Benign astrocytic and oligodendrocytic tumors of the cerebral hemispheres in children

Transcription

1 J Neurosurg 70: , 1989 Benign astrocytic and oligodendrocytic tumors of the cerebral hemispheres in children JEAN-FRAN{~OIS HIRSCH, M.D., CHRISTIAN SAINTE ROSE, M.D., ALAIN PIERRE- KAHN, M.D., ANDR~ PFISTER, M.D., AND ELIZABETH HOPPE-HIRSCH, M.D. Pediatric Neurosurgical Service, Hbpital Necker-Enfants Malades, Paris, France v," The authors review 42 consecutive cases of benign astrocytic and oligodendrocytic tumors of the cerebral hemispheres in children undergoing surgery in the pediatric service of the H6pital des Enfants Malades between 1975 and Epilepsy was the presenting sign in 76% of the children and remained the only clinical sign at diagnosis in 62%. Partial or complex partial seizures were observed in half of the cases, but other seizure types were also frequent. Diagnosis of the tumor as the etiological agent rested upon the results of computerized tomography or magnetic resonance imaging. Postoperative mortality (5%) and morbidity (16%) rates were low. The postoperative intelligence quotient was above 80 in 71% of the patients, and 77% of the children had no major problem in school. Although only two of the 42 patients were given postoperative radiotherapy, there were no recurrences in 82% of the survivors. The actuarial probability of nonrecurrence of the tumor was 95% at 5 years. Three patients with recurrent tumor underwent further surgery and are, at the present time, recurrence-free. The incidence of epilepsy fell from 76% before surgery to 19% after removal of the tumor alone; therefore, intraoperative electrocorticography and resection of the electrically abnormal cortex at the time of the first surgical procedure do not appear necessary. It is possible that tumor removal restores a mechanism that limits the spread of seizures and, thus, the clinical manifestations of epilepsy. Benign astrocytic and oligodendrocytic tumors of the cerebral hemispheres in children should not be treated with adjuvant radiotherapy, at least initially. KEY WORDS 9 brain neoplasm 9 astrocytoma 9 oligodendroglioma - epilepsy 9 outcome 9 children B ENIGN supratentorial glial tumors in children inelude astrocytomas grades I and II, 23 oligodendrocytomas, and oligoastrocytomas. Several types of tumor have been excluded from this study: benign supratentorial ependymomas, since they might not share the same oncological prognosis; ganglioneuromas, which involve neuronal cells; meningiomas, which are extracerebral; and thalamic tumors and gliomas of the optic pathways, since they present different surgical problems. Benign gliomas of the cerebral hemispheres in children raise three concerns. 1) They may be misdiagnosed since most often epilepsy is the only presenting symptom. 1'2'192) There are conflicting reports as to the efficacy and necessity of postoperative radiation therapy in astrocytic and oligodendrocytic tumors, 4-7A~ 14.16,17,19-22 so it is difficult to determine the role of postoperative radiotherapy in these patients. 3) At the present time, it is unclear whether tumor removal alone is sufficient to treat epilepsy or whether it is necessary to excise the electrically abnormal cortex using electrocorticography to guide the cortical excision. These questions are of importance because it is likely that the answers are not the same in children as in adults, 13 and because radiation therapy on a developing brain always involves a risk of mental retardation.11 In an attempt to answer these questions, we reviewed a series of 42 benign astrocytic and oligodendrocytic tumors of the cerebral hemispheres in children operated on from 1975 to 1987 in the pediatric neurosurgical service in the Hrpital des Enfants Malades. Clinical Material and Methods The series includes 22 patients with astrocytoma (44%), 12 with oligoastrocytoma (24%), and eight with oligodendrocytoma (16%). Twenty tumors were grade 1 and 22 were grade 2 according to Kernohan's criteria. 23 Among the astroeytomas, 15 were protoplasmic, six were pilocytic, and one was gemistocytic. Fifteen (36%) of these 42 benign tumors were associated with 568 J. Neurosurg. / Volume 70~April 1989

