Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Monday, December 4, 2006 Poster Session IV 7:30 a.m. – 4:30 p.m.

Monday, December 4, 2006 Poster Session IV 7:30 a.m. – 4:30 p.m. 1 Véronique M. André, 1 Carlos Cepeda, 1,3 Vinters V. Harry, 1,2 Mathern W. Gary, and 1 Levine S. Michael ( 1 Mental Retardation Research Center, David Geffen School of Medicine at UCLA, Los Angeles, CA ; 2 Division of Neurosurgery ; and 3 Department of Neuropathology, UCLA, Los Angeles, CA ) Rationale: Based on histopathology, cortical dysplasias (CD) are classified into mild CD if the cortex shows architectural abnormalities only and severe CD if the cortex also presents cytomegalic neurons and balloon cells in addition to architectural changes. Both types induce intractable seizures in children. We showed that GABAergic cell numbers and GABA terminals were altered in severe CD but not in mild CD. GABA peak currents, densities and desensitization time constants were also differentially altered in mild and severe CD, suggesting that mechanisms of seizure induction might be different in those two types of pediatric CD (Andre et al., Epilepsia 46 Suppl. 8: 5, 2005). The present study characterized further postsynaptic GABAA receptor function in non‐CD, mild and severe CD human brain. Methods: Cortical samples resected for the treatment of pharmaco‐resistant epilepsy were collected from non‐CD (n = 10), mild (n = 6) and severe CD (n = 10) patients. Brain tissue slices were acutely dissociated and electrophysiological recordings were performed on isolated pyramidal neurons. Patch electrodes were filled with N‐methyl‐D‐glucamine for whole‐cell voltage clamp recordings. Different concentrations of GABA and GABAA receptor modulators were applied to the cells. Results: EC50 values were similar in non‐CD and severe CD but higher in mild CD cells suggesting less sensitive GABAA receptors. Zinc reversibly decreased GABA peak currents in all cells. However, severe CD cells had a smaller sensitivity to zinc compared to non‐CD cells. Zolpidem, a benzodiazepine (BZ) of type I that binds specifically to α1 subunits, reversibly enhanced GABA peak currents in all cells. Zolpidem sensitivity was significantly smaller in severe CD cells compared to mild and non‐CD cells. Conclusions: GABA sensitivity determined by EC50 values was similar in non‐CD and severe CD cells. However, GABAA receptors responded differently to modulators in severe CD vs non‐CD. Sensitivity to zinc was smaller in severe CD, suggesting higher expression of γ2 subunits. Zolpidem effect was also decreased in severe CD vs non‐CD indicating a decreased expression of α1. As α1 and γ2 subunits are part of the BZ binding site, alteration in their expression is likely to change the efficacy of BZ used to treat seizures in severe CD. In mild CD, effects of modulators binding to α1‐ and γ2‐containing receptors were not different from non‐CD indicating those subunits are not altered. However, GABA sensitivity was lower in mild CD, which could implicate a loss of inhibition and lead to seizures. Together, these results indicate that GABAA receptor subunit composition is differentially altered in mild and severe CD and suggest that different drugs targeting GABA function might be needed to treat the two types of CD. (Supported by NIH NS38992.) 1 Karin Boer, 1 Sandra Redeker, 2 Wim G.M. Spliet, 3 Peter C. Van Rijen, 1 Dirk Troost, and 1 Eleonora Aronica ( 1 Department of (Neuro) Pathology, Academic Medical Center, University of Amsterdam, Amsterdam, Amsterdam, Netherlands ; 2 Department of Pathology, University Medical Center Utrecht, Utrecht, Utrecht, Netherlands ; and 3 Department of Neurosurgery/Rudolf Magnus Institute for Neuroscience, University Medical Center Utrecht, Utrecht, Utrecht, Netherlands ) Rationale: Developmental glioneuronal lesions, such as gangliogliomas (GG) are increasingly recognized causes of chronic pharmaco‐resistant epilepsy. The cellular mechanism(s) underlying the epileptogenicity of GG remain largely unknown. It has been postulated that chronic epilepsy in patients with malformations of cortical development is associated with dysfunction of the inhibitory GABAergic system. We aimed to identify the subtypes of interneurons present within GG specimens and the expression and cellular distribution patterns of GABA receptors (GABAR) and GABA transporter 1 (GAT1). The expression of the various components of the GABAergic system were also analysed in the perilesional cortex. Methods: We investigated immunocytochemically the expression of parvalbumin, calbindin, calretinin, neuropeptide Y (NPY), GABAAR (a1 subunit), GABAB (R1 and R2) and GAT‐1 in 30 specimens of GG obtained during epilepsy surgery, including 10 cases with sufficient amount of both perilesional and adjacent normal control cortex. Results: Immunocytochemistry for parvalbumin, calbindin, calretinin, neuropeptide Y demonstrate the presence of inhibitory neurons of different subtypes within the GG specimens. Most interneurons were NPY positive, whereas fewer reacted to calretinin, calbindin and parvalbumin. Both GABAAR and GABABR (R1 and R2) subtypes were detected within the neuronal component of GG specimens. In addition, GABABR2 immunoreactivity (IR) was observed in glial cells. GG specimens displayed also expression of GAT‐1 IR. Compared to normal cortex, reduced perilesional IR for parvalbumin was observed in 70%, for calbindin in 60% and for GABAR and GAT‐1 in 30% of the specimens. Increased staining was detected for NPY in 80% of the cases. Conclusions: The cellular distribution of components of the GABAergic system in GG, together with the perilesional changes suggest that alterations of the GABAergic system may contribute to the complex abnormal functional network of these highly epileptogenic developmental lesions. (Supported by National Epilepsy Fund –“Power of the Small”/Hersenstichting Nederland (NEF 02–10; NEF 05–11, E. Aronica and K. Boer) and by the Epilepsy Institute of the Netherlands (Heemstede, The Netherlands; E. Aronica).) 1 Michael L. Bell, 2 Jeffery R. Buchhalter, 3 Eugene B. Morris, and 4 Joseph E. Parisi ( 1 Neurology, Mayo Clinic, Rochester, MN ; 2 Child and Adolescent Neurology, Mayo Clinic, Rochester ; 3 Division of Neuro‐Oncology, St. Jude Children's Hospital, Memphis, TN ; and 4 Pathology, Mayo Clinic, Rochester ) Rationale: Epilepsy surgery outcomes may differ among the focal malformations of cortical development (MCDs). Specifically, most pathologic studies have suggested that balloon cells portend a better prognosis compared to other “focal cortical dysplasias.” This could impact surgical decision making, since there are several MRI features highly suggestive of balloon cells. For example, focal cortical thickening with T2 signal abnormality extending towards the ventricle has been reported to be 94% specific for balloon cell pathology. We have previously proposed a categorization scheme for focal malformations of cortical development, discarding the term “focal cortical dysplasia” since these lesions are not precancerous. This study correlates the histopathologic type of focal malformation with epilepsy surgery outcomes. Outcome of each focal MCD category Pathologic subtype Engel class I outcome (free of disabling seizures) Architectural disorganization 1 of 6 (17%) Cytoarchitectural dysmorphism without balloon cells 9 of 12 (75%) Cytoarchitectural dysmorphism with balloon cells 12 of 20 (60%) Polymicrogyria 4 of 8 (50%) Indeterminate 0 of 1 (0%) p = 0.12 Methods: Pathologic specimens were retrospectively reviewed from 49 consecutive focal MCDs resections at Mayo Clinic Rochester. These cases were all removed for intractable epilepsy. Specimens were systematically classified using microscopy and standardized immunohistochemical techniques, according to our recently reported classification scheme. Clinical outcomes were compared between different histopathologic subtypes. Results: Of the 49 cases, three were excluded since there was no post‐operative follow‐up. Of the remaining 46, 26 patients (57%) remain free of disabling seizures (Engel class I). Average follow‐up was 34 months. Patients with cytoarchitectural dysmorphism (with or without balloon cells) trended towards better outcomes compared to architectural disorganization and polymicrogyria. However, balloon cells did not portend the best prognosis. Only 60% of patients with balloon cells (“cytoarchitectural dysmorphism with balloon cells”) were free of disabling seizures compared to 75% of those with cytologically abnormal neurons and architectural disorganization without balloon cells (“cytoarchitectural dysmorphism without balloon cells”). Conclusions: Epilepsy surgical outcomes remain suboptimal in focal MCDs. Other focal lesions causing epilepsy, such as cavernomas or mesial temporal sclerosis, are associated with excellent outcomes in up to 90% of patients. In contrast, we found that only 57% of patients with focal MCDs remained seizure‐free. In our cohort, balloon cells did not portend a better prognosis than other focal MCDs. 1 Kristina A. Fenoglio, 1 Timothy A. Simeone, 2 Frank Schottler, 1 Do Young Kim, 1 Harold Rekate, 1 Jack Kerrigan, and 1 Jong M. Rho ( 1 Barrow Neurological Institute, St Joseph's Hospital & Medical Center, Phoenix, AZ ; and 2 Alpha Med Sciences Co., LTD, Osaka, Japan ) Rationale: The hypothalamic hamartoma (HH) represents a rare but important model of subcortical epileptogenesis. Clinical studies, based primarily on intracranial electrode recordings have established that the HH itself is epileptogenic, but the mechanisms are unknown. In the present study, we examined the network properties of surgically‐resected human HH tissue using a planar multielectrode array recording system. Methods: HH tissue was obtained with patient consent and upon resection was immediately submerged in ACSF bubbled with 95% O2/5% CO2. Tissue slices (400 μm) were placed in a microelectrode dish (Alpha Med Systems, Osaka, Japan) and perfused with warmed (35°C) oxygenated ACSF. The electrodes were arranged in an 8 × 8 grid with 150 μm separation. Slices were arbitrarily placed over the entire electrode grid. Results: Paired‐pulse stimulation (50 ms interval) to random points within the tissue elicited small negative field potentials and stimulation‐evoked single unit firing within specific regions of the tissue. A 5–20% paired‐pulse depression of the field potential was evident in most areas, and was converted to a similar magnitude of paired‐pulse facilitation by picrotoxin (100 μM) suggesting the presence of both excitatory and inhibitory synaptic components. Furthermore, the stimulation‐evoked single unit firing increased (∼4‐fold) with the second pulse consistent with paired‐pulse facilitation of an excitatory component. Picrotoxin substantially increased evoked unit firing suggesting that endogenous GABAergic neuron activity is present normally, and inhibition or shunting of an excitatory component is prominent. In addition, we recorded spontaneous single unit firing of individual cells from multiple electrodes with frequencies ranging from 1–14 Hz and spontaneous negative slow wave activity with amplitudes ranging from 5–300 mV. Interestingly, the frequency of spontaneous slow wave activity decreased with application of picrotoxin and increased with muscimol (30 μM). Conclusions: Our observations indicate that HH tissue is composed of neurons that are strongly modulated by GABAA receptor‐mediated mechanisms. This is consistent with earlier observations demonstrating positive immunoreactivity of HH tissue to glutamic acid decarboxylase (GAD65/67). A novel finding in this study is that populations of neurons also give rise to spontaneous slow wave discharges. It is of interest to determine if the spontaneous activity will become synchronous across the tissue and develop into epileptiform activity in response to provocation. Future studies will characterize the spontaneous activity and the functional architecture of HH tissue. (Supported by NIH and the Barrow Neurological Foundation.) 