Epilepsy Surgery for Tumors, Vascular Malformations, Trauma and Cerebrovascular Disease

Emad N. Eskandar, M.D.
MGH Epilepsy Center, Neurosurgical Service,
Massachusetts General Hospital
Harvard Medical School, Boston, Massachusetts

Address for Correspondence:
Emad N. Eskandar, M.D.
Massachusetts General Hospital
15 Parkman St. ACC # 331
Boston, MA 02114

E-mail: eeskandar@partners.org
Patient Appointments: 617.724.6590
FAX: 617.724.0339

Referrals | Stereotactic Surgery | Parkinson's Disease | Intractable Epilepsy | Movement Disorder Surgery
Guestbook | Selected Publications | Links

Introduction

The incidence of lesional epilepsy has been estimated to represent between 20-30% of cases with intractable seizures. (1) This incidence appears to be increasing due to the widespread availability of advanced magnetic resonance imaging (MRI) and its ability to identify subtle lesions that were previously undetectable by computed tomography (CT). Many patients with a first seizure will undergo a MRI scan and a "symptomatic lesion" is discovered. In these cases, therapeutic strategies are typically directed against the specific pathology and seizure control is rarely a problem. In other cases, patients with seizure disorders of long duration may eventually undergo neuro-imaging that detects a causative lesion. This chapter will only discuss those cases of intractable and recurrent seizures associated with such lesions.
The ability to identify a lesion as the cause of intractable epilepsy generally signifies a more favorable surgical outcome, however two important clinical issues remain. The first involves the natural history of the underlying lesion itself and the risk that such pathology imparts to the patient. The second involves the localization of the presumed epileptic focus and its relationship to the lesion. The surgical management of seizure patients with such intracranial lesions can therefore consist of either simple excision of the structural lesion itself (lesionectomy) or excision of the structural lesion and surrounding epileptogenic cortex (lesionectomy plus corticectomy) . The debate over whether one surgical approach has specific advantages over the other remains unresolved and forms the basis of this chapter.

Clinical and Pathological Issues

Risk of Pathology

First and foremost, careful consideration must be given to the natural history of the lesion identified and the risk that such a lesion imparts to the health of the patient. These issues especially predominate with arteriovenous malformations and cavernomas which have a small (1-3%) but real risk of hemorrhage each year. (2) This risk is cumulative over the life expectancy of the patient and is therefore more important in younger individuals. Most intrinsic brain tumors associated with intractable epilepsy of long duration tend to be low grade and indolent (i.e. pilocytic astrocytoma, DNET, ganglioglioma). Some other glial tumors such as astrocytomas and oligodendrogliomas may also be associated with a long history of recurrent seizures but can show progressive growth or radiographic changes suggesting malignant transformation. These tumors must be treated with appropriate neuro-oncological therapies. In cases of post-traumatic epilepsy or seizures related to cerebral infarction, the pathology is stable and non-progressive and in these situations the intractability of the seizure disorder dominates.

Risk of Epilepsy

The risk of developing seizures from a structural lesion depends on a wide variety of factors including lesion type, lesion location and involvement of cortical gray matter. In general, slow growing primary neoplasms of the brain tend to be associated with the highest risk of seizures. Seizures are the primary presenting symptom in 80-90% of patients with gangliogliomas. (3,4) Dysembryoplastic neuro-epithelial tumors (DNET's) and pilocytic astrocytomas are also associated with a high incidence of intractable epilepsy. (5) More aggressive, rapidly growing tumors such as glioblastoma multiforme (GBM) and cerebral metastases are associated with a lower risk of seizures in the range of 20 to 30%. (6) Oligodendrogliomas which may be either very slow growing and indolent or have more anaplastic features are classically associated with epilepsy in 50-75% of cases with a lower incidence in the more aggressive tumors and older patients. (7) Supratentorial meningiomas are associated with about a 50% risk of seizures although they tend to be less refractory to medical therapy. (8)

