1% in the study by Marret et al,33 and0% and 87% in the study by Rouse etl,35 respectively), followed by antepar-
um hemorrhage and chorioamnionitis m98 American Journal of Obstetrics & Gynecology
14% and 14% in the study by Crowthert al,31 and 19% and 11% in the study by
arret et al,33 respectively). In the studyy Crowther et al,31 15% of women en-olled had preeclampsia. One study onlyncluded women in active preterm laborith cervical dilation � 4 cm (tocolytic
rm) or � 4 cm (neuroprotectiverm).29 The Magpie trial32 included onlyomen with preeclampsia. Three trials
xcluded women with preeclamp-ia.29,33,35 The definition and diagnosticriteria of cerebral palsy were clearly re-orted in 4 trials.31,32,34,35 The study byittendorf et al30 did not report a defi-
ition or details of the diagnosis. A pedi-trician made the diagnosis of cerebralalsy in all trials at a corrected age of at
east 18 months30,32 or 24 months.31,34,35
The studies included a range of dosingchedules. In 5 trials,29,31-33 women re-eived a loading dose of 4 g of intrave-ous magnesium sulfate. In the trial byouse et al,35 women received a loading
nfusion of 6 g. There was no mainte-ance infusion of magnesium sulfate in 2
rials (neuroprotective arm of the studyy Mittendorf et al29 and the study byarret el al33). The maintenance doses
f magnesium sulfate used in 4 trialsere 1 g/h,31 2 g/h,35 2-3 g/h (tocolytic
rm of the study in the study by Mitten-orf et al29), and either 1 g/h intrave-ously or 5 g every 4 hours intramuscu-
arly.32 The maintenance infusion wasontinued until birth (if it occurredithin 24 hours) or up to 24 hours in the
tudy by Crowther et al,31 or until 24ours in the Magpie trial.32
In the study by Rouse et al,35 if deliveryad not occurred after 12 hours and waso longer considered imminent, the in-
usion was discontinued and resumedhen delivery was deemed imminent
gain. If at least 6 hours had passed sincehe discontinuation of the study medica-ion, another loading dose was given. Re-reatment with magnesium sulfate wasot allowed in 3 trials31-33 and in theeuroprotective arm of the study by Mit-
endorf et al.30 There was no informationn duration of maintenance infusionnd whether retreatment was permittedn the tocolytic arm of the study by Mit-endorf et al.30 The median total dose of
agnesium sulfate received by women v
JUNE 2009
n the magnesium group ranged betweenand 49.8 g.The quality assessment of trials is
hown in Table 2. The overall quality ofhe trials was relatively good. All studiesere randomized with the appropriateethod to generate the sequence of ran-
omization (computer-generated). Fiverials, including the neuroprotective armf the study by Mittendorf et al,29,30 wereouble blinded in which a placebo wassed, and the outcomes were evaluatedy individuals who were blinded to treat-ent group allocation. The tocolytic
rm of the study by Mittendorf et al,30
as not double blinded because theomen were randomly assigned to re-
eive magnesium sulfate or other toco-ytic therapy such as ritodrine, terbutal-ne, indomethacin, or nifedipine,lthough diagnosis of cerebral palsy wasade by a pediatrician who was masked
o the antenatal treatment. Four trialsad adequate methods of allocation se-uence concealment.31,32,34,35 The toco-
ytic arm of the study by Mittendorf etl30 had no concealment of allocation,hereas the neuroprotective arm did not
eport on this item. Information onithdrawals and dropouts was available
or 4 trials.31,32,34,35 In the 2 trials per-ormed by Mittendorf et al29,30 there waso statement on numbers and reasons
or withdrawals in each group. Three tri-ls reported assessment of primary out-omes in � 95% of randomly assignedetuses.31,34,35 The Magpie trial32 evalu-ted the primary outcomes in only7.4% of fetuses of all gestational agesandomly assigned before delivery. Fourrials had a modified Jadad score of 7 or
ore.31,32,34,35 The 2 trials performed byittendorf et al29,30 had scores of 2 (toco-
ytic arm) and 4 (neuroprotective arm).
rimary outcomesix trials (5357 infants) reported cerebralalsy, total pediatric mortality, and theombined outcome of death or cerebralalsy. The risk of giving birth to an infantubsequently receiving a diagnosis of ce-ebral palsy was significantly lower in theroup of women who received magne-ium sulfate than among women whoid not receive magnesium sulfate (3.9%
s 5.6%; RR, 0.69; 95% CI, 0.55-0.88)(da(avmtCgt
ttntcrosta
cia(af��tw0gmeIpgctsusstfsafge
ci
pg(0
bgrn
C
www.AJOG.org Obstetrics Reviews
Figure 2 and Table 3). There was evi-ence of low statistical heterogeneitymong the trials included in this analysisI2 � 4.4%). The funnel plot of trials ofntenatal magnesium sulfate in the pre-ention of cerebral palsy appeared sym-etrical either visually or in terms of sta-
istical significance (intercept, 0.23; 95%I, -1.97 to 2.43; P � .79) (Figure 3) sug-esting that there was no evidence of ei-her publication or related biases.
The number of women at risk for pre-erm delivery less than 34 weeks of gesta-ion who needed to be treated with mag-esium sulfate rather than with placebo
o prevent 1 case of cerebral palsy in theirhildren was 52 (95% CI, 31-154). Theandom-effects analysis of the primaryutcome of cerebral palsy yielded effectizes similar in magnitude and directiono those obtained from the fixed-effectsnalysis (Table 4).
Antenatal magnesium sulfate signifi-antly decreased the risk of cerebral palsyn sensitivity analyses limited to the 4 tri-ls with modified Jadad quality score � 4RR, 0.69; 95% CI, 0.54-0.88; I2 � 0.0%)nd to the 3 trials with completeness ofollow-up of randomly assigned fetuses
95% (RR, 0.69; 95% CI, 0.54-0.88; I2
0.0%). The RR for the subgroup of 4rials (4446 infants) whose primary aimas neuroprotection was 0.71 (95% CI,.55-0.91; I2 � 25.2%) and for the sub-roup of 2 trials (911 infants) whose pri-ary aim was tocolysis or to prevent
clampsia was 0.37 (95% CI, 0.09-1.58;2 � 0.0%). The lowered risk of cerebralalsy was demonstrated even in the sub-roup of 3 trials (3981 infants) that in-luded women with gestational age lesshan 32 weeks’ at trial entry and in theubgroup of 4 trials (4610 infants) thatsed a median total dose of magnesiumulfate greater than 4 g. No statisticallyignificant differences between groupsreated with antenatal magnesium sul-ate and controls in cerebral palsy wereeen in the subgroups of 2 trials that usedmedian total dose of magnesium sul-
ate � 4 g and that included women withestational age less than 30 weeks’ at trialntry.
