Hippocampal volumetry in children 6 years or younger: assessment of children with and without complex febrile seizures

PURPOSE

To study the relationship of complex febrile seizures (CFS) in the evolution of mesial temporal sclerosis.

METHODS

We studied five children 22-68 (mean 44) months old with MRI volumetry 2 days-46 months after their first CFS, and compared total hippocampal volumes and right to left hippocampal volume ratios to those of 11 controls, 15-83 (mean 55) months old, who had MRI for complaints which turned out to be neurologically insignificant.

RESULTS

In control children, total hippocampal volumes increased linearly with age, while right to left hippocampal volume ratios tended to decrease with age. In children with CFS total hippocampal volumes tended to be smaller than in controls. Right to left ratios were greater than 1 in all five children with CFS compared to seven of 11 controls. Hippocampal asymmetry was noted in only one child, with the right to left volume ratio exceeding two standard deviations from the control mean. The MRI of this child also demonstrated a subarachnoid cyst in the left frontocentral region, ipsilateral to the smaller hippocampus. Visual inspection of the remaining patients revealed no definite structural cortical abnormalities. None of the children developed subsequent afebrile seizures during the brief follow-up period.

CONCLUSIONS

Hippocampal volumetry in controls revealed a linear increase in total hippocampal volumes and a statistically nonsignificant trend toward reduced right larger than left hippocampal ratios between 17 and 83 months old. The tendency for smaller total hippocampal volumes and larger right to left hippocampal volume ratios in children with CFS compared to controls could suggest a developmental abnormality, injury during CFS, or be age-related. The significant hippocampal asymmetry in a single child with CFS suggests that age may not be a factor in every case. Further studies are needed to collect control data in young children as well as prospectively follow children with CFS with serial imaging to better understand the relationship between CFS and the evolution of hippocampal atrophy.

Semiological seizure classification

We propose an epileptic seizure classification based exclusively on ictal semiology. In this semiological seizure classification (SSC), seizures are classified as follows: a. Auras are ictal manifestations having sensory, psychosensory, and experiential symptoms. b. Autonomic seizures are seizures in which the main ictal manifestations are objectively documented autonomic alterations. c. "Dialeptic" seizures have as their main ictal manifestations an alteration of consciousness that is independent of ictal EEG manifestations. The new term "dialeptic" seizure has been coined to differentiate this concept from absence seizures (dialeptic seizures with a generalized ictal EEG) and complex partial seizures (dialeptic seizures with a focal ictal EEG). d. Motor seizures are characterized mainly by motor symptoms and are subclassified as simple or complex. Simple motor seizures are characterized by simple, unnatural movements that can be elicited by electrical stimulation of the primary and supplementary motor area (myoclonic, tonic, clonic and tonic-clonic, versive). Complex motor seizures are characterized by complex motor movements that resemble natural movements but that occur in an inappropriate setting ("automatisms"). e. Special seizures include seizures characterized by "negative" features (atonic, astatic, hypomotor, akinetic, and aphasic seizures). The SSC identifies in detail the somatotopic distribution of the ictal semiology as well as the seizure evolution. The advantages of a pure SSC, as opposed to the current classification of the International League Against Epilepsy (ILAE), which is actually a classification of electroclinical syndromes, are discussed.

Olfactory epileptic auras

BACKGROUND

Olfactory epileptic auras are rare, constituting about 0.9% of all auras, and are typically described as unpleasant. They have usually been associated with tumors, but in some recent studies they have not.

METHODS

We identified 13 patients (7 male, 6 female) with olfactory epileptic auras from 1423 patients with partial epilepsy evaluated for intractable seizures between 1991 and 1996. All had routine EEGs and MRI. Twelve underwent prolonged video-EEG monitoring.

RESULTS

Olfactory sensations were of various types but unpleasant in only seven. Auras evolved to complex partial seizures (automotor, hypermotor, or dialeptic seizures) in 12 patients, with further evolution to generalized tonic-clonic seizures in three and aphasic seizures in one patient. The EEG focus was localized to the mesial temporal region in all. Ten patients had a mesial temporal tumor; in one patient, it extended to the superior temporal gyrus, and in another, the frontal lobe. The tumor involved only the amygdala in two patients and both amygdala and hippocampus in six; none had hippocampal involvement alone. Surgery was performed in nine patients. All except one with partial tumor resection had a seizure-free outcome. This patient also became seizure-free after repeat surgery to remove residual tumor tissue in the amygdala.