2 Benign tumors of cerebral hemispheres in children FIG. 1. Bar graph showing the age of the children in this series at the time of onset of seizures. FIG. 3. Bar graph showing the delay between the development of the first symptom of the disease and surgery. FIG. 2. Bar graph showing the follow-up period for the 42 patients in this series. a cyst. During the same 12-year period, two benign supratentorial ependymomas (4%) and six ganglioneuromas (12%) were operated on but have not been included in the series for reasons given above. The patients were under 15 years of age at the time of onset of symptoms (average 5 years 7 months, median 4 years 3 months) (Fig. 1). There were 26 boys (62%) and 16 girls (38%). The postoperative follow-up period (Fig. 2) varied from 6 months (for two patients) to 17 years. The tumors were located in the temporal lobe in 19 cases (45%), in the frontal area in nine cases (21%), and in a central location in eight cases (19%). All operations were performed without the aid of electrocorticography. The tumors were removed as completely as was feasible: tumor removal was considered macroscopically complete in 40 cases and incomplete in two. Taking into consideration the fact that all neurons have a functional value, the surrounding nontumoral cortex was always preserved as much as possible. Tumor removal was therefore preferred to lobectomy in all cases. Because it was believed that any recurrence of these slowly growing tumors could be FIG. 4. Diagram showing the incidence of clinical features preoperatively in the 42 patients in this series. The numbers within the circles printed outside their overlapping areas indicate the patients who showed isolated seizures, isolated neurological deficits, or isolated intracranial pressure (ICP). detected on postoperative computerized tomography (CT) scans, radiation therapy was given only in two cases of demonstrated recurrence. Thus, the results of this series reflect those which can be obtained in astrocytic and oligodendrocytic tumors of the cerebral hemispheres in children by tumor removal alone without additional resection of the surrounding cortex or adjuvant radiotherapy. Clinical Features Results Epileptic seizures were the single presenting symptom in 32 (76%) of these 42 children. Diagnosis was difficult before the era of CT scanning, which probably explains the rather long delay in some cases between the occurrence of the first symptom and surgery (mean delay 3 years 4 months, median delay 10 months) (Fig. 3). At the time of diagnosis, epilepsy was still the sole symptom in 22 patients; however, at that time 12 patients also demonstrated signs of increased intracranial pressure and eight had other neurological findings (Fig. 4). Epileptic seizures were of the partial complex type in 46.9% of cases; partial seizures were observed in 3.1%, J. Neurosurg. / Volume 70~April,

3 J. F. Hirsch, et al. FIG. 5. Graph showing the actuarial probability of nonrecurrence of the tumor in 42 patients (Kaplan-Meier life table). Vertical bars represent the 95% confidence interval. generalized seizures in 12.5%, several seizure types in 31.3%, and a Lennox Gastaut syndrome in 3.1%. In 62.1% of the patients with epilepsy, the seizures were not controlled satisfactorily and recurred more than once a week, eventually several times a day. In 17.2% of the patients seizures recurred more than once a month and in 20.7% more than once a year. Electroencephalography (EEG) was considered abnormal preoperatively in 87% of the cases, and remained abnormal after surgery in 84% of the patients, although 42% of them improved. Postoperative Mortality and Morbidity Two children died postoperatively, for a mortality rate of 5%. One was in a very poor condition before surgery and the other died of postoperative infection. The postoperative morbidity rate was low (16%). Eight patients presented with neurological signs before surgery and six continued to show signs after surgery. There was, therefore, no aggravation due to surgery. Recurrence Rate There were three recurrences (8%) in 40 children with a mean postoperative follow-up period of 4 years. In two of these cases, the initial tumor removal had been considered incomplete. In this series, the actuarial probability of nonrecurrence of the tumor at 5 years (Fig. 5) was 95% (95% confidence interval 87% to 100%) and at 12 years was 78.5% (95% confidence interval 54% to 92%). Postoperative Intelligence Quotient