Original Articles

MRI Following Severe Perinatal Asphyxia:Preliminary Experience

Maja Steinlin, MD*, Regula Dirr*, Ernst Martin, MDt, Chris Boesch, PhD, MDt,

-.

In 30 children suffering from severe perinatal asphyxia

an attempt was made to determine the early prognostic

signs of severe hypoxic-ischemic brain injury with

magnetic resonance imaging (MRI). Ten early (l-4 days

of age), 16 intermedi ate (2-4 weeks of age), and 38 late

MRI (older than I month of age) procedures were per-

formed on a 2.35 T MR-system. Severe cerebral necro-

sis was suspected by Tz hyperintensity of the white

matter. with blurred limits to the cortex in early MRI'

and was confirmed by T1 hyperintensity of the cortex

in intermediate MRI. Severe cerebral necrosis was

established at 3 months of age. Of the ll children with

this pattern (group A),8 had severe and 3 had moder-

ate cerebral palsy on subsequent examination. Thirteen

children (group B) had normal late MRI scans; none

developed seyere cerebral palsy or marked mental re-

tardation. Two children (group C) had focal ischemic

lesions. Four children had intracranial hemorrhage(group D). Groups A and B did not differ in the severity

of their perinatal histories and findings, suggesting that

MRI during the first 3 months is of significant prog-

nostic value.

Steinlin M, Dirr R, Martin E, Boesch C, Largo RH,

Fanconi S, Boltshauser E. MRI following severe perinatal

asphyxia: Preliminary experience. Pediatr Neurol l99l;1 :

,^""

Introduction

Perinatal asphyxia still has an uncertain prognosis. lt

may cause hypoxic-ischemic brain injury and lead to later

neurologic and mental deficits. Many studies have re-

ported the lack of a clear relationship between biochemical

and most clinical factors and the outcome of asphyxiated

children. Although limited, the best prognostic indicator

appears to be the duration and severity of encephalopathy

during the first days of life and the neurologic findings on

discharge [1-5]. Several studies have assessed computed

tomographic and/or sonographic findings during the first

weeks of life in asphyxiated newborns and have demon-strated their prognostic value [6-9].

Experience with magnetic resonance imaging (MRI) in

hypoxic-ischemic encephalopathy of the human newbom,however, is limited U0l; most studies demonstrated radio-logic and clinical findings in preterm infants with periven-

tricular leukomalacia ll1,l2l. We are aware of only 3

studies that compared MRI findings in term and preterm

infants with their ult imate outcomes [3-15].We report our experience using serial MRI to study

severe perinatal asphyxia. We analyzed MRI pattems and

their possible prognostic value by comparing the imagingfindings with subsequent neurologic and developmentaloutcomes.

Methods

During 19t16 to 1989, 30 chi ldren (13 gi r ls , l7 boys) wi th h istor ies of

severe perinatal asphyxia underwent MRI scans. During the last 15

months of the period a study was begun to include all children referred

to our neonatal intensive care unit (NICU) after severe perinatal as-

phyxia. The examinations were performed with the approval of the

local ethics committee and the informed consent of the parents. Twen-

ty-eight children were referred to our NICU immediately after birth;

the remaining 2 were examined at our outpatient clinic because of

neurologic problems. Twenty-three were bom at term and 7 were pre-

tem (mean gestational age: 37.8 weeks; range: 28-42 weeks). The rela-

tively small number of preterm infants is a consequence of our exclu-

sion criteria (see below).Asphyxia is determined by ischemia and hypoperfusion and leads to

a great variability of clinical symptoms which can be divided into the

following:(l) Intrauterine signs of asphyxia: bradycardia (heart rate < tt0/min)

limited beat to beat variability, late decelerations, or meconium-stained

amniotic fluid;(2) Peripartum signs: an Apgtr score < 5 at 5 min and < [J at l0 min,

umbilical artery pH < 7.1, and base excess < -10 mmol,/L; and,

(3) Postpartum signs of encephalopathy during the first 48 hours,

such as seizures, lethargy, or pathologic spontaneous movemcnts.