CONCLUSIONS

Olfactory auras are not necessarily unpleasant. The amygdala is the most likely symptomatogenic zone of olfactory auras. Tumors are the commonest etiology; mesial temporal sclerosis is a relatively rare cause.

[Multi-national therapy study for Hodgkin's disease in children and adolescents GPOH-DH 95. Interim report after 2 1/2 years]

Based on concepts of the successful German-Austrian pediatric Hodgkin studies DAL-HD 78 until-90, a new trial was initiated addressing the question whether radiotherapy can be further reduced or can be omitted in case of complete remission after initial chemotherapy, aiming at reduction of sequelae after radiotherapy, especially radiogenic second malignancies. In respect to CHEMOTHERAPY patients are stratified into 3 therapy groups (TG) according to stage and gender: 2 courses of OPPA (girls) or OEPA (boys) in TG1 (stage IA/B, IIA), and in addition 2 (TG2: stage IEA/B, IIEA, IIB, IIIA) or 4 (TG3: stage IIEB, IIIEA/B, IIIB, IVA/B) COPP courses. Boys with stage IIIB and IIIEB receive OPPA instead of OEPA. RADIOTHERAPY is administered according to response to chemotherapy independent of stage: patients with complete remission or minimal residues do not receive irradiation; patients with more than 75% tumor regression are irradiated to involved fields at a dose of 20 Gy. Doses of 30 or 35 Gy are given to regions with tumor regression below 75% or residual bulky tumor of > 50 ml, respectively. INTERIM RESULTS: From 8/95 till 1/98 we registered 385 patients under the age of 18 years from Germany, Austria, Switzerland, Sweden and the Netherlands. Therapy has been completed in 334 patients. Three patients with solitary nodular paragranuloma were treated with surgery only. Out of 331 patients 89 (26.9%) achieved a complete remission with chemotherapy. Tumor regression of more than 75% was seen in 193 (58.3%) patients and below 75% in 39 (11.8%) patients. Tumor progression during chemotherapy occurred in 1 (0.3%) patient. Response after chemotherapy was not evaluable for 9 (2.7%) patients. Radiotherapy was omitted in 91 (27.1%) patients: in TG1 50 of 142 (34%) patients, TG2 24 of 98 (24.5%) patients and TG3 18 of 94 (19.2%) patients. Initially involved regions were irradiated at a dose of 20 Gy in 164 of 334 (49.1%) patients. Doses up to 30 Gy or 35 Gy were given to 19 (5.7%) or 57 (17.1%) patients respectively. Events (tumor progression, relapse or death) occurred in 23 of 334 patients until now. The event-free survival rate is 0.91 at 2 1/2 years for all study patients and 0.89 for patients without radiotherapy. Six relapses occurred in 91 patients without radiotherapy. No relapse occurred in TG1 (n = 49), but in 5 of 24 TG2-patients, and in 1 of 18 TG3 patients without radiotherapy. As yet, the results are not significantly inferior compared with trial DAL-HD 82. Therefore this trial aiming at omitting radiation therapy in patients with complete remission after a short lasting chemotherapy will be continued. Longer follow up is necessary for final evaluations and conclusions.

Ictal motor signs and interictal regional cerebral hypometabolism

Early motor manifestations are the main components of focal seizures involving the frontal lobe. We examined the relationship between the initial ictal motor manifestations and interictal abnormalities of cerebral glucose consumption (rCMRGlc) as assessed by PET in 48 consecutive patients with focal seizures of neocortical origin. Group data analysis revealed that patients with predominantly unilateral clonic seizures had a significant contralateral perirolandic hypometabolism and to a lesser degree a contralateral frontomesial hypometabolism. Patients with predominantly focal tonic manifestations showed a hypometabolism within the frontomesial and perirolandic regions that was unilateral in all patients with lateralized tonic seizures. Patients with versive seizures had mainly contralateral metabolic depressions without a consistent regional pattern. Patients with hypermotor seizures had metabolic depressions involving frontomesial, anterior cingulate, perirolandic, and anterior insular/frontal operculum areas. In all patient groups, bilateral and symmetric hypometabolism of the thalamus and cerebellum was observed. We propose that this pattern of distinctly abnormal metabolic brain regions demonstrates not only possible epileptogenic zones but also symptomatogenic brain regions as shown by the associations between clinical manifestations and sets of abnormal brain regions, particularly if epileptogenic zones are in a clinically silent neocortical brain region. The detection and possible differentiation of symptomatogenic and epileptogenic zones might improve the effectiveness of presurgical noninvasive studies.

