0148-396X/80/0704-0391$02.00/0 NEUROSURGERY Copyright © 1980 by the Congress of Neurological Surgeons Arteriovenous Malformation Presentin Vol. 7, No. 4, 1980 Printed in U.S.A. as a First Seizure in a 13-Year-Old Child: Surgical Indications Nancy Epstein, M.D. and Fred Epstein, M.D. Department of Neurosurgery, New York University Medical Center, New York, New York The authors report a 13-year-old child with an arteriovenous malformation (AVM) of the left middle cerebral artery. She presented with only a single seizure. Total removal of the lesion was accomplished through a left frontotemporal craniotomy. One year after operation, the patient has a residual right arm paresis. The authors review the literature to analyze the risks of the operative treatment of critically located AVMs as compared to the risks of the natural history of the lesion. (Neurosurgery, 7: 391-394, 1980) Key words: Arteriovenous malformation, Childhood, Middle cerebral artery, Seizure INTRODUCTION This report concerns a 13-year-old child with a left middle cerebral artery (MCA) arteriovenous malformation (AVM) who, like 25% of patients with hemispheric AVMs, presented with a single seizure in the absence of hemorrhage or neuro- logical deficit. A sizable proportion of the pediatric AVM population presents in this fashion, and the treatment of AVMs in this age group poses a serious therapeutic dilemma. When an AVM is easily accessible, the indications for surgical inter- vention are clear. However, in a strategically located AVM in a child, one confronts either the cumulative nonsurgical mor- bidity and mortality or the surgical risk of producing a per- manent neurological deficit or even death in a previously intact patient. CASE REPORT Two weeks before admission, this 13-year-old girl had her first seizure, which was characterized by speech arrest, right facial grimacing, and clonic movement of the right upper extremity, The seizure became generalized, lasted for 2 minutes, and responded to routine anticonvulsant therapy. The postictal neurological examination was normal. A com- puted tomographic (CT) scan revealed an enhancing left pos- terior frontal lesion. Angiograms disclosed this to be an AVM supplied by one anterior and two middle cerebral arterial branches and drained by a single vein into the superior sagittal sinus (Fig. 1). The approximate center of the lesion was above Broca’s area and abutted on the motor cortex. After much consideration, total removal via a left frontotemporal crani- otomy was accomplished without intraoperative complications. Postoperatively, the patient was transiently aphasic, with a right hemiplegia. The speech disorder and right lower extremity weakness resolved over the ensuing few weeks. The residual right upper extremity paresis improved slowly. The postoper- ative CT scan confirmed total excision of the lesion. On follow-up examination | year postoperatively, there was no dysphasia and the right arm paresis had improved. How- ever, the dexterity of the right hand had not recovered suffi- ciently to permit writing or other complex motor acts. REVIEW OF THE LITERATURE Surgical excision of an AVM after a single seizure is contro- versial. In the present case, the age of the patient, the critical location of the lesion, and the surgical risks were weighed 391 against the likelihood of future spontaneous hemorrhage with resultant disability or death. In this case, as in any instance when one is treating a patient with an AVM, the following issues must be taken into account: (a) the likelihood of a future hemorrhage and its accompanying morbidity and mortality, (b) the likelihood of progressive neurological deterioration without hemorrhage, and (c) the morbidity and mortality of surgery (16). Likelihood of future hemorrhage The likelihood of future hemorrhage in an unruptured AVM is a time-linked event; the longer the follow-up, the greater the chance of bleeding (3). In the cooperative study, Perret and Nishioka documented a 10% incidence of hemorrhage during the first 5 years; this increased to 24% between 5 and 20 years and to 30% after 20 years (12). Others have concluded that, over 10 to 40 years, the likelihood of sustaining a first hemor- rhage is between 20 and 36% (4, 5, 8, 11, 17). Morbidity of hemorrhage Although there is a correlation between the location of a ruptured AVM and the resultant neurological deficit (13), only a few papers in the neurosurgical literature identify and classify these lesions by location. In one such series (5), 16 patients presented with seizures alone in association with unruptured middle cerebral artery AVMs (8). Four of these patients expe- rienced hemorrhage within 8 years of the onset of seizure symptoms (8), all sustaining a major neurological deficit (hem- iplegia). Another series reported 