EEG diagnoses of neonatal seizures: Clinical correlations and outcome

EEG Diagnoses of Neonatal Seizures: Clinical Correlations and Outcome Mark

S. S c h e r , M D * , M i c h a e l J . P a i n t e r , M D * , Ira B e r g m a n , M D *

M a m d o u h a A. B a r m a d a , M D t , a n d J a m e s B r u n b e r g , M D ~

Electroencephalographic seizures were evaluated in 112 neonates. The first portion of the study involved 80 neonates with clinically identified abnormal movements, 8 of whom (10%) had electroencephalographic evidence of seizures coincident with this activity. Patients with abnormal movements (90 %) had no concurrent electrical seizures. In the second part of the study, 40 infants who had electrical seizures were investigated. Eight of these infants had been identified during the first part of the study. Two-thirds of the patients (25) were premature. Sixteen patients (40%) died; 90% had brain lesions documented by computed tomography and/or postmortem study. Cerebral infarction and intraventricular hemorrhage were the most common lesions. One-third of the survivors (8 of 24 patients) were normal at a mean age of 3 years, while two-thirds had significant neurologic and developmental abnormalities. Neonatal seizures often are subtle, not associated with observable clinical expression, and associated with adverse development. Electroencephalographic confirmation is important in the evaluation of neonatal seizures. Scher MS, Painter M J, Bergman I, Barmada MA, Brunberg J. EEG diagnoses of neonatal seizures: Clinical correlations and outcome. Pediatr Neurol 1989;5:17-24.

Introduction Neonatal seizures are associated with high mortality and a significant incidence of neurodevelopmental delay [ 1-9]. Prompt therapeutic intervention may be important [4], but recognition of seizures is difficult because of the nature of neonatal convulsive activity. Prompt recognition is hampered by highly variable clinical expression [10-13], the frequent use of paralytic agents [14-17], and the absence of clinical signs [ 18-20]. Certain neurologic events previously considered to be seizures also have been questioned as being either brainstem release phenomena [21-23] or movement abnormalities [23[, rather than seizures.

From the Departments of *Pediatrics and tPathology; Magee-Women's Hospital and Children's Hospital of Pittsburgh; :~Department of Radiology; Presbyterian-University Hospital and Children's Hospital of Pittsburgh; Pittsburgh, Pennsylvania.

Abnormal neonatal movements were correlated with the absence or presence of coincident electroencephalographic (EEG) seizure activity to establish the frequency of seizures as a cause of abnormal neonatal movements. Neonates with seizures by EEG criteria, with or without coincident movement abnormalities, were also studied to assess the clinical, radiographic, autopsy, and outcome characteristics of neonates with EEG seizures.

Methods In the first part of the study, EEG records obtained between July, 1983 and December, 1984 from infants with abnormal movements in the neonatal intensive care unit (NICU) of the Magee-Women's Hospital were tabulated.The suspicious movements were described on the EEG request form and identified at the time of the recording by the electroencephalographer (MSS). Clinical behaviors identified during each recording were classified according to a commonly used classification system [12,131. An EEG evaluation of certain abnormal movements, such as tremors and myoclonus, were routinely obtained to exclude seizures, although this diagnosis may not have been strongly suspected. A second group of infants with electrographic seizures was identified during this same 18 month period. The clinical, radiographic, neuropathologic, and outcome characteristics of neonates with documented electrical seizures were studied. The infants in this second group had EEGs performed in the presence of pharmacologic paralysis lor ventilatory care or for evaluation of an encephalopathic process. Clinical signs observed prior to or coincident with electrical seizures were not necessarily the reason for obtaining the initial EEG. Neonatal hospital records for each of the 40 patients in this second group were retrospectively reviewed by a pediatric neurologist (MJP); details of the hospital course, including neurologic status, major medical conditions, and laboratory and radiologic studies were recorded. Progress notes were reviewed for clinical observations made by nursing personnel and physicians regarding abnormal activity prior to EEG. The reason lot obtaining the EEGs also was recorded in each instance. Antiepileptic drug (AED) use was recorded but AED blood levels were not always obtained coincident with EEGs. The interval between the suggestion of clinical seizure activity and the performance of EEG was noted as either less than or greater than I day. EEGs were obtained with either a Nihon-Kohden model 42(111. 21channel machine, or a 16-channel synchronized video-EEG machine (Telefactor Corporation). Depending on the equipment used, II or 16 cerebral and 5 noncerebral electrodes were applied according to the international 10-20 system. Double interelectrode distances were used and all recordings were obtained at a 15 ram/see paper speed, with low-frequency adjustment at 0.3 Hz. Paper recordings were at least 60 rain in duration; video-EEG recordings were 6-15 hours in length.