1 Jan A. Gorter, 2 Karin Boer, 1 Erwin Van Vliet, 2 Sandra Redeker, 3 Fernando H. Lopes da Silva, 3 Wytse J. Wadman, and 2 Eleonora Aronica ( 1 Swammerdam Institute for Life Sciences/Stichting Epilepsie Instellingen Nederland, University of Amsterdam, Amsterdam, Amsterdam, Netherlands ; 2 Department of (Neuro)Pathology, Academic Medical Center, University of Amsterdam, Amsterdam, Amsterdam, Netherlands ; and 3 Swammerdam Institute for Life Sciences, University of Amsterdam, Amsterdam, Amsterdam, Netherlands ) Rationale: The complement cascade has been suggested to be involved in the inflammatory responses observed in a variety of disorders of the central nervous system. Recent evidence suggests that complement activation may contribute to seizure activity and promote neurodegeneration. We investigated the possible involvement of the complement cascade during epileptogenesis in a rat model of temporal lobe epilepsy (TLE) and in the chronic phase of spontaneous seizures in both experimental as well as human TLE. Methods: RNA was obtained from three different brain regions (CA3, entorhinal cortex and cerebellum) at three different time points (1 day = acute phase; and 1 week = latent phase; 3–4 months = chronic epileptic phase) after electrically induced status epilepticus (SE) in rat. Gene expression analysis was performed using the Affymetrix Gene Chip System (230A). The immunohistochemical pattern of C1q, C3c and C3d was assessed in human hippocampal specimens from controls and medically intractable TLE with or without hippocampal sclerosis (HS). Results: Genes involved in the complement pathway (e.g. C1qa, C1qc, C3, C4a, Cfh) were upregulated after induction of SE in rat and remained elevated in the chronic phase. C1qa and C1qc were already induced in the acute phase. C3, C4a and C8b were induced in the latent phase and showed the highest expression at the end of the latent period. In human TLE with HS, glial expression of C1q, C3c and C3d was observed particularly within regions where neuronal cell loss occurs (e.g. CA1 and hilus). Conclusions: Our data indicate that within the immune response the complement activation represents a prominent process during epileptogenesis. The persistence of complement activation is supported by the data on human HS specimen showing increased expression of complement factors in activated glial cells. Although the complement system may be useful in eliminating aggregated and harmful proteins associated with epilepsy, abnormal activation of the complement system can also have damaging effects through activation of inflammatory processes. (Supported by National Epilepsy Fund –“Power of the Small”/Hersenstichting Nederland (NEF 02–10; NEF 05–11, E. Aronica and K. Boer; 03–03 J.A. Gorter).) 1 Stephen T. Hantus, 1 Dileep Nair, 1 Richard Burgess, and 1 Imad Najm ( 1 Neurology, Cleveland Clinic, Cleveland, OH ) Rationale: One third of epilepsy patients are refractory to pharmacologic treatment. Malformations of cortical development (MCD) are a well‐recognized cause of refractory epilepsy that can often be cured by surgical resection. The presence of periventricular heterotopia (PVH) is associated with the development of epilepsy, although the role of these lesions in the generation of epileptogenicity and their cellular/molecular characteristics remain unclear. Methods: In this report, we present the histological and immunocytochemical (ICC) characteristics of a periventricular nodule that was resected from a 17 years old female with a history of pharmacoresistant epilepsy, a lateral temporal MCD and multiple PVH. Invasive video electrographic encephalography (VEEG) with subdural electrodes and a depth electrode targeted at the PVH did not confirm the PVH as the seizure onset zone. The patient underwent a temporal lobectomy that included the removal of a subcortical nodule. She remained seizure free on no antiepileptics for four years, then presented with a single generalized seizure. Results: Histological and ICC examinations of the resected nodule showed dysmorphic neurons lacking any laminar or columnar organizations. These neurons were intermixed with GFAP positive glial cells, immature neurons (TUJ1 positive) and immature astrocytes (vimentin positive). Some subcortical neurons expressed doublecortin, a marker of migrating cells. The expression of GABA (Calbindin, parvalbumin) and NMDA (NR1, NR2A, NR2B) receptors were also examined. Conclusions: Our results suggest that PVH is constituted of a mixture of neuronal and glial cells at various levels of differentiation that failed to mature and migrate to their final position in the neocortex. It is unclear if these nodules are epileptogenic. Further studies of the role of PVH in epilepsy are needed in order to further characterize the electrophysiology and molecular biology of the constituent cells of these lesions and their overlying neocortical areas. 1 David C. Henshall, 2 Clara K. Schindler, 2 Akitaka Yamamoto, 1 Carmen Bellver‐Estelles, 1 Brona M. Murphy, 1 Seiji Hatazaki, 2 Norman K. So, 2 Roger P. Simon, and 2 Robert Meller ( 1 Physiology & Medical Physics, Royal College of Surgeons In Ireland, Dublin, Ireland ; 2 RS Dow Neurobiology Laboratories, Legacy Research, Portland, OR ; and 3 Oregon Comprehensive Epilepsy Program, Neurological Sciences Center, Portland, OR ) Rationale: The role of tumor necrosis factor receptor 1 (TNFR1) signaling in temporal lobe epilepsy (TLE) is incompletely understood. Experimental modeling suggests seizures activate TNFR1 and downstream pro‐ and anti‐apoptotic signaling cascades, although little is known about whether these pathways function in human TLE brain. Methods: We examined TNFR1 signaling in hippocampus surgically obtained from refractory TLE patients (n = 10), and compared results to matched autopsy controls (n = 6). TNFR1 signaling was also examined in hippocampus obtained from mice following focally‐evoked limbic seizures. Samples were processed for detection of protein in whole cell lysates or subcellular fractions, immunoprecipitation studies and immunohistochemistry. Results: Western blotting established total protein levels of the TNFR1 proximal signaling adaptor TNFR‐associated protein with death domain (TRADD) and cleaved caspase‐8 were higher in TLE samples than matched autopsy controls. Intracellular distribution analyses determined cytoplasmic levels of TNFR1, TRADD and the caspase‐8 recruitment adaptor Fas‐associated protein with death domain (FADD), and microsomal levels of TRADD, FADD and cleaved caspase‐8, were raised in TLE samples. Immunoprecipitation studies, supported by fluorescence microscopy revealed increased TNFR1‐TRADD and TRADD‐FADD association in TLE samples. Expression and localization of these proteins was similar in mouse hippocampus. Seizures evoked by intraamygdala kainic acid increased microsomal levels of TNFR1 signaling components and triggered TNFR1‐TRADD and TRADD‐FADD complex formation and caspase‐8 processing. Conclusions: These data show seizures in mice activate hippocampal TNFR1 signaling and reveal this pathway to be engaged in patients with temporal lobe epilepsy. Accordingly, the TNFR1 pathway may be a potential target for adjunctive neuroprotective therapy following status epilepticus or in the setting of recurrent epileptic seizures. (Supported by NIH/NINDS grants NS39016 and NS41935, Science Foundation Ireland, Health Research Board, Marie Curie Foundation and the Program for Human Genomics, PRTLI & Higher Education Authority, Ireland.) 1 Ludmyla Kandratavicius, 1 Jaime Eduardo Hallak, and 1 Joao Pereira Leite ( 1 Neurology, FMRP‐USP, Ribeirao Preto, Sao Paulo, Brazil ) Rationale: Heat shock protein 70 (HSP70) is a stress inducible protein. In animal models of epilepsy, increased HSP70 expression is proportional to seizure severity. HSP70 neuroprotective mechanisms are not yet elucidated, but HSP70 can attenuate intracellular calcium influx and diminish seizure‐induced apoptosis. HSP90 has received little attention in epilepsy research. HSP90 interacts with calcium and calmodulin, and regulates nitric oxide synthase, proteasome, Tau, actine and tubulin, among others. There are few studies about the role of HSP70 in epilepsy and so far nothing regarding to HSP90. Due to the probable HSP70 protective role and the HSP90 relation with key‐factor proteins in epileptogenesis, we looked at HSP70 and HSP90 immunohistochemical expression in temporal lobe epilepsy (TLE). Methods: 47 hippocampi were obtained from medically intractable TLE patients: 16 TLE, 18 TLE+depression and 13 TLE+psychosis. 6 control hippocampi were from necropsy cases without past history of neurological disorders, with pos‐mortem delay no longer than 4 hours. Specimens were equally treated and submitted to imunohistochemistry to HSP70 and HSP90. Hippocampal regions examined were fascia dentata, hilum, CA4, CA3, CA2, CA1, prosubiculum, subiculum, parahippocampal gyrus and entorhinal cortex. Positive immunoreactivity (puncta) was estimated using the software ImageJ. Results: Analyzes of variancy (ANOVA) of our results showed significant differences. Dunn's post‐hoc tests (p < 0.05) of our ANOVA showed significant lower expression of HSP70 and HSP90 in epileptic patients when compared to controls in all hippocampal regions, except fascia dentata and subiculum. Regarding HSP70, the same occurred in all subicular subfields. As to HSP90, there are also differences between TLE+depression and TLE+psychosis at hilum, prosubiculum and hippocampal gyrus. TLE+psychosis always showed values closed to controls. Conclusions: Unlike the reports in animal models the present results indicate that chronic seizures in TLE patients are not sufficient to induce HSP70 and HSP90 activation. Neuronal loss, gliosis and synaptic reorganization, which are typical attributes of TLE, may be determinants of low HSP expression. Inherent to our findings and conclusions, it is important to notice methodological differences between experimental and control groups. Although an effort was made to include only necropsy patients with short post mortem delay, the cause of patients death could be, at least partially, responsible to some degree of HSP activation. Nevertheless, the neurological complications on this group led us to suggest that low expression of HSPs in epileptic groups is related to seizure maintenance. We found no explanation for the differential HSP90 expression in TLE patients with psychosis. (Supported by Fapesp (CInAPCe project # 05/56447–7), CNPq and FAEPA.) 