Three trials (4387 infants) reportederebral palsy according to sever-
ty.31,34,35 Moderate or severe cerebral 9alsy was significantly reduced in theroup who received magnesium sulfate2.1% vs 3.2%; RR, 0.64; 95% CI, 0.44-.92; I2 � 0.0%; NNT for benefit, 74;
FIGURE 1Study selection process
239 Ex titl
87 Ex 15 11 10 5 3 43
5 Studies (6 trials) include in metaanalysis
92 Retrieved for more detailed evaluation 85 From electronic sear 4 From reference lists 3 From conference
proceedings
331 Studies identified and screened for retrieval from searches
onde-Agudelo. Antenatal magnesium sulfate for preventing
5% CI, 41-373). The risk of mild cere- 0
JUNE 2009 Am
ral palsy was reduced by 26% in theroup allocated magnesium sulfateather than placebo, although this didot achieve statistical significance (RR,
ded after screening and/or abstracts
uded bservational studies eviews or metaanalysis etters to the editor ditorials uplicate publication andomized controlled
rials on magnesium ulfate for preventing reterm birth or eclampsia ith no data on cerebral alsy or pediatric mortality
bral palsy in preterm infants. Am J Obstet Gynecol 2009.
clues
cl O R L E D Rts
pw
p
d
ch
cere
.74; 95% CI, 0.52-1.04).
erican Journal of Obstetrics & Gynecology 599
TABLE 1Characteristics of studies included in the systematic review
Study Location Inclusion/exclusion criteria
No. of infants
Magnesium sulfate
Cerebral palsy
Magnesiumgroup
Nomagnesiumgroup
Loadingdose (g)
Maintenancedose
Median totaldose (g)received bywomen inmagnesiumgroup
Definition and/or diagnosticcriteria
Health professional who madethe diagnosis and age atdiagnosis
Mittendorfet al29,30
US, single center 1. Neuroprotective arm inclusion:women with single or twinpregnancy in preterm laborbetween 25 and 33 wk ofgestation with or without PROM,reassuring fetal assessment, andcervical dilatation � 4 cm.Exclusion: preeclampsia,infection.
30 29 4 None 4 Not reported Developmental pediatrician after18 mo of corrected age
.........................................................................................................................................................................................................................................................................................................................................................................................................................
2. Tocolytic arm inclusion:women with single or twinpregnancy in preterm laborbetween 25 and 33 wk ofgestation with or without PROM,reassuring fetal assessment, andcervical dilatation � 4 cm.Exclusion: preeclampsia,infection.
55 51 4 2-3 g/h;duration notreported
498 Not reported Developmental pediatrician after18 mo of corrected age
................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................
Crowtheret al31
Australia and NewZealand, 16 centers
Inclusion: women with single,twin, triplet, or quadrupletfetuses at risk of pretermdelivery before 30 wk gestationbecause of planned or expectedbirth within 24 h.Exclusion: women in secondstage of labor, if they hadreceived magnesium sulfate inthis pregnancy, andcontraindications to magnesiumsulfate.
629 626 4 1 g/h untilbirth (ifoccurredwithin 24 h)or up to 24 h
�10.5 Abnormalities of muscle toneand loss of motor function. Mildcerebral palsy: disability inambulant children thatinterferred only slightly withnormal daily activities;moderate cerebral palsy:children attempting to walk at2 y with or without appliances;severe cerebral palsy: childrenpermanently nonambulant.
Developmental pediatrician andpsychologist at 24 mo ofcorrected age
................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................
Magpie etal32a
19 countries across5 continents, 125centers
Inclusion: women with singletonor multiple pregnancy withpreeclampsia who had not givenbirth or were � 24 h postpartumand uncertain about whether touse magnesium sulfate toprevent eclampsia, irrespectiveof whether they had receivedmagnesium sulfate or otheranticonvulsants previously.Exclusion: hypersensitivity tomagnesium, hepatic coma with arisk of renal failure, ormyasthenia gravis.
404 401 4 Either 1 g/hIV for 24 h or5 g every 4 hIM for 24 h
18.0b Severe cerebral palsy: notwalking or unlikely to walkunaided by 24 mo; nonseverecerebral palsy: not defined
Pediatrician at 18 mo ofcorrected age
................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................
Conde-Agudelo. Antenatal magnesium sulfate for preventing cerebral palsy in preterm infants. Am J Obstet Gynecol 2009. (continued )
Rev
iews
Obstetricsw
ww
.AJO
G.org
600Am
ericanJournalofObstetrics
&Gynecology
JUNE2009
TABLE 1Characteristics of studies included in the systematic review (continued)
Study Location Inclusion/exclusion criteria
No. of infants
Magnesium sulfate
Cerebral palsy
Magnesiumgroup
Nomagnesiumgroup
Loadingdose (g)
Maintenancedose
Median totaldose (g)received bywomen inmagnesiumgroup
Definition and/or diagnosticcriteria
Health professional who madethe diagnosis and age atdiagnosis
Marret etal33,34
France, 13 centers Inclusion: women with single,twin, triplet, or quadrupletfetuses at risk of pretermdelivery before 33 wk gestationbecause of planned or expectedbirth within 24 h. Exclusion: fetuswith severe malformations orchromosomal abnormalities andwomen with pregnancy-associated vascular disease(preeclampsia, growth restriction,HELLP syndrome, retroplacentalhematoma) or with at least 1 ofthe following criteria:hypotension, cardiac rhythmabnormalities, hydroelectrolyteabnormalities, renal insufficiency,ingestion during the last 24 h ofcalcium channel blockers,digitalins, or indomethacin,persistent signs of cardiovasculartoxicity or tachycardia � 1 hafter cessation of tocolyticintake, myasthenia, or indicationfor emergency cesarean section.
352 336 4 None 4 European Cerebral PalsyNetwork definition
Pediatrician by clinicalexamination (77% of survivors)or parent telephone interview(22% of survivors) at 24 mo ofage
................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................
Rouse etal35
US, 20 centers Inclusion: women carryingsingletons or twins at 24-31 wkof gestation and at high risk forspontaneous delivery because ofPROM, advanced preterm laborwith dilatation of 4-8 cm andintact membranes, or indicatedpreterm delivery anticipatedwithin 2-24 h. Exclusion: deliveryanticipated within � 2 h,cervical dilatation � 8 cm,rupture of the membranes before22 wk, unwillingness of theobstetrician to intervene for thebenefit of the fetus, major fetalanomalies or death, maternalhypertension or preeclampsia,maternal contraindications tomagnesium sulfate, and receiptof IV magnesium sulfate withinthe previous 12 h.
1188 1256 6 2 g/hc 31.5 Presence of � 2 of thefollowing 3 features: (1) a delayof � 30% in gross motordevelopmental milestones; (2)abnormality in muscle tone.4� or absent deep tendonreflexes, or movementabnormality; (3) persistence ofprimitive reflexes or absence ofprotective reflexes
Pediatrician or pediatricneurologist at or beyond 24 moof corrected age
................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................