Communications should be addressed to: Dr. Scher; Magee-Women's Hospital; Developmental Neurophysiology Laboratory; Forbes Avenue and Halket Street; Pittsburgh, PA 15213. Received September 14, 1988; accepted December 1, 1988.

Scher et al: EEG and Neonatal Seizures

17

IN Eh(is with A H ) s ~ i ¢~p~lticnI,, 24 EEGs with electroclinical seizures in 22 patients 6 EE(]s ~ilh n{, :\I:l)s {6 palicms 39 EEGs of 33 patients (not paralyzed)

r

10 EEGs with AI';Ds (8 p tientst

15 EEGs with electrical sei-/ zures in t 1 patients

[

53 EEGs with seizure activity in 40 patients

5 EEGs with no AEDs (3 patients)

14 EEGs of 14 patients (paralyzed)

Figure 1.

EEGs with seizure activity documented over an 18 month period (July. 1983 - December, 1984)." AEDs = antiepileptic drugs.

All EEGs were interpreted by the same electroencephalographer. The diagnosis of electrical seizures required an evolution of discharges in frequency, morphology, electrical field, and amplitude distinct from background EEG activity. Based on the previously described EEG criteria [24], a seizure duration of at least 10 sec was required for diagnosis. If multiple seizures occurred during EEG, but only 1 seizure had a time-locked clinical accompaniment, the entire EEG then was classified as electroclinical. In several EEGs, neonates had more than 1 type of abnormal movement; the most frequently observed movement was used to classify the clinical seizure type. When electrical seizures occurred without clinical signs, the EEG was classified as electrical without clinical accompaniment. According to previously published criteria [25], asphyxia was defined as an initial arterial pH < 7.20 with a base deficit > 10 mEq/L. The diagnosis of persistent pulmonary hypertension (PPHN) was made only in infants without congenital heart disease, pulmonary parenchymal disease, or significant right-to-left shunting at the ductal level. Estimation of gestational age was made by both the calculation of the last menstrual period and the age determined by clinical (Ballard) criteria [25]. In the rare instances when there was a discrepancy between these estimates, the date obtained by clinical criteria was chosen. Cranial computed tomography (CT) scans were obtained, when possible, during the neonatal period utilizing a GE 8800 with 5 or 10 mm axial planes. A neuroradiologist (JB) independently reviewed all 24 CT scans without knowledge of the clinical history. Gross and microscopic examinations of brain tissue were performed by the same neuropathologist (MB). All survivors received comprehensive follow-up examinations at the Neonatal Follow-up Clinic of Magee-Women's Hospital. Sequential neurologic examinations, developmental assessments, and language evaluations were performed by 1 of 3 pediatric neurologist (MSS, IB. MJP), a nurse practitioner, and a speech/language specialist, respectively, at 2, 4, 6, 9, 12, and ! 8 months of age. After 24 months of age, each child was evaluated annually.

patients with abnormal movements were examined. Abnormal movements were initially observed by nurses in 90% of patients and by physicians in 10%. Table 1 describes the clinical movement pattern associated with the presence or absence of electrical seizures. Fourteen EEGs (16.5%) of 8 patients (10%)had clearly defined electrical seizures at the time that the abnormal movements were identified (Table 1). Sustained clonic movements correlated best with electrographic seizures. Forty-four percent of EEGs obtained during this activity demonstrated electrical seizures. Abnormal activity observed during 52% of the records included tremors and tonic and myoclonic movements. Table I. Suspected seizure diagnosis Number of Records

EEGs with Electrical Seizures

Bicycling

5

0

B uccolingual

5

2

Eye deviation

2

I

Apnea

2

I

Autonomic*

4

0

Irritability

5

0

Suspicious Movements Subtle clinical signs

Results

Clonic

18

8

During the first and second parts of this study, 112 neonates were evaluated by EEG. Eight of 80 patients in the first group had EEG seizures when abnormal movements were observed. These 8 patients then were added to the 32 patients in the second group who had at least 1 EEG with seizures, with or without concomitant clinical signs. Correlation of Clinically-observed Movements and EEG Seizure Patterns. Eighty-five EEGs of 80 preterm and term

Tonic

12

[

Myoclonic

15

I

Tremor (jitteriness)

17

0

85

14

18 PEDIATRIC NEUROLOGY

Vol. 5 No. 1

Total:

* Includes sudden heart rate or blood pressure changes.