Because a clinical definition of asphyxia is not available [16], we

assumed asphyxia to be severe when at least I criterion was fulfilled in

2 of 3 subgroups. Patients with dysmorphic syndromes or malfbrma-

tions and those who required surgery during the neonatal period or who

had hyaline membrane disease were excluded from the study.

From the Departments of *Neurology, iMagnetic Resonance' rGrowth

and Development, and "Nlntensive Care; Children's University Hospital;Zurich. Switzerland.

Communications should be addressed to:Dr. Steinlin; Children's University Hospital; Steinwiesstrasse 75CH-80-12 Zur ich. Su i tzer landReceived October i9, 1990; accepted February 7,1991.

164 PEDIATRIC NEUROLOGY Vol.7 No.3

[26] Safdari H, Boluix B, Gros C. Multifocal brain radionecrosis [39] Ron E, Modan B, Boice JD, et al. Tumors of the brain andmasquerading as tumour dissemination. Surg Neurol 1984:'21:35-41. nervous system after radiotherapy in childhood. N Engl J Med 1988;

[27] Breuer AC, Blank NK, Schoene WC. Multifocal pontine le- 319:1033-9.sions in cancer patients treated with chemotherapy and CNS radiother- [40] Chung CK, Stryker JA, Cruse R, Vannuci R, Towfighi J.apy. Cancer 1978l'41:.2112-2O. Glioblastoma multiforme following prophylactic cranial irradiation and

[28] Crosley CJ, Rorke LB, Evans A, Nigro M. Central nervous intrathecal methotrexate in a child with acute lymphocytic leukemia.system lesions in childhood leukemia. Neurology 1978128:678-85. Cancer l98l;47:2563-6.

[29] Cohen ME, Duflher, PK, Terplan KL. Myelopathy and severe [41] Malone M, Lumley H, Erdohazi M. Astrocytoma as a secondstructural derangement associated with combined modality therapy. malignancy in patients with acute lymphoblastic leukemia. CancerCancer 1983;52:1590-6. 1986;57:1979-8-5.

[30] Locksmith JP, Powers WE. Pemanent radiation myelopathy. [42] Kitanaka C, Shitara N, Nakagomi T, et al. Postradiation astro-Am J Roentgenol 1968;102:916-26. cytoma. J Neurosurg 1989;70:469-74.

[31] Wara WM, Phillips TL, Sheline GE, Schwade JG. Radiation [43] Soffer D, Gomori JM, Seigal T, Shalit MN. Intracranialtolerance of the spinal cord. Cancer 1975;35:1558-62. meningiomas after high-dose irradjation. Cturcer 1989;63:1514-9.

[32] Pallis CA, Louis S, Morgan RL. Radiation myelopathy. Brain [44] Shapiro S, Mealey J Jr, Sanorius C. Radiation-induced intra-196l;84:460-78. cranial malignant gliomas. J Neurosurg 1989;11:11-82.

[33] Reagan TJ, Thomas JE, Colby MY Chronic progressive radia- [45] Raventos A, Duszynski DO. Thyroid cancer following inadia-t ion myelopathy. JAMA 1968;203:106-10. t ion formedul loblastoma. AJR 1963;89:17-5-81.

[34] ()lsen JH. Risk of second cancer after cancer in childhood. [46] Andrew DS, Kerr IF. Carcinoma of thyroid following iradia-Cancer l9tl6:57:2250-4. tion for medulloblastoma. Clin Radiol 1965:16:282-3.

[35] Farwell J, Flannery JT. Second primaries in children with cen- [47] Horowitz ME, Mulhern RK, Kun LE, et al. Brain tumors intra lnervoussystemtumors.JNeurooncol 1984;2:311-5. the very young chi ld: Postoperat ive chemotherapy in combineo-

[36] Meadows AT, Baum E, Fossati-Bellani F, et al. Second malig- modality treatment. Cancer 1988;61:428-34.nant neoplasms in children: An update from the Late Effects Study [48] Baram TZ, Eys JV, Pack B. Optimal surgery and MOPPGroup. J Clin Oncol 1985;3:532-8. chemotherapy in infants with primitive neuroectodennal tumors of the

[37] Haymaker W, Rubinstein LJ, Miquel J. Brain tumors in ir- posteriorfbssa. Ann Neurol 1986;20:423.radi ated monkeys. Acta Neuropathol l9l 2;2O:261 -1 1 .