IM, intramuscular; IV, intravenous; PROM, premature rupture of membranes; US, United States.a Data for children whose mothers were � 34 wk of gestation and undelivered at randomization extracted from the review by Doyle et al22; b for women of all gestational ages and undelivered at randomization; c if delivery had not occurred after 12 h and was no
longer considered imminent, the infusion was discontinued and resumed when delivery was deemed imminent again. If at least 6 h had passed since the discontinuation of the study medication, another loading dose was given.Conde-Agudelo. Antenatal magnesium sulfate for preventing cerebral palsy in preterm infants. Am J Obstet Gynecol 2009.
ww
w.A
JOG
.orgObstetrics
Rev
iews
JUNE2009
American
JournalofObstetrics&
Gynecology601
ricewR(otsfMq1cpt2c
STtnnrmc9
c
tydsmfRNTnocaab
mdrprw(tCtat11
tscsfl9vs5(3v2(Cmtt
IoTenU2tdamp
Reviews Obstetrics www.AJOG.org
6
There was no overall difference in theisk of total pediatric mortality, includ-ng fetal mortality and under 2 years oforrected age mortality, between infantsxposed to magnesium sulfate and thoseho were not exposed (15.1% vs 14.8%;R, 1.01; 95% CI, 0.89-1.14; I2 � 38.9%)
Table 3). The lack of evidence for anyverall effect of magnesium sulfate onotal pediatric mortality was found in allensitivity and subgroup analyses per-ormed, except 1 (subgroup of 2 trials by
ittendorf et al29,30 with modified Jadaduality score � 4 [RR, 6.61; 95% CI,.12-36.11; I2 � 31.0%]) (Table 4). Theombined outcome of death or cerebralalsy was slightly lower for children inhe magnesium sulfate group (19.0% vs0.4%), although this was not a statisti-ally significant difference.
econdary outcomeshere were no significant differences be-
ween the groups in the risk of adverseeonatal outcomes (Table 5), although aonsignificant increase was seen in theisk of necrotizing enterocolitis in theagnesium group compared with the
ontrol group (7.1% vs 5.9%; RR, 1.23;5% CI, 0.98-1.54; I2 � 0.0%).For infant neurodevelopmental out-
TABLE 2Modified Jadad score for assessme
Item
Mittendo
Tocolytic
Randomization Yes...................................................................................................................
Method to generate randomizationclear and appropriate
Yes
...................................................................................................................
Double blind No...................................................................................................................
Methods for blinding appropriate No...................................................................................................................
Method of allocationconcealmenta
No
...................................................................................................................
Description of withdrawal ordropout
No
...................................................................................................................
Completeness of follow-up ofrandomized fetuses (%)b
Unreporte
...................................................................................................................
Total score 2...................................................................................................................
Yes � 1 point; no � zero points; scores: 0 � lowest qualia Adequate � 2 points; no concealment of allocation or inade
of all gestational ages and undelivered at randomization.Conde-Agudelo. Antenatal magnesium sulfate for preven
omes, the risk of substantial gross mo- f
02 American Journal of Obstetrics & Gynecology
or dysfunction at the corrected age of 2ears was significantly lower among chil-ren whose mothers received magne-ium sulfate than among children whose
others did not receive magnesium sul-ate (3 trials; 4387 infants; 2.6% vs 4.2%;R, 0.60; 95% CI, 0.43-0.83; I2 � 0.0%;NT for benefit, 53; 95% CI, 32-146).here were no significant effects of ante-atal magnesium sulfate treatment onther infant neurodevelopmental out-omes such as major neurologic disability,ny neurologic impairment, Bayley mentalnd psychomotor development indexes,lindness, and deafness (Table 6).There was no evidence of an effect ofagnesium sulfate on the risk of maternal
eath (3 trials; 3867 women), cardiac orespiratory arrest (3 trials; 3867 women),ulmonary edema (1 trial; 2241 women),espiratory depression (2 trials; 3303omen), severe postpartum hemorrhage
2 trials; 1626 women), and cesarean sec-ion (3 trials; 3867 women) (Table 7).ompared with women receiving placebo,
hose exposed to magnesium sulfate hadbout a 50% increased risk of both hypo-ension and tachycardia (RR, 1.51; 95% CI,.09-2.09; NNT for harm, 30; 95% CI, 17-56 and RR, 1.53; 95% CI, 1.03-2.29; NNT
of methodologic quality of includedt al29,30
Crowther et al31 MaNeuroprotective arm
Yes Yes Yes.........................................................................................................................
Yes Yes Yes
.........................................................................................................................
Yes Yes Yes.........................................................................................................................
Yes Yes Yes.........................................................................................................................
Unreported Adequate Ad
.........................................................................................................................
No Yes Yes
.........................................................................................................................
Unreported 98.9 47
.........................................................................................................................
4 8 7.........................................................................................................................
� highest quality.
method or unreported � zero points; b follow-up � 95% � 1 p
erebral palsy in preterm infants. Am J Obstet Gynecol 2009.
or harm, 28; 95% CI, 14-379, respec- v
JUNE 2009
ively). Maternal side effects secondary totudy medication were significantly moreommon among women allocated magne-ium sulfate rather than placebo, includingushing (58.4% vs 8.3%; NNT for harm, 2;5% CI, 2-2), nausea or vomiting (16.3%s 3.9%; NNT for harm, 8; 95% CI, 7-10),weating (25.2% vs 3.4%; NNT for harm,; 95% CI, 4-5), problems at injection site37.6% vs 4.1%; NNT for harm, 3; 95% CI,-3), stopping of infusion because of ad-erse effects (7.5% vs 2.6%; NNT for harm,0; 95% CI, 16-29), and any side effect70.7% vs 17.6%; NNT for harm, 2; 95%I, 2-2). All funnel plots showed no asym-etry, either visually or in terms of statis-
ical significance (P �.10 for all, by Eggerest).
mpact and economic evaluationf the interventionhe United Cerebral Palsy Foundation hasstimated that about 8000 infants are diag-osed with cerebral palsy each year in thenited States6 of which about 25% (n �
000) are born before 34 weeks of gesta-ional age.5,8 Therefore, if all women whoeliver before 34 weeks of gestation receiventenatal magnesium sulfate, the esti-ated number of new cases of cerebral
alsy that hypothetically could be pre-
dies
e group32 Marret et al33,34 Rouse et al35
Yes Yes..................................................................................................................
Yes Yes
..................................................................................................................
Yes Yes..................................................................................................................
Yes Yes..................................................................................................................
ate Adequate Adequate
..................................................................................................................
Yes Yes
..................................................................................................................
98.5 95.6
..................................................................................................................
8 8..................................................................................................................
follow-up � 95% or unreported � zero points; c for fetuses
nt sturf e
gpiarm
......... .........
......... .........
......... .........
......... .........
equ
......... .........
......... .........
d .4c
......... .........
......... .........
ty, 8
quate oint;
ting c
ented annually is 620 (95% CI, 240-900).
ttswdr
taee2gl
ioC
CIs
C
www.AJOG.org Obstetrics Reviews
The number of women at risk of pre-erm delivery before 34 weeks of gesta-ion who needed to receive magnesiumulfate to prevent 1 case of cerebral palsyas 52 (95% CI, 31-154). A recent costecision analysis from the United Statesevealed that the total cost for adminis-
TABLE 3Effect of magnesium sulfate on cer
Outcome N
Cerebral palsy 63
...................................................................................................................
Moderate/severe cerebral palsy 33
...................................................................................................................
Mild cerebral palsy 33
...................................................................................................................
Total pediatric mortality 62
...................................................................................................................
Fetal mortality 52
...................................................................................................................
Under 2 y of corrected age mortality 52
...................................................................................................................
Death or cerebral palsy 63
...................................................................................................................
CI, confidence interval.
