*Riki Matsumoto, *Akio Ikeda, *Kiyohito Terada, *Takeshi Satow, *Masao Matsuhashi, *Susumu Miyamoto, and *Hiroshi Shibasaki *Departments of Neurology and Neurosurgery, and Human Brain Research Center, Kyoto University Graduate School of Medicine, Kyoto, Japan .
Objective: To delineate the clinical and electrophysiologic features of epileptic negative phenomena in partial seizures.
Case Reports: Case 1: A 30-year-old man with a low-grade glioma in the right postcentral gyrus underwent cortical mapping by implanted subdural electrodes as a part of the presurgical evaluation. His habitual seizure started with an aura of chest discomfort, followed by a left arm clonic seizure. During the monitoring, ictal monoparesis of the left arm was documented twice with video and electrocorticogram (ECoG). The ictal monoparesis was accompanied by chest discomfort without loss of consciousness. The muscle tonus was normal, and no definite convulsion was observed. ECoG revealed ictal epileptiform discharges in the positive arm motor area of the right precentral gyrus, but not in the negative motor area. The ictal monoparesis, chest discomfort, and epileptiform discharges were all resolved soon after the injection of diazepam. Case 2: A 38-year-old woman had stiffening of the left hand for ∼30 min, and then weakness of the left upper extremity developed for several hours, and she sought medical assistance. Mild to moderate weakness was observed in the left upper extremity without sensory disturbance. Scalp EEG showed ictal epileptiform discharges in the right central region, which disappeared after the weakness was resolved with antiepileptic medication. Magnetic resonance imaging (MRI) revealed thrombosis of the right central vein, and high-intensity signal was seen along the arm/hand portion of the central sulcus, consistent with venous congestion. Case 3: A 38-year-old man underwent video-EEG monitoring for frequent drop attacks. Clinically, seizures consisted of abrupt onset of loss of consciousness and facial grimacing, followed by a general atonic state. The drop took 2–3 s. Ictal EEG started with paroxysmal fast activity in the bilateral frontocentral regions, followed by repetitive spikes in the vertex area (Fz to C3-Cz). The drop attacks were completely controlled after switching medication from valproic acid to carbamazepine. Case 4: A 12-year-old developmentally delayed boy had intractable drop attacks. Ictal manifestation documented by video-EEG monitoring, consisted of sudden onset of unresponsiveness, arrest of motion, staring and head nodding, followed by slow development of a general atonic state, and the drop took 3–5 s. Ictal EEG showed regional onset of 4- to 5-Hz rhythmic spikes at the left centroparietal area, and finally evolved into 3- to 4-Hz generalized rhythmic activity, which was clinically associated with a sustained general atonic state.
Conclusions: Ictal monoparesis in cases 1 and 2 was associated with a lesion in the perirolandic area. It seems that epileptic discharges in the sensorimotor arm area selectively activate the regional inhibitory motor system, thus inhibiting the spinal motoneuron pool without eliciting excitatory activity in the corticospinal pathway. In cases 3 and 4, partial epilepsy manifested atonic seizures, characterized by slow falls. Epileptic activation of the negative motor area is likely responsible for the generalized atonic state. Epileptic negative phenomena in partial seizures seem to be due to the epileptic activation of inhibitory motor system in the primary or nonprimary motor area.
*Masako Kinoshita, *Akio Ikeda, ‡Masao Matsuhashi, ‡Junichi Yamamoto, †Takeshi Satow, ‡Tahamina Begum, ‡Keiko Usui, *Riki Matsumoto, †Susumu Miyamoto, and *‡Hiroshi Shibasaki Departments of *Neurology and †Neurosurgery, Graduate School of Medicine , and ‡Human Brain Research Center, Kyoto University, Kyoto, Japan .
Objective: To delineate how ictal activity affects the background cortical activity, as an index of cortical excitability in the epileptogenic focus and also in its surrounding area.
Methods and Subjects: Data were obtained from a 51-year-old male patient with extratemporal lobe epilepsy who had simple partial seizures followed by complex partial seizures. The patient had implantation of subdural electrode grids (3-mm-diameter electrodes arranged in 1-cm center-to-center distance as two sheets of 2 × 5 grids) for a preoperative evaluation. The ictal electrocorticogram (ECoG) showed bursts of fast (20–25 Hz) activities, clearly restricted to one electrode in the right parietal cortex at the beginning of seizures. The ECoG was recorded by a digital EEG machine (EEG2100, Nihon Kohden) with a sampling rate of 200 Hz, time constant of 10, and low-pass filter of 70 Hz. A continuously recorded ECoG (41 h 3 min) was reviewed. Sixteen electrographic seizures, each being >3 s and not followed or preceded by another seizure for ≥10 min, were selected for analysis. All the 16 seizures had the same ictal pattern and recorded at only one major electrode. Electrodes with significant artifacts were excluded, and thus ECoGs recorded from 16 electrodes were analyzed. The ECoG portions 5 min before and after seizures were subdivided into 1-min segments, and power spectra were calculated for each 1-min segment for each electrode. The averaged data of power spectra for every 2-Hz bin were adopted for comparison. The averaged data of the first and second segments during the 5-min preictal period served as baseline.
