Epilepsiu, 33(2):298-303, 1992 Raven Press, Ltd., New York 0 International League Against Epilepsy
Temporal Lobectomy in Early Childhood *$Michael Duchowny , $§Bonnie Levin, “Prasanna Jayakar, *$Trevor Resnick, *$Luis Alvarez, ?Glenn Morrison, and *Patricia Dean Comprehensive Epilepsy Center, Departments oj’ *Neurology and ?Neurosurgery, Miami Children’s Hospital, and $Department of Neurology and $Division of Neuropsychology, University o f Miami School o f Medicine, Miami, Florida, U .S .A .
Summary: Children with medically resistant temporal lobe seizures that persist into adolescence often experience psychosocial deterioration and medical morbidity. It is therefore especially important to evaluate the contribution of surgical therapy in preadolescent children. We describe our experience with temporal lobectomy in 16 children < I 2 years (mean age 7 years) who had intractable seizures of temporal lobe origin. Structural lesions were identified on neuroimaging studies in 11 patients. In all patients, the standard anterior temporal lobectomy was tailored according to the extent of the lesion and
epileptogenic field. At follow-up, I 1 children were seizure-free, three were 90% improved, one was 50% improved, and one was unchanged. Neuropathological abnormalities were identified in virtually all children. Prenatally acquired abnormalities of neurogenesis were the most common, whereas mesial temporal sclerosis was found in only two children. We conclude that tailored temporal lobectomy in the first decade of life is highly beneficial in carefully selected children with medically refractory seizures. Key Words: Epilepsy-Neurosurgery-Temporal lobe-Child.
Excisional surgery is being increasingly used to treat patients with intractable temporal lobe epilepsy (NIH, 1990). Most surgical candidates are adults who have already endured the stigma and disruptions of longstanding medically uncontrolled seizures. More recently, surgery has been shown to be successful in childhood (Duchowny, 1989). However, a majority of the reported pediatric epilepsy surgery patients are adolescents, and there is little information about children under the age of 12 years (Davidson and Falconer, 1975; Whittle et al., 1981). Since the continuation of temporal lobe seizures into adolescence is associated with irreversible psychosocial deterioration and morbidity (Lindsay et al., 1984), it is especially important to evaluate the role of surgery in the preadolescent population. We describe our experience with temporal lobectomy in patients under age 12 with medically refractory seizures of temporal lobe origin.
METHODS Subjects Sixteen children (nine male, seven female) under the age of 12 years (range 2-1 1 years; mean 7 years) underwent temporal lobectomy between 1987 and 1990 (Table 1). All had been experiencing seizures for at least 2 years despite administration of multiple antiepileptic drugs (AEDs) singly and in combination. The mean duration of epilepsy was 3.8 years (range 2-10 years). Twelve children were experiencing multiple daily attacks, and four children had one to five seizures per week. EEG, neuroimaging, and clinical data were reviewed prior to surgery. A pediatric pharmacokineticist assessed the efficacy of antiepileptic therapy and pharmacologic profiles. This consisted of a detailed family interview, a review of the child’s past treatment regimens, and pharmacokinetic studies of current AEDs. The medical and family backgrounds of all surgical candidates were also carefully evaluated to determine the presence of adverse medical factors that would contraindicate surgery. Families were further screened for their
Received May 1991; revision accepted August 1991. Address correspondence and reprint requests to Dr. M. Duchowny at Comprehensive Epilepsy Center, Miami Children’s Hospital, 3200 S.W. 60 Court, Miami, FL 33155, U.S.A.
298
299
TEMPORAL LOBECTOMY IN EARLY CHILDHOOD TABLE 1. Clinical, neurodiugnostic, und neuroputhology profiles of study putients Age (yr)/Sex
Seizure duration (yr)
Postoperative clinical presentation
1
61M
3
2
11IM
3
3
91M
4
4
4lF
3
5
91F
3
6
7lM
5
7
9/M
3
8
71F
6
9
lOlM
8
10
21F
2
Behavioral arrest, automatism Behavioral arrest, automatism, GTC Behavioral arrest, autornatisms, tonic convulsion, incontinence Behavioral arrest, vocalization, tonic convulsion Behavioral arrest, automatism Behavioral arrest, tonic versive convulsion, automatism Aura (bad smell), behavioral arrest, tonic convulsion Behavioral arrest, crying, automatism Aura (bright light), behavioral arrest, automatism, tonic convulsion Tonic versive convulsion
11
9lM
2
12
1OIF
10
13
3iF
3-5
14
51M
4
15
91M
8
16
2lM
2
Patient no
Seizure frequency
Seizure focus
I-7lday
T4, T6
5-l01day
T4, T6
Mlday
F8, T4
Neuroimaging
Neuropathology
Postoperative seizure status
R lateral temporal postsurgical defect Dilated apex of R sylvian fissure R temporal, L parietal encephalomalacia
Ectopic neurons, gliosis Ectopic neurons, gliosis Gliosis
None
T3
Multiple cortical tubers
None
3lday
T4, Sp2
Normal
llday
T4, Sp2
Normal
Dysplastic tissue compatible with tuberous sclerosis Ectopic neurons, gliosis Neuronal atrophy, migrational arrest
I- 151day
FU, T4
Ganglioneurocytoma
None
I-31week
T3, Spl
Focal signal abnormality R basal temporal region Normal
Gliosis
None
1-5lweek
F7, Spl
Hypodense lesion left parahippocam pal gyrus
Normal
None
5-l01day
T3, Spl
Ectopic neurons
90% reduction
Automatism, crying, tonic convulsion Tonic versive convulsion
34lweek
T3, TS
Ganglioglioma
None
1-31day
T4, T6
1&2Olweek
Focal cortical dysplasia, mesial temporal sclerosis Ectopic neurons
None
Automatism, crying, version Behavioral arrest, automatism, tonic convulsion Tonic versive convulsion, automatism Behavioral arrest, automatism
L frontal periventricular calcification L mesia temporal lesion Small R hemisphere, R frontal porencephal y Normal
Neuraliglial hamartoma, mesial temporal sclerosis Ectopic neurons
90% reduction
Ganglioglioma
None
1 Olday
T4, T6
w/ddy
F8, T4
Calcified lesion-R amygdaba and uncus
34lweek
T4
Normal
I-flday
T4, T6
Calcified lesions-R amygdala uncus
50% reduction
90% reduction
None Auras only
Unchanged
None
GTC, generalized tonic-clonic seizures; R, right; L, left.