8 patients with similarly lo- cated, unruptured arteriovenous malformations; 2 patients sus- tained fatal first hemorrhages 2 and 22 years after the initial seizure (14). A third patient suffered from episodic seizures and later died of an unknown cause. The remaining 5 patients had no clinical evidence of hemorrhage during the follow-up pe- riod. Similarly, significant neurological deficit may be incurred secondary to first hemorrhages from an AVM in any supraten- torial locale (9, 16). In the study of Perret and Nishioka, the morbidity rate for patients sustaining first hemorrhages, re- gardless of locale, was 53%; 50% of these patients became either hemiparetic or hemiplegic (12). Mortality of hemorrhage To assess prognosis fully, one must tabulate mortality figures over the patient’s expected life-span. In the Cooperative Study, it was found that over 25 years at least 25% of the patient 392 EPSTEIN AND EPSTEIN Neurosurgery, Vol. 7, No. 4 Fic. |. Left carotid arteriograms disclosing the strategically located arteriovenous malformations. Arterial phase, anteroposterior (4) and lateral (B) views. Venous phase, anteroposterior (C) and lateral (D) views. Note the relationship of the lesion to the motor cortex and speech areas. population would sustain a significant hemorrhage, 10% of which would be fatal (12). Of these patients, 23% would go on to have a second hemorrhage within 3 to 5 years, 12% of which would be fatal (12). Further hemorrhages, each with an asso- ciated 10% mortality (12, 18), were still possible Progressive ischemic deficit in the absence of hemorrhage Progressive nev’ ological deficit secondary to “vascular steal” in the absence of hemorrhage (15) is more likely to occur with a middle cerebral artery AVM. In Mackenzie’s series, 7 of 16 patients with seizures alone developed progressive deficit with- out hemorrhage; deterioration was slowly progressive for 5 patients, whereas a precipitous decline was observed in 2 patients (8). Surgical morbidity Few patients presenting with seizures alone have had middle cerebral artery AVMs resected totally. There are only 7 such October 1980 AVM PRESENTING AS A SEIZURE 393 TaBLe | Middle Cerebral Artery AVMs with or without Hemorrhage, Seizures, or Preoperative Deficit: Results of Total Surgical Excision of Lesions Postoperative Condition aso — a — " Better Same Worse Dead Wilson et al. (8) 65 21 22 18 4 Olivecrona and Riives (10) 24 10 8 3 3 Perret and Nishioka (12) 98 53 16 16 13 Forster et al. (4) 49 45 3 0 1 Total 236 129 49 37 21 % total 55% 21% 1S% 9% TABLE 2 DISCUSSION Supratentorial AVMs Presenting as Seizure Disorders: Results of Operation The natural history of an unruptured AVM in a young child 7 Postoperative Condition must be considered in the context of anticipated longevity. The No. _ average pediatric patient will be at risk for a minimum of 50 to Authors Patients Better/ 60 years. Given the risks of hemorrhage and of neurological — Worse Dead deterioration, there is a strong likelihood of future neurological = = catastrophe. Forster et al. (4); 7 6 l 0 Therefore, the very existence of an AVM in a child is an Olivecrona and indication for surgical extirpation if technically feasible. Be- Riives (10); cause many of these lesions become larger with time (2, 3), it Perret and is probably wise to intervene at a younger age rather than to Nishioka (12) defer operation to a time that is technically more hazardous. Kune (6) 18 14 2 2 Furthermore, operation is probably less likely to cause a per- Total 25 20 3 2 manent neurological deficit in children because the young % total 80% 12% 8% brain has a greater capacity for recovery. cases reported in three major studies composed of 614 patients: 4 of these patients remained intact postoperatively, 2 were hemiparetic, and 1 became hemiplegic secondary to a postop- erative clot (4, 10, 12). Surgical mortality The surgical mortality rates, variously reported between 2 and 11.7% (1, 4, 10, 12, 14), seem not to be influenced by prior hemorrhage, lesion locality, or the presence of a preoperative neurological deficit. Summary Irrespective of an antecedent history of hemorrhage, morbid- ity and mortality rates after total surgical extirpation of AVMs have been similar for patients with AVMs in any vascular distribution. For patients with lesions confined to the MCA sector, with or without antecedent hemorrhage, 76% were the same or better after operation, 15% were worse, and 9% died (Table 1). Similarly, data compiled from four large studies (Table 2) in which patients had variously located supratentorial AVMs (presenting as seizure disorders) and underwent total surgical resection indicated that 80% of patients were the same or improved after operation, 12% deteriorated, and 8% died (4, 6, 10, 12, 18). The surgical data may be compared