FIGURE 2Effect of magnesium sulfate on cer
Study
Rouse et al35
Marret et al34
Crowther et al31
Mittendorf et al29,a
Combined
Mittendorf et al29,b
Magpie32
Test for heterogeneity I2= 4.4%
0.1 0.2 0
Favors maNeuroprotective armbTocolytic arm
onde-Agudelo. Antenatal magnesium sulfate for preventing
Conde-Agudelo. Antenatal magnesium sulfate for preventing c
rating magnesium sulfate as a tocolyticgent, including costs attributable to thevaluation and treatment of its adversevents, was $197.90 per patient in005.36 Thus, if magnesium sulfate wasiven to all women at risk of preterm de-ivery before 34 weeks of gestation, the
ral palsy and pediatric mortality
f trials
No. of events/total number
Magnesium No magnesi2,34,35 104/2658 152/2699.........................................................................................................................,35 45/2169 72/2218
.........................................................................................................................,35 54/2169 74/2218
.........................................................................................................................,32,34,35 401/2658 400/2699
.........................................................................................................................,34,35 17/2254 22/2298
.........................................................................................................................,34,35 217/2254 220/2298
.........................................................................................................................2,34,35 505/2658 551/2699.........................................................................................................................
ral palsy
Relative risk (fixed) M(95% CI)
521 1
Favors no magnesiumesium
bral palsy in preterm infants. Am J Obstet Gynecol 2009.
erebral palsy in preterm infants. Am J Obstet Gynecol 2009.
JUNE 2009 Am
ncremental cost of preventing one casef cerebral palsy would be $10,291 (95%I, 6135-30,477).
ommentn this systematic review, we found per-uasive evidence that magnesium sulfate
Relative risk (95% CI) I 2 (%)
0.69 (0.55-0.88) 4.4..................................................................................................................
0.64 (0.44-0.92) 0.0..................................................................................................................
0.74 (0.52-1.04) 0.0..................................................................................................................
1.01 (0.89-1.14) 38.9..................................................................................................................
0.78 (0.42-1.46) 0.0..................................................................................................................
1.00 (0.84-1.19) 47.3..................................................................................................................
0.92 (0.83-1.02) 43.3..................................................................................................................
Weight Control Relative risk nesium(%)n/N (95% CI)n/N
0.66 (0.11-3.94)
0.13 (0.01-2.51)
0.59 (0.40-0.85)
0.70 (0.41-1.19)
0.85 (0.56-1.31)
6.77 (0.37-125.7)
0.69 (0.55-0.88)
2.0
100.0
2.4
47.4
20.2
27.7
0.33/30 0/29
0/55 3/51
6/629 42/626
2/404 3/401
2/352 30/336
1/1188 74/1256
4/2658 152/2699
eb
o. o um0-3
......... .........1,34
......... .........1,34
......... .........9,31
......... .........9,31
......... .........9,31
......... .........0-3
......... .........
eb
ag
3
2
4
10
.5 0
agn
cere
erican Journal of Obstetrics & Gynecology 603
addctesIamswoTrbsmntoowcwa
sascpocwstmstatsai
afffda
Csrs0tRsst0n
ivfbr0gCsmNbafpfwrbiwaMttfpnm2(tnaatImphsp
S
CC
Reviews Obstetrics www.AJOG.org
6
dministered to women at high risk ofelivery before 34 weeks of gestation re-uces the risk of cerebral palsy in theirhildren. The evidence was strongest forhe subgroup of trials that specificallyvaluated the use of antenatal magne-ium sulfate in preventing cerebral palsy.n addition, this therapy was also associ-ted with a significantly decreased risk ofoderate or severe cerebral palsy and
ubstantial gross motor dysfunctionithout any statistically significant effectn the risk of total pediatric mortality.his last finding suggests the reduced
isk for cerebral palsy does not appear toe due to selective mortality of magne-ium sulfate-exposed infants. Overall,
agnesium sulfate therapy had no sig-ificant effect on the risk of major ma-
ernal complications, adverse neonatalutcomes, and other infant neurodevel-pmental outcomes, although a trendas found toward increased risk of ne-
rotizing enterocolitis. About 70% ofomen reported any minor side effect
FIGURE 3Funnel plot of trials of antenatal min the prevention of cerebral palsy
0.0
0.5
1.0
-4.01.5
-2.4 -0
tandard error
ircles indicate log (relative risks) from trials ionde-Agudelo. Antenatal magnesium sulfate for preventing
ssociated with the infusion of magne- r
04 American Journal of Obstetrics & Gynecology
ium sulfate mainly flushing, problemst injection site, and sweating. Thetrength of our findings is based on itsompliance with stringent criteria forerforming a rigorous systematic reviewf randomized controlled trials, the in-lusion of a relatively large number ofomen and infants in these metaanaly-
es, the relatively narrow confidence in-ervals obtained that made our results
ore precise, the evidence of clinical andtatistical homogeneity in the results ofhe trials for cerebral palsy, the sensitivitynalysis restricted to high-quality trialshat upheld our main results, and theymmetrical funnel plots that suggestedbsence of publication and related biasesn our metaanalyses.
We were unable to determine whetherntenatal magnesium sulfate is more ef-ective to prevent cerebral palsy in in-ants of singleton pregnancies than in in-ants of multiple pregnancies. Specificata for the use of magnesium sulfateccording to plurality of pregnancy were
esium sulfate
0.8 2.4 4.0
Log (Relative risk)
ded in metaanalysis.bral palsy in preterm infants. Am J Obstet Gynecol 2009.
eported only by 2 authors.31,35 s
JUNE 2009
rowther et al31 found that magnesiumulfate had no significant effect on theisk of cerebral palsy in both infants ofingleton pregnancies (RR, 1.01; 95% CI,.61-1.68) and infants of multiple gesta-ions (RR, 0.52; 95% CI, 0.21-1.25).ouse et al35 reported that magnesium
ulfate reduced the risk of moderate orevere cerebral palsy in infants of single-on pregnancies (RR, 0.52; 95% CI, 0.27-.98) but not in infants of twin preg-ancy (RR, 0.73; 95% CI, 0.27-1.92).A recent metaanalysis22 reported sim-
lar findings to those found in our re-iew. In fact, antenatal magnesium sul-ate given to women at risk of pretermirth before 37 weeks of gestational ageeduced the risk of cerebral palsy (RR,.68; 95% CI, 0.54-0.87) and substantialross motor dysfunction (RR, 0.61; 95%I, 0.44-0.85) in their children without a
tatistically significant effect on pediatricortality (RR, 1.04; 95% CI, 0.92-1.17).otwithstanding, our findings need toe distinguished from those of this meta-nalysis. We specifically evaluated the ef-ect of antenatal magnesium sulfate inreventing cerebral palsy in preterm in-
ants whose mothers were less than 34eeks of gestational age at the time of
andom assignment. The metaanalysesy Doyle et al22 included 5357 preterm
nfants whose mothers were less than 34eeks of gestation at the time of random
ssignmnt plus 788 infants from theagpie trial32 whose mothers were be-
ween 34 0/7 and 36 6/7 weeks of gesta-ion when treated with magnesium sul-ate or placebo. In this last subgroup ofreterm infants, cerebral palsy was diag-osed in none of the 394 infants in theagnesium sulfate group (0.0%) and inof 394 infants in the placebo group
RR, 0.20; 95% CI, 0.01-4.15). Thus,here is no evidence that antenatal mag-esium sulfate administered to woment risk of preterm delivery between 34 0/7nd 36 6/7 weeks of gestation decreaseshe risk of cerebral palsy in their infants.n addition, the Magpie trial was not pri-
arily designed to evaluate the neuro-rotective role of magnesium sulfate andad a low rate of follow-up. Finally, theystematic review by Doyle et al22 did noterform an economic evaluation to as-
agn
.8
nclucere
ess the cost-effectiveness of this inter-
www.AJOG.org Obstetrics Reviews
TABLE 4Sensitivity and subgroup analyses of metaanalysis on effect of magnesium sulfate on cerebral palsy and pediatric mortalitySubgroup No. of trials Relative risk (95% CI) I 2 (%)
CEREBRAL PALSY
Statistical model.......................................................................................................................................................................................................................................................................................................................................................................