Results: At the ictal-onset electrode, ECoG power of 12- to 16-Hz activities increased significantly during the 0- to 2-min period before seizure onset (p < 0.05), and returned to the baseline level after seizure. In the surrounding area, ECoG power of 10- to 40-Hz activities decreased during the 1- to 5-min period after a seizure (p < 0.05). No differences in ECoG power were noted in the other three electrodes that were distant from the ictal-onset electrode.
Conclusions: Power-spectrum analyses showed that electrographic fast activities increased before seizures at the ictal focus, but they decreased after seizures in the surrounding area. These findings possibly suggest that cortical excitability at the epileptogenic focus increases before seizures and that cortical inhibition in the surrounding area increases after seizures, the latter being consistent with so-called “surround inhibition.”
*Aya M. Tokumaru, †Aihide Yoshino, ‡Osamu Kobayashi, *Masayoshi Yamamoto, *Tatsumi Kaji, *Ikuko Sakata, and *Shoichi Kusano Departments of *Radiology , †Psychiatry , and ‡Pediatrics, National Defense Medical College, Tokorozawa, Japan .
Purpose: Some patients with localization-related epilepsy have a comorbid venous anomaly. However, whether the venous anomaly is involved in the epileptogenicity remains uncertain. To examine the issue, we investigated the tomographic relation between epileptic focus and venous anomaly by using current source dipole analysis.
Methods: Eight of 258 epilepsy patients showed a venous anomaly on magnetic resonance imaging (MRI) in our hospital. Four of them consented to participate in the present study. Interictal spike discharges were recorded by 78-channel digital EEG with a sampling rate of 1,000 Hz. Spatiotemporal dipole resolution of spikes was obtained by using a standardized realistic three-shell model and mapped on three-dimensional MRI images of the brain by using BESA and BrainVoyager. Two neuroradiologists visually evaluated the tomographic relation between dipole and venous anomaly.
Results: In case 1 of occipital lobe epilepsy, both a dipole and venous anomaly were located in right temporooccipital area. In case 2 of occipital lobe epilepsy, a dipole was localized to the right occipital cortex; however, multiple venous anomalies were depicted in the right frontal and left occipital subcortical areas. In case 3, parietal lobe epilepsy, a venous anomaly was located in the right parietal subcortical area just behind the dipole. In case 4 of temporal lobe epilepsy, a dipole was localized to the medial temporal area, and the venous anomaly was observed in the right parietotemporal area.
Conclusions: A venous anomaly was observed in 3.3% of 258 patients with epilepsy in our hospital. In three of four cases, the laterality of the dipole corresponded to the laterality of the venous anomaly. In two of the three cases, the dipole was close to the location of venous anomaly. The present results suggest that venous anomaly may be partly involved in epileptogenicity.
*Minako Ito, *Harumi Yoshinaga, *Yoko Ohtsuka, and *Eiji Oka *Department of Child Neurology, Okayama University Medical School, Okayama, Japan .
Purpose: We performed a retrospective longitudinal clinicoelectroencephalographic study on 23 children who were diagnosed as having absence epilepsy on their initial visits to our facility, and we analyzed the factors associated with an unfavorable prognosis.
Subjects and Methods: The 23 patients were divided into three groups according to their clinical courses: group A: eight patients who responded well to therapy and became seizure free without relapse of epileptic discharges on EEGs; group B: 13 patients who had relapse of epileptic discharges on EEGs, despite clinical seizure cessation; and group C: two patients who continued to have seizures. Clinical information and EEG data analyzed from the initial EEG recordings were collected from all patients. Clinical data included age at onset of seizures, presence of generalized tonic–clonic convulsions (GTCs), presence of automatisms, family history of idiopathic generalized epilepsy, and drug dosages. Ictal EEG data included the presence of “lead in,” average duration of the generalized three per s spike–wave pattern, good or poor synchronization of the spike–wave burst, and whether a clinical absence seizure was induced by hyperventilation. Interictal EEG data included the presence of the occipital delta burst, presence and localization of isolated interictal spikes, and photosensitivity. The data were analyzed by using one-way analysis of variance.
Results: (a) Ages at onset did not differ between groups A and B, although there were more children in group B whose seizures started at a later age; (b) The number of the patients who had GTCs and/or a family history did not differ among the three groups; (c) There were more children with automatisms in group B (46.1%) than in group A (12.5%), although the difference was not statistically significant. Moreover, none was observed in group C; (d) Seizures were controlled in most group A patients (87.5%) by >20 mg/kg of valproic acid (VPA), whereas seizure control was achieved in 46% of patients in group B with <20 mg/kg of VPA. One patient in group C was treated irregularly, resulting in relapse of seizures; (e) The presence of focal onset in ictal EEG, or “lead in,” was observed more frequently in group B (46.1%). However, it also was observed in two patients in group A, but in none in group C. All patients showed absence seizures induced by hyperventilation. No statistically significant differences between the three groups were found regarding the number of patients with poor synchronization, frequency of seizures, and average duration of seizures; (f) Typical posterior rhythmic two to four per s delta bursts were observed in only one patient in group B. None of the patients showed photosensitivity on initial EEG recordings; (g) 56% of all patients had focal epileptic discharges, including a surprising 63% of patients in group A, in contrast to none in group C; (h) In one patient in group C, the epilepsy evolved into complex partial seizures during her clinical course. She appeared to have so-called “frontal absence,” although her initial EEG did not show any focal abnormalities, and she had no clinical focal signs such as ictal automatisms.