Table 2. Fifteen term neonates with electrical seizures Number and Type of Seizure Records

Computed Tomography

Asphyxia, PPHN

1 Electrical (paralysis), 1 Electroclinical (stare)

Multifocal infarction

Developmental delay, spastic quadriparesis, hearing loss; 30 mos

M/3.2

Normal history

2 Electrical, 1 electroclinical (tonic)

L hemisphere infarction

L hemiparesis; 36 mos

M/3.2

Glycine encephalopathy

I Electroclinical (myoclonic)

None

Died on day 5, spongy myelinopathy

F/3.3

Asphyxia

1 Electroclinical (clonic)

Periventricular leukomalacia

Spastic paraplegia; 36 mos

M/4.7

Asphyxia, PPHN

1 Electroclinical (clonic)

L hemisphere infarction

Spastic quadriparesis (R > L), seizures, developmental delay, hearing loss; 36 mos: meningitis; 3 mos

M/3.7

Asphyxia, PPHN

1 Electroclinical (myoclonic)

None

Died on day 3, temporal lobe infarction, cerebral edema

M/3.5

Asphyxia, PPHN

1 Electrical (paralysis), 1 Electroclinical (clonic)

Multifocal infarction

L hemiparesis, hearing loss, behavior problems, seizures; 27 mos

M/3.0

Asphyxia, PPHN

Electrical (paralysis)

None

Died on day 4, multifocal infarction

M/3.7

Normal history

Electroclinical (clonic)

R hemisphere infarction

L hemiparesis; 24 mos

F/3.9

Asphyxia, PPHN

Electrical, 1 electroclinical (buccolingual)

Multifocal infarction

Developmental delay, spasticity; 14 mos

M/3.0

Asphyxia. PPHN

Electrical (paralysis)

R hemisphere infarction

Normal; 14 mos

M/3.9

Asphyxia

Electroclinical (clonic)

L hemisphere infarction

Mild R hemiparesis: 29 mos

M/3.3

Asphyxia, PPHN

Electrical (paralysis)

Periventricular leukomalacia

Spastic quadriparesis, developmental delay; 24 mos

F/3.8

Asphyxia

2 Electrical

L hemisphere infarction

Resolved mild R hemiparesis; 12 mos

M/3.9

Asphyxia

2 Electrical

L hemisphere infarction

Resolved mild R hemiparesis; 9 mos

Sex/Birth Weight (kg)

Etiology

M/4.1

Outcome/Age

Abbreviations: L = Left PPHN = Persistent pulmonary hypertension of the newborn R = Right

Seventeen records were obtained from infants demonstrating t r e m o r s ; n o n e d e m o n s t r a t e d electrical seizures. Two o f the 27 r e c o r d s o b t a i n e d d u r i n g t o n i c or m y o c l o n i c activity d e m o n s t r a t e d electrical seizures. Subtle clinical signs were the m o s t f r e q u e n t l y o b s e r v e d a b n o r m a l activities, but electrical seizure d i s c h a r g e s w e r e o b s e r v e d in o n l y 17% o f E E G s o b t a i n e d d u r i n g this activity.

Electrical Seizures with and without Clinical Expression. All E E G s d e m o n s t r a t i n g seizure activity, with or w i t h o u t clinical e x p r e s s i o n , were studied o v e r this 18 m o n t h period. T h i s g r o u p o f 4 0 p a t i e n t s i n c l u d e d the 8 p a t i e n t s identified in the first group. Figure 1 details 53 records o f t h e s e 4 0 patients. E E G r e c o r d s a n d p a t i e n t data are t a b u l a t e d bec a u s e 7 p a t i e n t s h a d b o t h electrical a n d e l e c t r o c l i n i c a l seizures at different r e c o r d i n g sessions.

