What tissue should be resected?

The most common surgical intervention for intractable complex partial seizures associated with a temporal lobe mass is the temporal lobectomy, but techniques vary widely between institutions.11 Some centers advocate the removal of the tumor alone, whereas others stress the importance of resecting the ictal focus, if separate from tumor pathology.23,32,40,45,110–113Epilepsy surgery, rather than tumor resection, may be more likely to be considered in patients with a longer duration of epilepsy.

Gross total resections are approached with standard craniotomy techniques or via computer-assisted stereotactic cranial procedures. When removal of only the tumor is advocated, gross total resection has been reported as providing greater seizure control than partial resection, with an overall range of 80–88% of patients with temporal or extratemporal lesions achieving complete postoperative relief from seizures.32,112–116

Conversely, authors have long suggested that to control seizures it may be necessary to remove epileptogenic cortex beyond the tumor, because it is the cortex adjacent to the tumor margins that is most likely responsible for epileptogenesis.36,45,110,111,117,118 Seizure relief after resection for the tumor and the epileptogenic zone has been reported as ranging from 80% to 95% for temporal and extratemporal lesions. Surgical strategies are varied and include resection of epileptogenic cortex and mesial temporal structures.

Presurgical evaluation: EEG

Initially, an MRI is most sensitive for identification of a structural lesion. A routine electroencephalogram (EEG) may demonstrate correlating cerebral dysfunction and a potential epileptogenic zone. Interictal EEG and ictal video-EEG may show mirror foci, false localization, secondary epileptogenesis, or even false lateralization with ictal onset in tumor patients, however.102–104

In most cases, seizures originate in the vicinity of the tumor. Focal background abnormalities with polymorphic slow wave frequencies may be seen in only 32–44% of patients.47 Noninvasive video-EEG monitoring may identify ictal onset, but long-term invasive video-EEG recording with extraoperative epidural, subdural, or depth electrodes may be needed to define an epileptogenic zone using ictal data at seizure onset.47,105 These data are usually the single most reliable way of defining an epileptogenic focus via EEG.106

Presurgical evaluation: Mapping brain function

Neuropsychological testing can assess functional abilities and potential regions of dysfunction that may coincide with the location of a tumor. The Wada test, with intracarotid amobarbital injection, can determine lateralization of language and memory.107 After amobarbital injection into either internal carotid artery, unilateral hemispheric function is tested, allowing comparison of the language and memory function of each hemisphere independently. This knowledge may be crucial in determining whether resection will be performed near language areas. Functional magnetic resonance imaging (fMRI) is being studied as a possible noninvasive substitute for the Wada test.

SPECT and PET scans can confirm a functional relationship between the radiologic lesion and the epileptogenic area.

Electrocorticography (ECog) may be performed to more precisely identify dysfunctional regions, as well as the epileptogenic zone. ECog involves recording of electrical activity through epidural, subdural, or depth electrodes on the cortex. Cortical mapping with electrical stimulation of the cortex and cortical somatosensory evoked potentials may be required to more precisely determine areas of eloquent cortex, such as motor and language regions, if the ictal focus lateralizes to this region, or to guide resection of tumors involving eloquent cortex.108,109

The use of ECog data in modifying the extent of surgical resection is controversial. ECog may demonstrate slow wave activity over the tumor, whereas epileptiform activity may be seen from normal-appearing cortex. There is no clear correlation between the presence of spikes on ECog over the tumor and seizure outcome postoperatively.119 New postresection discharges are considered to be activation phenomena, unless these discharges are sustained, independent, or clearly epileptiform,120,121 but residual spikes do not necessarily correlate with poor postoperative seizure control.

Only one prospective study of ECog during tumor surgery was published. Tumor resection was performed and not altered by ECog findings before or after resection. The frequency of spike discharges on ECog before and after resection was equal between patients who were seizure-free after surgery and those with persistent seizures.119 ECog may not be necessary during tumor resection in patients with well-controlled preoperative seizures.45 Further controlled prospective analyses of ECog in tumor surgery are needed.

Postoperative follow-up

Postoperative follow-up is usually scheduled for 1 and 6 months after surgery. It includes a neuropsychologic assessment, EEG, and MRI.

Adapted from: Mangano FT, McBride AE, and Schneider SJ. Brain tumors and epilepsy. In: Ettinger AB and Devinsky O, eds. Managing epilepsy and co-existing disorders. Boston: Butterworth-Heinemann; 2002;175–194.
With permission from Elsevier (www.elsevier.com). 

Authored By: 
Steven C. Schachter MD
Authored Date: