Magnesium Prevents the Cerebral Palsy Precursor in Premature Infants

AMERICAN JOURNAL OF PERINATOLOGY/VOLUME 14, NUMBER 5 May 1997

EFFECT OF MAGNESIUM SULFATE ON THEDEVELOPMENT OF CYSTIC PERIVENTRICULAR

LEUKOMALACIA IN PRETERM INFANTS

Ross B. FineSmith, M.D., * Kevin Roche, M.D.,f Paul B. Yellin, M.D.,*Kevin K. Walsh, Ph.D.,§ Calvin Shen, M.D.,# Mark Zeglis, Atiya Kahn, M.D., *

and Irving Fish, M.D. *

ABSTRACT

To determine if magnesium sulfate has an effect on the development of cysticperiventricular leukomalacia in preterm infants, this retrospective case control studywas conducted. There were 23,382 infants born at three teaching hospitals in the met-ropolitan New York area from January 1992 to December 1994. Four hundred ninety-two infants met our entrance criteria. Criteria included a birth weight <1 750 g, sur-vival to at least 7 days of life and at least one cranial ultrasound after 7 days of life.Infants exposed to magnesium sulfate in utero were less likely to develop periventric-ular leukomalacia. Two of 18 (11%) infants with periventricular leukomalacia wereexposed to magnesium sulfate in-utero compared to 14 of 36 controls (39%) (p =0.035) (OR = 0.196, 95% Cl = 0.039-0.988). Pre-eclampsia as an independent factorwas not associated with a reduced risk (p = 0.251) (OR = 0.294, 95% Cl =0.033-2.65). Preterm infants exposed to antenatal magnesium sulfate were found tohave a reduced risk of developing cystic periventricular leukomalacia.

Keywords: Periventricular leukomalacia; preterm infants; neuroprotection; cerebralpalsy; magnesium sulfate; preterm labor

Each year, more than 50,000 babies with birthweights less than 1500 g are born in the UnitedStates.1 These very low-birth-weight infants are be-ing successfully resuscitated and treated as the re-sult of recent advances in obstetric and neonatalcare. In fact, neonatal mortality has declined 36%since 1980, and the current survival rate of babieswith birth weights less than 1500 g is over 85%.2 De-spite such improvements in survival, neurologicalmorbidity remains a major threat to their long-termoutcome.3-7 One of the most common neurologicalsequela in the preterm infant is cerebral palsy (CP).Many different factors appear to contribute to thedevelopment of CP but strongly associated factorsinclude preterm birth, hypoxia/ischemia and infec-tion.

Naulty et al8 compared recent epidemiologicalfindings on prematurity, low birth weight, and cere-

bral palsy to the findings of the National Collabora-tive Perinatal Project of 1959 to 1966.9 This con-firmed that there are now more infants survivingwith birth weights less than 1500 g and that these in-fants have a threefold increase in the incidence ofCP when compared to the same weight group in theearlier study. This suggests that advances in car-diopulmonary care that allow these infants to sur-vive have not been met with equal advances in ourability to protect the brain.

Cystic periventricular leukomalacia (cPVL) isan ischemic lesion that is a structural antecedent toCP,10 most frequently the spastic diplegic form.cPVL is associated with preterm birth and is attrib-uted in part to the preterm infants metabolicallyvulnerable white matter,11-13 immature vasculatureof the periventricular white matter,14 passive pres-sure cerebral blood flow, and poor cerebral au-

*Department of Neurology, ̂ Department of Pediatric Radiology, and ^Department of Pediatrics, New YorkUniversity, New York; developmental Disabilities Center and #Department of Pediatrics, MorristownMemorial Hospital, Morristown, New Jersey

Request reprints: Dr. FineSmith, Department of Neurology, NYU School of Medicine, 550 First Avenue,New York, NY 10016

Copyright © 1997 by Thieme Medical Publishers, Inc., 381 Park Avenue South, New York, NY 10016. Allrights reserved. 303

Dow

nloa

ded

by: N

YU

. Cop

yrig

hted

mat

eria

l.