Scher et al: EEG and Neonatal Seizures

19

Thirty-nine EEGs demonstrated electrical seizures in 33 patients who were not pharmacologically paralyzed. Fourteen EEGs demonstrated only electrical seizures in 14 paralyzed infants. Twenty-four EEGs (22 patients) were classified as corroborating electroclinical seizures. Clonic activity was the predominant abnormal activity observed during 11 of these records in 10 patients. Six EEGs (6 patients) contained seizures with "subtle" clinical movements, while 5 others (5 patients) demonstrated seizure discharges associated with myoclonus. One EEG documented coincident tonic movements and electrical seizure discharges in 1 patient. Ten EEGs revealed electrical seizure discharges without accompanying clinical activity in 8 patients receiving AEDs. Abnormal movements were observed by medical personnel in 18 patients prior to the initial EEG. Thirteen patients were therapeutically paralyzed. Nine of 40 patients (23%) did not have clinical activity suggestive of seizures; EEGs were obtained to evaluate encephalopathy in the context of each infant's illness (e.g., asphyxia, PPHN). An EEG was obtained within 1 day of these movements in 28 neonates, making it impossible to determine seizure duration. Eight neonates had electrographic seizures recorded at the onset of EEG. Six infants demonstrated status epilepticus. Of the 5 patients with video-EEG monitoring sessions, a single prolonged electroclinical seizure was documented in 2 patients during a 6 hour period. Three patients had multiple electrographic seizures during 9-15 hour recording epochs. One of these patients was paralyzed and 13 electrographic seizures were recorded over t5 hours. Clinical Correlations and Outcome. Tables 2and3 summarize the clinical, EEG, radiologic, autopsy, and followup findings of the 40 patients with electrical seizures. Sixteen patients (40%) died; 24 survivors were followed for 948 months (mean: 36 months). Twenty-five patients were premature, l 3 of whom (52%) died. Fourteen patients were less than or equal to 30 weeks estimated gestational age (EGA) when seizures were observed. Severe handicaps were observed in 9 patients (38% of survivors) consisting of spasticity, language delay, impaired social-adaptive skills, and sensory deficits in varying combinations. Two of these patients had severe central nervous system infections after the neonatal period, probably contributing to their neurologic deficits. Six patients (25%) had mild to moderate handicaps consisting of either spastic hemiparesis or paraparesis unassociated with other deficits. Five patients (23%) were developmentally normal. Four of these patients demonstrated spasticity and motor delay, but these abnormalities resolved by 1 year of age. Psychometric evaluations have not yet been performed on every survivor. Thirty-six of 40 patients (90%) had brain lesions documented by CT (24 patients) and/or autopsy (12 patients). Four patients had normal CT scans. Twenty-three patients, 14 term and 9 preterm, had either unifocal or multifocal ischemic cerebral lesions. Eight had subependymal hemor-

20

PEDIATRIC NEUROLOGY

Vol. 5 No. 1

rhage with intraventricular extension. T,a.,, patients had cerebral nlalfomlations; I patient had spong 3 in3clino p athy. Examples of clectrographic seizures associated with cerebral infarction are included in Figure,, 2A. 2B. 3A. and 3B.

Discussion Neonatal seizures indicate significanl neurologic dysfunction of the immature brain I 12,131 and are among the most common clinical manifestations of neonatal neurologic disease [131. Their occurrence is associated with significant morbidity and mortality [1-9], in large part because of underlying brain injury. Seizure activity, however, may be, per se, injurious to the developing brain. In laboratory animals, an increased vulnerability to brain injury may be related to seizure frequency during the neonatal period 126,271. Optimum therapy is facilitated by a rapid diagnosis, but several unique aspects of neonatal seizures impede prompt recognition. The clinical expression of seizures is variable [10-13] and commonly accepted clinical criteria [I2,131 may not adequately distinguish seizure movements. Clinical classifications in conjunction with electrical patterns will potentially improve the accuracy of observed clinical activity [22]. Our study stresses the importance of EEG criteria in the diagnosis of seizures for both preterm and term neonates. We demonstrated that seizures can either be misdiagnosed or remain undetected unless EEG is utilized [ 14-20]. Some of these patients were therapeutically paralyzed 114-17], while in others, clinical seizures were not observed l 18-20]. The number of paralyzed patients in our cohort who would have had clinical seizures, concomitant with electrical seizures, is unknown. Nonetheless, coincident clinical seizure criteria were met in only 22 of 40 patients (55%) with electrical seizures. Our findings also suggest that the combination of both EEG and clinical criteria more accurately diagnoses and classifies seizures in neonates, including preterm infants. Indeed, the frequency of electrical seizures associated with abnormal movements in preterm infants may be quite different from that in term infants. In our preterm population, for example, we did not observe a consistent electrical association with clonic movements, as was reported by Mizrahi and Kellaway using a predominantly tern1 population [22]. Synchronized video-EEG monitoring for long periods is preferable [22] to paper-recorded studies; however, EEG documentation of seizures in our study included a higher percentage of preterm neonates than the report by Mizrahi and Kellaway [22] (63% vs 31%), and more patients were < 30 weeks EGA (56% vs 9%). Although different combinations of medical conditions or brain disorders may be associated with seizures, a group of etiologies related to hypoxic-ischemic encephalopathy commonly occurred in our patients. For instance, 8 of 15 term neonates (53%) in our cohort developed PPHN after