[38] Krupp JH. Nine-year rnortality experience in proton-exposed

macaca mulatta. Radiat Res 1916:.61:244-51.

Cohen and Duffner: CNS Treatment and Cancer I 63

Table 1. Group A with diffuse MRI lesions

PatientNo./Sex/GA

UFI39

2lpl39

3E/41

4/p/38

5E/41

6/F14l

1lMl42

Early MRI

4 days: diffuse Trflzhyperintensity, cortexlimits disturbed

I day: cortex limitsdisturbed

I day: T2 hyperinten-sity frontally hypo-intense stripes in T2

Intermediate MRI

2 wks: Tr hyperintensitycortex, lack of myelina-tion capsula intema

2 wks: Tr hyperintensitycortex, cerebral lique-faction, lack of myelin-ation capsula intema

2 wks: Tr hyperintensitycortex, lack of myelin-ation capsula intema

2 wks: mild T1 hyperinten-sity cortex, T2 hyperinten-sities thalami

2 wks: mild Tr hyperinten-sity cortex

I wk: Tz hyperintensitiesfrontally

4 wks: Tr hyperintensitycortex, cerebral lique-faction, lack of myelin-ation capsula intema

Late MRI

l0 mos: severe cerebralnecrosis, thin corpuscallosum

l9 mos: severe cerebralnecrosis, thin corpuscallosum

2 mos: cerebral necrosis,moderately disturbedmyelination

26 mos: Tz hyperintensitiesthalami, pallida; frontallyenlarged fluid spaces

3 mos: severe frontal atrophy

3 mos: delay of myelinationoccipitaVfrontal; 9 mos:frontal periventriculardysmyelination

28 mos: severe cerebralnecrosis, degeneration ofdescending tracts

2 mos: mild T1 hyperintensityof the cortex; 5 mos: mod-erate, diffuse atrophy

75 mos: diffuse median cere-bral necrosis, thin corpuscallosum with disturbedmyelination

7 mos: diffuse subcorticalcerebral necrosis, thin cor-pus callosum with disturbedmyelination, lack of myelin-ation capsula intema

Outcome

Severe quadriplegia,microcephaly, hearingloss; DQ: 0.22 (9 mos)

Severe quadriplegia,blindness, microcephaly;DQ: 0.16 (24 mos)

Severe quadriplegia;DQ: < 0.3 (6 mos),?blindness

Opisthotonus, quadriple-gia; DQ: < 0.5 (3 mos)

Severe quadriplegia,good visual/socialcontact; DQ: 0.25(36 mos)

Mild quadriplegia,visual retardation (3mos)

Mild quadriplegia.DQ: 0.9 (9 mos)

Severe quadriplegia,epi lepsy; DQ: 0.1(31 mos)

Mild quadriplegia;DQ: 1.0 (5 mos)

Severe quadriplegia,blindness; DQ: < 0.2(75 mos)

Severe quadriplegia,epi lepsy, b l indness;DQ: < 0.2 (7 mos)

8lFl3e

9M128

t0Ml40

tUp136

Abbreviations:DQ = Development quotient = developmental age/chronologic age

GA = Gestational age

All children underwent l-4 MRI examinations between the second

day and sixth year of life. Not all MRI investigations were obtained at

the same pref'erred time, mainly due to circumstantial reasons (i.e.,

limited access to MRI, lack of nursing staff, clinical instability). The

timing of the investigations is defined as follows:

Early MRI - during the first 4 days of life;

Intermediate MRI -2-4 weeks after birth; and,

Late MRI - after the first month of life.