Arteriovenous malformations (AVMs) have also been associated with a risk of seizures in the 30 - 35% range. (9) However, even patients harboring an AVM that has never ruptured have a 1.0 - 2.3% risk of hemorrhage per year and this must also be taken into consideration. (2) Cavernous angiomas present with seizures in 39 - 55% of cases and are often refractory to medical management. (10,11) The risk of seizures from venous angiomas is small and estimated to be about 5%. (12)

Cerebral vascular disease resulting in stroke or intracranial hemorrhage represents an important cause of seizures especially in the elderly. Cerebral infarction in the older population (> 50 years of age) has been found to be the cause of new-onset seizures in approximately 20% of cases. (13) Seizures may complicate subarachnoid hemorrhage in 10 - 20% of cases and are probably more common in patients that harbor a peri-sylvian or temporal lobe hematoma. (14) Intracerebral hematomas, especially of the lobar type, are associated with seizures in about 28% of cases whereas deep basal ganglia hematomas have a much lower incidence. (15)

Craniocerebral trauma has also been associated with approximately a 20% risk of post-traumatic epilepsy. (16) Diffuse closed head injury has a lower incidence than compound fractures or penetrating cerebral injury or closed head injury associated with intracranial hematomas.
Mechanisms of Epilepsy

Although it is clear that seizures can be associated with structural cerebral lesions it is not known by what mechanisms these seizures occur. It is generally acknowledged that the seizures do not arise from the lesion itself but rather from the irritated cortex immediately adjacent to the lesion. On occasion, seizures may arise from within the border zone of an infiltrative lesion such as an oligodendroglioma or astrocytoma.
Exact mechanisms of epileptogenicity are unknown although a number of electrophysiologic and pathophysiologic theories have been proposed. These include changes in altered neurotransmitter levels (GABA and somatostatin), free radical formation and altered second messenger physiology. (17) Morphological changes have also been identified such as alterations in vascular supply, neuronal cell loss, glial proliferation and subtle subcortical disconnections. (18) In the case of AVMs and cavernous angiomas and intracerebral hemorrhage, it is suspected that the breakdown products of the blood deposit ferric ions into the surrounding cortex which is a known epileptogenic substance.

In order for lesions to cause seizures, however, they generally must involve gray matter or the cortical mantel. In addition, specific areas of the cortex are associated with higher risks of seizures. Lesions in the Rolandic or peri-Rolandic cortex as well as the limbic areas tend to be the most epileptogenic. Lesions involving the temporal lobe are also associated with a higher incidence of seizures as compared to extra-temporal locations. (19) Lesion size may also represent a factor in epileptogenicity but there may be a variety of unknown mechanisms for the different pathologies.

A final issue that must be considered is the possibility that the lesion identified on imaging is causing epileptiform abnormalities at a distant site. For example, a lesion in the temporal lobe may be associated with atrophy and gliosis in the hippocampus. Whether this represents true dual pathology or the consequence of repeated epileptogenic discharges is unclear. In rare instances, a structural lesion may be an incidental finding and not play any role in seizure onset. The implications for presurgical evaluation and subsequent surgical decision-making are obvious. Overall, however, the general underlying assumption is that the morphologic or physiologic changes in the cortex that result in the epileptic condition are somehow caused by the structural lesion and its relationship with the surrounding cortex.

Presurgical Evaluation

Although the approach to patients with lesional epilepsy is often simplified by the neuro-imaging data, the principles of sound presurgical evaluation should not be ignored. A detailed description of the clinical semiology of the habitual events is important to determine if the seizures are consistent with the lesion location or whether the lesion may be an incidental finding. At times, the clinical features of the seizure are so distinct that it may clearly localize the seizure onset to the identified lesion. This would be the case in a patient having simple partial seizures with episodes of speech arrest and a cavernous angioma in the dominate frontal operculum. (Figure 1) Similarly, the clinical/radiological correlation is extremely compelling in a patient with focal motor seizures of the right hand and a lesion involving the left Rolandic cortex. In both of these situations it might be reasonable to consider lesionectomy as an appropriate intervention because of the likelihood of successful control of the seizures.