psychosocial stability and the availability of rehabilitative services. Preoperative evaluation revealed a definitive etiology in only three children (tuberous sclerosis, bacterial meningitis, low-grade glioma). In 13 patients there was no specific cause. All children had a normal general physical examination at time of surgery. Eleven school-age children ( a 6 years) underwent preoperative psychometric testing. Wechsler Intelligence Scale for Children-Revised testing revealed above average intelligence in two children [Full Scale IQ (FSIQ) 118, 1121, average intelligence in three (FSIQ 90, 96, 90), low average intelligence in one (FSIQ 84), and borderline intelligence in two (FSIQ 70, 79); three formally scored in the mentally deficient range (FSIQ 56, 58, 62).
Attention deficit disorder with or without hyperactivity was the most frequent associated problem (n = 12); two children were language-impaired. No child had evidence of an interictal personality disorder. Seizures AEDs were reduced or discontinued as needed to provoke seizures. Ictal seizure patterns were captured in 15 children. One child did not have clinical seizures despite prolonged EEG recording and medication discontinuation. Since this child had an extremely active interictal focus convergent with imaging data, surgery was based exclusively on interictal information. Behavioral arrest, stereotyped automatisms, and tonic motor seizures were the most frequently observed clinical seizure manifes-
Epilepsia, Vol. 33, No. 2, 1992
300
M . DUCHOWNY ET AL.
tations. Two children experienced auras at seizure onset. EEG Ictal and interictal seizure data were recorded via 32- and 64-channel video EEG telemetry (Telefactor Corporation). Closely spaced and sphenoidal electrodes were used routinely. Chronic invasive EEG recording using subdural (n = 9) or depth electrodes (n = 1) was obtained when the scalp EEG failed to localize epileptogenic regions or when functional cortical mapping was indicated. In the remaining patients, the epileptogenic region was defined by intraoperative electrocorticography . Subdural strips and/or 20-contact grid electrodes were inserted under the base of the temporal lobes in order to monitor activity from the amygdala and hippocampal formation. The posterior basal aspect of the temporal lobe and the lateral convexity were monitored when indicated. Placement was guided by surface and intraoperative subdural EEG data. When indicated from the scalp EEG data, additional subdural strips were placed over the ipsilatera1 frontal and contralateral temporal lobes. Twelve children had a seizure focus in which the ictal and interictal data localized to only one temporal lobe. One child with unilateral temporal lobe ictal onset demonstrated independent bilateral interictal temporal discharges (patient 3 ) , and two children who had localized temporal discharges also had independent extratemporal foci (patients 10 and 12). One child additionally displayed bilateral frontal lobe discharges on the scalp EEG (patient 9). Seizure discharges were recorded at the sphenoidal electrode in five children, but the sphenoidal electrode never displayed the initial or maximal discharge pattern. In all patients recorded extraoperatively the ictal focus involved the base and/or the lateral convexity of the temporal lobe, including the inferior and middle temporal gyris. Interictal discharges, however, were more extensive involving the superior temporal gyrus and/or the posterior regions. For each patient, the plane of resection included the entire field of ictal onset. The resection also included 80% of the associated interictal epileptogenic zone by amplitude criteria. We modified the resection plane to avoid eloquent regions (i.e., Wernicke’s area). Neuroimaging studies Computed tomography (CT) scans with and without contrast enhancement were performed in all children. Fifteen children also had gadolinium-
Epilepsia, Vol. 33, N o . 2,1992
enhanced magnetic resonance imaging (MRI) studies. Abnormalities were detected in 11 patients (Fig. 1): nine had lesions demonstrated by both CT and MRI (five with temporal lobe lesions and four with both temporal and extratemporal lesions); one child with a normal MRI study had an extratemporal lesion on CT (patient lo); and one child with a normal CT scan had a region of increased signal in the basal temporal region on MRI (patient 7). Three patients demonstrated mesial temporal lesions restricted to limbic sites. No lesion exhibited mass effect. Surgical procedure As discussed in the section on EEG, the plane of excision was based on the preoperatively determined epileptogenic zone and the site of the lesion. The standard technique of temporal lobectomy as a two-stage subpial resection of lateral and medial temporal structures (Falconer, 1972) was never performed. A tailored temporal resection was performed on the left in five and on the right in 1 1 children. Surgery was carried out under general anesthesia except in one child, who received neuroleptic analgesia for purposes of intraoperative cortical mapping. Postexcision electrocorticography (ECoG) was performed in all cases. The ECoG was always recorded during the lightest stage of anesthesia. The tailored resections were further extended whenever residual spikes could be excised without disrupting eloquent cortex. Complications of surgery included two bone flap infections, one of which required cranioplasty . One child had a large but transient collection of subgaleal fluid. These three children had undergone chronic subdural implantation prior to excisional surgery. One child developed a transient hemiparesis, which resolved over a period of 2 months. This child had undergone radiotherapy for a malignant temporal glioma 3 years prior to epilepsy surgery. The hemiparesis was most likely due to vascular spasm. OUTCOME OF SURGERY Seizure status At follow-up from 1 to 4 years (mean 1.5 years) 11 (69%) children were seizure-free. Seizures were improved in four (25%) children: three experienced a 90% reduction in seizure frequency, and one experienced a 50% reduction. Seizures were unimproved in one child. AEDs were discontinued after a 2-year seizure-free interval. Preoperatively , nine of the seizure-free children had exclusively unitemporal foci, and two children
TEMPORAL LOBECTOMY IN EARLY CHILDHOOD
301
FIG. 1. Patient 9. Transaxial T2 weighted magnetic resonance image revealing an extensive region of increased signal involving the left mesial and lateral temporal lobe, which extends posteriorly. Scalp-recorded ictal and interictal discharges demonstrated a widespread electrographic field. Subdural EEG recording revealed maximum epileptogenicity in the posterior mesial temporal region
also demonstrated extratemporal discharges. Two of the children who improved had exclusively unitemporal foci, one had bilateral temporal discharges, and one had extratemporal discharges. In the one child who was unimproved (patient 13), a large anomalous draining vein was discovered at surgery over the region of maximal epileptogenicity. The decision to preserve its vascular integrity severely restricted the planned extent of the resection. Behavioral status Behavioral and academic status were evaluated in school-age children by means of a postoperative structured interview. Of the nine families contacted, seven reported that their child had shown substantial improvement in psychosocial function, including increased self-esteem, greater selfconfidence, and improved mood. Five parents reported that prior to surgery their child was socially isolated and had few friends. Postoperatively, these children played more cooperatively and appropri-
ately with children of similar age. Six parents reported modest academic gains. Other behavioral changes included improved attention (n = 4), communication skills (n = 2), and decreased aggressive behavior (n = 3). Two patients had no improvement in either psychosocial or academic function. One was a severely retarded and overactive girl with tuberous sclerosis who was rendered completely seizure-free following left temporal lobectomy (patient 4). Her behavioral status was unimproved by surgery. The other patient was an older school-age boy who was depressed, isolated, and withdrawn prior to undergoing right temporal lobectomy (patient 2). His parents were divorced and he lived with his mother, who had been diagnosed as a borderline personality. Psychosocial problems continued in this patient despite a 50% reduction in seizure frequency. Pathology The histopathologic abnormalities of the resected temporal specimens are shown in Table 1. Tissue Epilepsia, Vol. 33, N o . 2 , 1992
M . DUCHOWNY ET AL.