to those for patients treated medically. Of a medically treated population, 25% suffered spontaneous hemorrhages within 10 to 25 years, and 10% of these hemorrhages were fatal (2.5% mortality rate). Significant morbidity due to the first hemorrhage affected over 50% of these patients (12). Furthermore, over this same period, one study reported a 30% morbidity from unruptured AVMs in the distribution of the middle cerebral artery (7). When operation is not feasible, embolic obliteration via flow-guided catheters may be applicable to AVMs of the middle cerebral artery distribution. The evolving field of “in- terventional neuroradiology” (2) may be expected to make future contributions to the management of these complex problems. Received for publication, April 25, 1980; accepted, June 15, 1980. Reprint requests: Nancy Epstein, M.D., Department of Neurosur- gery, New York University Medical Center, 550 First Avenue, New York, New York 10016. REFERENCES |. Amacher, A. L., Allcock, J. M., and Drake, C, G. Cerebral angiomas: The sequelae of surgical treatment. J. Neurosurg., 37: 571-575, 1972. 2. Baker, H. L., Jr. The impact of computed tomography on neuro- logic practice. Radiology, 116: 637-640, 1975. 3. Dandy, W. E. Venous abnormalities and angiomas of the brain. Arch. Surg., 17: 715-793, 1928. 4. Forster, D. M. C., Steiner, L., and Hakanson. S. Arteriovenous malformations of the brain: A long-term clinical study. J. Neuro- surg., 37: 562-570, 1972. 5. Henderson, W. R. and Gomez, R. Natural history of cerebral angiomas, Br. Med. J.. 4: 571-574, 1967. 6. Kunc, Z. Surgery of arteriovenous malformations in the speech and motor-sensory regions. J. Neurosurg., 40: 293-303, 1974. 7, Luessenhop, A. J. Operative treatment of arteriovenous malfor- mations of the brain. In Current Controversies in Neurosurgery. edited by Morely, T. P., pp. 203-209, Philadelphia, W, B. Saunders Co., 1976. 8. MacKenzie, I. The clinical presentation of the cerebral angioma: A review of 50 cases. Brain, 76: 184-214, 1953. 9. McCormick, W. F. and Nofzinger, J. D. “Cryptic” vascular mal- formations of the central nervous system. J. Neurosurg., 24: 865- 875, 1966. 10. Olivecrona, H. and Riives. J. Arteriovenous aneurysms of the 394 EPSTEIN AND EPSTEIN brain: Their diagnosis and treatment. Arch. Neurol. Psychiatry, 59: 567-602, 1948. 11. Paterson, J. H. and McKissock, W. A clinical survey of intracranial angiomas with special reference to their mode of progression and surgical treatment: A report of 110 cases. Brain, 79: 233-266, 1956. 12. Perret, G. and Nishioka, H. Report on the cooperative study of intracranial aneurysms and subarachnoid hemorrhage: Section VI. Arteriovenous malformations: An analysis of 545 cases of crani- ocerebral arteriovenous malformations and fistulae reported to the cooperative study. J. Neurosurg., 25: 467-490, 1966. 13. Schultz, E. C. and Huston, W. A. Arteriovenous aneurysm of the posterior fossa in an infant: Report of a case. J. Neurosurg., 13: 211-214, 1956. 14. Svien, H. J., Olive, I., and Angulo-Rivero, P. The fate of patients who have cerebral arteriovenous anomalies without definitive sur- gical treatments. J. Neurosurg., 13: 381-387, 1956. 15. Ténnis, W., Schiefer, W., and Walter, W. Signs and symptoms of supratentorial arteriovenous aneurysms. J. Neurosurg, 15: 471- 480, 1958. 16. Troupp, H. Arteriovenous malformations of the brain: What are the indications for operation? In Current Controversies in Neuro- surgery, edited by Morely, T. P., pp. 210-216, Philadelphia, W. B. Saunders Co., 1976. 17. Walker, A. E. Clinical localization of intracranial aneurysms and vascular anomalies. Neurology (N.Y.), 6: 79-90, 1956. 18. Wilson, C. B., U, H. S., and Domingue, J. Microsurgical treatment of intracranial vascular malformations. J. Neurosurg., 51: 446-454, 1979. COMMENTS The authors present an analysis of the natural history risk vs. the surgical risk for a neurologically normal child presenting with a single seizure associated with a critically located cerebral arteriovenous malformation. This clinical problem is typical of the more difficult decisions in the management of cerebral AVMs and therefore is pertinent. With respect to the future incidence for bleeding, I agree with the authors’ assessment of the literature cited. Overall, there is a spontaneous bleeding rate of approximately 2.5%/ year after the diagnosis is established. Bleeding is the only common denominator for cerebral AVMs. The other present- ing symptoms, such as seizures, progressive neurological deficit, headaches, and hydrocephalus, are factors relating to specific anatomical features for each lesion. The problem is that the literature cited by the authors does not yield information pertinent to the surgical risk presented by their patient. Surgical risk is largely a factor