Advanced neuro-imaging remains one of the most important aspects in the presurgical evaluation by providing information about the exact location and extent of the lesion. Particular MR imaging characteristics may help identify the underlying pathology and can demonstrate associated features of mass effect, focal atrophy or calvarial molding. On occasion, it can identify obvious dual pathology such as distinct hippocampal atrophy associated with a neocortical temporal lesion. Recent advances in functional MRI (fMRI) can also provide important information regarding the localization of eloquent cortex adjacent to lesions or planned resection lines. In the future, MR spectroscopy may be able to provide distinct neurochemical information on mass lesions that might correlate with specific histopathology. Computerized tomography (CT ) scans can be helpful in determining the presence or absence of fine calcifications. Positron emission tomography (PET) scanning to demonstrate regional hypometabolism is generally not needed for patients with lesional epilepsy. The presence of the lesion can often make accurate interpretation of the PET scan difficult. Similarly, single photon emission computed tomography (SPECT) provides little useful information in the presence of a mass lesion. Cerebral angiography is only necessary for arteriovenous malformations.
Ictal video/EEG recordings are probably necessary in most cases of lesional epilepsy to confirm that the seizure onset is indeed localized to the area of the lesion. Demonstrating electrographically that the seizure onset is consistent with the lesion location confirms the role of the lesion in the patients seizure disorder and imparts an excellent prognosis in terms of surgical outcome. Consistent and convincing inter-ictal EEG evidence of epileptiform activity originating in the area of the lesion however may be adequate in some cases. For example, in a patient with intractable complex partial seizures and a low grade glioma in the mesial temporal lobe, consistent interictal epileptiform activity on scalp EEG from the same temporal lobe would make the patient a reasonable candidate for surgery even without ictal EEG recordings. (Figure 2) In some instances, ictal EEG data may be falsely localizing and geographically distant from the lesion. These patients can still be considered for surgery but may need more extensive presurgical evaluation.
Invasive intracranial monitoring has frequently been used to evaluate patients with lesional epilepsy. Generally, large subdural grid arrays are placed over the lesion and surrounding cortex. Ictal EEG recordings from these arrays can often define the exact region of seizure onset in relation to the lesion and reveal pathways of seizure propagation. This information is then used to devise a resection strategy that encompasses both resection of the lesion and the surrounding cortex. Intracranial grids are also useful for mapping cortical function extraoperatively prior to excision of the lesion and surrounding cortex. This is especially true in the pediatric population. Subdural strip electrodes and intracerebral depth electrodes can also be useful in localizing seizure onset but the intracranial investigation strategy must be carefully individualized. Occasionally even intracranial EEG investigation may be misleading if an inadequate number of electrodes are used. The mortality and morbidity associated with subdural grid electrodes has been estimated at @ 2 - 4% and therefore the use of invasive intracranial monitoring must be balanced by the expense and risk of these investigations. (20)

The presurgical evaluation of patients with intractable epilepsy generally requires detailed neuropsychological testing and assessment of a variety of psychosocial factors. While extremely important in certain cases as a baseline study, and as a predictor of the risk of resective surgery, it is not mandatory in patients undergoing limited lesionectomy.

Surgical Approaches

The primary goal of any surgical treatment for intractable epilepsy is to abolish the seizures and avoid any neurologic deficits associated with the resection. A secondary goal, in cases of tumors and vascular malformations is to entirely remove the lesion for a maximal therapeutic benefit. The major controversy in lesional epilepsy surgery is whether lesionectomy alone is adequate to achieve these goals or whether lesionectomy plus corticectomy provides better seizure control.

For the purposes of this chapter, a "lesionectomy" refers to the removal of a structural lesion using conventional neurosurgical operative techniques. This may entail resection of some overlying cortex in the approach to lesions but does not specifically or intentionally identify and resect surrounding epileptogenic cortex. "Lesionectomy plus corticectomy" refers to the surgical removal of the structural lesion as well as the identification and resection of surrounding epileptogenic cortex to improve seizure control. This can be accomplished using either acute intraoperative electrocorticography (EcoG) or with chronic extraoperative intracranial recordings.