AMERICAN JOURNAL OF PERINATOLOGY/VOLUME 14, NUMBER 5 May 1997

toregulation.15'16 PVL has been reported to occur in5-17% of all infants born weighing <1500 g.17~19 Be-tween 66-100% of infants that develop PVL will de-velop CP.16-19 Hence, the incidence of CP is 30-50per 1000/live births <1500 g20-23 compared to 3 per1000 in all births.24"26 The characteristic type of CPpreterm infants develop is spastic diplegia.2127 Thespastic diplegic form occurs in up to 80% ofpreterm infants that develop CP28 compared to16% of all children with CP.29

Nelson and Grether30 recently reported thatantenatal exposure to magnesium sulfate (MgSO4)is associated with a reduced risk of developing cere-bral palsy in very low-birth-weight infants. It was in-teresting to note that of the three preterm infantsthat developed CP in thier MgSO4 exposed group,two had the hemiparetic type that is typically not re-lated to the cPVL lesion and 1 (33%) had the spas-tic diplegic type. In comparison, 23 of the 39 (59%)infants found to have CP that were not exposed toMgSO4 had the spastic diplegic form. Because therewere not cranial ultrasounds reported in Nelsonand Grether's study it is not possible to explore therelationship between MgSO4, cPVL, and CP. How-ever, these data are suggestive that MgSO4 may bemore specifically protective against spastic diplegiaand its antecedent cPVL. In utero, fetal exposure toMgSO4 occurs in the treatment of maternal pre-eclampsia and when the agent is used as a tocolyticto delay labor.

In an effort to determine if antenatal exposureto MgSO4 is associated with any effect on the inci-dence or relative risk of developing cPVL in thepopulation of preterm infants treated at our institu-tions, we performed the following retrospective case-control study. Periventricular echodensities werenot included in this analysis because their signifi-cance are not well understood.

METHODS

The study period covered the years 1992 to1994. During this period all infants born at threeteaching institutions (NYU Medical Center, Morris-town Memorial Hospital, and Bellevue Hospital)were reviewed for entrance criteria for the study.Criteria included birth weights less than 1750 g, sur-vival for more than 7 days, at least one cranial ultra-sound after 7 days of life and must have had all ofthe following information in the infant's chart orthe labor and delivery room records: gestationalage; history of maternal complications and reasonfor prematurity; length of rupture of membranes;whether antenatal steroids, MgSO4, or other to-colytic agents were administered; mode of delivery;evidence of pre-eclampsia; Apgar scores; and num-ber of days intubated. Birth weights were recordedand categorized as between 500-700, 751-1000,1001-1250, 1251-1500, or 1501-1750 g.

All infants received care in one of the three ter-tiary level neonatal intensive care units (NICU).

304 Each center was equipped with a portable neonatal

cranial ultrasound machine with a 7.5-MHz trans-ducer producing transfontanel coronal and sagittalscans. One institution reproduced ultrasound im-ages on paper. Cranial ultrasounds from all infantswere collected and re-read by a single pediatric ra-diologist experienced in newborn diagnostic imag-ing. Cranial ultrasounds were classified as normal,grade I through III intraventricular hemorrhage,intraparenchymal hemorrhage, and cPVL. The ra-diologist was blinded to all clinical parameters. Ul-trasound quality was judged on a 4-point gradingsystem. A quality grade 1 study was a technicallygood study and a grade 4 was a poor-quality study.

Controls were selected from a pool of infantswith a normal cranial ultrasound or grade I intra-ventricular hemorrhage. Infants in the control poolwith birth weights outside the range found in thecPVL group were removed to make the control poolsimilar to the cPVL group. From this control pool,two controls were randomly selected for each cPVLcase.

