вход по аккаунту


Childhood brain tumors presenting as chronic uncontrolled focal seizure disorders.

код для вставкиСкачать
Chddhood Brain Tumors
Presenting as Chronic Uncontrolled
Focal Seizure Disorders
Warren T. Blume, MD, FRCP(C), John P. Girvin, MD, FRCS(C),
and John C. E. Kaufmann, MB, ChB, FRCPath
Sixteen of 35 patients (46%) 21 years old or less who underwent surgical resection of a longstanding epileptic focus
harbored tumors. The median duration of the seizure disorder prior to operation was 6.0 years for patients with and
7.6 years for those without tumor. Among the 35 patients, tumor was more common when intelligence and results
of neurological examination were each normal, a plausible cause for uncontrolled seizures was lacking, and persistent focal delta activity occurred in a majority of electroencephalograms (EEGs). The type of seizures and the
distribution of EEG spikes failed to distinguish patients with tumor from those without. Multifocal EEG spikes
appeared in a majority of each group.
Blume WT, Girvin JP, Kaufmann JCE: Childhood brain tumors presenting as chronic uncontrolled focal
seizure disorders. Ann Neurol 12:538-541, 1982
Children and adolescents with unremitting, longstanding focal seizure disorders present a diagnostic challenge. Elements of the neurological history
and examination often fail to explain the magnitude
of the disorder. Of 35 patients who underwent surgical resection of an epileptogenic focus between 1774
and 1980, tumors were discovered in 16 (46%). This
considerably exceeds the 9% incidence of tumors
from an earlier series of patients operated on at the
same age for intractable seizures [ 101. By comparing
clinical aspects of these 16 tumor patients to those of
the 17 patients with other lesions, this study provides
warning signals for the presence of supratentorial
neoplasm in this clinical context.
Material and Method
Patients less than 2 1 years old considered to be candidates
for surgical relief of uncontrolled seizure disorders comprise this study. Surgery was undertaken when (1) the incapacitating attacks resisted thorough medical management
and when all o r most of the seizures arose from a surgically
resectable area or (2) radiology disclosed evidence of a
neoplasm. Patients who presented primarily with brain
tumors (progressive neurological deficit o r raised intracranial pressure) were excluded.
Electroencephalograms (EEGs) were recorded by Grass
Model 8-16 electroencephalographs using collodionapplied scalp electrodes placed according to the international 10-20 system. We inserted sphenoidal leads when
From the Epilepsy Unit, Department of Clinical Neurological Sciences, University Hospital, University of Western Ontario, London, Ont, Canada N6A 5A5.
clinically needed to search for temporal or inferior frontal
foci. Hyperventilation and sleep were t h e main EEG activating agents. Intravenous pentylenetetrazol (Metrazol)
was employed in 1 patient to activate a partial seizure. EEG
data from such procedures were lumped with those from
routine awake recordings in determining distributions and
incidences of spike and delta activity.
Wechsler-Bellevue or Stanford-Binet assessment gave
measures of intelligence in 3 2 patients. Clinical evaluation
was sufficient to place an additional patient in a mentally
subnormal category.
Twelve of the 16 patients with tumors had astrocytomas, 2 had oligodendrogliomas, 1 had a glioblastoma multiforme, and 1 a ganglioglioma. The 19
nontumor lesions included neuronal or architectural abnormalities in 9 patients, glial scars in 5 , cortical malformations in 4, nonneoplastic cysts in 3, and
subacute meningoencephalitis in 2.
The mean age at seizure onset was 8.75 years for
the tumor group (median, 10.0 years; range, 11
months to 16 years), as compared to 6.7 years (median, 6.3 years; range, 9 months to 16 years) for the
nontumor group. The mean duration of seizure disorder before operation was 6.3 years for the tumor
group (median, 6.0 years) and 8.7 years (median, 7.6
years) for the nontumor group.
Received May 2 1, 1981, and in revised form Feb 1, 1982. Accepted for publication Feb 7, 1982.
Address reprint requests to D r Blume.