The long-standing debate over lesionectomy versus lesionectomy plus corticectomy began with the observation that many patients do become seizure free or have a dramatic reduction in seizures after simple excision of a structural cortical lesion. The patient may be cured of seizures even though scalp EEG abnormalities remain. Falconer described several patients with structural lesions and EEG foci distinct from the lesion who underwent lesionectomy but no attempt at resection of the EEG focus. Many of these patients not only had a dramatic reduction in the seizure frequency but also had disappearance of the scalp EEG focus. (21) The disappearance of this distinct EEG focus often took many months suggesting a "running down" phenomenon. In other cases, the scalp EEG abnormalities remained but no longer resulted in clinical seizures. These observations suggest that a structural lesion does not necessarily result in permanent epileptogenic changes in the surrounding cortex and that lesionectomy alone may be able to reverse the epileptic condition in certain cases.

On the other hand, there are cases where lesionectomy alone does not relieve the patient of their seizures. This may be due to the presence of independent epileptogenic cortex, inadequate resection of the original lesion, or postoperative scar formation. The epileptogenic cortex may be geographically distant from the structural lesion and can become functionally independent despite lesionectomy. (22) This concept of "kindling" is well described in animal models but less well understood in the human condition. In these cases, resection of the lesion alone without removal of the functionally independent epileptogenic cortex would result in a surgical failure. In order to accurately identify these areas however requires invasive cortical monitoring, the results of which would guide a tailored resection of the lesion and surrounding cortex. The possibility of independent epileptic foci is less likely with smaller well-circumscribed lesions and more likely with the diffuse cerebral pathology that can result from trauma or cerebrovascular disease.

One alternative surgical approach is to suggest that patients with lesional epilepsy should undergo resection of the structural lesion after appropriate non-invasive presurgical evaluation. In many instances, especially those patients with cavernous angiomas or low grade gliomas, 70 to 80% of patients will remain seizure free after surgery. If the seizures persist after lesionectomy, then a more detailed and comprehensive evaluation could be undertaken with subdural grids, strips or intracerebral electrodes. This approach would minimize the expense and risk of invasive intracranial monitoring in all patients and seems a cost-efficient and effective compromise.

Results

There are several studies in the literature that have attempted to compare lesionectomy to lesionectomy plus surrounding epileptic cortex. The earliest study was a retrospective review by Gonzales and Elvidge at the Montreal Neurological Institute of 100 patients with low grade astrocytomas and intractable seizures. (23) Fifty two patients underwent lesionectomy and 48 had lesionectomy plus resection of surrounding cortex guided by intraoperative EcoG. Of the 53 remaining patients available for follow up at 5 years, 9 of 25 (36%) of lesionectomy candidates continued to have seizures compared to 6 of 28 (21%) who had undergone lesionectomy plus. Goldring et al compared the results of lesionectomy versus lesionectomy plus in patients with a variety of structural lesions and concluded that simple excision of the lesion was favorable and provided better seizure control. (24,25) Spencer et al reported their experience in 27 patients with lesions primarily involving the temporal lobe and observed a dramatic advantage of lesionectomy plus tailored cortical resection over simple excision. (26) All of these studies were flawed by their non-randomized nature and a retrospective selection bias.

Modern series of lesionectomy alone have demonstrated acceptable seizure free rates after surgery. Cascino et al reported their experience from the Mayo Clinic of stereotactic lesionectomy without cortical resection and found that this approach provided satisfactory outcomes in over 80% of patients although significantly fewer were completely seizure free. (27) The majority of these cases were extra-temporal and additional cortical resection was limited by the location of the lesions in eloquent cortex. The same group subsequently reported their experience in 51 patients with low-grade gliomas using either lesionectomy (n=17) or lesionectomy plus corticectomy (n=34) and noted 34/51 of their patients were seizure free at > 2 years follow-up. (28) Patients in both groups did equally well but patients with complete tumor resection had the best outcome.

In a large series of patients harboring tumors in the temporal lobe, complete lesion resection was associated with a 81% seizure free rate with a failure of seizure control associated with incomplete resection of the lesion. (29) Resection was not guided by EcoG and seizure free rates were not related to extent of mesial resection.