Topography of interictal glucose hypometabolism in unilateral mesiotemporal epilepsy

We mapped the regional cerebral glucose metabolism (rCMRGlu) in 20 patients suffering from medically refractory focal epilepsy of either left or right mesiotemporal origin (mTLE) during resting wakefulness. After temporal lobectomy, histology demonstrated hippocampal sclerosis in 18 patients. Pixel-by-pixel comparisons with healthy control subjects showed significant (p < 0.001) depressions of the mean rCMRGlu ipsilateral to the epileptic focus in the mesiotemporal region, including the hippocampus and the parahippocampal gyrus and middle temporal gyrus. Additional remote rCMRGlu depressions occurred bilaterally in the fronto-orbital cortex and ipsilaterally in the posterior insula and the thalamus. Patients with left-sided mTLE had additional rCMRGlu depressions in the left inferior frontal gyrus (Broca's region) and superior temporal gyrus at the parietotemporal junction, whereas corresponding rCMRGlu depressions were not present in patients with right mTLE. Neuropsychological testing showed impaired verbal fluency, verbal intelligence, and verbal memory in the left mTLE patients. Correlations of the specific mean rCMRGlu depressions and the neuropsychological deficits suggest that impaired language functions in patients with left mTLE could result from functional changes beyond the temporal lobe.

Brain-stem auditory evoked potentials (BAEPs) from basal surface of temporal lobe recorded from chronic subdural electrodes

BAEPs were recorded from the basal surface of the temporal lobe by subdural electrodes chronically implanted in 6 patients who were evaluated for surgical management of intractable partial seizures. Near-field recordings were obtained by recording between the subdural electrode closest and most distant to the brain-stem. Far-field recordings were obtained by recording between the subdural electrodes and an indifferent electrode over the spinal process of the seventh cervical vertebrae. The recordings were compared with standard ear-vertex recordings. After ipsilateral ear stimulation, the subdural electrode closest to the brain-stem recorded large amplitude waves I and II followed by less well-defined waves of longer latencies. Recordings to contralateral stimulation showed no clearly defined waves I and II and a large amplitude wave Vn. Waves III, IV, V, Vn and VI were of opposite polarity after ipsi- and contralateral stimulation. These findings indicate that waves I and II are generated ipsilaterally to the stimulation side, whereas wave Vn has a contralateral origin. Wave Vn may be generated in the brachium of the inferior colliculus, as suggested from dipole configuration studies. This conclusion is consistent with the classical anatomical observations that the supracollicular auditory pathways are predominantly crossed.

Frontal lobe seizures: electroclinical syndromes

To define further the electroclinical manifestations of frontal lobe epilepsy (FLE), we studied 150 seizures manifested by 24 patients; 18 patients had subdural electrode arrays (SEA). The findings in these patients clearly overlapped presumably reflecting the interconnections between functionally related frontal zones; yet the manner in which the symptoms clustered and the sequence in which they occurred generally indicated the anatomic site of the epileptogenic zone. We divided the patients into three major groups: (a) those with supplementary motor seizures, (b) those with focal motor seizures, and (c) those with complex partial seizures (CPS, psychomotor seizures). Supplementary motor seizures began with tonic posturing of the extremities. Focal motor seizures generally began with conscious contralateral version or unilateral clonic focal motor activity; tonic posturing was noted only late in the seizure. CPS (psychomotor) began with unresponsiveness at onset, followed by staring or unconscious contraversion. We compared frontal lobe seizures with temporal lobe seizures reported previously; oral-alimentary automatisms, repetitive hand movements, or looking around, were more common in temporal lobe seizures, whereas tonic posturing and bicycling movements were more common in frontal lobe psychomotor seizures.

