Is an epilepsy presurgical evaluation necessary for mid-grade and high-grade brain tumors presenting with seizures?


  • Hajo M. Hamer,

    Corresponding author
    1. Department of Neurology, Epilepsy Center, University of Erlangen, Erlangen, Germany
    • Address correspondence to Hajo Hamer, Department of Neurology, Epilepsy Center, University of Erlangen, Schwabachanlage 6, 91054 Erlangen, Germany. E-mail:

    Search for more papers by this author
  • Seung Bong Hong

    1. Department of Neurology, Samsung Biomedical Research Institute, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea
    Search for more papers by this author


Patients with epilepsy caused by mid-grade and high-grade tumors do not usually undergo formal presurgical epilepsy evaluations before tumor resection. However, a minority of these patients may benefit significantly from just such a structured presurgical evaluation especially when seizure freedom or seizure reduction is a surgical aim in addition to total tumor resection. Typical cases comprise patients with multifocal tumors, tumors with bilateral extension, tumors over eloquent cortex, and the need for differentiation of spells of an uncertain nature, for example, epileptic versus psychogenic nonepileptic seizures. If they are epileptic, the definition of the epileptic lesion versus the epileptogenic zone and eloquent cortex can be another reason for monitoring. In addition to noninvasive recordings, invasive studies that use subdural or depths electrodes can be of special importance in these patients, leading to an exact delineation of the epileptogenic zone, usually extending beyond the epileptic lesion, and allow safe differentiation of epileptic from eloquent cortex.

Seizures are frequently the first sign of a mid-grade or high-grade central nervous system (CNS) tumor (van Breemen et al., 2007). Complete presurgical evaluation for epilepsy surgery is usually not obligatory in many of these patients, especially if seizure semiology and findings of routine electroencephalography (EEG) are in line with the localization of the tumor on imaging. The concepts of oncologic treatment prevail in these patients and are often sufficient for management.

However, there are several situations in which presurgical epilepsy evaluation is important and helpful in shaping further treatment. The following case reports reflect patients who benefited significantly from presurgical epilepsy evaluations in the presence of mid-grade or high-grade tumors.

Case Reports

Case 1: Multifocal glioma

A 29-year-old man presented with recent onset of dialeptic and secondary generalized tonic–clonic seizures. No lateralizing or localizing signs and symptoms were reported. Routine EEG showed bilateral intermittent temporal slowing without interictal epileptiform discharges (IEDs). Seizures continued despite treatment with carbamazepine and levetiracetam.

Two lesions that were suspicious for multifocal glioma were detected on magnetic resonance imaging (MRI) in the right temporoinsular region and at the left temporoparietal junction (Fig. 1). Patient management discussions resulted in a decision to first resect the epileptogenic lesion. The patient was therefore admitted for presurgical epilepsy assessment.

Figure 1.

MRI of a 29-year-old patient with multifocal astrocytoma grade II and epilepsy. (A) Right temporoinsular tumor; (B) left temporoparietal tumor.

During video-EEG monitoring, left temporal IEDs were recorded as well as typical dialeptic seizures with a left lateralized seizure pattern on EEG. Postictal dysphasia was noted. No IEDs were recorded on the right. Functional MRI demonstrated left hemispheric language dominance, and neuropsychological testing showed mild verbal memory deficits. Left temporoparietal tumor-related epilepsy (TRE) was diagnosed with language dominance on the left.

Based on the results of the presurgical epilepsy assessment, the left-sided lesion was resected without cognitive or language deficits. Histopathology revealed astrocytoma grade II. The patient remained postoperatively seizure-free without antiepileptic drug (AED) changes.

Two years later there was no recurrence of the astrocytoma on the left, but a slight increase of the right-sided tumor was noted. Seizures did not return at this point. The patient was scheduled for a right temporal lesionectomy.

This case demonstrates the value of a presurgical epilepsy evaluation in the identification and characterization of the epileptogenicity of CNS tumors, and the influence of this evaluation on the therapeutic approach adopted.