Chi-square analyses were used to compare inci-dence of cPVL in infants that received MgSO4 inutero with the control group. Odds ratios (OR)were calculated to estimate the relative risk of devel-oping cPVL in the two groups. An OR <1 representsa reduction of a disease outcome related to the in-tervention. We used a 95% confidence interval (CI)as a measure of statistical precision for each OR.

To determine if the cPVL and control groupshad a similar demographic make-up, two separateanalyses were carried out. First, a multivariate testfor group differences was carried out on a numberof noncategorical variables. Specifically, a one-waymultivariate analysis of variance (MANOVA) bygroups (cPVL vs. control) and a univariate analysiswas conducted. We also analyzed the data with logis-tic regression in an effort to identify whether or notthe variables in the study could predict cPVL. Inparticular, we were interested in the nature ofMgSO4 in the prediction model. Criteria used forincluding variables in the prediction model were(a) there was a statistically significant logistic re-gression, and (b) the variables in the model servedto maximize correct prediction of cases in the sam-ple. Because this is a preliminary investigation andthere were no clear a priori reasons to identify pri-macy among variables, hierarchical or stepwise pro-cedures were not used. Second, categorical vari-ables were analyzed by Chi-square analysis.

This protocol was approved by the institutionalreview boards of the participating hospitals.

RESULTS

During the 2-year study period there were23,382 infants born at the three institutions. Therewere 561 (2.4%) infants admitted to one of thethree NICUs with a birth weight of less than 1750 g.Of these infants, 492 met the entrance criteria ofthe study. Seventeen infants died prior to 7 days oflife, 23 infants did not receive cranial ultrasounds

Dow

nloa

ded

by: N

YU

. Cop

yrig

hted

mat

eria

l.

MAGNESIUM SULFATE AND CYSTIC PERIVENTRICULAR LEUKOMALACIA IN PRETERM INFANTS/FineSmith et al

after 7 days of life, and 29 infants had insufficientdata recorded in their charts. There were no casesof cPVL detected in any of the cases excluded fromthe study. A total of 857 ultrasound studies were re-viewed. There were 12 quality grade 2 and 2 qualitygrade 3 ultrasound studies. The remainder weregrade 1 studies.

There were 18 cases of cPVL that formed theexperimental group. Two infants in this group(11.1%) were exposed to MgSO4 in utero comparedto 14 of the 36 (38.8%). Table 1 shows the OR, sig-nificance and 95% CI for MgSO4 exposure. Becauseearlier studies have examined the relationship be-tween preeclampsia and C.P. (see, e.g., Ref. 30), wespecifically examined our data from this perspec-tive. Preeclampsia was diagnosed in 5.5% (1 out of18) of the mothers with children with cPVL com-pared to 14.7% (6 out of 36) of the controls. Toevaluate the possible effect of preeclampsia oncPVL, the number of infants born with cPVL tomothers with preeclampsia were compared be-tween the two groups with results shown in Table 1.We also compared the nonpreeclamptic infants ex-posed to MgSO4 to the nonpreeclamptic controls.In the cPVL group, 2 of the 17 nonpre-eclamptic in-fants were exposed and 9 of the 30 nonpre-eclamp-tics in the control group were exposed (11.8 vs.30%).

The PVL and control group were found to beequal in all variables except magnesium exposure.Analysis was conducted on the continuous and cate-gorical variables. The continuous variables (gesta-tional age, number of days intubated and 1- and 5-min Apgar scores) were found to be nonsignificant(p = 0.53) although days of intubation was found tocovary with each of the variables tested. In part thisis due to the fact that one infant in the cPVL group(mean = 16 days) was intubated for 100 days. With-out this one case, the mean for the group wouldhave been 11 days and the group difference wouldnot have been statistically significant (t= 0.53, p = 0.14).Second, several categorical variables were availablein the records of the infants. These variables in-cluded gestational weight, time of rupture of mem-branes, reason for prematurity, type of delivery,whether or not the mother received terbutaline, ri-todrine, steroids, or magnesium, and the numberof infants in each group that survived. There was

not a statistically significant increase in the risk ofdeveloping cPVL associated with any of these vari-ables except magnesium. The Chi-square tests werenonsignificant suggesting a general similarity be-tween the experimental and control groups exceptfor magnesium, the variable of interest. See Tables1 and 2.

The results of the logistic regression showedthe solution that best satisfied the criteria of statisti-cal significance and maximum predictability in-cluded six variables: gestational weight, antenatalexposure to MgSO4, 1- and 5-min Apgar scores,number of days intubated, reason for prematurity,and type of delivery. The logistic solution includingthese variables was significant (Chi-square = 23.4,df= 12, p = 0.014). Of all the variables in the logisticequation, MgSO4 had the highest value of the R sta-tistic, which in the present analysis is a measure ofthe partial correlation of the variables. The R statis-tic for MgSO4 was significant (p < 0.03, df= 1). The lo-gistic regression model was able to correctly predict67% (12 of 18) of the cPVL cases and 92% (33 of36) of the controls. When this same logistic solutionwas calculated without MgSO4, prediction in thecPVL group dropped to 44%, less than chance, andthe regression was no longer statistically significant.Therefore, inclusion of MgSO4 resulted in a 22% in-crement in prediction and increased prediction17% above chance levels. This bears out the find-ings in the analyses above that MgSO4 plays a role inthe prevention of cPVL.

DISCUSSION

In this study of preterm infants with a birthweight of <1750 g, we found that those infants withcPVL were less likely to have been exposed to ante-natal MgSO4. The groups were shown to be statisti-cally similar in all other variables recorded. Theseobservations are consistent with those of Nelsonand Grether30 who noted a reduction in the risk ofdeveloping CP in the group exposed to antenatalMgSO4. Our findings suggest that the reduction inCP may be related to a reduction in cPVL. However,a larger prospective randomized study is needed tosubstantiate these findings.

Table 1. Maternal and Fetal Variables and Risk of Developing cPVL

Condition

MgSO4 exposedPre-eclampsiaSteroid exposedFetal distressPremature rupturePreterm labor

cPVL casesN=18

No.

21614

10

(%)

11.15.6

33.35.6

22.255.6

ControlsN=36

No.

146

172

1219

(%)

38.916.747.2

5.633.352.8

pValue

>0.035>0.1>0.1>0.1

>.1>.1

OR

.19

.29

.591.0.57

1.1

95% CI

.039-.988

.033-2.65.17-1.8.08-1.18.15-2.12

.058-1.65305

Dow

nloa

ded

by: N

YU

. Cop

yrig

hted

mat

eria

l.

AMERICAN JOURNAL OF PERINATOLOGY/VOLUME 14, NUMBER 5 May 1997

Table 2. Comparison of Neonatal Variables in cPVL and Control Group

Variables

# days intu batedApgar at 1 minApgar at 5 minGestational agebirthweight (g)

cPVL(mean)

16.15.447.11

28.061105

Cases(SD)

27.92.682.222.04

254

Controls(mean)

3.926.317.97

28.081137

(SD)

6.01.821.001.92

279

t Value

0.08 (ns)0.23 (ns)0.13 (ns)0.92 (ns)0.80 (ns)

NS-not statistically significant.

306

Since pre-eclampsia is a common indicationfor antenatal exposure to MgSO4, it would be im-portant to consider the possibility that preeclamp-sia itself may confer some neuroprotection. How-ever, when we looked only at those infants born topreeclamptic mothers, we were unable to find anyevidence of such a primary protective effect with re-gard to cPVL. Conversely, the association betweenantenatal MgSO4 and the reduction in cPVL per-sisted in infants born to nonpreeclamptic motherswho received MgSO4 solely for tocolysis. This fur-ther supports the notion that it is the MgSO4 andnot preeclampsia that is protective.