In contrast, other groups have found that simple excision of lesions involving the temporal lobe is often not successful in relieving seizures. Lombardi et al reported their experience in 15 cases of temporal lobe lesion and in 7 case of extra-temporal lesions. (30) In the 8 cases in the temporal lobe that did not involve the amygdala -hippocampus, lesionectomy alone yielded a Engel Class I in 4 cases but no improvement in the remaining 4. These 4 failures were reoperated upon with resection of the amygdalo-hippocampus with good results. Of the 7 extratemporal cases, lesionectomy alone was completely successful (Engel Class I) in 5, partially successful (Engel Class II) in 1 and unsuccessful (Engel Class IV) in 1. The last patient underwent subsequent invasive intracranial monitoring and frontal lobectomy to become seizure free at two years. An alternative approach to tumors in the temporal region is to perform en-bloc resections of the temporal lobe along with the tumor. In 31 patients with low-grade gliomas (predominately DNETs), this technique made 81% of patients seizure free. (31)

In 47 patients with structural lesions evaluated at the Cleveland Clinic, the presurgical evaluation involved both non-invasive EEG recordings and invasive EEG recordings using implanted subdural grids. (32) These grids were placed over the lesion and ictal recordings obtained. The epileptic zone involved the lesion only in 11/47 cases and extended beyond the lesion in 18/47 cases. In the remaining 18 cases, remote and noncontiguous epileptic zones were discovered. Postoperative control of seizures was obtained in 17 of 18 patients who underwent complete lesion excision regardless of the extent of seizure focus excision. Postoperative control of seizures was obtained in 5 of 6 patients with incomplete lesion excision but complete seizure focus excisions but in only 12 of 23 with incomplete lesion excision and incomplete focus excision. The most important variable predicting success of the surgical intervention was the completeness of lesion resection.

In a large series of 146 patients with brain tumors and medically intractable epilepsy, lesionectomy plus corticectomy yielded an Engel Class I outcome in 82% of patients at 6 months. (33) Most lesions were located in the temporal lobe and most were indolent low grade primary brain tumors ( gangliogliomas, pilocytic astrocytomas, DNET). This experience reinforces the importance of complete lesion resection in seizure control.

The experience with lesionectomy for cavernous angiomas is similar to that with low grade tumors. Lesionectomy alone can lead to seizure free rates in 70 - 90 % of patients. (34) Lesionectomy plus more extensive cortical resection may be necessary in case where the is poor concordance between the location of the lesion and the localizing data. Lesionectomy plus guided by intraoperative EcoG also provides seizure free rates of 80 - 90 %. It is generally agreed upon that the hemosiderin ring around the cavernoma must also be removed in order to achieve complete relief of seizures but several recent studies have suggested that this is unnecessary. (35)

Seizures may occur secondary to a cortical AVM whether the lesion has produced a clinical SAH or not. Seizures are the initial manifestation in about a third of patients but are generally intractable and incapacitating in only 10-15% of patients. Larger AVMs tend to be associated with seizures whereas smaller AVMs tend to be associated with SAH. Surgery for those patients with intractable epilepsy can achieve excellent results in @ 70% of patients. Outcomes appear to be better in patients who are slightly older at presentation (>30 years of age) and who have had seizures for less than one year.(36)

A recent meta-analysis of all studies involving lesional epilepsy reported in the literature using inappropriate outcome criteria (study size > 5 patients, postoperative follow up of > 2 years, and Engel Classification of seizure outcome) demonstrated a clear advantage of lesionectomy plus corticectomy over lesionectomy alone.(37) The analysis revealed that while both surgical approaches clearly reduce seizures in many patients, a patient who had undergone lesionectomy alone was almost twice as likely to continue to experience seizures after surgery as a patient who had undergone lesionectomy plus surrounding cortex for the same pathological lesion. The problem with this type of analysis is that surgical techniques differ from center to center. Some centers perform tailored resections based upon prolonged intraoperative EcoG while others use chronically implanted electrodes. The variations in presurgical evaluation and surgical technique make direct comparisons of outcome difficult. Nevertheless, it does appear that lesionectomy plus resection of surrounding epileptic cortex can provide improved seizure control in the majority of cases of lesional epilepsy.