Case 2: “Funny spells” after diagnosis of glioblastoma with bilateral extension

A 40-year-old female patient was diagnosed with a right frontal lobe tumor extending into the left frontal lobe after a first generalized tonic–clonic seizure. Tumor biopsy revealed an astrocytoma grade II. Levetiracetam was started and the patient initially remained seizure-free. Four months later the patient developed almost daily flurries of spells consisting of loss of tone without loss of consciousness. Falls and the inability to open the eyes accompanied the spells. Nonepileptic seizures were assumed in the setting of an acute stress disorder. However, repeated imaging revealed hemorrhagic growth of the tumor.

The patient was admitted for presurgical epilepsy assessment to clarify the nature of the spells before further treatment. Video-EEG monitoring revealed right frontal IEDs and frequent atonic seizures with closed eyes accompanied by right frontocentral seizure patterns (Fig. 2). Nonepileptic seizures were not recorded. Therefore, right frontal lobe epilepsy was diagnosed, caused by the right frontal tumor.

Figure 2.

Ictal EEG of an atonic seizure of a 40-year-old patient with right frontal glioblastoma. Note the lack of muscle artifacts during the seizure and the right frontocentral seizure pattern (arrow).

Combination therapy was established consisting of levetiracetam 4,000 mg/day and lacosamide 400 mg/day, which rendered the patient seizure-free again. Based on the results of the presurgical evaluation, a subtotal right frontal tumor resection was preformed followed by radiotherapy and chemotherapy. Histopathology showed malignant transformation of the tumor to glioblastoma.

In this case, presurgical epilepsy assessment confirmed the epileptic nature of the spells and was able to lateralize the epileptogenic zone in the presence of bilateral tumor. This helped achievement of seizure freedom and in the designing of the resective procedure.

Case 3: Right frontal tumor, epilepsy, and right hemispheric dominance for language

A 38-year-old, strictly right-handed man had drug-refractory right frontal lobe epilepsy due to a right frontal lobe tumor (Haag et al., 2006). His seizures consisted of a nonspecific aura with a feeling of derealization, followed by a generalized tonic seizure. He reported speech arrest, but no loss of consciousness and unimpaired language comprehension during the seizure. MRI showed a right frontal tumor without contrast enhancement indicative of a low-grade tumor.

The patient was admitted for presurgical evaluation. Interictal EEG showed right frontotemporal sharp waves. Despite the withholding of AEDs, no seizures occurred throughout video-EEG monitoring. Neuropsychological testing revealed minor deficits in working memory, but attention, long-term verbal, and figural memory as well as language and executive function were unimpaired. Surprisingly, functional MRI (fMRI) showed profound right hemispheric language dominance (Fig. 3). To confirm this finding, the patient underwent functional transcranial Doppler sonography (fTCD) (Knecht et al., 1998; Knake et al., 2003), also indicating right hemisphere dominance for language (Haag et al., 2006). Additional invasive language testing, such as the Wada test or language mapping through subdural grid electrode stimulation was not performed because the risks of invasive language testing were felt to outweigh benefits in this case.

Figure 3.

Functional MRI activation map for word stem completion task in a patient with right frontal lobe epilepsy. Note the right frontal tumor in slices 26–32 (white arrow) and the right frontal activation during the language task (white triangle).

With the aid of neuronavigation including the fMRI findings, a gross total tumor resection was carried out sparing language areas. Histologic investigations revealed an anaplastic ganglioglioma grade III. Postoperative radiotherapy was therefore initiated. At 6 month follow-up, the patient was seizure-free without language impairment or other neurologic deficits.

This and another case report (Spreer et al., 2001) emphasizes that atypical language dominance has to be taken into account, even in right-handed patients without a family history of left handedness. Apart from the definition of the epileptogenic zone, another integral part of presurgical epilepsy evaluation is the identification of eloquent cortex that must remain in situ. This aspect of presurgical epilepsy assessment generated important information in this case for the design of the resection. Hemispheric language dominance should routinely be evaluated when resections are planned within or adjacent cortical areas that may be involved in language function.