There is convincing evidence in studies at thecellular level 31 as well as in ischemic animal modelsthat magnesium has neuroprotective properties.32"35

Magnesium markedly reduces infarct size in perina-tal rats. The mechanism of action may be related tothe inhibition of the iV-methyl D-aspartate (NMDA)receptor.36 The NMDA receptor is stimulated by ex-citatory amino acids (EAA) in hypoxic conditionsand this results in a massive inflow of calcium ionsinto the neuron. This has been described as the ini-tial phase in the cascade of neuronal cell death.Magnesium blocks the ion pore in the NMDA re-ceptor in a voltage-dependent manner. Even whenthe NMDA receptor is stimulated by EAAs, calciumcannot enter the cell until the magnesium ion is ex-truded, thereby preventing the initiation of the celldeath cascade.

Other postulated mechanisms may be relatedto the overall metabolic and functional depressiveeffects of magnesium on the brain. Magnesium mayalso act as an antagonist to the presynaptic adeno-sine a3 receptor and inhibit EAA release.

This study is limited by the small number of fol-low-up ultrasounds in those infants found to have anormal ultrasound early in their course. Also, inthis retrospective study we were unable to docu-ment the dose of medication, duration of in-uteroexposure, and serum levels of magnesium in infantsexposed to MgSO4. Additionally, the number ofmothers included in the preeclampsia comparisonswas small making it difficult to demonstrate a statis-tical relationship. As noted, the wide discrepancy inthe number of days intubated between the cPVLgroup (0 to 100 days) and the controls (0 to 20days) made it difficult to determine significance.The logistic regression analysis in this study, while

significant and showing improved predictability,contained more variables than may be practical touse in clinical settings. When we removed some ofthese variables from the regression equation, how-ever, predictability suffered, thus leading us to theconclusion that there are complex interactions inthese variables that we may not fully understand.Such interactions should be the focus of future re-search. Further, the number of subjects in this studyis limited and it may be that a larger sample, withmore stable parameters, may yield more parsimo-nious regression solutions. Finally, in this prelimi-nary analysis, the method of the regression was sim-ply to enter all the variables into the equation at thesame time. Hierarchical and stepwise methods maybe able to refine models in future studies.

Neonatal resuscitation has been greatly im-proved resulting in more timely restoration of cere-bral oxygenation and perfusion in preterm infantswith evidence of hypoxic-ischemic disease (HID).Unfortunately, HID may occur pre- or perinatallywhen the neonate is not accessible to resuscitation.However, the fetus is accessible to in-utero pharma-cological therapy. The occurrence of the HID ismuch greater in the lower birth-weight and gesta-tional age infants.37 Many of these infants have tran-sient respiratory and cardiac complications thateither resolve or are amenable to medical manage-ment. However, these nonpermanent cardiorespira-tory deficiencies may result in permanent braindamage. The outcome and quality of life of theseinfants is often dramatically altered.

The etiology of CP is clearly heterogeneous,but two prominently identifiable and frequently as-sociated risk factors are preterm birth and HID.38 Itis therefore pertinent that the first step in preven-tion be aimed at the treatment of those infants thatare either at greatest risk of, or that have alreadysuffered HID. Comprehensive prenatal care, fetalmonitoring, and improved neonatal resuscitationare currently utilized but this has not been enoughto improve the incidence of CP in the preterm in-fant. Therefore, the next frontier is to learn to pro-tect the neonates brain against cell death and injuryin the face of the HID. With appropriate treatment,and ultimately prophylactic treatment of pretermneonates, the occurrence of neurological sequelaand the quality of life of many thousands of individ-uals could be vastly improved.

Dow

nloa

ded

by: N

YU

. Cop

yrig

hted

mat

eria

l.