Fixed effects 630-33,35 0.69 (0.55-0.88) 4.4.......................................................................................................................................................................................................................................................................................................................................................................
Random effects 630-33,35 0.70 (0.54-0.90) NA................................................................................................................................................................................................................................................................................................................................................................................
Modified Jadad quality score.......................................................................................................................................................................................................................................................................................................................................................................
� 4 431-33,35 0.69 (0.54-0.88) 0.0.......................................................................................................................................................................................................................................................................................................................................................................
� 4 230a,b 0.95 (0.02-44.91) 71.1................................................................................................................................................................................................................................................................................................................................................................................
Completeness of follow-up of randomized fetuses.......................................................................................................................................................................................................................................................................................................................................................................
� 95% 331,34,35 0.69 (0.54-088) 0.0.......................................................................................................................................................................................................................................................................................................................................................................
� 95% or unreported 330,32 0.83 (0.28-2.43) 43.5................................................................................................................................................................................................................................................................................................................................................................................
Aim of the treatment.......................................................................................................................................................................................................................................................................................................................................................................
Neuroprotective 430a,31,33,35 0.71 (0.55-0.91) 25.2.......................................................................................................................................................................................................................................................................................................................................................................
Other aims 230b,32 0.37 (0.09-1.58) 0.0................................................................................................................................................................................................................................................................................................................................................................................
Median total dose of magnesium sulfate.......................................................................................................................................................................................................................................................................................................................................................................
� 4 g 230a,33 1.37 (0.18-10.70) 56.5.......................................................................................................................................................................................................................................................................................................................................................................
� 4 g 430b,31,32,35 0.67 (0.51-0.88) 0.0................................................................................................................................................................................................................................................................................................................................................................................
Gestational age at trial entry.......................................................................................................................................................................................................................................................................................................................................................................
� 34 wk 630-33,35 0.69 (0.55-0.88) 4.4.......................................................................................................................................................................................................................................................................................................................................................................
� 32 wk 331,32,35 0.69 (0.52-0.91) 0.0.......................................................................................................................................................................................................................................................................................................................................................................
� 30 wk 231,32 0.86 (0.56-1.31) 0.0................................................................................................................................................................................................................................................................................................................................................................................
PEDIATRIC MORTALITY
Statistical model.......................................................................................................................................................................................................................................................................................................................................................................
Fixed effects 629,31-33,35 1.01 (0.89-1.14) 38.9.......................................................................................................................................................................................................................................................................................................................................................................
Random effects 629,31-33,35 0.99 (0.82-1.19) NA................................................................................................................................................................................................................................................................................................................................................................................
Modified Jadad quality score.......................................................................................................................................................................................................................................................................................................................................................................
� 4 431-33,35 0.99 (0.87-1.11) 27.9.......................................................................................................................................................................................................................................................................................................................................................................
� 4 229a,b 6.61 (1.21-36.11) 31.0................................................................................................................................................................................................................................................................................................................................................................................
Completeness of follow-up of randomized fetuses.......................................................................................................................................................................................................................................................................................................................................................................
� 95% 331,34,35 0.94 (0.79-1.12) 40.7.......................................................................................................................................................................................................................................................................................................................................................................
� 95% 329,32 1.10 (0.93-1.30) 49.5................................................................................................................................................................................................................................................................................................................................................................................
Aim of the treatment.......................................................................................................................................................................................................................................................................................................................................................................
Neuroprotective 429a,31,33,35 0.95 (0.80-1.12) 19.6.......................................................................................................................................................................................................................................................................................................................................................................
Other aims 229b,32 2.83 (0.21-38.80) 72.9................................................................................................................................................................................................................................................................................................................................................................................
Median total dose of magnesium sulfate.......................................................................................................................................................................................................................................................................................................................................................................
� 4 g 229a,33 0.88 (0.57-1.35) 0.0.......................................................................................................................................................................................................................................................................................................................................................................
� 4 g 429b,31,32,35 1.01 (0.80-1.28) 59.3................................................................................................................................................................................................................................................................................................................................................................................
Gestational age at trial entry.......................................................................................................................................................................................................................................................................................................................................................................
� 34 wk 629,31-33,35 1.01 (0.89-1.14) 38.9.......................................................................................................................................................................................................................................................................................................................................................................
� 32 wk 331,32,35 1.03 (0.83-1.28) 65.5.......................................................................................................................................................................................................................................................................................................................................................................
� 30 wk 231,32 0.97 (0.67-1.41) 83.0................................................................................................................................................................................................................................................................................................................................................................................
CI, confidence interval; NA, not applicable.a Neuroprotective arm of study by Mittendorf et al29,30; b tocolytic arm of study by Mittendorf et al.29,30
Conde-Agudelo. Antenatal magnesium sulfate for preventing cerebral palsy in preterm infants. Am J Obstet Gynecol 2009.
JUNE 2009 American Journal of Obstetrics & Gynecology 605
vptcnesa
mpdl
tvdtcwiestppa
fggtpasmhbchc
ting c
Reviews Obstetrics www.AJOG.org
6
ention for the prevention of cerebralalsy or evaluate some important neona-al and maternal outcomes such as ne-rotizing enterocolitis, maternal pulmo-ary edema, and some maternal sideffects of the infusion (eg, flushing, nau-ea or vomiting, sweating, and problemst injection site).
The molecular mechanisms by whichagnesium sulfate can prevent cerebral
alsy in preterm infants are not fully un-erstood. The most prevalent pathologic
esion seen in cerebral palsy is periven-
TABLE 5Effect of magnesium sulfate on neo
Outcome
Intraventricular hemorrhage (all grades)...................................................................................................................
Grade III/IV intraventricular hemorrhage...................................................................................................................
Periventricular leukomalacia...................................................................................................................
Apgar score �7 at 5 min...................................................................................................................
Neonatal seizures...................................................................................................................
Respiratory distress syndrome...................................................................................................................
Need for supplemental oxygen at 36wk...................................................................................................................
Bronchopulmonary dysplasia...................................................................................................................
Mechanical ventilation...................................................................................................................
Necrotizing enterocolitis...................................................................................................................
CI, confidence interval; NA, not applicable.Conde-Agudelo. Antenatal magnesium sulfate for preven
TABLE 6Effect of magnesium sulfate on inf
Outcome
Substantial gross motor dysfunction...................................................................................................................
Major neurologic disability...................................................................................................................
Any neurologic impairment...................................................................................................................
Bayley mental development index � 85...................................................................................................................