Ten children were examined with early MRI scans, 14 were exam-

ined with 16 intemediate scans, and 29 were examined with 37 late

scans. The children were sedated for the examination using chloral hy-drate (50-100 mg/kg body weighQ and/or flunitrazepam (0.05 mg,&gbody weight) administered either orally or rectally. Tr- (TR: 500; TE:30) and Tz- (TR: 3,000; TE: 120) weighted images were performed ona 2.35 T superconducting MR-system, using a 256 x 256 irnagingmatrix !71. All MRI scans were evaluated for gross morphology, stageof myelination !81, or bleeding [9] and ventricular size.

Twenty-seven of 30 infants were examined l-3 times in neurologicfollow-up. Twenty-five of these patients underwent developmental test-ing (Griffiths' test l20l or Snijder's Omen nonverbal intelligence test

Steinlin et al: MRI and Perinatal Asohvxia 165

[21]). Seven of them were evaluated between 3-6 months of age and 8

between 6-12 months. Nine children were tested during the second

year of life and 1 child between 2 and 3 years of age. Two children

died during the fi$t year of life; autopsy was permitted only in I

patient.

Results

Based on MRI findings, the patients were4 groups:

Group A - with diffirse brain injury;Group B - with normal MRI findings

pathology;Group C - with focal pathologic findings;Group D - with intracranial hemorrhages.

divided into

or transient

and,

Detailed analysis of perinatal data of groups A and B didnot reveal significant differences (i.e., children in group Awere not more severely affected).

G r o u p A ( N = 1 1 )

The findings are summa"rized in Table 1. Eleven childrenhad diffuse MRI abnormalities. Three early MRI scansdemonstrated hyperintensities of the white matter in Tz-weighted images with bluned cortical margins (Fig 1).

Of 7 intermediate MRI scans, hyperintensity of the cor-tex in Tt-weighted images was evident in Patients 1-3 and8 (Fig 2) and diencephalicTz hyperintensities in I patient.

Ten children had late MRI scans performed. Six MRI

scans disclosed severe, diffuse necrosis of the cerebralhemispheres (Patients 1-3,8,10,11; Fig 3); 5 were predom-

inantly in the occipital arca (Fig ). Myelination was dis-turbed or severely delayed in all 6 children. Lack of myel-ination of the intemal capsule was manifest in all after thefirst month of life (Fig 4).

In i child with frontal hyperintensity visualized duringintermediate MRI, retardation of myelination at 3 monthsled to patchy dysmyelination at 9 months (Fig 5).

Patient 5 exhibited T2 hyperintensities in the thalamus

and globus pallidum (Fig 6).All 6 children with diffuse cerebral necrosis suffered

from severe spastic quadriplegia and developmental delay(Patients 1 -3, 8, 10, 1 1). Both children with alterations in thediencephalon (Patients 4,5) already had severe spasficquadriplegia during the neonatal period.

G r o u p B ( N = 1 3 )

Relevant findings are summarized in Thble 2.All 13 children had, by definition, normal late MRI

scans.Three of 6 early MRI scans were norrnal (Patients 12,13,

15). Two had fine, hypodense centrifugal stripes periven-

tricularly (Fig 7) and in the centrum semiovale (Fig 8) on

Tz-weighted images (Patients 16,18). Myelination was

normal in all 6.Only 1 of 6 intermediate MRI scans in 5 children was

judged to be normal (Patient 19); the others demonsffatedeither myelination problems or in 1 patient hypodense

stripes.

166 PEDIATRIC NEUROLOGY Vo1.7 No.3

Figure 1. Diffuse hyperintensity of the white matter with blurredmargin to the cortex. Tz-weighted (TR: 3,000; TE: 120) MN of PatientI (group A) at 4 days of age-

Figure 2. Relative hyperintercity of the cortex. Tyweighted (TR: 500;TE: 23) MRI of Patient 2 (group A) at I I days of age.

Figure 3. Extensive cerebral necrosis with enlarged ventricles andsubarachnoidal spaces. Tz-weighted (TR: 3,000; TE: 120) MRI of Pa-tient 8 (group A) at 27 months of age.