1 Orfa Y. Galvis‐Alonso, 1,2 Jose E. Peixoto‐Santos, 3 David Jr. Araujo, 3 Renata Caldo‐Standiuzzi, and 3 Joao P. Leite ( 1 Molecular Biology, Medicine School of Sao Jose do Rio Preto (FAMERP), Sao Jose do Rio Preto, Sao Paulo, Brazil ; 2 Biology, University of Sao Paulo State (UNESP), Sao Jose do Rio Preto, Sao Paulo, Brazil ; and 3 Neurology, University of Sao Paulo at Ribeirao Preto (USP), Ribeirao Preto, Sao Paulo, Brazil ) Rationale: Patients with temporal lobe epilepsy (TLE) usually present hippocampal sclerosis that can be detected by magnetic resonance imaging (MRI). Most of the resected hippocampi of patients with TLE resistant to pharmacological treatment are characterized by neuronal loss and abnormal mossy fiber sprouting in the inner molecular layer of the fascia dentata. The main goal of this work was to study neuron density and mossy fiber distribution in the hippocampi of TLE patients with different MRI volumetric hippocampal profiles. Methods: Using MRI hippocampal volumetry, patients were classified as: 1) normal volume (NV; n = 8; hippocampal volume with at least 2.5 cm3); 2) mild atrophy (MA; n = 9; hippocampal volume between 2 and 2.5 cm3); and 3) severe atrophy (SA; n = 10; hippocampal volume with less than 2 cm3). Corresponding hippocampal histological sections were stained with cresylecht violet or hematoxylin/eosin for hippocampal cell counts and neo‐Timm histochemistry for mossy fiber detection. Histological controls were autopsy patients (n = 13; without antemortem clinical seizures or evidence of brain pathology on postmortem examination). Results: 1) Compared to autopsies, all three TLE groups showed loss of granular cells (NV: 44%; MA: 43%; SA: 48%) and CA1 (NV: 83%; MA: 69%; SA: 72%), CA2 (NV: 39%; MA: 39%; SA: 47%); CA3a (NV: 64%; MA: 70%; SA: 62%) and CA4 (NV: 55%; MA: 63%; SA: 64%) pyramidal cells, p < 0.001; 2) all TLE groups presented higher neo‐Timm staining in the inner molecular layer of the fascia dentata than their autopsy controls (gray level: Control: 7,6 ± 4; NV: 72,6 ± 8; MA: 90,8 ± 10,8; SA: 95,5 ± 8; p < 0.001). Conclusions: Atrophy profile observed on MRI does not necessarily reflect either cell loss intensity or axonal reorganization in patients with TLE. (Supported by BAP‐FAMERP; FAPESP – CInAPCe Project # 05/56447–7; CNPq.) 1 Mireille Lerner‐Natoli, 1,2 Valerie Rigau, 1 Melanie Morin, 1,3 Arielle Crespel, 1 Marie‐Claude Rousset, 3 Philippe Coubes, 1,3 Michel Baldy‐Moulinier, and 1 Joel Bockaert ( 1 Neurobiology, Institute of Functional Genomics, CNRS, INSERM, University of Montpellier, Montpellier, France ; 2 Unit of Pathological Anatomy and Cytology, CHU Gui de Chauliac, Montpellier, France ; and 3 Epilepsy Unit, CHU Gui de Chauliac, Montpellier, France ) Rationale: In adult patients with intractable temporal lobe epilepsy (TLE), we previously described inflammatory processes and neurogenesis; unexpectedly, we observed the presence of immature micro‐vessels. The present study was designed to demonstrate: i) a pathological vascularization associated with TLE, ii) angiogenic processes active in chronic focus, iii) an impact of vascular remodeling on epileptogenicity. Methods: Hippocampi were collected from 40 patients after surgery for intractable TLE with various etiologies and from 5 non epileptic (NE) patients (autopsies or hippocampectomy for parahippocampal tumor). By immunohistochemistry or immunoblot, we evaluated: microvessel density, expression of markers of immature endothelial cells, of angiogenic factors and their receptors. We also checked impairments of blood brain barrier (BBB). In parallel, we studied the kinetics of these processes in the rat model of lithium‐pilocarpine‐induced limbic epilepsy. Results: We observed division markers in vascular cells in neurogenic areas of epileptic hippocampi. The microvessel density was significantly higher in TLE patients than in NE patients. More, there was no quantitative differences between TLE with hippocampal sclerosis and TLE with other etiologies. Most of these microvessels were positive for markers of immature endothelial cells. The angiogenic factor VEGF was expressed mainly by neurons and often by astrocytes, and its receptor Flk‐1 was localized on thin vascular sprouts. In all cases of TLE, we observed impairments of BBB, allowing IgG leakage and blood cell extravasation. In rats, the degree of vascularization increased progressively from 1 week after status epilepticus. VEGF was rapidly and strongly overexpressed after seizures, whereas the other markers appeared during the silent period. The BBB was early damaged. In the chronic stage, these phenomena were still obvious, as for human TLE. Conclusions: For the first time, we demonstrate a neo‐vascularization of the TLE focus, independently of etiology. We suggest that the pathological angiogenic processes are related to the persistent inflammation and are maintained by recurrent seizures. The consequences of this vascular remodeling may concern hemodynamics, inflammation, neurotoxicity, BBB permeability and drug refractoriness. (figure 1) (Supported by French Fondation for Research on Epilepsy.) 1 Michael Majores, 1 Susanne Schoch, 2 Matthias Simon, 1 Gudrun Engels, and 1 Albert J. Becker ( 1 Department of Neuropathology, University of Bonn Medical Center, Bonn, Germany ; and 2 Department of Neurosurgery, University of Bonn Medical Center, Bonn ) Rationale: Focal cortical dysplasias (FCD) are characterized by a localized malformation of the neocortex and the underlying white matter. Balloon cells, similar to those observed in tuberous sclerosis (TSC) are present in a significant number of cases (FCDIIb). Histopathological similarities indicate that FCDIIb may be pathogenetically related to the tuberous sclerosis complex (TSC), caused by mutations in either TSC1 or TSC2, that normally constitute a tumor suppressor mechanism. Accumulation of coding allelic variants affecting exons 5 and 17 of TSC1 as well as loss of heterozygosity of the second TSC1 allele have been observed in many FCDIIb patients (Becker et al., Ann Neurol 2002). Here, we have studied the potential functional relevance of coding allelic variants of TSC1 found in FCDIIb with respect to the interaction of the TSC1/TSC2 complex. Methods: Sequence variants of the TSC1 allele found in FCDIIb patients were introduced in the full‐length TSC1 cDNA (kindly provided by Dr. M. Nellist, Amsterdam) by site‐directed mutagenesis. We have further generated TSC1 cDNA variants with mutations found in TSC patients, i.e. by introducing premature stop codons. Protein interaction assays were performed to analyze a potentially compromised interaction between the TSC1/2 proteins hamartin and tuberin. Results: TSC1 sequence variants were successfully introduced by site‐directed mutagenesis. Compromised interaction of the hamartin‐tuberin complex was most pronounced in TSC1 mutants bearing the premature stop codon at position R692X (p < 0.01). Sequence alterations of TSC1 present in FCDIIb (H732Y) also resulted in significantly compromised TSC1/2‐interaction compared to wild‐type TSC1 (p < 0.05). The latter was comparable to findings obtained by the expression of the TSC‐determining allele variant R786X. Conclusions: Our data suggest, that allelic variants with amino acid exchange of the TSC1 gene in FCDIIb may be associated with compromised TSC1/TSC2 interaction. The functional relevance is currently studied in cell culture and animal systems. (Supported by DFG (SFB TR3, Emmy‐Noether‐Program), BONFOR, the BMBF German Israeli program and the Deutsche Krebshilfe.) 1 Shailaja Srinath, 2 Richard Prayson, 3 William Bingaman, and 1 Prakash Kotagal ( 1 Department of Neurology, Cleveland Clinic Foundation, Cleveland, OH ; 2 Department of Neuropathology, Cleveland Clinic Founation, Cleveland, OH ; and 3 Department of Neurosurgery, Cleveland Clinic Foundation, Cleveland, OH ) Rationale: The cause of seizures in patients with arachnoid cyst (AC) is thought to be secondary to the compression of adjacent cortex, associated cortical dysplasia (CD) or subpial gliosis according to anecdotal reports in the literature. The aim of our study is to systematically review the pathology around the cyst in patients with epilepsy treated with AC and cortical resection. Methods: We retrospectively reviewed patient data from epilepsy surgeries at the Cleveland Clinic Foundation between 1978 and 2006. Only those patients treated with resection of cyst and adjacent cortex were included in the study. Patients' records were reviewed for demographic data, follow‐up duration, EEG, MRI or CT imaging, and histopathology reports. Results: Of 1880 patients, only 10 patients had AC and six were treated with cyst resection. There were five males and one female patient. The locations of the AC were in the left temporal fossa in four, left peri‐rolandic region and right sylvian fissure in one each. Ictal onset zone was localized to the AC region in all patients. The mean duration of epilepsy was 12.2 + 8.8 years and the mean cyst size was 5.7 × 4.2 × 4 cms. Review of pathology showed mild subpial gliosis around the AC in all patients. Four of these showed normal adjacent cortex. However, one patient showed mild cortical dysplasia and another showed hippocampal sclerosis. At follow‐up ranging from 9 ± 0.24 months, all 6 patients were seizure free (one patient had 3 seizures at 2 months and none thereafter). Patients with abnormal cortical pathology were found to have earlier onset, longer duration, and lower frequency of seizures. Conclusions: A cortical abnormality either independently or in association with subpial gliosis appears to be the cause of seizures in patients with AC. Identifying the cortical pathology preoperatively with the help of high‐resolution MRI and a dedicated epilepsy imaging protocol could influence the type and extent of resection required. It is important to have proper orientation of the cyst and adjacent cortex in order to verify the pathological substrate for the seizures. 