302
abnormalities were identified in 15 cases. It should be noted, however, that the single “normal” specimen was removed from a child who had a gross structural lesion on preoperative imaging studies. Thus, despite the absence of neuropathological confirmation in this one case, all children demonstrated evidence of structural abnormality. On histologic analysis, seven children had findings consistent with disordered neuronal migration. Three children harbored ganglion cell tumors, and two children exhibited gliosis. In three other children the pathology revealed focal cortical dysplasia (one), tuberous sclerosis (one), and a hamartoma (one). Mesial temporal sclerosis was noted in two children, but not as an isolated finding. DISCUSSION
Our experience demonstrates that temporal lobectomy in the first decade of life is a safe and effective procedure for children with intractable temporal lobe epilepsy. The majority of our patients were rendered seizure-free, an outcome comparable to that reported in adults (Engel, 1987). Since delaying surgery does not result in superior seizure control, it may be argued that earlier intervention is always indicated in suitable young candidates. Despite anecdotal evidence, there is little direct proof that surgery in children is ultimately more successful than surgery in adults. In the only study to actually compare the results of temporal lobe surgery as a function of age, Jensen (1976) concluded that surgery in childhood resulted in superior psychosocial rehabilitation. Very few patients in this series were operated on in the first decade of life. Although we do not have long-term data and our findings await systematic verification through standardized psychosocial measures, the majority of our parents described significant improvement in interpersonal skills and social functioning in their children regardless of cognitive level. These observations are encouraging and hold promise for longer term gains in psychosocial development. The presurgical evaluation of young children differs from that of adults in several important respects. Young children are rarely able to understand or cooperate with the complex nature of the presurgical evaluation, and intensive parental support is a prerequisite to helping them through the evaluation and surgery. Families must be assessed critically for their capacity to understand the presurgical evaluation and their ability to cooperate with postoperative follow-up. Obtaining informed consent is only the beginning: we believe that surgery will be successful only if seizure remission is Epilepsin, Vol. 33, N o . 2, 1992
achieved in the context of healthy family dynamics and adequate rehabilitative resources. Children with temporal lobe seizures also differ in that they rarely harbor mesial temporal lesions as described in adults (Babb and Brown, 1987). Our experience suggests that temporal lobe lesions of children are rarely restricted to mesial temporal cortex and are unlikely to reside within the standard 4-5-cm resection plane used in adults. These larger and more poorly circumscribed lesions are also unlikely to conform to the electrical data. Thus, even though the process of lateralizing temporal lobe seizure origin is generally straightforward in childhood, the localization and characterization of the entire epileptogenic region often proves extremely difficult. Furthermore, in contrast to older patients (Risinger et al., 1989), sphenoidal electrodes are not as helpful for localizing seizure origin in children, possibly because of the rarity of localized anteromesial foci. These technological issues underscore the complexity of the workup of very young temporal lobe surgical candidates. In the past, children with normal imaging studies or widespread electrographic fields were regarded as unsuitable surgical candidates. The introduction of intensive monitoring and subdural recording techniques has altered this view dramatically, and virtually all children with medically unresponsive localization-related seizures are now considered potential candidates for excisional procedures (Duchowny, 1985; Wyllie et al., 1988). Subdural monitoring techniques are of particular importance in the evaluation of complicated pediatric seizure patterns and are likely to continue to play a role in the future. Both the high incidence and nature of the neuropathological findings in the temporal lobes of young children deserve comment. All of our patients demonstrated histologic abnormalities, even the five who had normal imaging studies. The uniform absence of normal findings suggests that a high proportion of temporal lobe seizures that are already intractable by early childhood are lesional in origin. This early seizure onset is fully consistent with the high proportion of cases showing abnormalities of neurogenesis. Ganglion cell lesions are a particularly common cause of early-onset temporal lobe seizures. We previously reported a high incidence of these lesions in young patients with extratemporal epilepsy as well (Duchowny et al., 1989). Ganglion cell lesions are also disproprotionately represented in other pediatric epilepsy surgery series where they give rise to seizures of very early onset (Whittle et al., 1981; Drake et al., 1987). In contrast, isolated
TEMPORAL LOBECTOMY IN EARLY CHILDHOOD