Table 3. Twenty-five preterm neonates with electrical seizures Sex/GA

(wks)/Birth Weight (kg)

Etiology

Number and Type of Seizure Records

Computed Tomography

Outcome/Age (mos)

F/34/1.9

PPHN

1 Electrical (paralysis)

None

Died on day 4, multifocal infarction

F/25/0.8

Asphyxia, RDS

I Electroclinical (clonic), 2 electrical

Grade 4 IVH

Spastic quadriparesis; 30 mos: ventriculitis; 3 mos

M/25/0.8

RDS

Electroclinical (myoclonic)

Normal

Died on day 15, no autopsy

F/35/2.3

PPHN

Electrical (paralysis)

Normal

Normal; 18 mos

F/26/0.9

RDS

Electroclinical (clonic)

Grade 2 IVH

Normal; 18 mos

M/37/2.2

PPHN

Electrical (paralysis)

None

Died on day 6, muhifocal infarction

F/26/0.8

Asphyxia

Electrical

Normal

Died on day 18, no autopsy

F/25/0.7

RDS

Electrical, 2 electroclinical (clonic)

Grade 4 IVH

Died on day 56, no autopsy

F/26/0.9

RDS

1 Electroclinical (myoclonic)

None

Died on day 107, grade 4 IVH

1=/26/1.0

Asphyxia, RDS, BPD

1 Electrical

Grade 4 IVH

L hemiparesis; 24 mos

F/35/2.5

Trisomy 13

1 Electroclinical (staring)

Holoprosencephaly

Died on day 3

F/23/0.5

Necrotizing enterocolitis

1 Electrical

None

Died on day 11, infarction

M/28/1.0

Herpes encephalitis

2 Electroclinical (staring)

Multifocal infarction

Spastic quadriparesis; 24 mos: blind

M/34/1.7

Asphyxia

1 Electroclinical (hiccough)

None

Died on day 4, multifocal infarction

M/31/2.0

PPHN

1 Electrical (paralysis)

Grade 4 IVH

Resolved spastic paraparesis; 24 mos

M/30/1.2

PPHN

1 Electrical (paralysis)

None

Died on day 9, multifocal infarction

M/26/0.7

Asphyxia

1 Electrical

None

Died on day 17, grade 4 intraventricular hemorrhage

M/26/0.8

RDS, BPD

I Electrical

Normal

Resolved spastic paraparesis; 24 mos

F/35/4.0

IDM, PPHN

1 Electrical (paralysis)

Multifocal infarction

Spastic quadriparesis; 14 mos: blind

F/29/1.3

PPHN

1 Electrical (paralysis)

Mal formation

Spastic paraparesis; 14 mos

M]34/1.5

Twin-to-twin transfusion

1 Electroclinical (apnea)

Periventricular leukomalacia

Normal; 24 mos

M/30/1.6

PPHN

1 Electrical (paralysis), I electroclinical (nystagmus)

Grade 4 IVH

Died on day 72, no autopsy

M/36/3.1

PPHN

1 Electrical (paralysis)

Normal

Normal; 13 mos

M/35/2.0

Meningitis

1 Electroclinical (clonic)

Multiple abscesses

Spastic quadriparesis, developmental delay; 24 mos

F/33/1.9

Asphyxia

1 Electroclinical (clonic)

None

Died on day 4, infarction

Abbreviations: BPD = Bronchopulmonary dysplasia GA = Gestational age IDM = Infant of diabetic mother IVH = lntraventricular hemorrhage