1 Mary T. Silvia, 3 Alexander R. Judkins, 2 Phillip B. Storm, 1 Robert G. Kalb, and 1 Brenda E. Porter ( 1 Department of Pediatrics and Neurology, Children's Hospital of Philadelphia, Philadelphia, PA ; 2 Department of Neurosurgery ; and 3 Department of Neuropathology ) Rationale: Cortical dysplasia (CD) is the cause of intractable pediatric epilepsy in up to 60% of surgical case series.1 Previously identified histologic features of CD include large dysmorphic neurons and prominent disorganized neurites. We hypothesize that there are differences in dendrite morphology (total dendrite length and branching patterns) between CD tissue and controls. To better understand the cellular morphology of CD, we quantitatively characterized the dendritic tree of dysmorphic neurons from Palmini grade II A and II B dysplasia from children with refractory epilepsy.2 Methods: Tissue from pediatric CD patients was collected at the time of epilepsy surgery and stained using the Golgi‐cox method. We have analyzed dendrite length and branching order from 3 refractory patients. Dendrites attached to the cell body are first order dendrites and those arising from the first order are called second order dendrites, and so on. Branching order is a measure of the neuron's dendrite tree complexity. Dendrite length refers to the summed total of the basilar dendritic tree. Slides were analyzed for total dendrite length, and dendrite order for at least 10 neurons per patient and compared against literature autopsy controls.2,3 Results: Patients 1, 2, and 3 were 11% (461μm ± 74.7 SEM), 30.3% (1271 μm ± 182.3 SEM) and 17.0% (714.8 μm ± 121.5 SEM) of the total dendrite tree length compared to literature controls (4193.4 μm ± 250 SEM). The most complex of our patients neurons branched to the 5th order, compared to the 8th order in controls. The largest difference in branching order occurred in 3rd and 4th branches, with Patients 1, 2, and 3 at 8%, 22%, and 2.7% of the control values, respectively, for number of dendrites per cell that branched to a 4th order. Conclusions: Pediatric CD tissue varied significantly from literature autopsy controls', with shorter total dendrite length per cell and decreased branching, indicating that CD dendrite trees are smaller and simpler in structure. Our findings suggest that the large and disorganized appearance of neurites in type II CD is associated with lack of dendrite organization rather than dysmorphic neurons with enlarged dendrite trees. Our data also suggest the dysmorphic neurons are similar to immature neurons with less complex dendrite trees. We are analyzing autopsy specimens from multiple ages to address this issue. References 1. Neurology. 2003 Aug 12; 61(3)365–8. 2. Neurology. 2004 Mar 23; 62(6 Suppl 3):S2 3. Cereb Cortex. 2001 Jun; 11 (6):558–71. (Supported by: This work was funded by grants from the Epilepsy Foundation of America for Silvia MT and by the CURE Foundation for Porter BE.) 1 Hidenori Sugano, 1 Madoka Nakajima, 1 Akihide Kondo, 1 Ikuko Ogino, 1 Hajime Arai, and 2 Tatsunori Seki ( 1 Neurosurgery, Juntendo University, Hongo Bunkyo‐ku, Tokyo, Japan ; and 2 Anatomy, Juntendo University, Hongo Bunkyo‐ku, Tokyo, Japan ) Rationale: It is known that neurogenesis still continues in the dentate gyrus even after reaching adulthood. Several reports have described the neurogenesis from surgical specimens of temporal lobe epilepsy, not only in an animal epilepsy model. However, the function of neurogenesis remains to elucidate. We herein examined the neurogenesis in the human hippocampus from temporal lobe epilepsy patients by immunohistochemical staining, and compared the neurogenesis between the specimens from hippocampal sclerosis and non‐hippocampal sclerosis. Methods: Six surgical specimens were examined by immunohistochemical staining using NeuroD, PSA‐NCAM, GFAP, Ki‐67. Three cases thus demonstrated hippocampal sclerosis in both magnet resonance imaging and pathological evaluations, and another 3 showed non‐hippocampal sclerosis. The cell counting of positive cells of each type of staining was applied in the ammon horn and the dentate gyrus. We also observed the nature of the cell and axon with positive staining. Results: PSA‐NCAM positive cells were mostly detected in the hilus, and also in the granule cell and CA1 pyramidal cell layers in both the hippocampal sclerosis and non‐hippocampal sclerosis groups. The PSA‐NCAM positive immature neurons were co‐stained with NeuN, but not with either NeuroD or Ki‐67. The PSA‐NCAM positive immature neurons were frequently detected in the non‐hippocampal sclerosis group. Conclusions: The PSA‐NCAM and NeuN co‐stained immature neurons were located in the hilus of human temporal lobe epilepsy. However, the ability of neurogenesis was found to be higher in the non‐hippocampal sclerosis group than in the hippocampal sclerosis group. Figure 1 (figure 1) (Supported by Juntendo University.) 1 Kiyotaka Suwa, 2 Thomas L. Beaumont, 2 Bin Yao, 3 Eishi Asano, 1 Aashit Shah, 4 William J. Kupsky, and 1,2 Jeffrey A. Loeb ( 1 Department of Nuerology, Wayne State University School of Medicine, Detroit, MI ; 2 Center for Molecular Medicine and Genetics ; 3 Department of Pediatrics ; and 4 Department of Pathology ) Rationale: Cortical malformations are one of the most common causes of medically intractable epilepsy. One of the more common malformations is polymicrogyria (PMG), where up to 80% of patients have epilepsy. The underlying mechanisms responsible for this high incidence of epilepsy are not known, nor is it known if epileptic regions of neocortex differ between patients with PMG and patients with non‐lesional neocortical epilepsy. Recently, we used a functional genomic strategy to identify a small group of activity‐depdnent genes are induced in neocortical epileptic foci, regardless of the underlying cause. Here, we hypothesize that both common and as well as unique patterns of gene expression can be identified in children with PMG compared to patients with non‐lesional neocortical epilepsy. Localization of these induced genes can identify neuronal populations involved human epilepsy associated with malformations. Methods: Electrically mapped neocortical tissue was obtained from 8 children with intractable epilepsy (4 with PMG, and 4 with gliosis) who underwent a two‐stage surgical resection. The seizure onset region and a nearby area without spontaneous epileptiform activity as a “control” were compared by histology, real‐time qPCR, and Agilent genome wide microarrays. Genes identified as differentially expressed were characterized further by in situ hybridization to identify populations of activated neurons. Results: Despite a marked difference in tissue histology between PMG patients and those with gliosis, there was a remarkably similar pattern of gene induction at regions of seizure onset and maximal interictal spiking. Patients without malformations expressed these activity‐dependent genes predominantly in normal‐appearing layers II‐IV pyramidal neurons. In patients with PMG, these same genes were also restricted to layer II‐IV neurons, in highly epileptic regions directly adjacent to thickened layer I regions in the polymicrogyri. In addition to common gene expression changes, we also identified genes uniquely altered in PMG neocortex that could underlie some of the morphollgical features of this malformation. Conclusions: Results genomic studies of PMG suggest that similar molecular pathways are activated in PMG as in other non‐lesional forms of neocortical epilepsy, however, the expression of some of these activity‐dependent genes in PMG localize to neurons directly adjacent to malformations. In addition, there are unique patterns of gene expression in PMG that are likely to relate directly to the malformation. (Supported by: 1. NIH NINDS R01 NS45207.; 2. Postdoctoral fellowships from the Center for Molecular Medicine and Genetics, Wayne State University School of Medicine, Detroit, MI, USA; 3.Japan research foundation for clinical pharmacology, Tokyo, Japan.) 1 Delia M. Talos, 2 Joseph R. Madsen, 2 Peter M. Black, and 1 Frances E. Jensen ( 1 Neurology, Children's Hospital/Harvard Medical School, Boston, MA ; and 2 Neurosurgery, Children's Hospital/Harvard Medical School, Boston, MA ) Rationale: The most common neurological symptom of human neocortical dysplasias is early‐life epilepsy, including infantile spasms. Frequently, these patients develop pharmaco‐resistant seizure disorders, which require surgical excision. The exact mechanisms of epileptogenesis and high incidence of pharmaco‐resistance to conventional antiepileptic drugs (AED) in cortical dysplasias are poorly understood. GABA receptors (GABARs), which are normally hyperpolarizing, can become depolarizing under certain circumstances and actually increase neuronal excitation. This depends on the intracellular chloride concentration, regulated by the chloride‐importer NKCC1 and the chloride‐exporter KCC2. Increased NKCC1 and decreased KCC2 expression in hippocampal subiculum from temporal lobe epilepsy (TLE) patients renders GABARs depolarizing (PNAS 103:8465;2006). We hypothesized that similar to TLE, neurons from neocortical dysplastic tissue would demonstrate higher NKCC1 expression and lower KCC2 levels relative to control neurons, a profile consistent with the presence of depolarizing GABARs. Methods: Cortical tissue from 6 patients ages 1–6 years were obtained during brain surgery for drug‐resistant epilepsy. In all cases, malformations of cortical development were confirmed by histopathological examination (tuberous sclerosis, n = 3; focal cortical dysplasia, n = 2; hemimegalencephaly, n = 1). Age‐matched autopsy cortical samples from cases with normal neurologic history were used as controls (n = 4). Tissue was collected and handled in accordance with the Clinical Research Committee at Children's Hospital, Boston. 4% paraformaldehyde‐fixed tissue blocks were cut at 50 microns and immunohistochemically double labeled with cell markers (NeuN, non‐phosphorylated neurofilament, vimentin, GFAP) in combination with transporter antibodies NKCC1 (1:100, Chemicon) and KCC2 (1:500, Upstate). Results: Staining of dysplastic tissue with neuronal and glial markers demonstrated loss of cortical lamination, the presence of disoriented, misshaped neurons, an increase in number of white matter neurons and a variable degree of gliosis. In addition, in cortical tubers and hemimegalencephaly tissue, cytomegalic dysplastic neurons could be easily identified. In all cases, increased NKCC1 expression associated with a moderate decrease in KCC2 levels was observed in most normal‐size and cytomegalic dysplastic neurons, compared to controls. Notably, strong NKCC1 immunoreactivity was observed in glial cell population as well. Conclusions: Our results demonstrate an imbalance in neuronal NKCC1 and KCC2 expression (increased NKCC1: KCC2 ratio) in human neocortical dysplasias that may render GABARs depolarizing. This mechanism may play a critical role in increased seizure susceptibility and relative refractoriness to GABAR agonists in these patients. (Supported by TS Alliance, Boston Neurosurgical Fdn, NIH NS31718.) 