Bayley mental development index � 70...................................................................................................................
Bayley psychomotor development index � 85...................................................................................................................
Bayley psychomotor development index � 70...................................................................................................................
Blindness...................................................................................................................
Deafness...................................................................................................................
CI, confidence interval; NA, not applicable.
Conde-Agudelo. Antenatal magnesium sulfate for preventing c06 American Journal of Obstetrics & Gynecology
ricular white matter injury resulting fromulnerability of the immature preoligo-endrocytes before 32 weeks of gesta-ion.37 Preoligodendrocytes are the pre-ursors of myelinating oligodendrocytes,hich constitute a major glial population
n the white matter. Oxidative stress andxcitotoxicity resulting from excessivetimulation of ionotropic glutamate recep-ors on preoligodendrocytes are the mostrominent molecular mechanisms foreriventricular white matter injury.38,39 Inddition, recent studies have identified
tal outcomes
. of trials
No. of events/total number
Magnesium No magnes,31,33,35 467/2254 493/2298
.........................................................................................................................,31,35 74/1902 91/1962
.........................................................................................................................,31,33,35 71/2254 76/2298
.........................................................................................................................,33,35 351/2169 351/2218
.........................................................................................................................,33,35 55/2169 70/2218
.........................................................................................................................,35 730/1540 779/1592
.........................................................................................................................,33 220/981 195/962
.........................................................................................................................
213/1188 218/1256.........................................................................................................................,33,35 1381/2169 1446/2218
.........................................................................................................................,33,35 155/2169 131/2218
.........................................................................................................................
erebral palsy in preterm infants. Am J Obstet Gynecol 2009.
neurodevelopmental outcomes
No. of trials
No. of events/total numbe
Magnesium No magn
331,34,35 56/2169 94/2218.........................................................................................................................
231,32 93/1033 85/1027.........................................................................................................................
231,32 198/1033 194/1027.........................................................................................................................
331,34,35 639/2169 660/2218.........................................................................................................................
135 165/1188 171/1256.........................................................................................................................
135 299/1188 315/1256.........................................................................................................................
135 134/1188 144/1256.........................................................................................................................
231,34 2/981 2/962.........................................................................................................................
231,34 8/981 11/962.........................................................................................................................
erebral palsy in preterm infants. Am J Obstet Gynecol 2009.
JUNE 2009
unctional N-methyl-D-aspartate (NMDA)lutamatergic receptors in oligodendro-lial injury processes.40,41 Antagonists ofhe NMDA receptors for glutamate areotent neuroprotective agents in severalnimal models of perinatal brain le-ions.42 Growing evidence suggests that
agnesium sulfate may reverse thearmful effects of hypoxic/ischemicrain damage by blocking the NMDA re-eptors and, as a calcium antagonist,indering calcium influx into theells.43,44 In addition, magnesium sulfate
Relative risk (95% CI) I 2 (%)
0.96 (0.86-1.08) 20.1..................................................................................................................
0.83 (0.61-1.11) 0.0..................................................................................................................
0.93 (0.68-1.28) 0.0..................................................................................................................
1.03 (0.90-1.18) 7.3..................................................................................................................
0.80 (0.56-1.13) 0..................................................................................................................
1.01 (0.85-1.19) 65.8..................................................................................................................
1.12 (0.95-1.32) 23.1
..................................................................................................................
1.03 (0.87-1.23) NA..................................................................................................................
0.99 (0.89-1.09) 82.1..................................................................................................................
1.23 (0.98-1.54) 0.0..................................................................................................................
Relative risk (95% CI) I 2 (%)um
0.60 (0.43-0.83) 0.0..................................................................................................................
1.09 (0.83-1.43) 15.3..................................................................................................................
1.02 (0.86-1.20) 0.0..................................................................................................................
1.00 (0.91-1.09) 0.0..................................................................................................................
1.02 (0.84-1.24) NA..................................................................................................................
1.00 (0.88-1.15) NA..................................................................................................................
0.98 (0.79-1.23) NA..................................................................................................................
0.97 (0.14-6.90) 0.0..................................................................................................................
0.51 (0.05-4.96) 58.7..................................................................................................................
na
No ium
530
......... .........
430
......... .........
530
......... .........
331
......... .........
331
......... .........
233
......... .........
231
......... .........
135
......... .........
331
......... .........
331
......... .........
antr
esi
......... .........
......... .........
......... .........
......... .........
......... .........
......... .........
......... .........
......... .........
......... .........
coa
spapthtttwnttsstmwiyeht
gwfnc(duvictdrdaWwc
ndc2ep
tttt1(neta
cdmmpaoctoamtft
www.AJOG.org Obstetrics Reviews
ould also reverse the destructive actionf oxygen radicals and excitatory aminocids.45,46
The finding that antenatal magnesiumulfate significantly reduces cerebralalsy compared with placebo is a strongrgument for its use in women at risk ofreterm delivery before 34 weeks of ges-ation. Some limitations must be noted,owever, when interpreting the results ofhis review. First, there were variations inhe inclusion and exclusion criteria, ges-ational age at which magnesium sulfateas administered, loading and mainte-ance doses, duration of the interven-
ion, and use of retreatment. In addition,he optimal time to administer magne-ium sulfate was unclear. Second, 2 of thetudies included in our review30,32 madehe assessment of cerebral palsy at 18
onths of corrected age, and another34
as made by parent telephone interviewn 22% of surviving infants. For reasonset poorly understood, many children,specially those born very preterm, ex-ibit a variety of neurologic abnormali-
TABLE 7Effect of magnesium sulfate on ma
Outcome
Death...................................................................................................................
Cardiac or respiratory arrest...................................................................................................................
Pulmonary edema...................................................................................................................
Respiratory depression...................................................................................................................
Hypotension...................................................................................................................
Tachycardia...................................................................................................................
Severe postpartum hemorrhage...................................................................................................................
Cesarean section...................................................................................................................
Clinical and self-assessed maternal side effecthe infusion
..........................................................................................................
Flushing..........................................................................................................
Nausea or vomiting..........................................................................................................
Sweating..........................................................................................................
Problems at injection site..........................................................................................................
Stopping of infusion because of adverse e..........................................................................................................
Any side effect...................................................................................................................
CI, confidence interval; NA, not applicable.Conde-Agudelo. Antenatal magnesium sulfate for preven
ies in the first 18 months of life that sug- w
est the diagnosis of cerebral palsy,hich then can resolve later.47 There-
ore, it has been suggested that the diag-osis of cerebral palsy should be assignedautiously before the age of 24 monthsor 24 months after the expected date ofelivery in the case of preterm infants),nless the disorder is exceptionally se-ere.48 Third, not all of the trials reportedncluded important secondary out-omes, such as adverse neonatal and ma-ernal outcomes, as well as infant neuro-evelopmental outcomes. Finally, theesults of the studies we summarized areirectly applicable only to the womennd infants included in those studies.
hether the results are applicable toomen and infants with other clinical
onditions is unknown.The antenatal administration of mag-
esium sulfate to women at high risk ofelivery before 34 weeks of gestationould be a cost-effective intervention. In003, Honeycutt et al2 estimated the av-rage lifetime costs (direct and indirect)er person were $921,000 for persons
nal outcomes
No. of trials
No. of events/total num
Magnesium No ma
331,33,35 0/1917 1/19.........................................................................................................................