Figure 4. Occipitally pronounced cerebral necrosis, deficient myelin-ation of the internal capsule. T2-weighted (TR: 3,000; TE: 120) MRI ofPatient 2 (group A) at 17 months of age.

Figure 5. Patchy, irregular myelination of the frontal white matter.Tz-weighted (TR: 3,000; TE: 120) MRI of Patient 7 (group A) at 9months of age.

Figure 6. Hyperintensities of the lateral thalamus and striatum. Ir-regular appearance of the frontal white matter. T2-weighted (TR:3,000; TE: 120) MRI of Patient 5 (group A) at 22 months of age-

Sixteen late MRI scans were performed in 9 children. In4 children we found diencephalic or occipital myelinationproblems resolving during subsequent examinations. Allother MRI scans were normal.

Nine group B patients were neurologically abnormal ondischarge from the NICU; 4 were considered to be normal.

Four children were nofinal during subsequent neuro-logic examinations (Patients 12,19,20,24). The others suf-fered from minimal signs of cerebral palsy.

Seven children (Patients 12,13,16,17,20,23,24) werenolmal and 2 (Patients 14,19) had minimally retarded de-velopment. In Patients 15, 21, and 22, retardation was notmanifest, but testing was refused by the parents. Patient 18has not yet been reassessed.

GroupC(N=2)

Only 2 children had focal ischemic infarcts.

GroupD(N=4)

Four children, 3 of whom were preterm, suffered fromintracranial hemorrhage. Localizatron and degree of hem-orrhage were variable.

Groups C and D were too small and too variable to drawany conclusions and will not be discussed further.

Discussion

Children of group A (i.e., diffuse hypoxic-ischemicbrain injury) for whom early or intermediate MRI exam-inations were possible (N = 8) were bom at term. They hadthe characteristic pattern of necrosis in the cortical andsubcortical watershed regions, following the age-de-pendent development of the ventriculofugal arteries(Barkovich and Truwit [3]). Early MRI scans were char-acterized by diffuse hyperintensity of the white matter onT2-weighted images with blurred limits to the cortex (Fig1) and hypointensity of the white matter on T1-weightedimages with a sharp demarcation to a small cortex. Thesefindings are probably due to increased water content as aconsequence of ischemia and following severe edema andnecrosis. In our experience, differentiation between edemaand necrosis was not possible.

However. in all 6 intermediate MRI scans of childrenwith verified severe cerebral necrosis, marked hyperinten-sity of a thin cortex on T1-weighted images was manifest(Fig2), probably because ofthe relative hypointensity ofthe white matter as a result of increased water content. Wehave never seen this pattern in the other groups; mildhyperintensity of the cortex in intermediate MRI may bea predictor of moderate cerebral necrosis.

Late MRI demonstrated that severe brain injury wasestablished at 3 months of age (Fig 3). A lack of myelina-tion of the intemal capsule (Fig a) was observed after thefirst month, while deficient myelination of the colpus cal-losum had become evident during the first year of life.

Of 9 children,5 had parieto-occipitally (Fig a) and I hadfrontally pronounced alterations, sustaining the notion of

Steinlin et al: MRI and Perinatal Asphyxia 167

Table 2. Group B with normal MRI findings or transient pathology

PatientNo./Sex/GA

tz^4l4l

t3Ml40

t5A4l40

t6Ml42

z0M/39

t7lpl38

t8Ml31

t4A4/40

Early MRI

2 days: normal

4 days: normal

3 days: normal

2 days: Tz hypoin-tense srlpes

2 days: Tz hypoin-tense stnpes

lntermediate MRI

2 wks: diencephalon,delayed myelination

I wk: Tz hypointensestripes

I wk: normal

2 wks: diencephalon,delayed myelination

I wk: premature gyralpattem, hyperintensewhite matter

Late MRI

9 mos: normal

3 mos: normal

6 mos: normal

3 mos: normal

3 mos: normal

28 mos: normal

10 mos: normal

7 wks: normal

46 mos: ?Tz hyperin-tensity thalamus

l8 mos: minimalfrontal atrophy

3 mos: Tz hyperin-tensities pallidum;9 mos: normal

Outcome(age at follow-up)