538 0364-5 1341821 120538-04$01.25 @ 1982 by the American Neurological Association
Aspects Differentiating Tumor Patients
from Patients without Tumor
Several features distinguished patients with tumor
from those without:
1. Neurological examination: The neurological
examination was normal in 12 of 16 patients (75%)
with and 6 of 19 (32%) without tumor. Alternatively,
tumors were discovered in 12 of 18 patients (67%)
with normal neurological examinations as compared
to only 4 of 17 (24%) with abnormal examinations (p
< 0.005). Among those with normal intellect, tumors
were found in 11 of 15 whose neurological examination was normal and in 2 of 6 with abnormal examinations.
Excluding patients with temporal lobe lesions, who
are less likely to have abnormalities on neurological
examination (Fig l ), tumors appeared in 4 of 6 patients with normal neurological examinations and 2
of 14 with abnormal examinations. Clinical signs of a
progressive lesion occurred in only 1 of our 16 tumor
patients, a 15-year-old boy who became progressively demented over nine years.
2. Intelligence: Thirteen of 15 patients (87%)
with tumors had normal intelligence ( I Q above 79)
compared to 8 of 18 patients (44%) without tumors.
Intelligence could not be measured or estimated in 2
patients. Alternatively, tumor was found in 13 of 21
patients (62%) with normal intelligence and in 2 of
12 (17%) with mental subnormality (p < 0.005).
Among those with a normal neurological examination, tumors appeared in 11 of 15 with normal intellect and none of 2 with subnormal intelligence. For
the 16 patients with abnormal examinations, tumors
occurred in 2 of 6 with normal intellect and 2 of 10
with subnormal intellect.
3. Plausible reason for intractable focal seizures:
When the clinical picture, including history and
neurological examination, provided experienced
neurologists with a plausible explanation for the
F i g 1 . Diagram of the left hemisphere depicts the site of operation by lobe for both hemispheres for 16 patients with tumor
(numbers in boxes) and 17 wtthout tumor (numbers i n circles).
Hemispherertomy was performed i n 2 additional patitwts
without tumor.
presence of intractable focal seizures, a tumor was
present in only 2 of 16 patients (12.5%). However,
when sufficient cause was not clinically apparent,
tumors were found in 13 of 19 patients (68%) ( p <
Among patients without a plausible reason for uncontrolled seizures, a tumor was found in 10 of 12
with both normal intellect and a normal neurological
4. Persistent EEG delta activity: Persistent delta
activity in a majority of EEGs proved to be an ominous sign: a tumor was lurking in 10 of 14 such instances (7 1%).When persistent delta activity was not
seen in a majority of recordings, tumor was found in
only 6 of 20 cases (30%) ( p < 0.005). One patient
without tumor had only a single preoperative recording.
Aspects Failing t o Differentiate Tumor Patients
from Patients without Tumor
Two features were not helpful in distinguishing between patients with and patients without tumor:
1. Nature of the seizure disorder: All patients had
focal (partial) seizures, and a majority of each group
had generalized motor seizures as well.
2. EEG spikes: Multiple independent spikes,
defined as spikes arising from at least three noncontiguous electrode positions with at least one focus
in each hemisphere [3], occurred on at least one
recording in 12 of 16 patients with tumor and in
12 of 19 without tumor. Generalized epileptiform
paroxysms, usually spike and wave, occurred more
commonly in patients without tumor (10 of 19, 53977)
than in patients with tumor (4 of 16, 25%), but the
difference attains weak statistical significance ( p <
Accuracy of Clinical and E E G Data
i n Localizing Tumor
Seizure manifestations accurately localized the lobe
and hemisphere of the lesion in 4 patients and
identified the lobe but not the hemisphere in an additional 10 patients. Ictal features implicated a lobe
adjacent to that harboring the tumor in 2 patients:
temporal for a frontal tumor in 1 and vice versa in I.
Of the 10 patients with tumor who had persistent
delta waves on a majority of EEGs, 1 patient had
diffuse delta activity with predominance shifting
from one hemisphere to the other on sequential recordings. However, persistent focal delta activity was
localized to the lobe containing tumor in all 9 patients in whom it was present.