mesial temporal sclerosis is relatively uncommon in early childhood. Admittedly, in our protocol the hippocampus and other mesial structures were partially removed by suction rather than en bloc, and it is therefore possible that mesial temporal sclerosis may have gone undetected. However, we rarely encountered this lesion in any of the intact hippocampal tissue submitted for pathologic analysis. Considering the remarkably high incidence of mesial temporal sclerosis in older patients (Falconer et al., 1964), we speculate that in the adult epileptic mesial temporal sclerosis may be a secondary consequence of longstanding illness. In conclusion, the neuropathological findings in young children undergoing temporal lobectomy differ considerably from the findings in older patients. Younger children have a higher incidence of abnormalities of neurogenesis, which may explain why the temporal lobe seizures of young children are so often catastrophic at an early age. These developmental lesions are poorly circumscribed, which also explains why the epileptic field is so often widespread in young children and why surgery and the presurgical evaluation must be individualized. Nonetheless, if proper attention is directed toward these factors, temporal lobectomy in early childhood is at least as successful as surgery at a later age. Furthermore, earlier surgery is more likely to reduce the long-term medical morbidity of chronic epilepsy and lead to improved psychosocial outcome. REFERENCES Babb TL, Brown WJ. Pathological findings in epilepsy. In: Engel J Jr, ed. Surgical treatment of the epilepsies. New York: Raven Press, 1987:51140. Davidson S, Falconer MA. Outcome of surgery in 40 children with temporal-lobe epilepsy. Lancet 1975;l: 1260-3. Drake J, Hoffman HJ, Kobayashi J, Hwang P, Becker LE. Surgical management of children with temporal lobe epilepsy and mass lesions. Neurosurgery 1987;21:792-7. Duchowny M. Intensive monitoring of the epileptic child. J Clin Neurophysiot 1985;2:221-9. Duchowny M. Surgery for intractable epilepsy: issues and outcome. Pediatrics 1989;84:88&94. Duchowny M, Resnick TJ, Alvarez LA. Dysplastic gangliocytoma and intractable partial seizures in childhood. Neurology 1989;39:602-4. Engel J Jr. Outcome with respect to epileptic seizures. In: Engel J Jr, ed. Surgical treatment of the epilepsies. New York: Raven Press, 198733-71. Falconer MA, Serafetinides EA, Corsellis JAN. Etiology and pathogenesis of temporal lobe epilepsy. Arch Neurol 1964; 10:23348. Falconer MA. Temporal lobe epilepsy in children and its surgical treatment. Med J A u s t 1972;1:1117-21.
303
Jensen I. Temporal lobe epilepsy. Types of seizures, age, and surgical results. Acta Neurol Scand 1976;53:335-57. Lindsay J, Glaser G , Richards P, Ounsted C. Developmental aspects of focal epilepsies of childhood treated by neurosurgery. Dev Med Child Neurol 1984;26:574-87. National Institutes of Health consensus development conference statement: surgery for epilepsy, March 19-21, 1990. Epilepsia 1990;3 I : S O 6 1 2. Risinger MW, Engel J Jr, Van Ness PC, Henry TR, Crandall PH. Ictal localization of temporal lobe seizures with scalp/ sphenoidal recordings. Neurology 1989;39:1288-93. Whittle IR, Ellis HJ, Simpson DA. The surgical treatment of intractable childhood and adolescent epilepsy. Aust NZ J Surg 1981;S1: 109-96. Wyllie E , Liiders H, Morris HH, et al. Subdural electrodes in the evaluation for epilepsy surgery in children and adults. Neuropediatrics 1988;19:8M.
RESUMEN Los nihos con ataques del 16bulo temporal resistentes al tratamiento mtdico que persisten durante la adolescencia presentan con frecuencia un deterioro psicosocial y morbilidad mCdica. Es pues especialmente importante la valoracion de la contribucion de la terapia quirdrgica en 10s niri6s preadolescentes. Los autores describen su experiencia con lobectomia temporal en 16 niiios por debajo de 10s 12 anos (edad media = 7 anos) que tenian ataques intratables originados en ldbulo temporal. Las lesiones estructurales se identificaron mediante estudio de neuroimagen en 11 pacientes. En todos 10s enfermos se practico una lobectomia temporal anterior standard cuantificada de acuerdo con la extension de la lesion o del foco epileptogtnico. Durante el seguimiento, 11 nifios no presentaron mas ataques, 3 mostraron una mejoria del SO%, 1 mejoro en un 50% y en 1 no se observ6 ningun cambio. Las anomalias neuropatologicas fueron identificadas en la mayoria de 10s ninos. Las anomalias prenatales de neuroginesis fueron las mas comunes mientras que la esclerosis temporal medial se encontr6 solamente en 2 ninos. Los autores concluyen que la lobectomia temporal adecuada a cada caso y practicada en la primera dtcada de la vida es altamente beneficiosa en ninos cuidadosamente seleccionados con ataques refractarios a tratamiento mCdico. (A. Portera-Sanchez, Madrid)
ZUSAMMENFASSUNG Kinder mit therapieresistenten Temporallappenanfallen, die bis zur Adoleszenz fortbestehen, erfahren haufig einen psychosozialen Abbau. Es ist deshalb besonders wichtig, den Beitrag der chirurgischen Therapie bei praeadoleszenten Kindern zu werten. Wir schildern unsere Erfahrung mit temporaler Lobektomie bei 16 Kindern unter 12 Jahren (mittleres Alter 7 Jahre) mit therapieresistenten Anfallen temporalen Ursprungs. Bei 1 1 Patienten wurden strukturelle Lasionen durch bildgebende Verfahren festgestellt. Bei allen Patienten wurde eine iibliche anterior-temporale Lobektomie entsprechend der Ausdehnung der Lasion und des epileptogenen Feldes durchgefiihrt. Bei der Folgeuntersuchung waren 11 Kinder anfallsfrei, 3 zeigten eine 80%-ige Besserung, 1 Kind eine 50%-ige Besserung und 1 Kind zeigte keine Anderung. Neuropathologische Auffalligkeiten waren praktisch bei allen Kindern vorhanden. Pranatal entstandene Abnormalitaten der Neurogenese waren am haufigsten, eine mesiale temporale Sklerose wurde nur bei 2 Kindern gefunden. Wir schlupfolgern, dap eine angepasste temporale Lobektomie in der ersten Lebensdekade von gropem Nutzen bei sorgfaltig ausgewahlten Kindern mit therapie resistenten Anfallen ist. (C. Benninger, Heidelberg)
Epilepsia, Vol. 33, No. 2 , 1992
Temporal Lobectomy in Early Childhood *$Michael Duchowny , $§Bonnie Levin, “Prasanna Jayakar, *$Trevor Resnick, *$Luis Alvarez, ?Glenn Morrison, and *Patricia Dean Comprehensive Epilepsy Center, Departments oj’ *Neurology and ?Neurosurgery, Miami Children’s Hospital, and $Department of Neurology and $Division of Neuropsychology, University o f Miami School o f Medicine, Miami, Florida, U .S .A .