L = Left PPHN = Persistent pulmonary hypertension of the newborn RDS = Respiratory distress syndrome

Scher et al: EEG and Neonatal Seizures

21

Fp~-T:

~v,m

T, -0~

~

T4 " 0 2

F.-C~ C, -0~ F~-C. C 4 -~ T~ c, -c z Cz-C, c,-~, Fz -Cz

Cz-P~ T,-C~

~'~,,,-J.,-~--~--.'~

......... -4,.,,"

d" 'vi 7~. ~'V"',? ~.,'/' V" ~..~v"'4 I~ v~,b

,;yev v ~ } ¢ ~ ~ ? , 4 / l , h ' m e ~ , J t v , ? ~ ' , ? , ~

, - ~ , r - ...... J;v~

7~

",~,,'~.~,~ :I, ?', ~ '\ 71 ,'~, ,~,/~:" t ~''~'¢~,'~¢W~'~'~'W~?t~tt,

~,-Cz

Figure 2. (A) EEG of a term male documents an electrographic seizure in progress when reeording began. No clinical movements were observed. The duration of this seizure was 7.2 rain, and largely involved the sagittal-parasagittal regions. The phenobarbital level was 36.8 at this time. (B) CT of the same patient. There is a large infaJz't in the distribution ~'the right middle cerebral artel T. There is invoh'ement of the right parasagittal region.

significant asphyxial insult. In our group of preterm infants, 9 of 25 (36%) also had a diagnosis of PPHN when EEG seizures were documented. Asphyxia without PPHN was the principal diagnosis determined in 9 of 40 preterm and term patients (23%). Determining the specific cause of seizures may be essential either for treatment or prediction of outcome [5,7,8,13]. EEG seizures correlated significantly with high mortality (40%) and high morbidity in the survivors (66%), based in part on underlying brain lesions. Previously reported associations between neonatal seizures and cerebral infarction [28-31] also are supported by the 58% incidence of cerebral infarction in our patients with EEG seizures. This percentage is higher than has been previously reported [2,5,7,11-13,22], in part because of more frequent use of CT scans or postmortem examinations. The present retrospective study cannot adequately address a number of issues. Some of our patients with only electrographic seizures were pretreated with AEDs prior to EEG; therefore, we are unable to ascertain whether concomitant clinical seizure activity would have accompanied EEG seizures if no medication had been administered. Similarly, pretreatment of patients with suspicious clinical activity before EEG may have suppressed both EEG and clinical seizure activity. Inadequate documentation of AED levels during EEGs prevented verification of the efficacy of treatment. Although several reports indicated that recommended loading doses of phenobarbital, for example, failed

22

PEDIATRIC NEUROLOGY

Vol. 5 No. 1

I

to stop all clinical seizure activity [32-34], studies of EEGs and simultaneous drug level samplings have not been reported. Prospective, randomized clinical trials are necessary to ascertain the effects of AEDs on the clinical and electrical expression of neonatal seizures. The 3 year follow-up evaluations do not reflect the longterm outcome of these patients. Conclusions based on early evaluation of surviving high-risk neonates may be misleading, either by over-emphasizing abnormalities which are transient, or under-emphasizing more subtle neurocognitive and/or behavioral signs which may become function-

Fpo'TI

~--------"T--~~

~--"'----'-'~--r-~x/

~

.t '~' ' - ~ ' Z ~ - ~ - ~

Ts -0, FN-T4

T,% F~'C s

c, -0, Fp4"C4

c,.o, T: .c, C,-T~

EKO.

~,m,w. ==============================

w,e m

~-f-;;;;;;~-~r~dlrlevv~er

::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::

A

Figure 3. (A) EEG of a premature female, therapeutically paralyzed, demonstrates an elec'trographic seizure without clinical accompaniment. The duration of this seizure was 1 rain with greater involvement c)]"the left temporal-central regions. The phenobarbital level was 27 at this time. (B) Brain of the same patient showing a large hemorrhagic' infarction in the left hemisphere with extravasation of blood into the left ventricle.

ally apparent at older ages [35,36]. Assessments made at successively older ages using several psychometric measures are planned to evaluate learning and behavior patterns of survivors during the early school years. It is unknown what contributions specific cortical and/or subcortical regions have on the initiation and propagation of seizure activity in neonates who have no EEG seizures on scalp recordings. For instance, an isolated report docu-

ments subcortical seizures in an atelencephalic infant [37]. Unlike the mature brain, the propagation of seizure discharges may be dependent on subcortical as well as cortical connections [38,39]. Because neonatal seizures are commonly localized, the number of seizures that fail to propagate to the surface may be significant; therefore, the role of subcortical seizure activity is, as yet, poorly described and requires greater attention.