Case 4: Pericentral tumor

A 55-year-old man has very frequent partial seizures consisting of fear or somatosensory aura (tingling sensation in right arm) and then right arm clonic seizures, >10 times per day. His seizures rarely progressed to generalized tonic–clonic seizures and were resistant to AED treatment (valproate 900 mg, oxcarbazepine 900 mg, and clobazam 20 mg/day). The patient's brain MRI showed a rounded mass lesion with calcification in the left pericentral region (Fig. 4). The brain 18F-fluorodeoxyglucose positron emission tomography (FDG-PET) revealed increased glucose metabolism over that area indicating a high grade malignant tumor. The patient was admitted to the epilepsy monitoring unit to localize the seizure focus. He had 39 seizures consisting of right wrist clonic seizures followed by right shoulder clonic seizures; 23 seizures were preceded by an aura (fear of having a convulsion). No secondarily generalized tonic–clonic seizures was observed. No interictal spikes were recorded, but ictal EEG recordings showed nonlateralized, diffuse fast activity with bi-frontocentral maximum (Fig. 4). Ictal single photon emission computed tomography (SPECT) showed focal hyperperfusion in the brain regions adjacent to the tumor. Brain fMRI showed sensorimotor activation in brain regions around tumor, rather than right on the tumor.

Figure 4.

A rounded left pericentral tumor is seen (A), with increased glucose metabolism (B) and focal ictal hyperperfusion on subtraction ictal single-photon emission computed tomography (SPECT) coregistered to MRI (SISCOM). (C) over that region. fMRI shows peritumoral activation with right hand sensory stimulation (D). Ictal EEG onset shows nonlateralized diffuse fast activities (E, black arrow).

Intracranial grid electrodes were placed over the tumor and adjacent brain regions to further localize the epileptic focus and to perform functional mapping by direct electrical stimulation on the cortex surrounding the tumor. Twenty-two seizures (right wrist clonic seizure and then right shoulder clonic seizure) were recorded during the invasive EEG recording. Poorly formed interictal spikes were recorded in the left inferior parietal region far from the tumor location. But ictal EEG on intracranial grid electrodes showed seizures starting from one electrode (I-14) overlying the tumor with spread to the adjacent electrodes (Fig. 5). Tumor with overlying cortex and adjacent cortex showing EEG seizure onset was resected. The pathology of the tumor was glioblastoma. The patient had a mild right hand weakness and received radiation therapy and chemotherapy postsurgically. After the surgery, the patient was seizure-free for 4 months when his right leg focal motor seizures recurred but occurred less frequently at one per one or 2 months. At that time, the follow-up brain MRI did not show residual tumor or tumor recurrence.

Figure 5.

Ictal EEG on intracranial grid electrodes showed seizures starting from one electrode (I-14, black arrow) with spread to adjacent electrodes (left). Results of brain mapping with electrical stimulation are shown in relation to the tumor and resection margins (right).

Glioblastoma recurred at 10 months after surgery, and the patient received gamma knife surgery. However, the postoperative 16-month brain MRI showed increased extent of enhancing tumor lesion in the left frontal lobe. The patient died of infection at 21 months after the surgery. In summary, the patient survived 21 months after epilepsy surgery with initial seizure freedom and later much less frequent seizures. Although the patient survived <2 years after surgery, invasive EEG monitoring and resection of seizure-onset zone greatly reduced seizures and increased quality of life.


The case reports underscore the importance of presurgical epilepsy evaluation in a significant minority of patients with epilepsy caused by mid-grade or high-grade tumors especially when seizure freedom is a goal of surgery in addition to total tumor resection. A structured presurgical epilepsy evaluation includes identification of eloquent cortex to be spared in addition to the definition of the epileptogenic zone to be removed (Rosenow & Luders, 2001). Therefore, some patients benefit from presurgical evaluation in this respect even if seizure semiology and routine EEG findings are concordant with radiologic localization.

In general, case characteristics that make it advisable to perform a presurgical epilepsy evaluation include cases with multifocal tumors, tumors with bilateral extension, tumors over motor or language cortex, the need for clarification of unclear spells, as well as identification of the epileptic lesion versus the epileptic zone and eloquent cortex.

In addition to noninvasive testing, invasive recordings with subdural or depths electrodes have the great advantage of accurately delineating the epileptogenic zone, which usually extends beyond the borders of the epileptogenic lesion. Furthermore, cortical mapping enables differentiation of epileptic from eloquent cortex so as to design a safe and successful resective procedure. This may lead to postoperative seizure freedom in addition to tumor removal without inducing cognitive or other neurologic deficits.


The authors have no conflict of interest to disclose. The authors confirm that they have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.