Summary: Children with medically resistant temporal lobe seizures that persist into adolescence often experience psychosocial deterioration and medical morbidity. It is therefore especially important to evaluate the contribution of surgical therapy in preadolescent children. We describe our experience with temporal lobectomy in 16 children < I 2 years (mean age 7 years) who had intractable seizures of temporal lobe origin. Structural lesions were identified on neuroimaging studies in 11 patients. In all patients, the standard anterior temporal lobectomy was tailored according to the extent of the lesion and
epileptogenic field. At follow-up, I 1 children were seizure-free, three were 90% improved, one was 50% improved, and one was unchanged. Neuropathological abnormalities were identified in virtually all children. Prenatally acquired abnormalities of neurogenesis were the most common, whereas mesial temporal sclerosis was found in only two children. We conclude that tailored temporal lobectomy in the first decade of life is highly beneficial in carefully selected children with medically refractory seizures. Key Words: Epilepsy-Neurosurgery-Temporal lobe-Child.
Excisional surgery is being increasingly used to treat patients with intractable temporal lobe epilepsy (NIH, 1990). Most surgical candidates are adults who have already endured the stigma and disruptions of longstanding medically uncontrolled seizures. More recently, surgery has been shown to be successful in childhood (Duchowny, 1989). However, a majority of the reported pediatric epilepsy surgery patients are adolescents, and there is little information about children under the age of 12 years (Davidson and Falconer, 1975; Whittle et al., 1981). Since the continuation of temporal lobe seizures into adolescence is associated with irreversible psychosocial deterioration and morbidity (Lindsay et al., 1984), it is especially important to evaluate the role of surgery in the preadolescent population. We describe our experience with temporal lobectomy in patients under age 12 with medically refractory seizures of temporal lobe origin.
METHODS Subjects Sixteen children (nine male, seven female) under the age of 12 years (range 2-1 1 years; mean 7 years) underwent temporal lobectomy between 1987 and 1990 (Table 1). All had been experiencing seizures for at least 2 years despite administration of multiple antiepileptic drugs (AEDs) singly and in combination. The mean duration of epilepsy was 3.8 years (range 2-10 years). Twelve children were experiencing multiple daily attacks, and four children had one to five seizures per week. EEG, neuroimaging, and clinical data were reviewed prior to surgery. A pediatric pharmacokineticist assessed the efficacy of antiepileptic therapy and pharmacologic profiles. This consisted of a detailed family interview, a review of the child’s past treatment regimens, and pharmacokinetic studies of current AEDs. The medical and family backgrounds of all surgical candidates were also carefully evaluated to determine the presence of adverse medical factors that would contraindicate surgery. Families were further screened for their
Received May 1991; revision accepted August 1991. Address correspondence and reprint requests to Dr. M. Duchowny at Comprehensive Epilepsy Center, Miami Children’s Hospital, 3200 S.W. 60 Court, Miami, FL 33155, U.S.A.
298
299
TEMPORAL LOBECTOMY IN EARLY CHILDHOOD TABLE 1. Clinical, neurodiugnostic, und neuroputhology profiles of study putients Age (yr)/Sex
Seizure duration (yr)
Postoperative clinical presentation
1
61M
3
2
11IM
3
3
91M
4
4
4lF
3
5
91F
3
6
7lM
5
7
9/M
3
8
71F
6
9
lOlM
8
10
21F
2
Behavioral arrest, automatism Behavioral arrest, automatism, GTC Behavioral arrest, autornatisms, tonic convulsion, incontinence Behavioral arrest, vocalization, tonic convulsion Behavioral arrest, automatism Behavioral arrest, tonic versive convulsion, automatism Aura (bad smell), behavioral arrest, tonic convulsion Behavioral arrest, crying, automatism Aura (bright light), behavioral arrest, automatism, tonic convulsion Tonic versive convulsion
11
9lM
2
12
1OIF
10
13
3iF
3-5
14
51M
4
15
91M
8
16
2lM
2
Patient no
Seizure frequency
Seizure focus
I-7lday
T4, T6
5-l01day
T4, T6
Mlday
F8, T4
Neuroimaging
Neuropathology
Postoperative seizure status
R lateral temporal postsurgical defect Dilated apex of R sylvian fissure R temporal, L parietal encephalomalacia
Ectopic neurons, gliosis Ectopic neurons, gliosis Gliosis
None
T3
Multiple cortical tubers
None
3lday
T4, Sp2
Normal
llday
T4, Sp2
Normal
Dysplastic tissue compatible with tuberous sclerosis Ectopic neurons, gliosis Neuronal atrophy, migrational arrest
I- 151day
FU, T4
Ganglioneurocytoma
None
I-31week
T3, Spl
Focal signal abnormality R basal temporal region Normal
Gliosis
None
1-5lweek
F7, Spl
Hypodense lesion left parahippocam pal gyrus
Normal
None
5-l01day
T3, Spl
Ectopic neurons
90% reduction
Automatism, crying, tonic convulsion Tonic versive convulsion
34lweek
T3, TS
Ganglioglioma
None
1-31day
T4, T6
1&2Olweek
Focal cortical dysplasia, mesial temporal sclerosis Ectopic neurons
None
Automatism, crying, version Behavioral arrest, automatism, tonic convulsion Tonic versive convulsion, automatism Behavioral arrest, automatism
L frontal periventricular calcification L mesia temporal lesion Small R hemisphere, R frontal porencephal y Normal
Neuraliglial hamartoma, mesial temporal sclerosis Ectopic neurons
90% reduction
Ganglioglioma
None
1 Olday
T4, T6
w/ddy
F8, T4
Calcified lesion-R amygdaba and uncus
34lweek
T4
Normal
I-flday
T4, T6
Calcified lesions-R amygdala uncus
50% reduction
90% reduction
None Auras only
Unchanged
None
GTC, generalized tonic-clonic seizures; R, right; L, left.