Normal; DQ: l l (9 mos)

Minimal hypotonia,limited visual contact;DQ: 1.0 (3 mos)

Minimal hemisyndrome;DQ:0 .95 (18 mos )

Normal by history

Minimal quadriplegia;DQ: 1.0 (3 mos)

Minimal hemisyndrome;DQ: 1.0 (3 mos)

Too young for follow-up

Neurologically normal;DQ: 0.85 (27 mos)

Neurologically normal;DQ: 1.0 (10 mos)

Lost to follow-up

Minimal hemiparesis

Minimal quadriplegia;DQ: 1 .0 ( 18 mos )

Neurologically nomal;DQ: Ll (9 mos)

r9Ml38

2tMl38

22M139

23lMl3s

21lFl39 4 days: white mat-terl T: hYPerin-tensity, Tr hypo-intensitY

Abbreviations:DQ = Developmental quotient = developmental age/chronologic age

GA = Gestational age

high sensitivity to ischemia of the posterior part of the

brain.In 2 children, diencephalic isohemic lesions were de-

tected, combined with mild cortical necrosis (Fig 6) in I

child confirmed later by autopsy findings. The typical clin-

ical pattern was severe spastic quadriplegia evident soon

after birth and less pronounced mental deficit. Only I of

2 children suffered complete circulatory arrest which

could explain damage in the area with the highest meta-

bolic requirements because of preserved vascular autoreg-

ulation, as observed by Barkovich and Truwit [14].Cerebral myelination was delayed or even disturbed

(i.e., patchy dysmyelination) in all 9 children with severe

damage. Barkovich and Truwit reported that in severe

insults, myelination delay can be detected in children up

to 4 years of age [14]. Our oldest patient was 6 years of

agel it appears that, with very severe damage, myelination

retardation can persist. In I patient delayed myelination at

3 months of age was a predictor of periventricular patchy

disturbed myelination in later scans.Only 2 children in group A were preterm. The one bom

at 36 weeks gestation demonstrated the typical pattern of

diffuse subcortical cerebral necrosis and less affected cor-

168 PEDIATRIC NEUROLOGY Vol.7 No. 3

tex. The one bom at 28 weeks gestation had moderate,diffuse atrophy with diminished white matter.

In all patients, the cerebellum and the brainstem werenormal; infratentorial structures are most resistant tohypoxemia.

All 7 children with severe, diffuse cerebral necrosissuffered from prominent spastic quadriplegia and markedgeneral developmental delay. Spastic quadriplegia waseven more prominent in children with diencephalic le-sions; general developmental delay was similar, but socialand visual contact was better. Moderate atrophy andpatchy disturbed myelination were combined with mild tomoderate spastic quadriplegia and only mild developmen-tal problems.

Early or intermediate MRI scans were found to be nor-mal in 9 of 13 children in group B (by definition normallare MRI).

In 2 children we found moderate, diffuse T2 hyperinten-sity of the white matter with early MRI, a sign of increasedwater content because of resolving edema.

Three patients had early or intermediate MRI hypoin-tense centrifugal stripes, periventricularly near the frontalhoms or in the centrum semiovale (Figs 7,8), which dis-

Figure 7. Centrifugal hypointense stipes in the deep frontal whitematter. Tz-weighted (TR: 3,000; TE: 120) MRI of Patient 16 (group B)at I day of age.

appeared without any sequelae in further studies. The ana-tomic pattem and intensity of these stripes on MRI arecompatible with minimal intracerebral bleeding by diape-desis due to venous congestion after acute asphyxia.