MAGNESIUM SULFATE AND CYSTIC PERIVENTRICULAR LEUKOMALACIA IN PRETERM INFANTS/FineSmith et al

REFERENCES

1. Guyer B, Strobino DM, Venttura SJ, Singh GK. Annual sum-mary of vital statistics—1994. Pediatrics 1995;96:1029-39

2. Wagner ME. Annual summary of statistics—1993. Pediatrics1995;94:792-803

3. Horbar JD, McHuliffe TL, Adler SM, Albersheim S, CassadyG, Edwards W, Jones R, Kaftwinkel J, Kraybill EN, KrishnanV, Raschko P, Wilkinson AR. Variability in 28 day outcomesfor very low birthweight infants: An analysis of 11 neonatalintensive care units. Pediatrics 1988;82:554-59

4. Hagberg B, Hagberg G, Olow I. The changing panorama ofcerebral palsy in Sweden. Acta Paediatr Scand 1984;73:433-400

5. Pharoah PD, Cooke T, Rosenbloom I, Cooke RI. Trends inbirth prevalence of cerebral palsy. Arch Dis Child 1987;2:379-84

6. Hagberg B, Hagberg G, Zetterstrom R. Decreasing perinatalmortality—Increase in cerebral palsy morbidity. Acta Paedi-atr Scand 1989;78:664

7. Krudrjavcev T, Schoenberg BS, Kurland LT, Groover, RV.Cerebral palsy—Trends in incidence and changes in concur-rent neonatal mortality: Rochester, MN, 1950-1976. Neurol-ogy 1983;33:1433-8

8. Naulty C, Long L, Pettett G. Prevalence of prematurity, lowbirth weight, and asphyxia as perinatal risk factors in a cur-rent population of children with cerebral palsy. Am J Perina-tol 1994; 11:377-81

9. Nelson KB, Ellenberg JH. Antecedents of cerebral palsy. I.Univariate analysis of risk. AmJ Dis Child 1985;139:1031-8

10. Banker BQ, LarrocheJC. Periventricular leukomalacia of in-fancy: A form of neonatal anoxic encephalopathy. Arch Neu-rol 1962;7:386-410

11. de Vries L, Levene M. Cerebral ischemic lesions. In LeveneM, Lilford R, eds. Fetal and Neonatal Neurology and Neurosur-gery. London: Churchill Livingstone; 1995:323

12. Young RS, Hernandez MJ, Yagel SK. Selective reduction ofblood flow to white matter during hypotension in newborndogs: A possible mechanism of periventricular leukomalacia.Ann Neurol 1982;12:445-8

13. Duffy TE, Cavazzuti M, Cruz NF, Sokoloff L. Local cerebralglucose metabolism in newborn dogs: Effects of hypoxia andhalothane anesthesia. Ann Neurol 1982;11:233-46

14. Cavazzuti M, Duffy TE. Regulation of local cerebral bloodflow in normal and hypoxic newborn dogs. Ann Neurol1982;11:247-57

15. Takashima S, Tanaka K. Development of cerebrovascular ar-chitecture and its relationship to periventricular leukomala-cia. Arch Neurol 1978;35:11-16

16. Volpe, JJ. Neurology of the Newborn. Philadelphia: Saunders;1994:296-9

17. de Vries LS, Eken P, Groendaal F, van Haastert IC, MeinersLC. Correlation between the degree of periventricularleukomalacia diagnosed using cranial ultrasound and MRIlater in infancy in children with cerebral palsy. Neuropedi-atrics 1993;24:263-9

18. Rodgers B, Msall M, Owens T, Guernsey K, Brody K, Buck G,Hudak M. Cystic periventricular leukomalacia and type ofcerebral palsy in preterm infants. J Pediatr 1994;124:Sl-8

19. Fujimoto S, Yamaguchi N, Togari H, Wada Y, Yokochi K.

Cerebral palsy of cystic periventricular leukomalacia in low-birth-weight infants. Acta Paediatrica 1994;83:397-401