331,33,35 0/1917 0/19.........................................................................................................................
135 8/1096 3/11.........................................................................................................................
231,35 41/1631 31/16.........................................................................................................................
231,33 80/821 52/80.........................................................................................................................
131 56/535 36/52.........................................................................................................................
231,33 28/821 26/80.........................................................................................................................
331,33,35 822/1917 834/19.........................................................................................................................
f
.........................................................................................................................
331,33,35 1119/1917 162/19.........................................................................................................................
331,33,35 312/1917 76/19.........................................................................................................................
231,35 411/1631 57/16.........................................................................................................................
231,35 614/1631 68/16.........................................................................................................................
ts 231,35 123/1631 44/16.........................................................................................................................
331,33,35 1356/1917 343/19.........................................................................................................................
erebral palsy in preterm infants. Am J Obstet Gynecol 2009.
ith cerebral palsy. We have estimated t
JUNE 2009 Am
hat if magnesium sulfate is adminis-ered to all women at high risk of pre-erm delivery before 34 weeks of gesta-ion, the incremental cost of preventingcase of cerebral palsy would be $10,291
95% CI, $6135-$29,685). It should beoted that we did not estimate the cost-ffectiveness of this intervention inerms of the marginal cost per quality-djusted life year (QALY) gained.
Antenatal magnesium sulfate should beonsidered for use in women at high risk ofelivery before 34 weeks of gestation,ainly in those with premature rupture ofembranes, labor in active phase, and
lanned delivery within 24 hours. Loadingnd maintenance doses, and the durationf the treatment should not normally ex-eed 6 g, 1-2 g/h, and 24 hours, respec-ively. Further studies and/or metaanalysisf individual patient data from the avail-ble trials are required to assess the mini-um effective dose of magnesium sulfate,
he optimal time to administer it, the needor retreatment, the efficacy of interven-ion in singleton and multiple pregnancies,
r
Relative risk (95% CI) I 2 (%)sium
0.32 (0.01-7.92) 0.0..................................................................................................................
Not estimable NA..................................................................................................................
2.79 (0.74-10.47) NA..................................................................................................................
1.31 (0.83-2.07) 0.0..................................................................................................................
1.51 (1.09-2.09) 3.6..................................................................................................................
1.53 (1.03-2.29) NA..................................................................................................................
1.06 (0.63-1.79) 0.0..................................................................................................................
1.00 (0.93-1.07) 21.6..................................................................................................................
..................................................................................................................
7.56 (3.39-16.88) 93.8..................................................................................................................
4.60 (1.54-13.75) 91.5..................................................................................................................
6.37 (1.96-20.68) 94.6..................................................................................................................
9.12 (7.19-11.57) 0.0..................................................................................................................
2.81 (2.01-3.93) 0.0..................................................................................................................
5.05 (2.06-12.39) 98.3..................................................................................................................
terbe
gne
50......... .........
50......... .........
45......... .........
72......... .........
5......... .........
7......... .........
5......... .........
50......... .........
ts o
......... .........
50......... .........
50......... .........
72......... .........
72......... .........
ffec 72......... .........
50......... .........
ting c
he cost-effectiveness of intervention, the
erican Journal of Obstetrics & Gynecology 607
etp
aubwcismct
R1rb22GchS3sP4DMc225mb6lac27Ctir8Hol19AB1oo21sl2
1Doms1Bfmc21cmanfR1Is2G1MnwS1nl11Hclc1Dnpt12NrdA2MteG2raf12Rrto12
cT2iHbtTN2ma2Ms2n12nsp3Anhn3ASEtt3Gcfm3agba3mb23dfJ3ds3ciR3c3p24
Reviews Obstetrics www.AJOG.org
6
ffect on the risk of necrotizing enterocoli-is, and the long-term consequences of ex-osure for the women and their children.In summary, our results suggest that
ntenatal magnesium sulfate could besed for the primary prevention of cere-ral palsy in preterm infants less than 34eeks of gestational age. However, be-
ause cerebral palsy is a result of multiplenteracting risk factors rather than of aingle cause, it is unlikely that antenatal
agnesium sulfate administration alonean prevent all cases of this illness in pre-erm infants. f
EFERENCES. Rosenbaum P, Paneth N, Leviton A, et al. Aeport: the definition and classification of cere-ral palsy April 2006. Dev Med Child Neurol007;109(Suppl):8-14.. Honeycutt A, Dunlap L, Chen H, al Homsi G,rosse S, Schendel D. Economic costs asso-iated with mental retardation, cerebral palsy,earing loss, and vision impairment: Unitedtates, 2003. MMWR 2004;53:57-9.. Paneth N, Hong T, Korzeniewski S. The de-criptive epidemiology of cerebral palsy. Clinerinatol 2006;33:251-67.. Yeargin-Allsopp M, Van Naarden Braun K,oernberg NS, Benedict RE, Kirby RS, DurkinS. Prevalence of cerebral palsy in 8-year-old
hildren in three areas of the United States in002: a multisite collaboration. Pediatrics008;121:547-54.. Moster D, Lie RT, Markestad T. Long-termedical and social consequences of pretermirth. N Engl J Med 2008;359:262-73.. United Cerebral Palsy. Press room. Vocabu-
ary tips. Cerebral Palsy - Facts & Figures. Avail-ble at: http://www.ucp.org/ucp_generaldoc.fm/1/9/37/37-37/447. Accessed Jan. 20,009.. Himpens E, Van den Broeck C, Oostra A,alders P, Vanhaesebrouck P. Prevalence,
ype, distribution, and severity of cerebral palsyn relation to gestational age: a meta-analyticeview. Dev Med Child Neurol 2008;50:334-40.. Himmelmann K, Hagberg G, Beckung E,agberg B, Uvebrant P. The changing pan-rama of cerebral palsy in Sweden: IX, preva-
ence and origin in the birth-year period 1995-998. Acta Paediatr 2005;94:287-94.. Petterson B, Blair E, Watson L, Stanley F.dverse outcome after multiple pregnancy.aillieres Clin Obstet Gynaecol 1998;12:1-17.0. Scher AI, Petterson B, Blair E, et al. The riskf mortality or cerebral palsy in twins: a collab-rative population-based study. Pediatr Res002;52:671-81.1. Nelson KB, Grether JK. Can magnesiumulfate reduce the risk of cerebral palsy in very
ow birthweight infants? Pediatrics 1995;95:
63-9. s08 American Journal of Obstetrics & Gynecology
2. Hauth JC, Goldenberg RL, Nelson KG,uBard MB, Peralta MA, Gaudier FL. Reductionf cerebral palsy with maternal MgSO4 treat-ent in newborns weighing 500-1000 g [ab-
tract]. Am J Obstet Gynecol 1995;172:419.3. Schendel DE, Berg CJ, Yeargin-Allsopp M,oyle CA, Decoufle P. Prenatal magnesium sul-
ate exposure and the risk for cerebral palsy orental retardation among very low-birth-weight
hildren aged 3 to 5 years. JAMA 1996;76:1805-10.4. Wiswell TE, Graziani LJ, Caddell JL, Vec-hione N, Stanley C, Spitzer AR. Maternally ad-inistered megnesium sulphate protects
gainst early brain injury and long-term adverseeurodevelopmental outcomes in preterm in-ants: a prospective study [abstract]. Pediatres 1996;39:253.5. Matsuda Y, Kouno S, Hiroyama Y, et al.