of patient selection. For example, in the most recent publication referred to by the authors (Wilson et al.), the overall surgical mortality and morbidity rates for total excision are given for 64 patients. However, these 64 cases include a wide assortment of lesions; one was not even intracranial, others were confined to the dura, and many were minute, offering surgical difficulty only by location. It is meaningless to cite surgical mortality and morbidity rates for aneurysms without reference to clinical grading. Sim- ilarly, it is meaningless to cite surgical mortality and morbidity rates for cerebral AVMs without reference to a system of anatomical grading. In a recent publication, a scheme for anatomical grading correlating with surgical risk was described (1). By this scheme the patient presented by the authors would be anatomically Grade III. The surgical mortality and morbidity rates for this grade approximate those for untreated lesions becoming symp- tomatic in midlife. However, I would agree with the authors that diagnosis in childhood carries with it a longer period of natural history risk and, therefore, surgical excision was indi- cated in their patient. The paper could have been enhanced with a more detailed description of the ultimate neurological handicap so that the final tradeoff, neurological deficit vs. natural history and risk, Neurosurgery, Vol. 7, No. 4 would be more evident to the readers. Also, postoperative angiography would have been more accurate than CT for determining the completeness of the excision. Alfred J. Luessenhop, M.D. Washington, District of Columbia 1. Luessenhop, A. J. and Gennarelli, T. A. Anatomical grading of supratentorial arteriovenous malformations for determining oper- ability. Neurosurgery, 1: 30-35, 1977. The authors have made a strong case for the surgical removal of angiomas that present in the pediatric age group as an initial seizure. They base their decision to excise the lesion principally upon the risk of hemorrhage in the future, as well as on the potential for development of the so-called “steal syndrome,” especially with angiomas in the middle cerebral artery territory. I agree strongly with this rationale. It is worth adding that angiomas have a definite tendency to enlarge with time, thereby making their removal a more difficult technicai problem (1). Furthermore, the best time to prevent a seizure disorder from becoming firmly established when the etiology is an accessible angioma is after the first seizure. Then, removal of the angioma itself will probably prevent the occurrence of further seizures. Once the seizure disorder has become firmly established, simple removal of the angioma may not suffice to stop the seizure disorder. In that case, formal cortical resection utilizing intra- operative cortical recordings may be required. A surgeon experienced in the surgical excision of angiomas should be able to remove accessible angiomas such as the one illustrated, even when quite large, without significant morbidity of a permanent nature and with negligible mortality, particu- larly in the pediatric age group. Our pediatric series of operated angiomas, Botterell Grades 1 to 3, now numbers 40, with no operative mortality and no patient made permanently worse by operation (2). When the angioma is huge or has an inacces- sible blood supply, techniques presently still in the develop- mental stage may warrant serious consideration. Such tech- niques are embolization and collimated radiotherapy as being developed by Leksell. A. Loren Amacher, M.D. London, Ontario 1. Amacher, A. L., Allcock, J. J., and Drake, C. G. Cerebral angio- grams; the sequellae of surgical treatment. J. Neurosurg., 37: ST1- 575, 1972. 2. Amacher, A. L., Drake, C. G., and Hovind, L. The results of operating upon cerebral aneurysms and angiomas in children and adolescents: Part 2. Cerebral angiomas. Childs Brain, 5: 166-173, 1979, For recent assessments of the natural history of intracranial arteriovenous malformations, the interested reader is referred to four other sources (1-4). Robert H. Wilkins, M.D. Durham, North Carolina 1. Drake, C. G. Cerebral arteriovenous malformations: Considera- tions for an experience with surgical treatment in 166 cases. Clin. Neurosurg., 26: 145-208, 1979. 2. Michelsen, W. J. Natural history and pathophysiology of arterio- venous malformations. Clin. Neurosurg., 26: 307-313, 1979. 3. Pellettieri, L. Surgical versus conservative treatment of intracranial arteriovenous malformations: A study in surgical decision-making. Acta Neurochir. (Suppl.), 29: 1-86, 1980. 4. Wilkins, R. H. Multiple aneurysms and associated arteriovenous malformations: Operative considerations. In The Aneurysm Patient Preoperative and Postoperative Care, edited by Hopkins. L. N.. New York, Raven Press. In press.