In these 9 patients, normal early and intermediale myel-ination assessed by MRI was within the normal timetable,but most manifested slowed subsequent myelination.Myelination abnormalities were pronounced occipitally,although the myelination process in the occipital lobe isusually more advanced than in the frontal lobe. In 2 lateMRI scans we found diencephalic T2 hyperintensities thatwere not evident on repeated tests. These findings point to

Figure 8. Hypointense stripes in the central corona radiata. Tz-weighted(TR:3,000;TE: 120) MRI of Patient 16 (group B) at I dayoJ age.

the probability that perinatal asphyxia may lead only totransient retardation or disturbance of myelination withsubsequent recovery. Our findings confirm those ofJohnson et al. [13], although their myelination stages weredefined in T1-weighted images; however, in other respectscomparison of our results with those of Johnson et al. isnot possible. Their series included only 4 patients withhypoxic-ischemic encephalopathy and early and inter-mediate MRI scans usually were not performed.

In group B only 3 of I I children were normal in alllongitudinal neurologic examinations. Three patients hadtransient minimal cerebral paresis, while 5 had mild neu-

Birth Asphyxia

group A group BIIRI

ear ' ly:

interediate:

late; 3 m

9-12 rc

2 years

0uTc0!tE

Ii,::::":::;::'11.,."IIV

T1 hype.intensity cortex

III

establ ished necrosis

Il ack o f nye l ina t ionof Cl and cc

IY

' lack of myel inat ion

descending tracts

mi ld T1 diencephal ic ischsia

hyperintensity cortex ni ld Tl hyperintensity

l rl r

n o d e r a t e n e c r o s i s d i e n c e p h a l i c i s c h f l i a

mderate necrosis

*i't

T2 hyperintensity noml

I / Id e l a y e d n y e l i n a t i o n n o r n a l

I \ lpatchy myel inat ion noml

i s c h e m i c l e s i o n s

/nyper

intense str ipes

intense str ipes

cerebra l severe 6/6

pares is:

oeveloprent severe 6/6

retardat ion:

noml 1/1 severe 2/2

m i l d 1 / 1 serere 212 ni ld 2/2

m i l d 1 / 2

normal 1/2

mi ld 5/13

nom'l 5/13

unknorn 2/13

n o m a l l 0 / 1 3

n i l d l / 1 3

unknorn 2/13

Figwe 9. Surnmary of MN patterns and outcotnes following asphyxia in our series.

Steinlin et al: MRI and Perinatal Asphyxia 169

rologic abnormalities or developmental delay, but normalMRI at the last evaluation. An explanation for this obser-vation could be the disappearance of lesions or minimalcerebral damage not visible by MRI. Transient pathologicfindings could not be correlated with neurologic or devel-opmental problems. Normal developmental assessments in8 of I I children should be taken with caution because ofearly testing age.

Our findings are comparable with the study by Byrneet al. [15]. The authors performed serial MRI scans inpatients with neonatal hypoxic-ischemic encephalopathy.ln 6 of 8 infants with abnormal MRI scans at 4 months ofage, cerebral palsy was found at l8 months; one of theremaining patients had findings suggestive of cerebralpalsy, while the other one was normal. Three of 4 infantswith normal MRI scans at 8 months had normal outcomes.Our observations are summarized in Figure 9. With con-siderable assurance MRI allows recognition of severe en-cephalopathy as early as the first week of life. This flndingcan be confirmed l0-14 days postpartum. The predictivevalue of early and intermediate MRI has to be confirmedin a larger patient sample.

Supported by the Swiss National Foundation, grant nos. 3.917-0.88 and

3824-0.87.

References

[] Robertson C, Finer N. Term infants with hypoxic-ischemic en-

cephalopathy: C)utcome at 3.5 years. Dev Med Child Neurol 198-5;27:

473-84.

[2] Aylward GP, Verhulst SJ, Bell S. Correlatiolt oi asphyxia and

other risk lactors with outcome: A contemponuv view. Dev Med Child

Neurol 19u9:3 I :329-40.

t3l Vieneta R. Perinatal asphyxia, biochemical parameters as in-

dices of asphyxia at birth and predictors ol brain damage, zrnd a trial of

preventing damage by phenobarbital. Academic thesis, Helsinki, 1988.