1 Cristiane Q. Tilelli, 2 Jean A. Tkach, 2 Yuguo Li, 1 Roger Oghlakian, 2 Mark Pagel, and 1 Imad M. Najm ( 1 Department of Neurology, Section of Adult Epilepsy, Cleveland Clinic, Cleveland, OH ; and 2 Department of Biomedical Engineering, Case Western Reserve University, Cleveland, OH ) Rationale: Magnetic Resonance Imaging (MRI) is a non‐invasive method often used for the assessment of patients with epilepsy. More than a quarter of patients who present with drug‐resistant epilepsy due to cortical dysplasia (CD) show no MRI abnormalities. The role of differential expression of glutamate receptors as a substrate for hyperexcitability in epilepsy has been shown. We intended to develop a non‐invasive and sensitive imaging technique to be used with specific biomarkers for epileptic tissue. Methods: Anti‐NMDA receptor subunit 1 (NR1) antibody (Chemicon AB1516) was labeled with Gd‐DOTA (NR1/Gd‐DOTA) using an activated carboxyl to create an amide coupling. In a first study, ex vivo human brain slice tissue was maintained alive immediately following surgical resection, and optimized T1‐weighted MRI was acquired using a 9.4T MRI scanner. Following baseline MRI, each tissue was treated with NR1/Gd‐DOTA or NR1 without Gd‐DOTA for two hours and washed for one hour to remove non‐specific binding. In a second study, MRI (7.0T) was acquired before and after ex vivo live rat brain slices were treated with NR1/Gd‐DOTA or control treatments (vehicle only, Gd‐DOTA only or anti‐NR1 only). Additional 3‐hour washing followed by MRI was conducted in order to verify signal variation over time. Results: The first study (Figure) demonstrated that optimized high‐field MRI and antibody‐based MRI probes can be used to non‐invasively detect cell receptors within freshly resected brain tissue. The second study validated that MRI contrast was due to specific binding of the labeled antibody to cell receptors. Fading in the signal detected after additional washing revealed a temporal dependence on the amount of antibody probe bound to tissue slices. Conclusions: Our preliminary data show that NR1 protein can be detected in freshly harvested human and rat brain tissue samples using high field MRI techniques. The development and application of new antibody‐based MRI probes that can be selectively detected and quantified will provide a new platform technology for visualizing molecular substrates of epileptogenic brain areas and other pathologies. (figure 1) (Supported by Northeastern Ohio Animal Imaging Resource Center, an NIH funded program #R24CA110943, part of the Case Center for Imaging Research; and R21 NS42354 from NINDS to IN.) 1 W. Saskia Van Der Hel, 2 Peter C. Van Rijen, 2 Cees W.M. Van Veelen, and 1 Pierre N.E. De Graan ( 1 Pharmacology and Anatomy, University Medical Center Utrecht, Utrecht, Netherlands ; and 2 Neurosurgery, University Medical Center Utrecht, Utrecht, Netherlands ) Rationale: Elevated levels of extracellular glutamate have been linked to the initiation and maintenance of seizures in patients with pharmaco‐resistant temporal lobe epilepsy (TLE). Impairment of the glutamate‐glutamine cycle, responsible for the inactivation and recycling of extracellular glutamate, has been implicated.One of the key neuronal components of this cycle is the vesicular glutamate transporter (VGLUT), which mediates the uptake of glutamate into synaptic vesicles. Three types of VGLUTs have been identified so far. VGLUT1 is the major type in the hippocampus. Little is known about VGLUT expression in the human hippocampus. The aim of this study is to investigate changes in VGLUT expression in the hippocampus of patients with TLE. Methods: We examined the hippocampal distribution of the VGLUT subtypes by immunohistochemistry, immunofluorescence and in situ hybridisation in 7 μm paraffin sections. Protein levels were assessed by semi‐quantitative western blotting. Measurements were performed in hippocampal biopsies obtained during surgical resection as treatment for TLE. Comparisons were made between hippocampi of TLE patients with (HS group) and without hippocampal sclerosis (nonHS group) and autopsy controls without neurological symptoms. Results: VGLUT1 immunoreactivity (IR) showed a diffuse neuropil staining of the subiculum, the pyramidal cell layer of all CA fields, the hilus, and the inner molecular layer of the dentate gyrus (DG). Large punctate structures around neuronal cell bodies and axons were observed in CA4 and CA3. No IR was detected in cell bodies. VGLUT1 mRNA was abundant in granule cells and pyramidal neurons. In nonHS and HS hippocampi VGLUT1 mRNA levels were decreased compared to controls, especially in areas with neuron loss. However, VGLUT1 protein levels were significantly increased in the nonHS hippocampus compared to both the HS and control hippocampus. This upregulation indicates translational regulation within hippocampal neurons. In addition, VGLUT1 mRNA and protein were upregulated in the DG of the HS hippocampus, consistant with the increased number of mossy fiber synapses onto granule cells due to mossy fiber sprouting. VGLUT2 IR was very weak in controls, but it was upregulated in the molecular layer of the DG in TLE patients. In the nonHS group VGLUT2 IR was only found in the outer molecular layer, whereas in the HS group VGLUT2 was also present in the inner molecular layer. These data indicate an upregulation of VGLUT2 associated with mossy fiber sprouting. Conclusions: Hippocampal mRNA expression of VGLUT1 is decreased and correlates with loss of glutamatergic synapses due to neuron loss. The increase in VGLUT1 and VGLUT2 protein found in TLE patients is most likely due to new glutamatergic synapse formation. This increase in glutamatergic signalling may contribute to the progression and pathology of TLE. 1 Jie Wu, 1 Jamie DeChon, 2 Nasim Bahadorani, 3 Harold Rekate, 1 John F. Kerrigan, and 2 Yongchang Chang ( 1 Neurology, Barrow Neurological Institute, St. Joseph's Hospital and Medical Center, Phoenix, AZ ; 2 Neurobiology, Barrow Neurological Institute, St. Joseph's Hospital and Medical Center, Phoenix, AZ ; and 3 Neurosurgery, Barrow Neurological Institute, St. Joseph's Hospital and Medical Center, Phoenix, AZ ) Rationale: Human hypothalamic hamartomas (HH) are associated with precocious puberty, gelastic seizures and severe refractory epilepsy. Emerging evidence suggests that the HH itself plays a pivotal role in this syndrome. γ‐aminobutyric acid is the major inhibitory neurotransmitter in the brain and plays an important role in modulation of fast inhibition through GABAA receptors. Abnormal GABAA receptor function is an important mechanism of epileptogenesis in various types of human epilepsy and animal epileptic models. However, it is unknown whether functional GABAA receptors are natively expressed in HH neurons – and if so, what are the functional, pharmacological and molecular properties of these receptors. Methods: Acutely enzymatic/mechanical dissociation of HH neurons was performed. Fresh resected HH tissues were immediately placed in ice‐cold dissection solution containing (in mM): 136.7 NaCl, 5 KCl, 0.1 NaH2PO4, 9.84 HEPES, 16.6 glucose and 21.9 sucrose. After incubation at 22 ± 1 °C for at least 1 h, tissue sections were treated with papain (4–6 mg/ml at 31 °C for 50–60 min), and then one tissue fragment was mechanically dissociated. Isolated single HH cells usually maintained function for 2–6 h. Perforated patch‐clamp whole‐cell recording combined with U‐tube drug application was applied to characterize function and pharmacology of GABAA receptors. Results: Under perforated patch‐clamp recording in voltage‐clamp mode, GABA induced an inward current (IGABA) at a holding potential of –50 mV. The IGABA was mimicked by a GABAA receptor agonist (muscimol) and blocked by GABAA receptor antagonist (bicuculline), suggesting that the IGABA was mainly mediated through the activation of the GABAA receptor. The concentration‐response relationship curve of IGABA showed that the EC50 and Hill coefficient were 1.3 mM and 0.9, respectively. The current‐voltage relationship current was linear at a reversal potential close to zero with a symmetric external and internal chloride concentration. HH neurons exhibited heterogeneity for allosteric modulations. Although most HH neurons presented an enhancement of IGABA by pentobarbital and pregnenalone, only ∼60% of neurons showed potentiation of IGABA by diazepam. RT‐PCR showed normal GABAA receptor subunit expression in HH tissues. Conclusions: We are the first to describe the functional, pharmacological and molecular properties of the GABAA receptor expressed in human HH. An understanding of functional GABAA receptors in HH neurons provides new insights into the epileptogenesis of gelastic seizures. (Supported by Women's Board Fund of the Barrow Neurological Institute and NS056104.) 1 Irene Yamazaki, 1 Véronique M. André, 1 Nanping Wu, 2 Harry V. Vinters, 3 Gary W. Mathern, 1 Michael S. Levine, and 1 Carlos Cepeda ( 1 Mental Retardation Research Center, David Geffen School of Medicine, UCLA ; 2 Department of Neuropathology ; and 3 Division of Neurosurgery, David Geffen School of Medicine, UCLA, Los Angeles, CA ) Rationale: Tuberous Sclerosis Complex (TSC) is an autosomal dominant disease caused by mutations in TSC genes that encode hamartin (TSC1) and tuberin (TSC2). These proteins affect cellular migration, proliferation, and differentiation in multiple organs, including the brain. In particular, it has been suggested that these proteins may represent growth suppressors. Epilepsy is one of the major complications of TSC and affects about 85% of patients. TSC shares histopathologic similarities with severe forms of focal cortical dysplasia (CD), specifically the presence of cytomegalic neurons and balloon cells. Methods: In this study we examined the morphological and electrophysiological properties of cells in cortical tissue samples from pediatric cases with TSC (n = 11, ages 1.8–10.1 yr) or severe CD (n = 15, ages 0.2–14 yr), that underwent surgery for the treatment of pharmacoresistant epilepsy. Normal‐ and abnormal‐appearing cells in slices were identified morphologically by infrared videomicroscopy and biocytin labeling (n = 72 in TSC and n = 151 in severe CD). Electrophysiological membrane properties were assessed with whole‐cell patch clamp recordings in current and voltage clamp modes. Results: In both TSC and severe CD cases, abnormal‐appearing cells were observed in the cortical mantle and classified into four principal groups. The first consisted of very large, pyramidal‐shaped cells corresponding to cytomegalic neurons. The second group included large, non‐pyramidal cells with atypical somatodendritic morphology that correspond to balloon cells. The third group included misoriented and dysmorphic pyramidal neurons, and the fourth consisted of immature‐looking pyramidal cells. Morphologically, abnormal cells from TSC and severe CD cases were very similar. Electrophysiologically, the most abnormal properties were found in cytomegalic neurons and balloon cells. Cytomegalic neurons were hyperexcitable, whereas balloon cells had no active inward currents and lacked synaptic inputs. Balloon cells were found more frequently in TSC cases, whereas cytomegalic pyramidal neurons occurred more frequently in severe CD cases. Conclusions: Besides the frequency of occurrence, there were no clear morphological or electrophysiological differences in cytomegalic neurons and balloon cells from TSC and CD patients, suggesting a link between the lesions produced by TSC gene mutations and the cellular alterations observed in severe forms of CD. (Supported by NIH NS38992.) 