Scher et al: EEG and Neonatal Seizures 23

117] Tharp BR, Latx~y]ic PMThcincidcncc~,{ titial~umumlituThis study was supported in part by a Clinical Investigator Development Award to Mark S. Scher from the National Institute of Neurological and Communicative Disorders and Stroke (NS0I 110}, anti a research grant from the Twenty-five Club of Magec-Women's Hospital. The authors wish to thank Dr. Robert Guthrie and the Neonatology staff at Magee-Women's Hospital, Georgia Cook and Pal O'I)onoghue of the Neonatal Follow-up Clinic, Marquita Beggarly for the EEG recordings, and Diane Minsterman and Sharon Vogel for assistance in preparing the manuscript.

References [1] Burke JB. Prognostic significance of neonatal convulsions. Arch Dis Child 1954;29:342-5. [2] Rose AL, Lombroso CT. Neonatal seizure states: A study of clinical, pathological, and electroencephalographic features in 137 full-term babies with a iong-term follow-up. Pediatrics 1970;45:404-25. [3] Brown JK, Cockburn E Forfar JO. Clinical and chemical correlates in convulsions of the newborn. Lancet 1972;1:135-8. [4] Holden KR, Freeman JM. Neonatal seizures and their treatment. Clin Perinatol 1975;2:3-13. [5] Seay AR, Bray PE Significance of seizures in infants weighing less than 2,500 gm. Arch Neurol 1977;34:381-2. [6] Knauss TA, Marshall RE. Seizures in a neonatal intensive care unit. Dev Med Child Neurol 1977;19:719-28. [7] Ment LR, Freedman RM, Ehrenkranz RA. Neonates with seizures attributed to perinatal complications. Am J Dis Child 1982: 136:548-50. [8] Holden KR, Mellits ED, Freeman JM. Neonatal seizures, 1. Correlation of prenatal and perinatal events with outcomes. Pediatrics 1982; 70:165-76. [9] Bergman I, Painter MJ, Hirsch RP, Crumrine PK, David R. Outcome in neonates with convulsions treated in an intensive care unit. Ann Neurol 1983; 14:642-7. [10] Dreyfus-Brisac C, Monod N. Electroclinical studies of status epilepticus and convulsions in the newborn. In: Kellaway P, Petersen I, eds. Neurological and electroencephalographic correlative studies in infancy. New York: Grune and Stratton, 1964;250-72. [11] Watanabe K, Hara K, Miyazaki S, et al. Electroclinical studies of seizures in the newborn. Folia Psychiatry Neurol Jpn 1977;31:383-9Z [12] Fenichel GM. Neonatal neurology. New York: ChurchillLivingstone, 1985;25-52. [13] Volpe JJ. Neonatal seizures. In: Volpe JJ, ed. Neurology of the newborn, Vol XXII; Major problems in clinical pediatrics. Philadelphia: W.B. Saunders, 1987,129-44. [14] Staudt F, Roth JC, Engel Re. The usefulness of electroencephalography in curarized newborns. Electroencephalogr Clin Neurophysiol 1981 ;51:205-8. [15] Eyre JA, Oozen RC, Wilkinson AR. Continuous electroencephalographic recording to detect seizures in paralyzed newborns. Br Med J [983;286:1017-8. [16] Goldberg RN, Goldman SL, Ramsay RE, Feller K. Detection of seizure activity in the paralyzed neonate using continuous monitoring. Pediatrics 1982;69:583-6.

24

PEDIATRIC NEUROLOGY

Vol. 5 No. I

anti ou|collle o f inf:_nlts p a r a l y z e d wilh neHrcq1111scllJ;it h]l~( I
Cril Care Med 1983:l 1:926-1L 1181 Cnen RW, McCutcheon ('B. Wctmcl I). >,,u dcl ! (ih