Conclusions: We conclude that it is common for patients with absence epilepsy to show focal abnormalities on EEGs and clinical ictal automatisms. Clinical ictal automatisms and focal EEG findings are not sufficient predictors of the treatment outcome. Furthermore, regular and adequate drug therapy appears to be important for a favorable outcome.
*†‡Hideaki Shiraishi, †Seppo P. Ahlfors, ‡Kyoko Takano, ‡Maki Okajima, ‡Naoko Asahina, ‡Akira Sudo, ‡Shinobu Kohsaka, †Susanne Knake, †Steven M. Stufflebeam, †Eric Halgren, §Keisaku Hatanaka, and ‡Shinji Saitoh *Department of Pediatrics, Teine Keijinkai Hospital, Sapporo, Japan ; †MGH/MIT/HMS Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital/Harvard Medical School, Boston, Massachusetts, U.S.A. ; ‡Department of Pediatrics, Hokkaido Graduate School of Medicine, Sapporo ; and §Department of Applied Physics, Okayama University of Science, Okayama, Japan .
Purpose: We evaluated a novel dynamic statistical parametric mapping (dSPM) method for localizing the source of epileptiform activity recorded with magnetoencephalography (MEG) in patients with epilepsy. We compared two methods of localizing the source of epileptiform activity recorded with MEG, equivalent current dipoles (ECDs) and dynamic statistical parametric mapping (dSPM), and evaluated the results by comparison with clinical symptoms and other information.
Subjects and Methods: Fourteen children aged 3–15 years (mean, 8.9 years) were studied. Eight patients had symptomatic localization-related epilepsy (SLRE); two had idiopathic localization-related epilepsy (ILRE), one of whom had benign partial epilepsy with centrotemporal spikes (BECT), and the other had atypical benign partial epilepsy in infancy (ABPE); two patients with symptomatic generalized epilepsy (SGE); one patient with idiopathic generalized epilepsy (IGE); and one patient with undetermined epilepsy (UDE) with Landau–Kleffner Syndrome (LKS). MEG was collected by using a 204-channel helmet-shaped sensor array (Vectorwiew; Neurromag Ltd., Helsinki, Finland.). The raw MEG data were bandpass filtered between 0.03 and 133 Hz and sampled at 400 Hz. EEG was recorded simultaneously for visual screening by using 20 scalp electrodes placed according to the international 10-20 system. The raw MEG data were digitally filtered with a bandpass of 3–30 Hz for offline analysis. Segments containing abnormal paroxysms were selected manually. ECDs were calculated by using dipole-fit software (Neuromag Ltd.). The single-dipole model was used, and ≥20 ECDs were obtained to determine the epileptogenic regions. The dSPM movies were based on the minimum-L2 norm estimate, minimizing the square sum of the dipole element amplitudes. The dSPM represented spatiotemporal maps (or movies) of brain activity with millisecond time resolution (Dale et al., Neuron 2000). In dSPM, a further step of normalizing the spatiotemporal estimates for noise sensitivity yields dynamic statistical parametric maps of brain activity. We selected representative MEG signals for 500–1,000 ms at the vicinity of initial ictal discharges to make dSPM movies.
Results: The localization of ECDs was consistent with dSPM activity in three patients with SLRE, one patient with BECT, and one patient with LKS. In these patients, ECDs were clustered in a specific area, and the activity depicted in the dSPM movie of the initial interictal discharge was generated from and remained in the same area as ECDs. In other patients, dSPM provided data that were more reasonable and informative than the calculated ECDs. In one patient with SLRE, ECDs could not be calculated because the epileptiform discharges were only slow waves, whereas dSPM movies suggested broad activity in a specific region. In four patients with SLRE and one patient with ABPE with ECDs suggesting multifocal activity, dSPM movies demonstrated the propagation of activity from the beginning of initial ictal discharge to multiple foci over a broad area. In these patients, dSPM demonstrated the course of change in epileptiform activity. Furthermore, dSPM depicted bilateral discharges and their propagation in patients with SGE and IGE.
Conclusions: The dSPM method demonstrated the course of propagation of epileptiform activity with high temporal resolution, simultaneously with multifocal activities in the cerebral cortex. The dSPM technique also depicted bilateral discharges and slow-wave activities in patients with GE, which could not be analyzed by conventional methods with the ECDs.