Discussion

It is clear that lesionectomy can provide excellent seizure control in many patients. This is especially true for those patients with lesions who have infrequent or rare seizures of less than a years duration. Even patients with a long history of intractable epilepsy may be relieved of their seizures by simple resection of the lesion. However, in order to select those patients that might benefit from lesionectomy and those that might need lesionectomy plus corticectomy, the presurgical evaluation must be individualized.

In certain cases, it may be adequate to select patients for surgery simply on the basis of MRI identification of the lesion and the clinical semiology of the seizures. If the clinical features of the patients seizures are entirely consistent with the location of the lesion, then lesionectomy alone has a reasonable chance of success. This is especially true for lesions that are small and are highly epileptogenic such as cavernous angiomas and low grade gliomas. In these situations, resection of the lesion is generally associated with a good prognosis in terms of seizure control. If seizures are not well-controlled, then a more rigorous and intensive presurgical investigation could be carried out.

In most cases, however, inter-ictal and ictal EEG recordings are necessary to better define the zone of seizure onset in relationship to the lesion. This is especially true in patients with complex partial seizures in whom the clinical semiology of the seizure often does not provide convincing localizing information. Inter-ictal EEG recordings alone may be sufficient in cases where there is good correlation between the lesion location and inter-ictal EEG evidence of epileptogenicity. In most cases however, the principles of multidisciplinary presurgical evaluation are followed by attempting to localize the seizure onset to the site of the lesion. Ictal-EEG recordings that demonstrate that the seizures arise from the same region as the lesion identified on neuroimaging, implies that resection of the lesion will impart substantial therapeutic benefit. Invasive EEG recordings may be required when a lesion is identified but seems anatomically distinct from the area of seizure onset. Intracranial recordings with subdural strip or intracerebral electrodes may demonstrate that the lesion is an incidental finding and unrelated to seizure onset or that seizures actually do arise in close proximity to the lesion but were not detectable on scalp electrodes. Intracranial EEG investigation may also be necessary in cases of diffuse lesions involving large areas of cortex in order to identify the cortical regions adjacent to the lesion where seizures arise. This is particularly true when investigating lesions associated with ischemic cerebrovascular disease and trauma.

One issue that is extremely important in terms of a favorable surgical outcome is the completeness of resection of the lesion. Surgical approaches that do not entirely remove the offending lesion are generally less successful. In certain situations, the location of the lesion and its involvement of essential cortex prevents the complete removal of the lesion because of the risk of irreversible neurological deficit.
Specific pathological issues are of great importance in that certain lesions are associated with a high incidence of surgical success after lesionectomy. These include cavernous angioma which is associated with a greater than 80% seizure free rate. The low grade gliomas including pilocytic astrocytoma, ganglioglioma and DNET are also associated with an extremely high success rate after lesionectomy alone. In these instances, lesionectomy is an appropriate first intervention.

Overall, it appears that the surgical results following lesionectomy plus corticectomy are superior to lesionectomy alone. This may be due to a variety of factors including patient selection, differences in presurgical evaluation, center experience and surgical technique. In general, a tailored resection guided by detailed presurgical evaluation, intracranial or intraoperative EEG recordings is more likely to remove the causative lesion as well as surrounding epileptogenic cortex. This approach is more likely to remove not only the regions of seizure onset but also the areas involved in seizure propagation and areas of secondary epileptogenesis. In the temporal lobe, lesionectomy alone is often not successful because there are additional regions in the involved temporal lobe that are fundamentally involved in seizure onset and propagation. Lesionectomy alone is rarely successful in the temporal lobe where dual pathology is often present and pathways of propagation often have kindled nearby areas.

Conclusions

There is no question that modern neuroimaging has detected an increasing number of lesions. The presence of a lesion certainly increases the localization accuracy of seizure onset. The presurgical evaluation and treatment however must be individualized in all circumstances. Lesionectomy is often successful in certain circumstances including cavernous angioma and low grade gliomas. If lesionectomy is unsuccessful in these patients, a more detailed and comprehensive evaluation is often necessary. Lesionectomy plus corticectomy is generally more successful than lesionectomy alone in most patients with lesional epilepsy. Overall, the prognosis for seizure control is excellent in patients undergoing surgical treatment of seizures related to tumors, vascular malformations and cerebral vascular disease.