1 Andreas V. Alexopoulos, 1 Cristiane Q. Tilelli, 1 Roger Oghlakian, and 1 Imad M. Najm ( 1 Neurology, Cleveland Clinic Epilepsy Center, Cleveland, OH ) Rationale: Synaptic vesicle protein 2A (SV2A) was recently identified as the specific binding site for levetiracetam in the brain and as a promising target for drug development. We have previously reported the differential expression of SV2A in the neocortex of patients with cortical dysplasia (CD)‐associated pharmacoresistant epilepsy (Alexopoulos et al., Epilepsia 46 Suppl. 8:273, 2005). The objective of this study is to investigate the expression of the same protein in a well‐characterized animal model of CD. Methods: Sprague‐Dawley time‐pregnant rats received 145cGy whole body irradiation at embryonic day 17, resulting in pups with multifocal hippocampal (HIP) and cortical abnormalities. These animals exhibit recurrent, spontaneous seizures following “a second hit” in the form of a single dose of pentylenetetrazol (PTZ 30 mg/kg i.p.; postnatal day 60). Occurrence of seizures was confirmed by video‐EEG and behavioral monitoring. Age‐matched, nonirradiated controls received the same subconvulsive dose of PTZ. Video‐EEG recordings were obtained for 1 month before and after PTZ injection in all animals. Brains were examined using Cresyl Violet and immunostaining for SV2A and neuron‐specific nuclear protein. Results: Irradiated rats (n = 8), but not control rats (n = 3) developed spontaneous behavioral and electrographic seizures following PTZ (“second hit”), characterized by extended periods of arrest, facial or forelimb clonus and occasionally rearing and falling. Immunocytochemical (ICC) analysis in the cortex and HIP of non‐irradiated controls showed SV2A to be localized in regions surrounding cell bodies. Punctate staining was observed in HIP areas occupied by mossy fiber terminals. Histological examination of stained slices of irradiated animals confirmed the presence of multifocal CD and neuronal dispersion in HIP. ICC analysis of irradiated epileptic animals displayed a redistribution of the SV2A protein from the cell bodies to the neuropil (figure 1). SV2A in HIP of irradiated rats occupied the area of the hilus and apical dendrites in the dentate gyrus and cornus ammonis. Conclusions: The differential changes in SV2A immunoreactivity following recurrent seizures in dysplastic but not control animals are reminiscent of those observed in human epileptic and dysplastic tissue. Further studies will examine the influence of seizures in the SV2A redistribution, and its potential role in the expression of epileptogenesis and pharmacoresistance. (figure 1) (Supported by Investigator‐initiated study grant to A.V.A. sponsored by UCB, Inc.) 1 Atif Ali, 1 Yashomati Dua, and 1 F. Edward Dudek ( 1 Physiology, School of Medicine, University of Utah, Salt Lake City, UT ) Rationale: Animal models of chronic epilepsy with spontaneous seizures may be useful for antiepileptic (AED) drug development. Convenient and effective methods for testing potential AEDs over days and weeks should facilitate many experiments. This study investigated the effect of carbamazepine (CBZ) administered in food pellets on rats with kainate‐induced epilepsy. Methods: Adult male rats (Sprague‐Dawley; 150–250 g) were given repeated low‐dose kainate treatment (5 mg/kg ip hourly until convulsive status epilepticus persisted >3 h). Food pellets with CBZ were provided (30, 100 or 300 mg/kg/day) in three 2‐week trials (n = 7–9 rats) involving 5 days of CBZ or control treatment, separated by 2 recovery days. The total amount of food provided corresponded to normal caloric diet of 60 gm/kg/day. Results: All animals consumed the entire daily drug‐in‐food dose intermittently but continuously throughout the 24‐h day. CBZ significantly reduced mean seizure frequency in a dose‐dependent manner. At 30 and 100 mg/kg/day, CBZ treatment consistently failed to block all motor seizures, but no motor seizures were observed in any of the rats at 300 mg/kg/day. Conclusions: This AED‐in‐food protocol appears to be an effective means of chronic drug administration. When administered over 5‐day periods, CBZ reduced the frequency of spontaneous motor seizures in a dose‐dependent manner, but very high doses were required to block all motor seizures. The effects on electrographic seizures remain to be determined. (Supported by NS 049620 (FED).) 1 Michael R. Kasten, 1 Sanghoon Lee, and 1 Matthew P. Anderson ( 1 Neurology, Harvard Medical School, Beth Israel Deaconess Medical Center, Boston, MA ) Rationale: Recent studies have found epilepsy‐associated mutations in non‐ion channel genes. LGI1, a secreted protein of unknown function, was found to be mutated in patients with autosomal dominant lateral temporal lobe epilepsy (ADLTE). To investigate the functional effects of this protein on native neural circuits, we developed methods to study wild‐type and mutant LGI1 protein in living brain tissue. Methods: Transgenic mice were created using a full‐length LGI1 gene engineered to contain the epilepsy‐associated LGI1 gene mutations. We also created transgenic mice using the wild‐type, full‐length LGI1 gene. In order to measure the effect of this mutant protein on neuron excitability and synaptic communication, we prepared acute brain slices from adult LGI1 transgenic and control mice and compared the electrophysiological properties of individual neurons. Because ADLTE is characterized by auditory hallucinations during seizures, we studied neurons in a second order thalamic nucleus that relays auditory sensation to the cortex, the dorsal medial geniculate nucleus (DMGn). Results: Mutant LGI1 transgene increased the intrinsic excitability of DMGn thalamic neurons. Mutant LGI1 transgene also increased phosphorylation of a number of protein kinases. Conclusions: These findings support our current hypothesis that mutant LGI1 causes epilepsy through a unique mechanism involving altered signal transduction. (Supported by: Funded by National Institute of Neurologic Disease (MPA), Burrough Wellcome Fund Career Award (MPA), and Beth Israel Deaconess Medical Center.) 1 Ramin Atefy, 1 Sanghoon Lee, 1 Takatoshi Mochizuki, 1 Thomas E. Scammell, 1 Donald L. Schomer, and 1 Matthew P. Anderson ( 1 Neurology, Harvard Medical School, Beth Israel Deaconess Medical Center, Boston, MA ) Rationale: Patients with autosomal dominant lateral temporal epilepsy (ADLTE) with auditory features were recently found to harbor mutations in the non‐ion channel protein LGI1. To determine if mutations of this protein are sufficient to produce epilepsy, we created a transgenic mouse from the full‐length LGI1 gene engineered to contain an epilepsy‐associated gene mutation. As a control, other transgenic mice were also created that carry extra copies of the wild‐type, full‐length LGI1 gene. We predicted that ADLTE mutant LGI1 will produce epilepsy in mice. Methods: Mice of each genotype were implanted with electroencephalogram (EEG) and electromyogram (EMG) electrodes for polysomnogram recording. After > 9 days of recovery from surgery, they were transferred to recording cages in a sound‐attenuated chamber with a 12 hr light/dark (LD) cycle (30 lux; lights on at 7:00 A.M. and off at 7:00 P.M.) and a constant temperature (22–24°C) for 5 days of continuous recording. Video recordings were also performed during the last day in a subset of mice. The signals were digitally filtered (EEG, 0.3–30 Hz; EMG, 2–100 Hz) and semi‐automatically scored in 10 sec epochs as wake, NREM sleep, REM sleep, or seizures. Sleep analyses were focused on the last recording day. The preliminary scoring was visually inspected and corrected when appropriate. Results: Mutant LGI1 transgenic (n = 5), full‐length LGI1 transgenic (n = 4), and wild‐type littermate (n = 5) mice showed no significant differences in the hourly amounts of NREM, REM, or wakefulness. Seizures induced by pentylene‐tetrazol (60 mg/kg, i.p.) were also examined. Conclusions: ADLTE mutant LGI1 transgene does not significantly influence the pattern of sleep. We are currently testing the hypothesis that mutant LGI1 is sufficient to cause epilepsy. (Supported by: Funded by National Institute of Neurologic Disease (MPA), Burrough Wellcome Fund Career Award (MPA), and Beth Israel Deaconess Medical Center.) 1 Paven K. Aujla, 2 Michael Fetell, and 1 Frances Jensen ( 1 Neurology, Children's Hospital and Harvard Medical School, Boston, MA ; and 2 IVAX Pharmaceuticals, Key Biscayne, FL ) Rationale: Hypoxic encephalopathy is a major cause of neonatal seizures, and can lead to long term neurologic deficit and epilepsy. Neonatal seizures can be refractory to conventional AED medication. AMPARs are overexpressed on cortical neurons in the neonatal rat and term human, when the susceptibility to hypoxic seizures is highest (J. Comp. Neur. 2006; 497:42). We have shown that the AMPA subtype of neuronal glutamate receptors in cortex and hippocampus play a critical epileptogenic role in a rodent model of neonatal seizures (J. Neurosci. 2005; 25:3442), and AMPAR antagonists such as NBQX and topiramate effectively suppress the acute and long term epileptogenic effects of hypoxia in the perinatal rodent (Ann. Neurol. 2001; 50:366; Epilepsia 2004; 45:569). Here we evaluated the efficacy of talampanel (GYKI‐53773), a novel orally‐available non‐competitive AMPAR antagonist in our rodent model. Methods: P10 rat pups were exposed to global hypoxia per our previous protocols (J. Neurosci. 