Spikes arose principally or exclusively from the
tumorous lobe in 12 of the 16 patients. Profuse
spiking arose from the frontal lobe in association with
Bfume et al: Seizures and Childhood Tumors
tumors confined to the temporal lobe in 2 patients
and to the parietal lobe in 1. One patient had no
Spikes occurring at a distance from the site of
tumor were often prominent and moderately frequent (Fig 2 ) . Such independently occurring spikes
persisted on postoperative recordings in 11 of the 13
patients in whom they had been present preoperatively.
Electrographic seizures were recorded in 8 patients
and originated from the same hemisphere as the
tumor in 7. The seizure arose from the lobe containing the tumor in all 4 of these 7 patients in whom
the lobe of origin could be determined electrographically. Seizures clearly arose from the hemisphere contralateral to the tumor in 2 patients; 1 of
these patients is also among the 7 with an ipsilaterally
originating seizure.
Radiological Findings
Initial computerized tomographic (CT) scans revealed neoplasm in 6 patients. In the 5 patients in
whom the neoplasm was discovered only on a subsequent scan, the first scan was done on an earliergeneration CT scanner or contrast was not injected.
Tumor was discovered at operation in 3 patients.
One had had a CT scan performed by an earlygeneration machine. The other 2, who had temporal
lobe tumors, had scans with late-model units plus
contrast injection. Isotope brain scan and air encephalography identified tumors in 2 additional patients.
Early Results of Surgery
"Complete" removal of the tumor was obtained in 8
of 16 patients. Seven of the 8 have been free from
seizures over a postoperative period of 8 to 38
Fig 2. The righl frovtalspike (Fp,, F4) occurred z('ith moderate
freyzlency in several recordings from this patient with a h$t
temporal tumor. Note the arrbythmic left temporal delta activity at F;, which spread t o T:,with an accompanyirig sharp
w a v e (F;, TJ.
months. The other has had fewer attacks. Six of the 8
patients whose tumors were incompletely removed
returned for follow-up. Four had fewer than half the
number of seizures as preoperatively, and 2 are seizure free. The follow-up period has been 17 to 45
months. Thus, all 24 patients with tumor who have
been followed have at least temporarily benefited
from either complete or partial surgical removal of
tumor. Most are taking less anticonvulsant medication than preoperatively; none is taking more.
Supratentorial brain tumor is one of the rare causes
of epilepsy in children, accounting for an estimated 1
to 29% of cases [I]. However, tumor of the brain is
the second most common type of neoplasm in this
age group [4, 51. Approximately 40% of brain
tumors in children are supratentorial 12,51, and the
associated incidence of seizures ranges from 2 5 to
40% [ I , 2, 91. Moreover, seizures began in the first
or second decade of life in 20% of a large series of
patients with glioma [ 8 ] .
In our series of children with intractable focal seizures, tumor was the most common cause when (1)
neurological examination, intelligence, or both were
normal, (2) no other plausible cause of seizures
existed, or (3) focal EEG delta activity appeared in
most recordings. Studies of brain tumor manifestations in children and adults have similarly concluded
that seizures may be the only symptom of brain
540 AnndlsofNeurology Vol 12 No 6 December 1982
tumor for several years [2, 8, 11, 121. Other reviews
of childhood brain tumors warn of their likelihood
when focal seizures are idiopathic [ I , 71. The high
incidence of neurological findings in some studies of
seizure disorders with hemispheric tumors in childhood [I, 9, 121 may reflect the inclusion of patients
who presented with the classic signs of an expanding
intracranial neoplasm.
Grey Walter [ 131 first demonstrated the association of focal arrhythmic delta activity with cerebral
hemispheric tumors. Other causes of such activity,
including a recent focal seizure, are acute events
whose associated delta activity would diminish in
subsequent EEGs. In the present clinical context,
sequential EEGs of varying duration from the last
seizure may be required to separate seizure-related
and tumor-related delta activity. The large number of
preoperative EEGs performed on our patients allowed the association of tumor with persistent delta
activity to emerge.
Other studies have shown a good but incomplete
correlation between the lobe implicated by clinical
features of focal seizures and the tumor site [ 111.