psychosocial stability and the availability of rehabilitative services. Preoperative evaluation revealed a definitive etiology in only three children (tuberous sclerosis, bacterial meningitis, low-grade glioma). In 13 patients there was no specific cause. All children had a normal general physical examination at time of surgery. Eleven school-age children ( a 6 years) underwent preoperative psychometric testing. Wechsler Intelligence Scale for Children-Revised testing revealed above average intelligence in two children [Full Scale IQ (FSIQ) 118, 1121, average intelligence in three (FSIQ 90, 96, 90), low average intelligence in one (FSIQ 84), and borderline intelligence in two (FSIQ 70, 79); three formally scored in the mentally deficient range (FSIQ 56, 58, 62).
Attention deficit disorder with or without hyperactivity was the most frequent associated problem (n = 12); two children were language-impaired. No child had evidence of an interictal personality disorder. Seizures AEDs were reduced or discontinued as needed to provoke seizures. Ictal seizure patterns were captured in 15 children. One child did not have clinical seizures despite prolonged EEG recording and medication discontinuation. Since this child had an extremely active interictal focus convergent with imaging data, surgery was based exclusively on interictal information. Behavioral arrest, stereotyped automatisms, and tonic motor seizures were the most frequently observed clinical seizure manifes-
Epilepsia, Vol. 33, No. 2, 1992
300
M . DUCHOWNY ET AL.
tations. Two children experienced auras at seizure onset. EEG Ictal and interictal seizure data were recorded via 32- and 64-channel video EEG telemetry (Telefactor Corporation). Closely spaced and sphenoidal electrodes were used routinely. Chronic invasive EEG recording using subdural (n = 9) or depth electrodes (n = 1) was obtained when the scalp EEG failed to localize epileptogenic regions or when functional cortical mapping was indicated. In the remaining patients, the epileptogenic region was defined by intraoperative electrocorticography . Subdural strips and/or 20-contact grid electrodes were inserted under the base of the temporal lobes in order to monitor activity from the amygdala and hippocampal formation. The posterior basal aspect of the temporal lobe and the lateral convexity were monitored when indicated. Placement was guided by surface and intraoperative subdural EEG data. When indicated from the scalp EEG data, additional subdural strips were placed over the ipsilatera1 frontal and contralateral temporal lobes. Twelve children had a seizure focus in which the ictal and interictal data localized to only one temporal lobe. One child with unilateral temporal lobe ictal onset demonstrated independent bilateral interictal temporal discharges (patient 3 ) , and two children who had localized temporal discharges also had independent extratemporal foci (patients 10 and 12). One child additionally displayed bilateral frontal lobe discharges on the scalp EEG (patient 9). Seizure discharges were recorded at the sphenoidal electrode in five children, but the sphenoidal electrode never displayed the initial or maximal discharge pattern. In all patients recorded extraoperatively the ictal focus involved the base and/or the lateral convexity of the temporal lobe, including the inferior and middle temporal gyris. Interictal discharges, however, were more extensive involving the superior temporal gyrus and/or the posterior regions. For each patient, the plane of resection included the entire field of ictal onset. The resection also included 80% of the associated interictal epileptogenic zone by amplitude criteria. We modified the resection plane to avoid eloquent regions (i.e., Wernicke’s area). Neuroimaging studies Computed tomography (CT) scans with and without contrast enhancement were performed in all children. Fifteen children also had gadolinium-
Epilepsia, Vol. 33, N o . 2,1992
enhanced magnetic resonance imaging (MRI) studies. Abnormalities were detected in 11 patients (Fig. 1): nine had lesions demonstrated by both CT and MRI (five with temporal lobe lesions and four with both temporal and extratemporal lesions); one child with a normal MRI study had an extratemporal lesion on CT (patient lo); and one child with a normal CT scan had a region of increased signal in the basal temporal region on MRI (patient 7). Three patients demonstrated mesial temporal lesions restricted to limbic sites. No lesion exhibited mass effect. Surgical procedure As discussed in the section on EEG, the plane of excision was based on the preoperatively determined epileptogenic zone and the site of the lesion. The standard technique of temporal lobectomy as a two-stage subpial resection of lateral and medial temporal structures (Falconer, 1972) was never performed. A tailored temporal resection was performed on the left in five and on the right in 1 1 children. Surgery was carried out under general anesthesia except in one child, who received neuroleptic analgesia for purposes of intraoperative cortical mapping. Postexcision electrocorticography (ECoG) was performed in all cases. The ECoG was always recorded during the lightest stage of anesthesia. The tailored resections were further extended whenever residual spikes could be excised without disrupting eloquent cortex. Complications of surgery included two bone flap infections, one of which required cranioplasty . One child had a large but transient collection of subgaleal fluid. These three children had undergone chronic subdural implantation prior to excisional surgery. One child developed a transient hemiparesis, which resolved over a period of 2 months. This child had undergone radiotherapy for a malignant temporal glioma 3 years prior to epilepsy surgery. The hemiparesis was most likely due to vascular spasm. OUTCOME OF SURGERY Seizure status At follow-up from 1 to 4 years (mean 1.5 years) 11 (69%) children were seizure-free. Seizures were improved in four (25%) children: three experienced a 90% reduction in seizure frequency, and one experienced a 50% reduction. Seizures were unimproved in one child. AEDs were discontinued after a 2-year seizure-free interval. Preoperatively , nine of the seizure-free children had exclusively unitemporal foci, and two children
TEMPORAL LOBECTOMY IN EARLY CHILDHOOD
301
FIG. 1. Patient 9. Transaxial T2 weighted magnetic resonance image revealing an extensive region of increased signal involving the left mesial and lateral temporal lobe, which extends posteriorly. Scalp-recorded ictal and interictal discharges demonstrated a widespread electrographic field. Subdural EEG recording revealed maximum epileptogenicity in the posterior mesial temporal region
also demonstrated extratemporal discharges. Two of the children who improved had exclusively unitemporal foci, one had bilateral temporal discharges, and one had extratemporal discharges. In the one child who was unimproved (patient 13), a large anomalous draining vein was discovered at surgery over the region of maximal epileptogenicity. The decision to preserve its vascular integrity severely restricted the planned extent of the resection. Behavioral status Behavioral and academic status were evaluated in school-age children by means of a postoperative structured interview. Of the nine families contacted, seven reported that their child had shown substantial improvement in psychosocial function, including increased self-esteem, greater selfconfidence, and improved mood. Five parents reported that prior to surgery their child was socially isolated and had few friends. Postoperatively, these children played more cooperatively and appropri-
ately with children of similar age. Six parents reported modest academic gains. Other behavioral changes included improved attention (n = 4), communication skills (n = 2), and decreased aggressive behavior (n = 3). Two patients had no improvement in either psychosocial or academic function. One was a severely retarded and overactive girl with tuberous sclerosis who was rendered completely seizure-free following left temporal lobectomy (patient 4). Her behavioral status was unimproved by surgery. The other patient was an older school-age boy who was depressed, isolated, and withdrawn prior to undergoing right temporal lobectomy (patient 2). His parents were divorced and he lived with his mother, who had been diagnosed as a borderline personality. Psychosocial problems continued in this patient despite a 50% reduction in seizure frequency. Pathology The histopathologic abnormalities of the resected temporal specimens are shown in Table 1. Tissue Epilepsia, Vol. 33, N o . 2 , 1992
M . DUCHOWNY ET AL.