Propagation of interictal epileptic activity in temporal lobe epilepsy

We recorded interictal spikes with closely spaced scalp electrodes and sphenoidal electrodes in four patients with temporal lobe epilepsy. We used multiple dipole modeling to study the number, three-dimensional intracerebral location, time activity, and functional relationship of the neuronal sources underlying the epileptic spike complexes. In all patients, we found two significant sources generating the interictal spikes which showed considerable overlap in both space and time. Source 1 was located in the mesiobasal temporal lobe and generated a restricted negativity at the ipsilateral sphenoidal electrode and a widespread positivity over the vertex. Source 2 could be attributed to the lateral temporal neocortex and was associated with a relatively restricted negativity at the ipsilateral temporal electrodes and a more widespread positivity over the contralateral hemisphere. The sources were well separated in space, with an average distance of 45 mm between them. The time activities of both sources showed similar biphasic patterns, with the mesial source leading the lateral source by approximately 40 msec, suggesting propagation of interictal epileptic activity from the mesiobasal to the lateral temporal lobe.

Automatisms with preserved responsiveness: a lateralizing sign in psychomotor seizures

This is a report of a 1-year prospective study to investigate how often automatisms occur with preserved responsiveness in psychomotor seizures. Responsiveness is usually impaired or lost when automatisms occur during psychomotor seizures. However, there are several anecdotal reports in the literature of patients who have automatisms with preserved responsiveness (APRs). We evaluated 123 patients with temporal lobe epilepsy (57 patients [46%] left-sided, 48 patients [39%] right-sided, and 18 patients [15%] bitemporal) with video/EEG monitoring, testing responsiveness by asking the patient to respond verbally and to follow motor commands. Seven patients (5.6%) had preserved responsiveness in the presence of prominent automatisms (lip smacking, swallowing). In 15 seizures, the responsiveness was adequately tested (3.6 questions per period of automatism). Average seizure duration was 71.6 +/- 14.8 seconds (range, 45 to 100 seconds). Average duration of automatisms was 59.5 +/- 13.5 seconds (range, 40 to 80 seconds). Ictal EEG was localized over the right temporal area in nine seizures, over the right hemisphere in five, and was nonlocalizable in one seizure. APRs never occurred in left-sided psychomotor seizures and occurred in 10% of the right temporal cases. In conclusion, APRs reliably lateralized to the right side in temporal lobe epilepsy.

[Classification of epileptic seizures. Comparison of two systems]

In spite of its universal acceptance, the international classification of epileptic seizures suffers from certain limitations. The fundamental divisions between partial and generalized seizures on the one hand, and between partial simple and complex seizures on the other, are not always practical, nor useful. The terminology is often cumbersome, and does not contain essential localizing information. Finally, the seizure classification is sometimes dependent on ancillary testing, particularly the EEG. We propose a different seizure classification, which answers the above shortcomings of the international classification, and which has been used for years in major epilepsy centers. In this system, the seizure classification is based exclusively on seizure symptomatology. The terms are generally more concise than those of the international classification, and yet convey more precise information. For example, the term left visual aura-->automotor seizure-->generalized tonic clonic seizure would be equivalent to simple partial seizure with visual symptom evolving into complex partial seizure evolving into generalized tonic clonic seizure. The international classification is longer and omits essential later-alizing information. This classification is easy to apply, and can be an extremely useful complement to the international seizure classification, especially for centers whose emphasis is on surgical treatment of epilepsy.

Quantitative analysis of seizure frequency 1 week and 6, 12, and 24 months after surgery of epilepsy

We made quantitative analysis of seizure frequency 1 week and 6, 12, and 24 months after seizure surgery. Seizure recurrence was significantly higher when seizures occurred in the first postoperative week. Seizure recurrence increased progressively with longer follow-ups, but the 6 month postoperative follow-up period was an excellent index of long-term outcome. In operative follow-up studies, seizure frequency should be reported at fixed follow-up periods, e.g., at 6 months and 1, 2, 5, and 10 years. Meaningful comparison of outcomes between different studies is possible only when reports include outcome at fixed postoperative follow-up periods (as opposed to ranges of follow-up periods).