Postoperative intelligence quotients were above 80 in 71% of these patients (Fig. 6). These good results are reflected by the children's academic ability: 77 % attend a normal school. Within this group, 19% are more than 2 years behind for their age, but this rate is very close to that observed in a normal population. Finally, 23% are entered in a specialized school and have a major schooling problem. Operative Outcome Surgical removal of the tumor alone was an effective treatment of epilepsy (Fig. 7). Postoperatively, 81% of FIG. 6. Pie chart showing the postoperative intelligence quotient (IQ) in 40 survivors: 70% of the IQ's are above 80. the patients had no epileptic seizure activity and 57% did not require medical therapy. For various reasons, anticonvulsant drugs were still given to 24% of these children. Nineteen percent of the patients were not cured of their epilepsy. This failure was more likely to occur when the duration of preoperative epilepsy was long (Fig. 8). The management of a recurrent tumor in two patients demonstrates several points. One child underwent surgery and was free of seizures for 4 years. Recurrence of seizure activity heralded recurrence of the tumor. The surgical removal of this recurrent tumor again eliminated the clinical manifestations of epilepsy. In the second case, tumor removal was incomplete and the epilepsy was not cured. The remaining tumor, a small mural nodule within a cyst, was removed several years later with successful ablation of epileptic activity. Discussion The association of brain tumors with childhood epilepsy is rare, and varies in different series from 0.2% to 1.4%. 2'9'15'18 Therefore, when a child presents with epileptic seizures alone, brain tumor is not the first diagnosis considered. The seizure type is nothing more than a diagnostic orientation. In our series, partial seizures (whether complex or not) were observed in half of the patients. However, as in the series of Aicardi, et al.,"2 other seizure types were not uncommon, and frequently several seizure types occurred in the same patient. Moreover, epileptic seizures are sometimes related to nontumoral focal lesions. During the period of study (1975 to 1987), epileptic seizures were the presenting sign in two patients with cavernous angiomas and six with thrombosed arteriovenous malformations (AVM's). Computerized tomography scans are necessary to detect brain tumors in childhood epilepsy. Even so, cavernous angiomas and thrombosed AVM's will usually be diagnosed after surgery. Since the removal of 570 J. Neurosurg. / Volume 70/April, 1989

4 Benign tumors of cerebral hemispheres in children FIG. 7. Bar graph showing the postoperative incidence of epilepsy: 81% of the children experienced no further seizures. FIG. 8. Correlation of operative results with preoperative duration of epilepsy. The duration of preoperative epilepsy was almost significantly shorter in patients who were cured than in those with persistent postoperative seizures. The bars represent the 95% confidence interval for mean. these lesions is also effective in treating the associated epilepsy, they should be treated surgically. Several authors have already pointed out that the prognosis of benign supratentorial gliomas is different in children and adults. 7'w'1~'16 The 10-year survival rate was 75.8% in the pediatric series of Mercuri, et al., 16 and 87% in that of Laws, et al. 13 This series confirms their results. There are only three recurrences in 40 survivors; two of these recurrences were in children whose original tumor had not been totally removed. The actuarial probability of nonrecurrence at 5 years is close to 95%. Thus, the necessity (at least initially) of adjuvant radiotherapy in benign astrocytic and oligodendrocytic tumors of the cerebral hemispheres in children has not been demonstrated. Moreover, its efficacy has been questioned in several studies. 6'13'19 Studies restricted to adult series have demonstrated its relative usefulness. 