Figure Legends

Figure 1 Axial T1-weighted MRIwith gadolinium demonstrates a heterogenous lesion in the left temporal lobe consistent with a cavernous angioma. Note the heterogenous signal of the lesion with a surrounding ring of low signal likely representing hemosiderin. This young women had intractable simple and complex partial seizures characterized by speech arrest which were completely cured by lesionectomy despite the fact that EEG recordings were non-localizing.

Figure 2 Coronal T1-weighted MRI demonstrates a low grade cystic tumor in the left mesial temporal lobe. Interictal EEG recordings demonstrated frequent left temporal spike discharges and the patient underwent lesionectomy plus anterior temporal corticectomy and has been seizure free for over 5 years. Pathology demonstrated an xanthoastrocytoma.

References

1. Brooks BS, King DW, Gammel T, Meador K, Yaghmai F, Gay JN, Smith JR, Flanigan HF: MR Imaging in patients with intractable complex partial seizures. Am J Roentgenol 1990;154:577-583.

2. Forster DMC, Steiner L, Hakanson S: Arteriovenous malformations of the brain. A long term clinical study. J Neurosurg 1972;37:562-571.

3. Silver JM, Rawlings CE, Rossitch E, et al: Ganglioglioma: A clinical study with long term follow-up. Surg Neurol 1991;35:261-267.

4. Zentner J, Wolf HK, Ostertun B, Hufnagel A Campos MG, Solymosi L, Schramm J: Gangliogliomas: clinical, radiological and histopathological findings in 51 patients. J Neurol Neurosurg Psychiatry 1994;57:1497-1502.

5. Daumas-Duport C, Scheithauer BW, Chodkiewicz JP, Laws ER, Vedrenne C: Dysembryoplastic neuroepithelial tumor: A surgically curable tumor of young patients with intractable partial seizures. Neurosurgery 1988;23:545-556.

6. Penfield W, Erickson TC, Tarlov I: Relation of intracranial tumors and symptomatic epilepsy. Arch Neurol Psychiatry 1940;300-312.

7. Whittle IR, Beaumont A: Seizures in patients with supratentorial oligodendroglial tumours. Clinicopathological features and management considerations. Acta Neurochir 1995;135:19-24.

8. Ramamurthi B, Ravi R, Ramachandran V: Convulsions with meningiomas: Incidence and significance. Surg Neurol 1980;14:415-422.

9. Stein BM, Wolpert SM: Arteriovenous malformations of the brain. I: current concepts and treatment. Arch Neurol 1980;37:1-5.

10. Simard JM, Garcia-Bengochea F, Ballinger W, et al: Cavernous angioma: A review of 126 collected and 26 new cases. Neurosurgery 1986;18:162-169.

11. Farmer J-P, Cosgrove GR, Villemure J-G, Meagher-Villemure K, Melanson D: Intracerebral cavernous angiomas. Neurology 1989;38:1699-1704.

12. Garner TB, Curling DO, Kelley DL, et al: The natural history of intracranial venous angiomas. J Neurosurg 1991;75:715-721.

13. Sayette V de la, Cosgrove GR, Melanson D, Ethier R: CT findings in late-onset epilepsy. Can J Neurol Sci 14:286-289,1987.

14. Murphy MJ. Long-term follow-up of seizures associated with cerebral arteriovenous malformations. Results of therapy. Arch Neurol 1985;42:477-479.

15. Cervoni L, Artico M, Salvati M, Bristot R, Franco C, Delfini R: Epileptic seizures in intracerebral hemorrhage: a clinical and prognostic study of 55 cases. Neurosurg Rev 1994;17:185-188.

16. Annegers JF, Hauser WA, Coan SP, Rocca WA: A population-based study of seizures after traumatic brain injuries. N Engl J Med 1998;338:20-24.

17. Haglund MM, Berger MS, Kunkel DD, et al: Changes in gamma-aminobutyric acid and somatostatin in epileptic cortex associated with low-grade gliomas. J Neurosurg 1992;77:209-214.

