Free Papers: Neurosurgery


*Hiroyuki Shimizu, *Kensuke Kawai, and *Shigeki Sunaga *Department of Neurosurgery, Tokyo Metropolitan Neurological Hospital, Fuchu, Tokyo, Japan .

Background: Despite good surgical outcomes after temporal lobectomy, postoperative impairment of verbal memory remains a serious problem to be solved, especially when surgery is performed on the left (dominant) side in patients without hippocampal atrophy on preoperative magnetic resonance imaging (MRI). To solve this long-standing problem, we developed a new surgical technique for the treatment of left temporal lobe epilepsy that preserves verbal memory.

Surgical Technique: An ∼2-cm corticotomy is placed on the anterior superior temporal gyrus along the sylvian veins. After aspiration of the gray matter along the sylvian fissure, the inferior horn is accessed through the temporal stem. Intraoperative electrocorticogram (ECoG) is recorded over the surface of the hippocampus and amygdala, and the extent of epileptic areas is determined. After cutting the alveus, the pyramidal layer is transected with a 2-mm ring transector. The transection lines are ∼5 mm apart and parallel to the alveus fibers to preserve neuronal connections. A 4-mm ring transector is used at both ends to transect the CA4 and the transitional zone from the CA1 to the subiculum.

Subjects and Methods: For the past 40 months, a total of 19 patients underwent hippocampal transaction; right side in six, left side in seven, and bilateral in six. For all patients, complex partial seizures were the most disabling habitual seizures. Preoperative MRI demonstrated no hippocampal atrophy and no structural lesions. In most of the patients, intracranial electrodes were placed, and prolonged ECoG and video monitoring were carried out to confirm the exact distribution of epileptic areas. Neuropsychometry was performed preoperatively and postoperatively. For evaluation of verbal memory, an auditory verbal learning test (AVLT) was performed. In the AVLT test, a patient tried to memorize 15 different words, and the ability to recall these words immediately was assessed. This test was repeated 5 times. After that, a different list of 15 words was presented, and immediate recall was again tested. Twenty minutes after this interference test, the patient was asked to recall the first list of 15 words. The score of this delayed recall was used to examine verbal memory function of the hippocampus.

Results: The seizure outcome after hippocampal transection was excellent in cases with unilateral focus. All 13 patients with unilateral transection achieved a seizure-free outcome for >6 months after surgery. On the contrary, various outcomes were obtained in patients with bilateral foci; two cases became seizure free, another two cases showed >90% and 50–90% reduction of seizures, and the remaining two developed new types of seizures after surgery. The WAIS-R and Benton visual retention tests showed no remarkable changes in any of the cases. Verbal memory function evaluation by the AVLT showed no significant changes in the right-sided cases. Of the left-sided cases, four cases with only hippocampal transection showed preservation of the AVLT scores immediately after surgery; the remaining three cases with additional excision of the temporal tip showed transient decreases of the AVLT scores after surgery and recovered to the preoperative level within 6 months. In cases with bilateral transections, all cases recovered to a level near the preoperative score within 5–9 months after the last surgery.

Conclusions: These results show that our new method of hippocampal transection is very effective for controlling seizures in temporal lobe epilepsy. Verbal memory can be preserved even in left (dominant) temporal lobe cases. This technique also opens a possibility for surgical treatment of bilateral temporal lobe epilepsy.


*Tomonori Ono, *Hiroshi Baba, *Keisuke Toda, and †Kenji Ono *Department of Neurosurgery, National Nagasaki Medical Center, Omura ; and †Yokoo Hospital, Isahaya, Japan .

Purpose: Whereas the callosal participation in some generalized epilepsies has long been presumed, selective activities of the callosal neurons have not been studied in relation to the epileptic seizures. Our recent study (Ono et al. Epilepsia 2002; 43:1536–42) of intraoperative electrocorticograms (ECoGs) and callosal compound action potentials (CCAP) revealed that the corpus callosum engages in the interhemispheric recruitment of bilateral epileptogenesis. In other words, callosal and other cortical neurons in both hemispheres are simultaneously recruited, thereby resulted in bilaterally synchronous spike–wave discharges (BSSWs). In this study, asymmetrical recruitment of the callosal neurons associated with BSSWs was evaluated and correlated with the EEG outcome after callosotomy.