Functional anatomy of the human supplementary sensorimotor area: results of extraoperative electrical stimulation

Electrical stimulation studies have demonstrated that a "supplementary motor area" (SMA) exists in humans. However, its precise functional organization has not been well defined. We reviewed the extraoperative electrical stimulation studies of 15 patients with intractable epilepsy who were evaluated with chronically implanted interhemispheric subdural electrodes. SMA-type positive motor responses were elicited not only from the mesial portion of the superior frontal gyrus but also from its dorsal convexity, and from the paracentral lobule, cingulate gyrus, and precuneus. Sensory symptoms, that could not be attributed to stimulation of the primary sensory area, were elicited from the superior frontal and cingulate gyri in addition to the precuneus. Therefore, human SMA, as defined by electrical stimulation, is not always confined to the mesial portion of the superior frontal gyrus as described previously. It is also not strictly "motor" but "sensorimotor" in representation. We propose referring to this region as the "supplementary sensorimotor area" (SSMA). We observed a somatotopic organization within the SSMA with an order of lower extremity, upper extremity, and head from posterior to anterior. Sensory representation in an individual was either anterior or posterior to the positive motor representation but never both. There was a supplementary eye field within the head representation. A supplementary negative motor area was noted at the anterior aspect of the SSMA. No language area was demonstrated within the SSMA. The physiologic significance of the SSMA and functional consequences of its resection must be addressed in further studies.

Pattern visual evoked potentials recorded from human occipital cortex with chronic subdural electrodes

Pattern evoked potentials to full- and partial-field stimulation were recorded simultaneously from scalp electrodes and from subdural electrodes located over the temporal and occipital cortex, including electrodes placed over or close to the lower lip of the calcarine fissure. High-amplitude pattern evoked potentials were recorded exclusively from electrodes localized in the vicinity of the calcarine fissure and showed a positive-negative deflection in phase with surface recordings, followed by a second negative peak phase reversed with respect to the major surface positive peak ("P100"). The findings suggest that the initial component is an expression of the afferent volley and that the second component (equivalent of the surface "P100") is most probably generated as a dipole strictly localized to the visual cortex in close proximity of the calcarine fissure (area 17 and/or area 18).

Detection of changes in material-specific memory following temporal lobectomy using the Wechsler Memory Scale-Revised

To determine the utility of the Wechsler Memory Scale-Revised (WMS-R) in measuring material-specific memory changes, within-subject comparisons of the Verbal-Visual Memory Index discrepancy and discrepancy scores using short-term and delayed Logical Memory and Visual Reproduction subtests from the WMS-R were studied prior to and following temporal lobectomy among 30 patients with left temporal lobectomy, 30 with right temporal lobectomy, and 50 epileptic, non-surgical controls. The groups were matched on age, sex, handedness, age at seizure onset, duration of epilepsy, and presurgical Verbal and Performance IQ; the right temporal group had a higher mean educational level (p <.05). All surgical patients were left hemisphere dominant for speech; those who had persistent postoperative seizures were excluded from study. On retesting, left temporal lobectomy was associated with a marked change in short-term and delayed memory discrepancy scores primarily due to a drop in verbal memory. Right temporal lobectomy was not associated with a drop in visual memory, suggesting that the WMS-R appears to reflect decrements in material-specific memory following left but not right temporal lobectomy. The nonsurgical controls showed increases in both short-term and delayed memory discrepancy scores due to increases in short-term and delayed verbal memory. Relative to these controls, the absence of comparable increases in verbal memory among the right temporal patients suggests that right temporal lobectomy may be associated with risk to verbal memory.

Temporal lobe epilepsy in early childhood

To explore the electroclinical features of temporal lobe epilepsy (TLE) in early childhood, we studied results of video-EEG and other tests of 14 children aged 16 months to 12 years selected by seizure-free outcome after temporal lobectomy. Four children had mesiotemporal sclerosis, 1 had cortical dysplasia, and 9 had low-grade temporal neoplasms. The children had complex partial seizures (CPS) with symptomatology similar to that of adults with TLE, including decreased responsiveness and automatisms. Automatisms tended to be simpler in the younger children, typically limited to lip smacking and fumbling hand gestures. Scalp/sphenoidal EEG showed anterior/inferior temporal interictal sharp waves and unilateral temporal seizure onset in the 4 children with mesiotemporal sclerosis and in the child with cortical dysplasia, but EEG findings in 9 children with low-grade temporal tumors were complex, including multifocal interictal sharp waves or poorly localized or falsely lateralized EEG seizure onset. In children without tumors, video-EEG was critical to localization of the epileptogenic zone for resection, but in patients with tumors video-EEG was less localizing and its main value was to confirm that the reported behaviors were epileptic seizures with semiology typical of temporal lobe onset.