18. Kraemer DL, Awad IA: Vascular malformations and epilepsy: clinical considerations and basic mechanisms. Epilepsia 1994;35:S30-S43.

19. Rasmussen T: Surgery of epilepsy associated with brain tumors. In Purpura DP, Penry JK, Walter RD (eds): Advances in Neurology, Vol 8. New York, Raven Press 1975, p227-239.

20. Van Buren JM: Complications of surgical procedures in the diagnosis and treatment of epillepsy. In Engel JJ (ed): Surgical Treatment of the Epilepsies. New York, Raven Press, 1987, p465-475.

21. Falconer MA, Serafetinides EA: A follow-up study of surgery in temporal lobe epilepsy. J Neurol Neurosurg Psychiatry 1963;26:154-161.

22. Morrell F: Secondary epileptogenesis in man. Arch Neurol 1985;42:318-325.

23. Gonzalez D, Elvidge AR: On the occurrence of epilepsy caused by astrocytomas of the cerebral hemispheres. J Neurosurg 1962;19:470-475.

24. Goldring S, Gregorie EM: Experience with lesions that mimic gliomas in patients presenting with a chronic seizure disorder. Clin Neurosurg 1984;33:43-56.

25. Goldring S, Rich KM, Picker S: Experience with gliomas in patients presenting with a chronic seizure disorder. Clin Neurosurg 1984;33:15-26.

26. Spencer DD, Spencer SS, Mattson RH, et al: Intracerebral masses in patients with intractable partial epilepsy. Neurology 1984;34:342-349.

27. Cascino GD, Kelly PJ, Hirschorn KA, Marsh WR, Sharbrough FW: Stereotactic resection of intra-axial cerebral lesions in partial epilepsy. Mayo Clin Proc 1990;65:1053-1060.

28. Britton JW, Cascino GD, Sharbrough FW, Kelley PJ: Low-grade glial neoplasms and intractable epilepsy: efficacy of surgical treatment. Epilepsia 1994;35:1130-1135.

29. Fried I, Kim JH, Spencer DD: Limbic and neocortical gliomas associated with intractable seizures: A distinct clinicopathological group. Neurosurgery 1994;34:815-824.

30. Lombardi D, Marsh R, de Tribolet N: Low grade glioma in intractable epilepsy: lesionectomy versus epilepsy surgery. Acta Neurochir 1997;68:70-74.

31. Kirkpatrick PJ, Honavar M, Janota I, Polkey CE: Control of temporal lobe epilepsy following en bloc resection of low-grade tumors. J Neurosurg 1993;78:19-25.

32. Awad IA, Rosenfeld J, Hahn JF, Luders H: Intractable epilepsy and structural lesions of the brain: mapping, resection stratagies and seizure outcome. Epilepsia 1991;32:179-186.

33. Zentner J, Hufnagel A, Wolf HK, Ostertun B, Behrens E, Campos MG, Elger CE, Wiestler OD, Schramm J: Surgical treatment of neoplasms associated with medically intractable epilepsy. Neurosurgery 1997;41:378-386.

34. Casazza M, Broggi G, Franzini A, Avanzini G, Spreafico R, Bracchi M, Valentini MC: Supratentorial cavernous angiomas and epileptic seizures: preoperative course and postoperative outcome. Neurosurgery 1996;39:26-32.

35. Zevgaridis D, van Velthoven, Ebeling U, Reulen HJ: Seizure control following surgery in supratentorial cavernous malformations: a retrospective study in 77 patients. Acta Neurochir 1996;138:672-677.

36. Piepgras DG, Sundt TM, Ragoowansi AT, Stevens L: Seizure outcome in patients with surgically treated cerebral arteriovenous malformations. J Neurosurg 1993;78:5-11.

37. Weber JP, Silbergeld DL, Winn HR: Surgical resection of epileptogenic cortex associated with structural lesions. Neurosurg Clin N Am 1993;4:327-336.

Referrals | Stereotactic Surgery | Parkinson's Disease Intractable Epilepsy | Movement Disorder Surgery Guestbook | Selected Publications | Links