20. Pidcock FS, Graziani LJ, Stanley C, Mitchell DG, Merton P.Neurosonographic features of periventricular echodensitiesassociated with cerebral palsy in preterm infants. J Pediatr1990;116:417-22

21. Powell TG, Pharoah PD, Cooke RI, Rosenbloom L. Cerebralpalsy in low-birthweight infants. II. Spastic diplegia: Associa-tions with intrapartum stress. Dev Med Child Neurol 1988;30:19-25

22. Powell TG, Pharoah PD, Cooke, RI, Rosenbloom L. Cerebralpalsy in low-birthweight infants. I. Spastic hemiplegia: Asso-ciations with intrapartum stress. Dev Med Child Neurol1988;30:ll-18

23. Nelson KB, Ellenberg JH. Neonatal signs as predictors ofcerebral palsy. Pediatrics 1979;64:225-32

24. Naeye RL, Peters EC, Bartholomew M, Landis B. Origins ofcerebral palsy. AmJ Dis Child 1989; 143:1154-61

25. Boyle CA, Decoufle P, Yeargin-Allsopp M. Prevalence andhealth impact of developmental disabilities in US children.Pediatrics 1994;93:399-403

26. Murphy CC, Yeargin-Allsopp M, Decoufle P, Drews CD.Prevalence of cerebral palsy among ten-year-old children inmetropolitan Atlanta, 1985 through 1987. J Pediatr 1993;123:S13-20

27. Stanley FJ. Spastic cerebral palsy: Changes in birthweightand gestational age. Early Hum Dev 1981;51:167-78

28. McDonald AD. Cerebral palsy in children of very low birthweight. Arch Dis Child 1963;38:579

29. Jarvis S, Holloway J, Hey E. Increase in cerebral palsy in nor-mal birthweight babies. Arch Dis Child 1985;60:1113-21

30. Nelson KB, Grether JK. Can magnsium sulfate reduce therisk of cerebral palsy in very low birthweight infants? Pedi-atrics 1995;95:263-9

31. Ames A, Nesbett AM. Pathophysiology of ischemic celldeath: I. Time of onset of irreversible damage; Importanceof the different components of the ischemic insult. Stroke1983;14:219-26

32. Izumi Y, Roussel S, Pinard E, Seylaz J. Reduction of infarctvolume by magnesium sulfate after middle cerebral arteryocclusion in rats. J Cereb Blood Flow Metab 1991 ;11:1025-30

33. Schmitt H, Barth G, Thierauf P. Neuronal protection by isch-emic brain perfusion: An electron microscopy study in therat. J Neurosurg Anesthesiol 1994;6:265-74

34. Marret S, Gressens P Gadisseux J, Evrard P. Prevention bymagnesium of exitotoxic neuronal death in the developingbrain: An animal model for clinical intervention studies. DevMed Child Neurol 1995;37:473-84

35. Thordstein M, Bagenholm R, Thringer K, Kjellmer I. Scav-engers of free oxygen radicals in combination with magne-sium ameliorate perinatal hypoxic-ischemic brain damage inthe rat. Pediatr Res 1993;34:23-6

36. McDonald JW, Silverstein FS, Johnston MV. Magnesium re-duces iV-methyl-D-aspartate (NMDA)-mediated brain injuryin perinatal rats. Neurosci Lett 1990;109:234-8

37. McDonald HM, Mulligan, JC, Allen AC, Taylor PM. Neonatalasphyxia. I. Relationship of obstetric and neonatal complica-tions to neonatal mortality in 38,405 consecutive deliveries. JPediatr 1980;96:898-902

38. Nelson KB, Ellenberg, JH. Antecedents of cerebral palsy.Multivariate analysis of risk. NEJM 1986;315:81-6

307

Dow

nloa

ded

by: N

YU

. Cop

yrig

hted

mat

eria

l.