302
abnormalities were identified in 15 cases. It should be noted, however, that the single “normal” specimen was removed from a child who had a gross structural lesion on preoperative imaging studies. Thus, despite the absence of neuropathological confirmation in this one case, all children demonstrated evidence of structural abnormality. On histologic analysis, seven children had findings consistent with disordered neuronal migration. Three children harbored ganglion cell tumors, and two children exhibited gliosis. In three other children the pathology revealed focal cortical dysplasia (one), tuberous sclerosis (one), and a hamartoma (one). Mesial temporal sclerosis was noted in two children, but not as an isolated finding. DISCUSSION
Our experience demonstrates that temporal lobectomy in the first decade of life is a safe and effective procedure for children with intractable temporal lobe epilepsy. The majority of our patients were rendered seizure-free, an outcome comparable to that reported in adults (Engel, 1987). Since delaying surgery does not result in superior seizure control, it may be argued that earlier intervention is always indicated in suitable young candidates. Despite anecdotal evidence, there is little direct proof that surgery in children is ultimately more successful than surgery in adults. In the only study to actually compare the results of temporal lobe surgery as a function of age, Jensen (1976) concluded that surgery in childhood resulted in superior psychosocial rehabilitation. Very few patients in this series were operated on in the first decade of life. Although we do not have long-term data and our findings await systematic verification through standardized psychosocial measures, the majority of our parents described significant improvement in interpersonal skills and social functioning in their children regardless of cognitive level. These observations are encouraging and hold promise for longer term gains in psychosocial development. The presurgical evaluation of young children differs from that of adults in several important respects. Young children are rarely able to understand or cooperate with the complex nature of the presurgical evaluation, and intensive parental support is a prerequisite to helping them through the evaluation and surgery. Families must be assessed critically for their capacity to understand the presurgical evaluation and their ability to cooperate with postoperative follow-up. Obtaining informed consent is only the beginning: we believe that surgery will be successful only if seizure remission is Epilepsin, Vol. 33, N o . 2, 1992
achieved in the context of healthy family dynamics and adequate rehabilitative resources. Children with temporal lobe seizures also differ in that they rarely harbor mesial temporal lesions as described in adults (Babb and Brown, 1987). Our experience suggests that temporal lobe lesions of children are rarely restricted to mesial temporal cortex and are unlikely to reside within the standard 4-5-cm resection plane used in adults. These larger and more poorly circumscribed lesions are also unlikely to conform to the electrical data. Thus, even though the process of lateralizing temporal lobe seizure origin is generally straightforward in childhood, the localization and characterization of the entire epileptogenic region often proves extremely difficult. Furthermore, in contrast to older patients (Risinger et al., 1989), sphenoidal electrodes are not as helpful for localizing seizure origin in children, possibly because of the rarity of localized anteromesial foci. These technological issues underscore the complexity of the workup of very young temporal lobe surgical candidates. In the past, children with normal imaging studies or widespread electrographic fields were regarded as unsuitable surgical candidates. The introduction of intensive monitoring and subdural recording techniques has altered this view dramatically, and virtually all children with medically unresponsive localization-related seizures are now considered potential candidates for excisional procedures (Duchowny, 1985; Wyllie et al., 1988). Subdural monitoring techniques are of particular importance in the evaluation of complicated pediatric seizure patterns and are likely to continue to play a role in the future. Both the high incidence and nature of the neuropathological findings in the temporal lobes of young children deserve comment. All of our patients demonstrated histologic abnormalities, even the five who had normal imaging studies. The uniform absence of normal findings suggests that a high proportion of temporal lobe seizures that are already intractable by early childhood are lesional in origin. This early seizure onset is fully consistent with the high proportion of cases showing abnormalities of neurogenesis. Ganglion cell lesions are a particularly common cause of early-onset temporal lobe seizures. We previously reported a high incidence of these lesions in young patients with extratemporal epilepsy as well (Duchowny et al., 1989). Ganglion cell lesions are also disproprotionately represented in other pediatric epilepsy surgery series where they give rise to seizures of very early onset (Whittle et al., 1981; Drake et al., 1987). In contrast, isolated
TEMPORAL LOBECTOMY IN EARLY CHILDHOOD