Methods: In 16 patients undergoing corpus callosotomy for intractable epilepsy, intraoperative ECoGs and CCAPs simultaneously were recorded before callosal resection with the patients' consent. Detailed procedures of the intraoperative recordings and data analysis have been reported (Ono et al. Epilepsia 2002; 43:1536–42). We devised a novel method based on spike-locked averaging to estimate separately the component for each hemisphere (hemispheric CCAP) from the observed CCAPs synchronized to BSSWs (those essentially include bidirectional activation). The degree of asymmetry in callosal involvement was defined as a laterality index (LI) based on the amplitude difference between hemispheric CCAPs. According to our routine procedure, postoperative seizure frequency and EEG were assessed 1 month after the surgery. Laterality of hemispheric CCAP was correlated with clinical seizure outcomes and postoperative EEG findings.

Results: Seizure reduction of ≥80% was obtained in nine patients (excellent outcome group, group E). The remaining patients had unsatisfactory reduction <80% (not-excellent outcome group, group NE). LI was significantly higher in group E than in group NE (0.51 ± 0.08 vs. 0.16 ± 0.05; p < 0.01). Although preoperative BSSWs were common to all patients, the EEG outcome was categorized into either unilaterally lateralized epileptiform discharges (group U, eight patients) or bilaterally asynchronous epileptiform discharges (group B, eight patients). LI was significantly higher in group U than in group B (0.55 ± 0.10 vs. 0.20 ± 0.04; p < 0.05). In addition, higher amplitude of hemispheric CCAP coincided with the side of the persistent postoperative seizure discharges in all patients in group U, except one patient with the lowest LI.

Conclusions: Our results suggested that the amount of recruited callosal neurons may be hemispherically asymmetrical during bilateral synchrony (i.e., more callosal neurons are fired in the more epileptogenic hemisphere even if preoperative EEG findings are virtually symmetrical). CCAP recordings may provide some information to predict outcomes, although its clinical usability is limited. According to our results, excellent seizure outcome and lateralization of epileptiform discharges may be expected more in cases with higher LIs. Moreover, diagnosis of hemispheric asymmetry in epileptogenesis may help subsequent treatment (e.g., resection of epileptogenic foci that emerge after callosotomy).


*†Kensuke Kawai, †Hiroyuki Shimizu, and ‡Olivier Delalande *Department of Neurosurgery, University of Tokyo ; †Department of Neurosurgery, Tokyo Metropolitan Neurological Hospital, Tokyo, Japan ; and ‡Unité de Neurochirurgie Pédiatrique, Fondation Ophtalmologique A. de Rothschild, Paris, France .

Purpose: To compare various surgical procedures in hemispherotomy in terms of reliability of disconnection and occurrence of impaired cerebrospinal fluid circulation, and to report a new modification using the vertical approach.

Methods: Since 1994, 47 patients with intractable epilepsy have undergone hemispherotomy at Tokyo Metropolitan Neurological Hospital and one at University of Tokyo Hospital. Thirty-nine cases had malformation of cortical development, 20 of whom had hemimegalencephaly. The other eight cases had an atrophic hemisphere due to infarction or trauma. We reviewed and compared the various surgical procedures we used in hemispherotomy. In the most recent cases, we adopted a new modification of the vertical approach originally designed by Delalande for the purpose of reducing surgical incursion into the deep field and thus reducing blood loss. The frontal horizontal fibers were disconnected at a more posterior level than with the original method, by making the cutting plane between the most anterior point of the choroid plexus in the inferior horn and the foramen of Monro.