4'5'14"21'22 Considering the low recurrence rate and the dangers of radiotherapy to the cerebral hemispheres of infants and children, it is our belief that these benign astrocytic and oligodendrocytic tumors occurring during childhood should be surgically removed and no adjuvant radiotherapy should be given. Radiation therapy should be limited to recurrent tumors that cannot be removed because of their location. In those cases, it should be given as late as possible. Although some authors have suggested the necessity of resection of the electrically abnormal cortex to cure epilepsy, 3,~z Schisano, eta/., 2~ have shown that tumor removal alone reduced the incidence of epilepsy from 66.5% before surgery to 26% after excision. Our conclusion is the same. In this series, the incidence of epilepsy was 78% preoperatively while it was 19% after surgery. It is known that the clinical expression of epilepsy requires the breakdown of neuronal or glial mechanisms that limit the spread of seizures) There was a discrepancy between the obvious clinical improvement of our patients and the much more discrete amelioration in electrical activity (EEG showed improved results in 42% of the patients but remained abnormal in most of the cases that were abnormal before surgery). This finding indicates that the change after tumor removal depends more on the clinical expression of epilepsy and on the spread of seizures than on the condition of the epileptic neurons themselves. References 1. Aicardi J: Epilepsy in Children. International Review of Child Neurology. New York: Raven Press, 1986, 413 pp 2. Aicardi J, Praud E, Bancaud J, et al: Epil6psies eliniquement primitives et tumeurs crrrbrales chez l'enfant. Arch Fr Pediatr 27: , Bancaud J, Talairach J, Geier S, et al: EEG et SEEG dans les Tumeurs Crrrbrales et dans l'l~pilepsie. Paris: Edifor, 1973, p 9 4. Bloom HJG: Intracranial tumors: response and resistance to therapeutic endeavours Int J Radiat Oncol Biol Phys 8: , Chin HW, Hazel J J, Kim TH, et al: Oligodendrogliomas 1. A clinical study of cerebral oligodendrogliomas. Cancer 45: , Dohrmann G J, Farwell JR, Flannery JT: Oligodendrogliomas in children. Surg Neurol 10:21-25, Gol A: Cerebral astrocytomas in childhood. A clinical study. J Neurosurg 19: , Grisar MT: Neuron-glia relationships in human and experimental epilepsy: a biochemical point of view, in Delgado-Escueta AV, Ward AW, Woodbury DM, et al (eds): Basic Mechanisms of the Epilepsies. Molecular and Cellular Approaches. Advances in Neurology, Vol 44. New York: Raven Press, 1986, pp Harwood-Nash DC: Computed tomography and seizures in children. J Neuroradiol 10: , Heiskanen O: lntracranial tumors of children. Childs Brain 3:69-78, Hirsch JF, Renier D, Czernichow P, et al: Medulloblastoma in childhood. Survival and functional results. Acta Neuroehir 48:1-15, Hoffman H J: Epilepsy, in American Association of Neurological Surgeons (ed): Pediatric Neurosurgery. Surgery of the Developing Nervous System. New York: Grune & Stratton, 1982, pp Laws ER, Taylor WF, Clifton MB, et al: Neurosurgical management of low-grade astrocytoma of the cerebral J. Neurosurg. / Volume 70~April,

5 J. F. Hirsch, et al. hemispheres. J Neurosurg 61: , Leibel SA, Sheline GE: Radiation therapy for neoplasms of the brain. J Neurosurg 66:1-22, Livingston S: Comprehensive Management of Epilepsy in Infancy, Childhood and Adolescence. Springfield, II1: Charles C Thomas, Mercuri S, Russo A, Palma L: Hemispheric supratentorial astrocytomas in children. Long-term results in 29 cases. J Neurosurg 55: , Neumann J, Kimpel J, Gulotta F: Das oligodendrogliom. Der klinische Verlauf in bezug zum histologischen Grading. Neurochirurgia 21:35-42, Page LK, Lombroso CT, Matson DD: Childhood epilepsy with late detection of cerebral glioma. 3 Neurosurg 31: , Reedy DP, Bay JW, Hahn JF: Role of radiation therapy in the treatment of cerebral oligodendroglioma: an analysis of 57 cases and a literature review. Neurosurgery 13: , Schisano G, Tovi D, Nordenstam H: Spongioblastoma polare of the cerebral hemisphere. J Neurosurg 20: , Sheline GE, Wara WM: Radiation therapy of brain tumors, in Youmans JR (ed): Neurological Surgery, ed 2. Philadelphia: WB Saunders, 1982, Vol 5, pp Shenkin HA: The effect of roentgen-ray therapy on oligodendrogliomas of the brain. J Neurosurg 22:57-59, Svien HJ, Mabon RF, Kernohan JW, et al: Astrocytomas. Mayo Clin Proc 24:54-64, 1949 Manuscript received December 7, Accepted in final form August 15, Address reprint requests to: Jean-Franqois Hirsch, M.D., Service de Neurochirurgie, H6pital Necker-Enfants Malades, 149 rue de S6vres, Paris Cedex 15, France. 572 J. Neurosurg. / Volume 70/April 1989