Some studies have found that a majority of patients
have more than one type of seizure [ l , 91, which
would apparently lessen the tumor-localizing value of
ictal symptoms. However, it is our experience that
the several types of focal seizures almost always implicate narrowly separate foci within one region of
the brain and confirm further the localizing value of
each ictal symptom.
Our finding that focal delta activity has greater
tumor-localizing value than spikes parallels studies in
adults (see Goldensohn [6]) which have demonstrated delta activity at the center of a lesion and
spikes at its margin or at the homologous portion of
the contralateral hemisphere. Our study may be the
first to show multifocal spikes in association with
unifocal tumor, including discharges arising from
nonhomologous regions of the opposite hemisphere.
This study found a strikingly high incidence (46%)
of tumors among patients with uncontrolled focal
epilepsy who were candidates for surgery. In the
Montreal Neurological Institute’s series from 1961
to 1970, tumors were encountered at epilepsy
surgery in 9% of patients aged 20 years or less [lo].
This difference is unlikely to be due merely to differences in policy regarding surgical candidacy. Newer
anticonvulsants and knowledge of their serum levels
may be more effective in controlling focal epilepsy
due to nonprogressive lesions than epilepsy caused
by tumors, shifting the composition of surgical candidates toward a higher percentage with neoplasms.
In addition, surgery was prompted in some patients
with only moderate seizure disorders when CT revealed lesions which might have eluded previous
methods of neuroradiology.
We thank Mr Ian M. Fyfe and Mr John Koval for their statistical
review, and Mrs M. Raffa for carefully typing the manuscript.
1. Aicardi J, Praud E, Bancaud J, Mises J, Chevrie JJ: Epilepsies
cliniquement primitives et tumeurs c6rebrales chez I’enfant.
Arch Fr Pediatr 27:1041-1055, 1970
2. Backus RE, Millichap JG: The seizure as a manifestation of
intracranial tumor in childhood. Pediatrics 29378-984, 1962
3. Blume WT: Clinical and electroencephalographic correlates
of the multiple independent spike foci pattern in children.
Ann Neurol 4:541-547, 1978
4. Chutorian AM: Tumors of the central nervous system. In
Carter S, Gold AP (eds): Neurology of Infancy and Childhood. New York, Appleton-Century-Crofts, 1774, p 89
5. Gjerris F Clinical aspects and long-term prognosis of intracranial tumours in infancy and childhood. Dev Med Child
Neurol 18:145-159, 1976
6. Goldensohn ES: Use of the EEG for evaluation of focal intracranial lesions. In Klass DW, Daly D D (eds): Current Practice of Clinical Electroencephalography. New York, Raven,
1977, p 311
7. Gomez MR: Tumors of the brain and spinal cord: supratentorial tumors. In Swaiman KF, Wright FS (eds): The Practice of
Pediatric Neurology. St. Louis, Mosby, 1975, vol 2, pp 666683
8. LeBlanc FE, Rasmussen T: Cerebral seizures and brain
tumors. In Vinken PJ, Bruyn GW (eds): Handbook of
Neurology. Vol 15, The Epilepsies. New York, American
Elsevier, 1974, pp 295-301
9. Low NL, Correll JW, Hammill J F Tumors of the cerebral
hemispheres in children. Arch Neurol 13:547-554, 1965
10. Mathieson G: Pathologic aspects of epilepsy with special reference to the surgical pathology of focal cerebral seizures. I n
Purpura DP, Penry JK, Walter R D (eds): Neurosurgical Management of the Epilepsies. Adv Neurol 8:107-138, 1975
1 1. Millichap JG, Bickford RG, Miller RH, Backus RE: The electroencephalogram in children with intracranial tumors and
seizures. Neurology (Minneap) 12:329-336, 1962
12. Page LK, Lombroso CT, Matson DD: Childhood epilepsy with
late detection of cerebral gliama. J Neurosurg 31:253-261,
13. Walter W G The location of cerebral turnours by electroencephalography. Lancet 2:305-308, 1936
Blume et al: Seizures and Childhood Tumors
Без категории
Размер файла
386 Кб
focal, disorder, childhood, seizure, presenting, brain, chronic, tumors, uncontrolled
Пожаловаться на содержимое документа