[4] Lupton B, Hill A, Roland E, Whitfleld M, Flodmark O. Brain

swelling in the asphyxiated term newborn: Pathogenesis and outcome.

Pediatrics I 988:82: I 39-.16.

[5] Sarnat H. Siunat M. Neonatal encephalopathy fbllowing f'etal

distress. Arch Neurol 1976:33:696-705.

[6] Adsett DB, Fitz CR, Hill A. Hypoxic-ischemic cerebral injuryjn the term newbom: Correlation ol CT findings with neurological out-come. Dev Med Chi ld Neurol 1985127:155-60.

[7] Fitzharding PM, Flodmark O, Fitz CR, Ashby S. The prognos-tic value of computed tomography as ur adjunct to assessment of theterm infant wi th postasphyxia l encephalopathy. J Pediatr l98l ;99:111-8t .

[tt] Fkrdmark O, Fitz CR, Harwood-Nash DC. Computed tomo-graphic diagnosis and short-term prognosis of intracranial hemorrhageand hypoxic-ischemic brain damage in neonates. J Comput AssistTomogr I 980;4:775-87.

[ 9 ] L i pp -Zwah len AE , Deonna T , M i chek i JL , Ca lame A ,Chrzanowski R, C0tre E. Prognostic value of neonatal CT scans in as-phyxiated term babies: Low density score compared with neonatalneurological s igns. Neuropediatr ics l9U5;16:209-17.

[0] McArdle CB, Richardson CJ, Hayden CK, Nicholas DA,Amparo EG. Abnormalites of the neonatal brain: MR imaging, part lI.Hypoxic-ischemic brain injury. Radiology l9fl7;163:395-403.

[l] Fkrdmark O, Lupton B, Li D, et al. MR imaging of periven-t r icular leucomalacia in chi ldhood. AJR 1989: l -52:583-90.

[12] Dubowitz L, Bydder G, Mushin J. Developmental sequence ofperiventricular Ieukomalacia. Correlation of ultrasound, clinical and nu-clear magnetic resonance functions. Arch Dis Child 1985;60:349-5-5.

l l3 l . Iohnson MA, Pennock JM, Bydder GM, Dubowitz LMS,Thomas DJ, Young IR. Serial MR imaging in neonatal cerebral injury.AJNR l9tJ7:u:u3-92.

[4] Barkovich A, Truwit CH. Brain damage fiom perinatal as-phyxiir: Correlation of MR findings with gestational age. AJNR 1990;I l : I 087-96.

[5] Byrne P, Welch R, Johnson MA, Danah J, Piper M. Serialmagnetic resonance imaging in neonatal hypoxic-ischemic encephalop-athy. J Pediatr 1990;l l7:694-700.

116l Hal l D. Bir th asphyxia and cerebral palsy. BMJ 1989;299:2t9-82.

llTl Boesch C. Martin E. Combined application of MR imagingand spectroscopy in neonates and infants: Installation and operation ofa 2.3-5 Tesla system in a c l in ical set t ing. Radiology 1988; i68:481-8.

[8] Martin E, Kikinis R, Zuerrer M, et ai. Development stages ofhuman brain: An MR study. J Comput Assist Tomogr 1988;12:911-22.

[9] Ziirrer M. Bestimmung von Alter und Ausdehnung intrak-ranieller Blutungen bei Neugeborenen und Siuglingen mittels Mag-netresonanz. MD Thesis, Ziirich, 1989.

[20] Griffith R. Griffiths Entwicklungsskalen zur Beurteilung derElrtwicklung in den ersten beiden Lebensjahren/dt. Bearbeitung BrandtI . : Bel tz. Basel : Weinheim, 1983.

[21] Snijders J, Snijders-Omen N. Snijders-Omen nicht verbale In-telligenztestreihe S.O.N. 2 l/2 -7. Groningen: Wollers-Nordhoff , 19'71 .

170 PEDIATRIC NEUROLOGY Vol.7 No. 3