Usefulness of unilateral interictal sharp waves of temporal lobe origin in prolonged video-EEG monitoring studies

The value of EEG interictal epileptiform activity in predicting location of the seizure focus remains controversial. In 64 patients, scalp video-EEG monitoring studies showed one or two ipsilateral interictal foci in the temporal lobe. The site of these interictal foci correlated with location of the seizure focus recorded during prolonged video-electrocorticography (ECoG) with use of subdural grids placed under the mesiobasal temporal region and over the lateral temporal convexity. Our findings suggest that unilateral anterotemporal interictal foci can accurately predict location of seizure onset. This is also true in patients with two ipsilateral temporal interictal foci, provided that the dominant focus is localized in anterotemporal regions. We believe that in such patients invasive recordings are not warranted, but we caution against sole use of interictal epileptiform criteria for localization of the seizure focus. Correlation with clinical information, ictal EEG, neuropsychometric, and neuroimaging studies is required before performance of epilepsy surgery.

Comparison of scalp electroencephalogram with subdural electrocorticogram recordings and functional mapping in frontal lobe epilepsy

We compared the findings of scalp electroencephalogram with subdural electrode array (SEA) recordings in 19 patients with refractory frontal lobe epilepsy. Prolonged scalp interictal recordings localized the epileptogenic zone in 12 patients; seven had no interictal sharp waves. The SEAs showed multifocal interictal sharp waves in all patients. Seven patients with localized seizure onset on scalp recording showed extensive ictal onset on the SEA recording. Five patients with lateralized seizure onset to one hemisphere on scalp recording were found to have ictal onset on SEA restricted to a smaller area. Because of the large epileptogenic zone found on SEA recordings, a complete resection was possible in only five (33%) of the 15 patients who had resections. Eight (53%) of the 15 patients benefited from surgery (mean follow-up, 4.6 years). The SEAs also allowed functional localization in most patients. From these data, we suggest that a localizing scalp electroencephalogram in patients with frontal lobe epilepsy may be misleading because SEA recordings show larger epileptogenic zones than anticipated. Furthermore, we postulate that the larger extensive epileptogenic zone may account for the poorer surgical outcome in patients with frontal lobe epilepsy compared with patients with temporal lobe epilepsy.

Comparison of the intracarotid amobarbital procedure and interictal cerebral 18-fluorodeoxyglucose positron emission tomography scans in refractory temporal lobe epilepsy

The relationship between interictal focal hypometabolism determined by 18-fluorodeoxyglucose positron emission tomography (FDG-PET) scans and memory function with the intracarotid amobarbital procedure (IAP) was evaluated in 23 patients with temporal lobe epilepsy. All patients underwent prolonged EEG/video monitoring. The epileptogenic focus was defined by interictal epileptiform discharges and ictal onsets. All 23 patients had recorded seizures arising exclusively from one temporal lobe. PET showed temporal lobe hypometabolism ipsilateral to the epileptogenic focus in 86% (20 of 23) of patients; IAP showed impaired memory of the hemisphere of seizure onset in 65% (15 of 23). Sixty-five percent (13 of 20) of patients with focal hypometabolism had ipsilateral memory impairment. Memory impairment contralateral to the hypometabolic zone was not observed. Ninety-five percent (22 of 23) of patients demonstrated functional impairment by either PET or IAP (or both) on the epileptogenic side.

Use of intracranial neurophysiologic recording techniques in the evaluation for epilepsy surgery in children

Resective surgery is an accepted treatment modality for medically intractable focal epilepsy in children as well as in adult. During the presurgical evaluation processes, intracranial neurophysiologic recording of epileptiform abnormalities have been used much more commonly in adults and older adolescents than in infants and children. However, as infants and children are increasingly referred for early surgery in many centres, it may be necessary to study complex cases in some children with invasive electrodes in order to plan a safe and effective resection. This article gives first an overview of the rationale and indications, with case illustrations, for using these techniques. This is followed by general discussions on individual electrodes and their use in infants and children.

Recording of event-related potentials (P300) from human cortex

Auditory event-related potentials were recorded simultaneously from chronically implanted subdural electrodes and from scalp electrodes in three patients who were being evaluated for surgical treatment of epilepsy. These three cases showed clearly defined scalp-P300 and scalp-N300. A cortex-P300 was recorded from the midtemporal area, and a cortex-N300 was recorded from the inferior frontal area with some reflection at the basal temporal region. There were no potentials from an interhemispheric region. We could not observe any component from the cortex studied corresponding to scalp-recorded N200. Therefore, while the activity generated from the mesial temporal lobe may only make minor contributions to scalp-P300, that generated from the midtemporal area might make a major contribution to the scalp-P300. Additionally, generators of N200, P300, and N300 are different from each other. These findings, together with previous reports regarding the generator source of P300, also suggest that P300 is a complex arising from multifactorial generator sources, including the midtemporal and inferior frontal area.