ntrauterine infection, magnesium sulfate expo-ure and cerebral palsy in infants born between6 and 30 weeks of gestation. Eur J Obstetynecol Reprod Biol 2000;91:159-64.6. Paneth N, Jetton J, Pinto-Martin J, Susser. Magnesium sulfate in labor and risk of neo-
atal brain lesions and cerebral palsy in low birtheight infants. The Neonatal Brain Hemorrhagetudy Analysis Group. Pediatrics 1997;99:E1.7. O’Shea TM, Klinepeter KL, Dillard RG. Pre-atal events and the risk of cerebral palsy in very
ow birth weight infants. Am J Epidemiol998;147:362-9.8. Wilson-Costello D, Borawski E, Friedman, Redline R, Fanaroff AA, Hack M. Perinatalorrelates of cerebral palsy and other neuro-
ogic impairment among very low birth weighthildren. Pediatrics 1998;102(2 Pt 1):315-22.9. Boyle CA, Yeargin-Allsopp M, SchendelE, Holmgreen P, Oakley GP. Tocolytic mag-esium sulfate exposure and risk of cerebralalsy among children with birth weights lesshan 1,750 grams. Am J Epidemiol 2000;52:120-4.0. Grether JK, Hoogstrate J, Walsh-Greene E,elson KB. Magnesium sulfate for tocolysis and
isk of spastic cerebral palsy in premature chil-ren born to women without preeclampsia.m J Obstet Gynecol 2000;183:717-25.1. Costantine MM, How HY, Coppage K,axwell RA, Sibai BM. Does peripartum infec-
ion increase the incidence of cerebral palsy inxtremely low birthweight infants? Am J Obstetynecol 2007;196:e6-8.2. Doyle LW, Crowther CA, Middleton P, Mar-et S, Rouse D. Magnesium sulphate for woment risk of preterm birth for neuroprotection of theetus. Cochrane Database Syst Rev 2009;:CD004661.3. Moher D, Cook DJ, Eastwood S, Olkin I,ennie D, Stroup DF. Improving the quality of
eports of meta-analyses of randomised con-rolled trials: the QUOROM statement, Qualityf Reporting of Meta-analyses. Lancet999;354:1896-900.4. Jadad AR, Moore RA, Carroll D, et al. As-
essing the quality of reports of randomized pJUNE 2009
linical trials: is blinding necessary? Control Clinrials 1996;17:1-12.5. Deeks JJ, Higgins JPT, Altman DG. Analys-
ng data and undertaking meta-analyses. In:iggins JPT, Green S, eds. Cochrane hand-ook for systematic reviews of interven-ions.Version 5.0.1 [updated September 2008].he Cochrane Collaboration, 2008. Hoboken,J: Wiley; 2008.6. Higgins JP, Thompson SG, Deeks JJ, Alt-an DG. Measuring inconsistency in meta-
nalyses. BMJ 2003:327:557-60.7. Egger M, Davey Smith G, Schneider M,inder C. Bias in meta-analyses detected by a
imple graphical test. BMJ 1997:315:629-34.8. Altman DG. Confidence intervals for theumber needed to treat. BMJ 1998;317:309-12.9. Mittendorf R, Covert R, Boman J, Khosh-ood B, Lee KS, Siegler M. Is tocolytic magne-ium sulphate associated with increased totalaediatric mortality? Lancet 1997;350:1517-8.0. Mittendorf R, Dambrosia J, Pryde PG, et al.ssociation between the use of antenatal mag-esium sulfate in preterm labor and adverseealth outcomes in infants. Am J Obstet Gy-ecol 2002;186:1111-8.1. Crowther CA, Hiller JE, Doyle LW, et al;ustralasian Collaborative Trial of Magnesiumulphate (ACTOMg SO4) Collaborative Group.ffect of magnesium sulfate given for neuropro-
ection before preterm birth: a randomized con-rolled trial. JAMA 2003;290:2669-76.2. Magpie Trial Follow-Up Study Collaborativeroup. The Magpie trial: a randomised trialomparing magnesium sulphate with placeboor pre-eclampsia: outcome for children at 18
onths. BJOG 2007;114:289-99.3. Marret S, Marpeau L, Zupan-Simunek V, etl; PREMAG trial group. Magnesium sulphateiven before very-preterm birth to protect infantrain: the randomised controlled PREMAG tri-l*. BJOG 2007;114:310-8.4. Marret S, Marpeau L, Bénichou J. Benefit ofagnesium sulfate given before very pretermirth to protect infant brain. Pediatrics008;121:225-6.5. Rouse DJ, Hirtz DG, Thom E, et al. A ran-omized, controlled trial of magnesium sulfate
or the prevention of cerebral palsy. N EnglMed 2008;359:895-905.6. Hayes E, Moroz L, Pizzi L, Baxter J. A costecision analysis of 4 tocolytic drugs. Am J Ob-tet Gynecol 2007;197:383.e1-6.7. Back SA. Perinatal white matter injury: thehanging spectrum of pathology and emerging
nsights into pathogenetic mechanisms. Mentetard Dev Disabil Res Rev 2006;12:129-40.8. Folkerth RD. Neuropathologic substrate oferebral palsy. J Child Neurol 2005;20:940-9.9. Jensen FE. Role of glutamate receptors ineriventricular leukomalacia. J Child Neurol005;20:950-9.0. Salter MG, Fern R. NMDA receptors are ex-
ressed in developing oligodendrocyte pro-c44re44Mt2
4rcU4n4Hd
4Cmt24dD4
www.AJOG.org Obstetrics Reviews
esses and mediate injury. Nature 2005;38:1167-71.1. Micu I, Jiang Q, Coderre E, et al. NMDAeceptors mediate calcium accumulation in my-lin during chemical ischaemia. Nature 2006;39:988-92.2. Johnston MV, Nakajima W, Hagberg H.echanisms of hypoxic neurodegeneration in
he developing brain. Neuroscientist 2002;8:
12-20. n3. Antonov SM, Johnson JW. Permeant ionegulation of N-methyl-D-aspartate receptorhannel block by Mg(2�). Proc Natl Acad SciS A 1999;96:14571-6.
4. Gathwala G. Neuronal protection with mag-esium. Indian J Pediatr 2001;68:417-9.5. Golan H, Kashtutsky I, Hallak M, Sorokin Y,uleihel M. Maternal hypoxia during pregnancyelays the development of motor reflexes in
ewborn mice. Dev Neurosci 2004;26:24-9. bJUNE 2009 Am
6. Zylinska L, Gulczynska E, Kozaczuk A.hanges in erythrocyte glutathione and plasmaembrane calcium pump in preterm newborns
reated antenatally with MgSO4. Neonatology008;94:272-8.7. Bracewell M, Marlow N. Patterns of motorisability in very preterm children. Ment Retardev Disabil Res Rev 2002;8:241-8.8. Paneth N. Establishing the diagnosis of cere-
ral palsy. Clin Obstet Gynecol 2008;51:742-8.erican Journal of Obstetrics & Gynecology 609