2001; 21:8154). Rats were pretreated i.p. with talampanel or vehicle 30 minutes prior to hypoxia and seizure time and severity were recorded by videotape and scored blinded. A second outcome measure was evaluated by assessing susceptibility of rats to KA‐induced seizures 20 days (P30) after they were exposed to hypoxia. KA was administered i.p. (10 mg/kg) and seizure activity was recorded over a 3‐hour period. Rats given KA on P30 were killed on P33 by transcardiac perfusion with 4% paraformaldehyde. Each brain was removed, postfixed, cyroprotected in 20% sucrose overnight, and mounted onto a freezing microtome. To quantify DNA fragmentation, counts of positively stained cells per 12,500μm2 in CA1, CA3 and amygdala were made as previously described (Ann. Neurol. 2001; 50:366). Results: Acute hypoxia‐induced seizures at P10 were suppressed by talampanel in a dose‐related manner. Anticonvulsant activity was maximal at 7.5 and 10 mg/kg, where seizures were blocked 74.6% at 10 mg/kg (25.4 ± 7.3, n = 17; p < 0.001) and 86.7% at 7.5 mg/kg (13.3 ± 3.2, n = 17; p < 0.001). Rats pretreated with vehicle prior to hypoxia showed significantly more neuronal injury in amygdala following KA‐induced seizure at P30 than did nonhypoxic controls or rats pretreated with talampanel (7.5 and 10 mg/kg) prior to hypoxia at P10 (p < 0.001, one‐way ANOVA). Conclusions: Talampanel exhibited a dose‐dependent effect on hypoxia‐induced hypoxic seizures in the immature rat. Futhermore, acute talampanel treatment during hypoxia protected the rats from later life increases in seizure susceptibility that are observed in vehicle treated hypoxic controls. These data suggest talampanel may have clinical potential in the treatment of neonatal seizures. (Supported by Epilepsy Project, IVAX, and NINDS RO1 NS31718.) 1,2 Natacha Porta, 1,2 Stephane Auvin, 1,2 Cecile Lecointe, 2 Regis Bordet, and 1,2 Louis Vallee ( 1 Pediatric Neurology, Lille University Hospital, Lille, France ; and 2 EA 1046, School of Medicine, Lille, France ) Rationale: Peroxisome proliferator‐activated receptor α (PPARα) is one of the three subtypes of the nuclear receptor PPAR family. PPARα activation by fenofibrate can protect cerebral injury by anti‐oxidant and anti‐inflammatory mechanisms. In this study, we tested PPARα activation on the cerebral excitability before study its potent neuroprotective effects in status epilepticus. Methods: Wistar rats (280g‐320g) were separated in two groups (n = 8). Sham group (Sh) received a complete powder diet ad libidum, while experimental group (F) received powder diet containing 0.2% fenofibrate. After 14 days of treatment, pentylenetetrazole threshold (PTZth) was used to assess anti‐convulsive effect. PTZ was dissolved in a saline solution (10 mg/ml) and was infused i.v. at 5 ml/h. Two threshold doses (mg/kg) were studied: the first bilateral forelimb myoclonus and the start of generalized seizure. Liver weights were recorded in each animal to assess that fenofibrate was pharmacologically active. All values are given as mean ± sem. Results: Rats receiving diet with 0.2% Fenofibrate had an increase of liver weight (liver weight/body weight): 0.063 ± 0.004 versus 0.036 ± 0.002 (p = 0.001). Myoclonic and seizure PTZ thresholds were increased in F group in comparison to Sh group: myoclonic PTZth were 40.85 ± 3.06 mg/kg in F group vs. 34.28 ± 1.26 mg/kg (Sh) (p = 0.028); seizure PTZth were 46.85 ± 3.58 mg/kg in F group vs. 37.29 ± 1.8 mg/kg (Sh) (p = 0.021). Conclusions: A treatment with 0.2% fenofibrate significantly increases myoclonic and seizure PTZ thresholds suggesting anticonvulsive proprieties. The role of PPAR activation in brain excitability should be investigated. (Supported by AEAC association and NP was supported by the grant “Conseil Régional Nord‐Pas‐de‐Calais et CHRU de Lille”.) 1,2 Thomas L. Babb, 1 Weili Zhou, and 1 Thomas P. Miller ( 1 Pediatrics, Wayne State University, Detroit, MI ; and 2 Neurology, Wayne State University, Detroit, MI ) Rationale: We used the rat in utero radiation model of cortical dysplasia (CD), which is characterized postnatally (P) by neuronal dyslaminations, polarity disorientations, and daily increases in cortical thickness. At P30 these rats may exhibit cortical seizures from the region of the permanent CD. Because the CD is present at birth and increases daily in size with the growth of neurons and dendrites, we hypothesized that a permanent increase in NMDA receptor densities may develop early in the first week to enhance CD neurons' excitability. Methods: Dams were radiated with 145 Rads (cGy) at embryonic day 17 (E17), their offspring were studied at 24 hour intervals, postnatal ages P0–10 and P21–36. The development of CDs measured neuron sizes and neuron densities in cortical layers II‐V. Densities were compared to age matched controls using stereology. In siblings, expressions of NMDA receptor subunits NR1 and NR2B were measured by immunocytochemistry and Western blotting. Statistics between control and radiated age matched pups were made between treatments and across ages. Results: Daily increases in neuron sizes were found from P0 through P10 in both groups, and after that there were no greater cytoplasmic increases in either group. At P0 and P1 nearly all neurons were spherical or oval in shape, lacking dendritic processes. At P2 neurons developed larger cytoplasms and larger proximal dendrites, which gave the neurons a pyramidal shape in all cortical layers. Growth of dendritic and axonal processes began at P2 and continued after P10. The expression levels for NR1 and NR2B at P0 and P1 were low, but increased significantly by P2; however, this increase plateaued after P2 and through P10. It is important to note that these developmental patterns were the same for both radiated and nonradiated pups. Conclusions: We used the rat in utero radiation model to identify postnatal malformations of cortical neurons and their NMDA receptor subunit expressions. During the earliest postnatal period, P0‐P10 the cortex is expanding with increasing neuron sizes and neuritic outgrowths. Only P0 and P1 had significantly immature neurogenesis, with small, round neurons lacking dendrites and having the lowest NMDA receptor expressions. At P2‐P10 neuron sizes and dendritic lengths increased. Also, at P2 NMDA expressions were maximal, and neither NR1 nor NR2B expressions increased on subsequent days. These developmental patterns in neurons and receptors were the same for control and radiated rats that always had permanent CD starting at birth. Hence, the later (P30) process of epilepsy likely occurs when all the axo‐dendritic synapses, including aberrant excitatory connections within the CD, are fully mature. (Supported by: 1) NIH Grant NS41375; 2) The Frankel Family Endowed Chair in Pediatric Neuroscience.) 1 Mihyun Bae, 1 Melissa Brunckhorst, 4Wendy M. Mars, 4George K. Michalopoulos, 4Cristian Achim, and 1,2,3 Elizabeth M. Powell ( 1 Department of Anatomy and Neurobiology, University of Maryland, School of Medicine, Baltimore, MD ; and 3 Department of Pathology, University of Pittsburgh, School of Medicine, Pittsburgh, PA ) Rationale: Inhibitory neurotransmission functions in neural circuits by repressing of excitatory output. Local GABAergic interneurons are the main source of inhibition in the cerebral cortex. Loss of these neurons leads to epilepsy along with other neurological disorders. Interneuron development is tightly regulated by molecular cues including hepatocyte growth factor/scatter factor (HGF/SF). HGF/SF signals via its receptor MET to promote cellular survival, differentiation, mitogenesis and migration dependent upon cellular context. HGF/SF, which is secreted as a single chain precursor form, must be cleaved for biological function by serine proteases such as urokinase plasminogen activator (uPA). uPAR, the receptor of uPA, efficiently increases the protease cleavage by uPA. Adult uPAR−/− mice have decreased numbers of neocortical GABAergic interneurons and severe behavioral dysfunction such as spontaneously generated seizure and increased susceptibility to PTZ(pentylenetretrazol)‐induced seizures. These defects of GABAergic interneurons have been attributed to a reduction of HGF/SF. Methods: The uPAR−/− mouse was bred with transgenic mouse overexpressing HGF/SF under the astrocytic glial fibrillary acidic protein promoter (GFAP). Immunohistochemistry and seizure behaviors to PTZ (pentylenetetrazol) were used to evaluate anatomical and functional recovery of GABAergic interneuron in wildtype, GFAP‐Hgf, uPAR−/−, and uPAR−/−:GFAP‐Hgf mice. Results: The loss of interneurons in the uPAR−/− mouse is restored in uPAR−/−xGFAP‐Hgf mouse, in which the uPAR−/− mouse was bred with an HGF‐overexpressing mouse line (GFAP‐HGF). Reduction of PV(parvalbumin), a marker of GABAergic neuron, in the uPAR−/− mouse is rescued in uPAR−/−xGFAP‐Hgf mouse (Figure 1). The seizure susceptibility to PTZ of uPAR−/− mice is decreased in the uPAR−/−xGFAP‐Hgf mouse, to frequencies and severity observed in wildtype animals. The latency to seizure activity after PTZ treatment is also increased in the uPAR−/−xGFAP‐Hgf mouse as compared to uPAR−/− mouse. Conclusions: These data suggest that restoring HGF/SF levels can recover the anatomical interneuron defects and neocortical circuitry, leading to prevention of the seizure activity of uPAR−/− mice. This study will help uncover molecular mechanisms critical for seizure onset and their possible therapeutic targets for epilepsy. (figure 1) (Supported by Epilepsy Foundation (EMP and MB) and by the Women's Health Research Group (EMP).) 