Significance of sharp waves in routine EEGs after epilepsy surgery

We retrospectively analyzed the presence of sharp waves in 2-h EEGs performed 6 months after epilepsy surgery in 59 patients. To study the significance of the postoperative interictal epileptiform activity in the tissue remaining after resection, we included only patients with a single epileptic focus (as defined preoperatively by prolonged video/EEG recordings and subdural electrode arrays studies) and no progressive structural lesions. Temporal lobectomy was performed in 51 patients (86%); extratemporal resections were performed in the remainder. The epileptogenic focus was completely resected in 26 patients (44%). The immediate postoperative electrocorticograms (EcoG) showed spikes in 13 patients (22%). At 6-month follow-up, 43 patients (73%) were seizure-free or had auras only and 12 patients (20%) had epileptiform activity on EEG. A significant correlation was noted between presence of sharp waves in the 6-month postoperative EEG and recurrence of seizures (Fisher's exact test p = 0.011) and also with the extent of the resection (complete vs. incomplete p = 0.042). We noted no correlation between postoperative epileptiform activity and location of the resection (temporal vs. extratemporal), presence of spikes in immediate postoperative EcoG, or occurrence of auras only at 6-month follow-up.

Etiologic factors for unitemporal vs bitemporal epileptiform discharges

We compared the etiologic factors and clinical characteristics of 30 patients with unitemporal vs those of 30 patients with bitemporal independent (minimum 20% from one side) interictal epileptiform discharges on extracranial electroencephalograms. Febrile seizures occurred significantly more frequently in the unitemporal (40%) than in the bitemporal (17%) group. Mass lesions were more common in the bitemporal group, and seven of 10 patients with mass lesions showed bitemporal interictal epileptiform discharges. There were no statistically significant differences in age at onset, frequency of seizures, duration of epilepsy, and history of central nervous system infection or trauma between the two groups. A history of febrile seizures or central nervous system infection that may be expected to cause diffuse cerebral injury does not appear to be the major factor predisposing to the development of bitemporal interictal epileptiform discharges.

Experimental limbic epilepsy: models, pathophysiologic concepts, and clinical relevance

Complex partial seizures originating in the temporal lobe are one of the most common types of seizures in patients with epilepsy. They are frequently intractable to medical treatment and are increasingly considered for surgical therapy. These seizures are often associated with focal epileptogenicity in limbic structures (amygdala and hippocampus) or with rapid spread of seizure activity to these areas. Much research is being undertaken to better understand this disorder and to develop more effective approaches to diagnosis and treatment. Experimental work in animals has contributed to the understanding of epileptogenesis, the interictal state, and the homeostatic mechanisms that limit seizure activity.

Intracarotid amobarbital procedure: I. Prediction of decreased modality-specific memory scores after temporal lobectomy

To assess predictive value of the intracarotid amobarbital procedure (IAP) for decreased postoperative modality-specific memory, we studied 37 temporal lobectomy patients with intractable partial epilepsy who were selected for operation independent of preoperative IAP findings. When ipsilateral IAP failure was defined by an absolute method as a retention score less than 67%, the results were not associated with decreased modality-specific memory after operation. When ipsilateral IAP failure was defined by a comparative method as a retention score at least 20% lower after ipsilateral than contralateral injection, the results showed greater differences between groups, but differences still did not achieve statistical significance. Four left-resection patients who failed the ipsilateral IAP had a median postoperative change in the Wechsler Memory Scale-Revised (WMS-R) Verbal Memory Index score of -14%, whereas 16 left-resection patients who passed the ipsilateral IAP had a mean postoperative change in the WMS-R Verbal Memory Index score of -7.5% (p = 0.12). These results suggested that the IAP interpreted comparatively may be a helpful adjunctive test in assessment of relative risk for modality-specific memory dysfunction after temporal lobectomy, but larger series of operated patients are needed to confirm this possibility. In this series, complete amnesia was not noted after ipsilateral injection, even in patients with postoperative modality-specific memory decline.