Results: The anatomic components to be disconnected in hemispherotomy were (a) the corpus callosum (CC), and (b) the descending fibers (DFs) from hemispheric cortices. For disconnection of CC, the interhemispheric approach and the transventricular approach were compared. The interhemispheric approach enabled easier identification of the midline structures and reliable disconnection of CC, although it necessitated another route for disconnection of DFs. It was especially advantageous in anomalous and asymmetric CC frequently found in hemimegalencephaly. The transventricular approach enabled disconnection of CC and DFs by a single route, but identification of CC and reliable disconnection of the genu and splenium were difficult in some cases. In four of 19 patients who underwent transventricular disconnection of CC, follow-up MRI revealed residual callosal fibers in the genu. Three of them had hemimegalencephaly and had anomalous and asymmetric CC. They underwent additional callosotomy via the interhemispheric route. None of the patients who underwent interhemispheric disconnection of CC had residual callosal fibers. For disconnection of DFs, we originally used the periinsular transcortical approach. It necessitated a long temporoparietofrontal corticectomy around the insula and a long disconnection down to the inferior horn behind the insular cortex, which resulted in problematic blood loss in some cases of hemimegalencephaly. Removal of the frontoparietal operculum overcame the problem, as we previously reported. However, large opercular resection resulted in a higher rate of impaired CSF circulation, necessitating ventricular or subdural shunting. Shunting operation was performed in the postoperative follow-up periods in seven of 15 patients who underwent opercular resection, compared with one of 11 who underwent periinsular approach without opercular resection. Reviewing these results, we adopted interhemispheric disconnection of CC followed by vertical transventricular disconnection of DFs via a small parasagittal corticectomy. To disconnect the frontal horizontal fibers, we originally removed the hippocampal head and inferior amygdala and then removed the posterior part of orbitofrontal cortices, but had difficulties in obtaining a sufficient anterior view. The latest modification was to accomplish disconnection of DFs by the smallest length. After anteriorly cutting down the lateral aspect of the thalamus, we disconnected the frontal horizontal fibers between the foramen of Monro and the most anterior point of the choroid plexus in the inferior horn. This disconnection formed the anterior surface of the thalamus.

Conclusions: We reviewed and compared our surgical procedures of hemispherotomy. Interhemispheric callosotomy was reliable for the disconnection of CC in hemimegalencephaly. Parasagittal corticectomy and vertical disconnection of DFs downward along the lateral border and anterior border of the thalamus would be the most efficient approach in view of the small corticectomy and disconnection plane.


*Nobuhiro Mikuni, *Naoki Hayashi, *Namiko Nishida, †Akio Ikeda, *Susumu Miyamoto, and *Nobuo Hashimoto *Departments of Neurosurgery and †Neurology, Kyoto University Graduate School of Medicine, Kyoto, Japan .

Background: In intractable epilepsy patients with brain tumor, epileptogenic tissues are commonly peritumoral and identified as gliosis/siderosis or cortical dysgenesis. Lesionectomy including the peritumoral tissue cures or significantly improves seizures in 70–90% of patients. It is known that hippocampal sclerosis or cortical dysgenesis coexists with brain tumor. These histologic abnormalities may coexist as an epileptogenic area distant from the tumor, causing intractable seizure after tumor resection. In this study, the incidence and spatial relations between coexisting epileptogenic lesion and brain tumor were studied.

Subjects and Methods: In all 26 patients, epileptogenic areas determined by electrocorticogram were additionally excised after tumor resection. The tumor was in the temporal lobe in 17 and in the extratemporal lobe in nine patients. Low-grade astrocytoma (n = 8), meningioma (n = 4), and cavernoma (n = 5) were common pathologic categorizations of the tumors. In two patients (one with ganglioglioma and the other with cavernoma), epileptiform discharges disappeared soon after total removal of the tumor, without peritumoral resection. In all other patients, peritumoral tissues still generated epileptiform discharges after tumor resection and were diagnosed as gliosis/siderosis in 10 patients. Tumor coexisted (dual pathology) with hippocampal sclerosis in five patients, and with cortical dysgenesis in nine patients. In four patients, two with hippocampal sclerosis and two with cortical dysgenesis, these histologic abnormalities were separated from the tumor by more than one gyrus. Clinical features in these four patients were (a) tumor location in the temporal lobe; (b) frequent seizures; and (c) long duration before surgery. Twenty-one patients became seizure free, and four patients showed significant seizure improvement (Engel's class II).

Conclusions: Our results suggest that isolated pathology such as hippocampal sclerosis or cortical dysgenesis located at a distance from the tumor may acquire epileptogenicity during long-term frequent seizures. In these patients, evaluation of epileptogenicity extensively around the tumor is important for further understanding of tumoral epilepsy.