mesial temporal sclerosis is relatively uncommon in early childhood. Admittedly, in our protocol the hippocampus and other mesial structures were partially removed by suction rather than en bloc, and it is therefore possible that mesial temporal sclerosis may have gone undetected. However, we rarely encountered this lesion in any of the intact hippocampal tissue submitted for pathologic analysis. Considering the remarkably high incidence of mesial temporal sclerosis in older patients (Falconer et al., 1964), we speculate that in the adult epileptic mesial temporal sclerosis may be a secondary consequence of longstanding illness. In conclusion, the neuropathological findings in young children undergoing temporal lobectomy differ considerably from the findings in older patients. Younger children have a higher incidence of abnormalities of neurogenesis, which may explain why the temporal lobe seizures of young children are so often catastrophic at an early age. These developmental lesions are poorly circumscribed, which also explains why the epileptic field is so often widespread in young children and why surgery and the presurgical evaluation must be individualized. Nonetheless, if proper attention is directed toward these factors, temporal lobectomy in early childhood is at least as successful as surgery at a later age. Furthermore, earlier surgery is more likely to reduce the long-term medical morbidity of chronic epilepsy and lead to improved psychosocial outcome. REFERENCES Babb TL, Brown WJ. Pathological findings in epilepsy. In: Engel J Jr, ed. Surgical treatment of the epilepsies. New York: Raven Press, 1987:51140. Davidson S, Falconer MA. Outcome of surgery in 40 children with temporal-lobe epilepsy. Lancet 1975;l: 1260-3. Drake J, Hoffman HJ, Kobayashi J, Hwang P, Becker LE. Surgical management of children with temporal lobe epilepsy and mass lesions. Neurosurgery 1987;21:792-7. Duchowny M. Intensive monitoring of the epileptic child. J Clin Neurophysiot 1985;2:221-9. Duchowny M. Surgery for intractable epilepsy: issues and outcome. Pediatrics 1989;84:88&94. Duchowny M, Resnick TJ, Alvarez LA. Dysplastic gangliocytoma and intractable partial seizures in childhood. Neurology 1989;39:602-4. Engel J Jr. Outcome with respect to epileptic seizures. In: Engel J Jr, ed. Surgical treatment of the epilepsies. New York: Raven Press, 198733-71. Falconer MA, Serafetinides EA, Corsellis JAN. Etiology and pathogenesis of temporal lobe epilepsy. Arch Neurol 1964; 10:23348. Falconer MA. Temporal lobe epilepsy in children and its surgical treatment. Med J A u s t 1972;1:1117-21.
303
Jensen I. Temporal lobe epilepsy. Types of seizures, age, and surgical results. Acta Neurol Scand 1976;53:335-57. Lindsay J, Glaser G , Richards P, Ounsted C. Developmental aspects of focal epilepsies of childhood treated by neurosurgery. Dev Med Child Neurol 1984;26:574-87. National Institutes of Health consensus development conference statement: surgery for epilepsy, March 19-21, 1990. Epilepsia 1990;3 I : S O 6 1 2. Risinger MW, Engel J Jr, Van Ness PC, Henry TR, Crandall PH. Ictal localization of temporal lobe seizures with scalp/ sphenoidal recordings. Neurology 1989;39:1288-93. Whittle IR, Ellis HJ, Simpson DA. The surgical treatment of intractable childhood and adolescent epilepsy. Aust NZ J Surg 1981;S1: 109-96. Wyllie E , Liiders H, Morris HH, et al. Subdural electrodes in the evaluation for epilepsy surgery in children and adults. Neuropediatrics 1988;19:8M.
RESUMEN Los nihos con ataques del 16bulo temporal resistentes al tratamiento mtdico que persisten durante la adolescencia presentan con frecuencia un deterioro psicosocial y morbilidad mCdica. Es pues especialmente importante la valoracion de la contribucion de la terapia quirdrgica en 10s niri6s preadolescentes. Los autores describen su experiencia con lobectomia temporal en 16 niiios por debajo de 10s 12 anos (edad media = 7 anos) que tenian ataques intratables originados en ldbulo temporal. Las lesiones estructurales se identificaron mediante estudio de neuroimagen en 11 pacientes. En todos 10s enfermos se practico una lobectomia temporal anterior standard cuantificada de acuerdo con la extension de la lesion o del foco epileptogtnico. Durante el seguimiento, 11 nifios no presentaron mas ataques, 3 mostraron una mejoria del SO%, 1 mejoro en un 50% y en 1 no se observ6 ningun cambio. Las anomalias neuropatologicas fueron identificadas en la mayoria de 10s ninos. Las anomalias prenatales de neuroginesis fueron las mas comunes mientras que la esclerosis temporal medial se encontr6 solamente en 2 ninos. Los autores concluyen que la lobectomia temporal adecuada a cada caso y practicada en la primera dtcada de la vida es altamente beneficiosa en ninos cuidadosamente seleccionados con ataques refractarios a tratamiento mCdico. (A. Portera-Sanchez, Madrid)
ZUSAMMENFASSUNG Kinder mit therapieresistenten Temporallappenanfallen, die bis zur Adoleszenz fortbestehen, erfahren haufig einen psychosozialen Abbau. Es ist deshalb besonders wichtig, den Beitrag der chirurgischen Therapie bei praeadoleszenten Kindern zu werten. Wir schildern unsere Erfahrung mit temporaler Lobektomie bei 16 Kindern unter 12 Jahren (mittleres Alter 7 Jahre) mit therapieresistenten Anfallen temporalen Ursprungs. Bei 1 1 Patienten wurden strukturelle Lasionen durch bildgebende Verfahren festgestellt. Bei allen Patienten wurde eine iibliche anterior-temporale Lobektomie entsprechend der Ausdehnung der Lasion und des epileptogenen Feldes durchgefiihrt. Bei der Folgeuntersuchung waren 11 Kinder anfallsfrei, 3 zeigten eine 80%-ige Besserung, 1 Kind eine 50%-ige Besserung und 1 Kind zeigte keine Anderung. Neuropathologische Auffalligkeiten waren praktisch bei allen Kindern vorhanden. Pranatal entstandene Abnormalitaten der Neurogenese waren am haufigsten, eine mesiale temporale Sklerose wurde nur bei 2 Kindern gefunden. Wir schlupfolgern, dap eine angepasste temporale Lobektomie in der ersten Lebensdekade von gropem Nutzen bei sorgfaltig ausgewahlten Kindern mit therapie resistenten Anfallen ist. (C. Benninger, Heidelberg)
Epilepsia, Vol. 33, No. 2 , 1992