2 Justin M. Keener, 1 H. Steve White, and 2 Joanna C. Beachy ( 1 Pharmacology and Toxicology, University of Utah, Salt Lake City, UT and 2 Pediatrics, University of Utah, Salt Lake City, UT ) Rationale: Moderate to severe hypoxic‐ischemic encephalopathy (HIE) occurs in 1‐ 2/1000 live births. Approximately 40% of the infants with HIE and neonatal seizures develop epilepsy by 4 years of age. The Levine rat pup model of hypoxia/ischemia (H/I) insult is a well characterized model of HIE. Recently, it has been reported that 40% of rat pups exposed to H/I insult at 7 days of age exhibit spontaneous seizures by 6 months of age (Epilepsia 2004; 45(10):1210–1218). We hypothesize that a H/I insult will alter seizure threshold and, thus, predispose the rat to the development of spontaneous seizure activity. Methods: Seven day old Sprague‐Dawley rat pups were anesthetized with isoflurane and subjected to right carotid artery ligation. After recovery, the pups were returned to their dams for 2 hours. Pups were then exposed to 8% oxygen/92% nitrogen for 2 hours. Temperature was maintained at 34.5–36°C. Convulsive current curves for clonic (forebrain) and tonic (hindbrain) seizures were generated at 1 and 2 months after H/I insult and compared to those obtained from age‐ and sex‐matched naive control rats (> 25 rats per treatment group). Data were compared using Probit analysis (Minitab statistical software program). Results: Data for each seizure type is presented as CC50, the current at which 50% of the rats exhibited seizures ± standard error. A decrease in CC50 suggests increased seizure susceptibility. In naive male rats, CC50 significantly increased over time. Female naive rats had significantly deceased CC50 at 2 months of age (clonic 18.2 ± 0.4 mA, tonic 29.2 ± 0.9 mA) when compared to male naive rats (clonic 27.3 ± 0.9 mA, tonic 52.6 ± 1.4 mA). When compared to naive female rats, the CC50 for clonic seizures in female rats was significantly decreased (p = 0.005) at 1 month post H/I insult but increased at 2 months (p = 0.001). CC50 for tonic seizures in female naive and H/I rats did not differ. Male rats exhibited increased CC50 for tonic seizures (p = 0.007) at 1 month (35.5 ± 1.2 mA) and 2 months (61.6 ± 1.7 mA) when compared to naive male rats (28.3 ± 1.7 mA and 52.6 ± 1.4 mA, respectively). Conclusions: Alteration in seizure susceptibility is dependent on sex and time after H/I insult. Female rat pups exhibited enhanced clonic (forebrain) seizure susceptibility at 1 month after H/I insult but deceased susceptibility 2 months after insult. Interestingly, H/I insult in male rats decreased tonic seizure susceptibilty at 1 and 2 months after insult. Thus, H/I insult does not appear to consistently increase seizure susceptibility as evaluted by the MES method. Based on these results, it would appear that electrical population seizure threshold testing is not predictive of long‐term outcome (e.g. spontaneous seizure activity). (Supported by Primary Children's Medical Center Foundation Innovative Research Grant.) 1 Walter G. Besio, 1 Kanthaiah Koka, 2 Andrew J. Cole, and 1 Fei Zhu ( 1 Biomedical Engineering, Louisiana Tech University, Ruston, LA ; and 2 Epilepsy Service, Massachusetts General Hospital, Harvard Medical School, Boston, MA ) Rationale: We sought to evaluate the effect of transcutaneous electrical stimulation (TcES) via concentric ring electrodes on ictal electrographic and behavioral activity and mortality in rats with pilocarpine‐induced status epilepticus (SE). Methods: Male Sprague‐Dawley rats (290–330 g) were briefly anesthetized and 3 concentric ring electrodes were afixed to their scalps one day before the experiment. Scopolamine methylnitrate (2 mg/kg i.p.) was given 30 minutes prior to pilocarpine. Pilocarpine HCl (310 mg/kg i.p) was given to cause long lasting SE. Laplacian EEG was recorded from tri‐polar concentric electrodes on the scalp. TcES was applied 5 minutes after the onset of SE. Time‐frequency analysis was performed on the Laplacian EEG signals to compare the electrographic activity before and after the application of TcES. Behavior was monitored by inspection. Survival was assayed at 24 h after administration of pilocarpine. Results: Control rats (n = 8) followed the classic electrographic stages of pilocarpine‐induced status epilepticus described by Treimen (1987) and expired on average 15 hours after the pilocarpine injection. TcES‐treated rats (n = 8) lived significantly longer (p = 0.013, Two‐Sample t‐Test), which were euthanized on average at 48 hours. Twenty‐four hours after the pilocarpine injection, six (75%) TcES‐treated rats versus one (12.5%) control rat were alive (p = 0.041, Mann‐Whitney U test). Both electrographic and behavioral manifestations of seizure activity were reduced or abolished after the application of TcES. All surviving TcES‐treated rats recovered to baseline activity, including eating and drinking. By contrast, none of the control rats ate or drank after they entered SE. The time‐frequency analysis showed evident differences before and after TcES. Conclusions: The application of TcES increased survival of rats with pilocarpine‐induced SE. Positive TcES effects on electrographic and behavioral manifestations of seizure activity were significant and persistent. TcES may represent a novel and effective early treatment for SE. (Supported by Louisiana Tech University Center for Entrepreneurship and Information Technology and the Louisiana Board of Regents.) 1 Devin J. Cross, 1 Svitlana Kravchenko, and 1 Jose E. Cavazos ( 1 Neurology and Pharmacology, University of Texas Health Science Center at San Antonio, San Antonio, TX ) Rationale: Early life status epilepticus is detrimental to the behavioral and cognitive development in childhood and increases the risk to develop epilepsy in later life. Recurrent seizures during the first two weeks of life in rats induce limited neuronal injury and synaptic reorganization of the mossy fiber pathway in the CA3 region of the hippocampus (Holmes et. al., 1999; Cilio et. al., 2003). These neurobiological alterations enhance the degree of neuronal injury and the duration of status epilepticus experienced after a second excitotoxic injury later in life (Koh et al., 1999; Schmid et al., 1999). The present study examines the frequency of late spontaneous seizures one year after early life kainic acid induced status epilepticus. Methods: Kainic acid was administered in age specific doses for the following experiments. On postnatal day (P) 15, kainic acid was administered to induce convulsive status epilepticus and control rats received equal volumes of saline. On P45, half of rats from each group received kainic acid induced status epilepticus. Status Epilepticus was monitored for 6 hours and behavioral seizures were scored using a previously described scale. These four groups were allowed to mature for 1 year and then underwent surgical implantation of bipolar hippocampal electrodes to monitor the occurrence of late onset spontaneous seizures. Rats were then monitored for 8 hours to determine the frequency of late onset spontaneous seizures. After rats were monitored they were perfused with sodium sulphide and paraformaldehyde/gluteraldehyde for histological examination with Timm's reagent and Cresyl Violet. Results: After one year of maturation, rats that received kainic acid or saline on P15, did not demonstrate late onset spontaneous seizures over several 8 hour monitoring sessions, however mossy fiber sprouting was detected in the oriens of the CA3b region. Rats that receive saline on P15 and kainic acid on P45 developed late onset spontaneous seizures with an hourly seizure frequency of 0.22 ± 0.13 (mean ± SEM). This excitotoxic insult induced neuronal injury and synaptic reorganization of the mossy fiber pathway into the dentate gyrus and CA3b regions. Rats that received a “double hit” kainic acid on P15 and P45 developed severe late onset spontaneous seizures, with an hourly seizure frequency of 0.62 ± 0.31. These rats also developed widespread neuronal injury and synaptic reorganization of the mossy fiber pathway into the dentate gyrus and the CA3b regions. Conclusions: Early life status epilepticus induced by kainic acid induces limited neuronal injury and synaptic reorganization in the CA3b region, which is not sufficient to initiate the process of epileptogenesis. However, alterations after early life seizures shift the neurobiology in a pro‐epileptic direction that enhances the susceptibility to seizures and accelerates epileptogenesis after subsequent insults during adulthood. 1 Marion Czapp, 1 Jens P. Bankstahl, 2 Mark Helm, 3 Ulrich Massing, 3 Vittorio Ziroli, 4 Uta Huelsermann, 4 Gert Fricker, and 1 Heidrun Potschka ( 1 Department of Pharmacology, Toxicology, and Pharmacy, University of Veterinary Medicine Hannover, Hannover, Germany ; 2 Department of Chemistry, University of Heidelberg, Institute of Pharmacy and Molecular Biotechnology, Heidelberg, Germany ; 3 Clinical Research, Clinic for Tumorbiology, Freiburg, Germany ; and 4 Department of Pharmaceutical Technology and Pharmacology, University of Heidelberg, Institute of Pharmacy and Molecular Biotechnology, Heidelberg, Germany ) Rationale: Despite the launch of several new antiepileptic drugs (AED) during the last two decades pharmacoresistance of epilepsy remains an unsolved problem. It is hypothesized that local overexpression of multidrug transporters (MDT) in the epileptic tissue of pharmacoresistant individuals results in limited access of AED to the epileptic focus region. Downregulation of MDT by RNA interference (RNAi) would be a possible approach to overcome pharmacoresistance. Aim of the present study was to test if specific delivery of siRNA targeting the mRNA sequence of the major MDT P‐glycoprotein (Pgp) to brain‐capillary endothelial cells is possible. Methods: Liposomes were coupled with ApoE4 peptides which are discussed to mediate an uptake of liposomes into brain‐capillary endothelial cells via LDL receptors. ApoE4‐liposomes loaded with fluorescence‐labelled siRNA targeting the rat Pgp mRNA sequence were administered i.v. to female Wistar rats. Distribution of siRNA was investigated at different time points (4, 10, 24 h) post administration by confocal microscopy in rat brain sections. Based on a fluorescence resonance energy transfer (FRET) effect the integrity of siRNA was controlled (siRNA was labelled with two fluorescent dyes). The influence of siRNA molecules on Pgp expression was evaluated 24 h post administration by immunhistochemistry and compared with control groups. Results: Fluorescence microscopy of rat brain sections after siRNA administration demonstrated a successful targeting to brain‐capillary endothelial cells. An accumulation of labelled siRNA in the cells was obvious at all time points. The intensity of fluorescence increased from 4 to 24 h post administration. FRET analysis indicated that the siRNA was still intact 24 h following administration. Preliminary results of the immunhistochemical investigations gave evidence for a reduced Pgp expression in rats treated with siRNA in comparison with control groups. Conclusions: It was demonstrated that specific delivery of siRNA encapsulated in immunoliposomes to brain‐capillary endothelial cells can be achieved. Thus, the study substantiates the suitability of this novel approach for siRNA delivery. Reduction of Pgp expression in brain‐capillary endothelial cells will be investigated in further experiments to ensure that an RNAi‐based downregulation of Pgp takes place and enhances AED efficacy in the amygdala kindling model of temporal lobe epilepsy. (Supported by CURE (Citizens United for Research in Epilepsy).) 1 Veerle De Herdt, 2 Jan De Waele, 1 Jeroen Van Aken, 1 Kristl Vonck, 3 Jean Delbeke, 1 Robrecht Raedt, 1 Wytse Wadman, and 1 Paul Boon ( 1 Laboratory for Clinical and Experimental Neurophysiology, Ghent University Hospital, Ghent, Flanders, Belgium ; 2 Department of Intensive Care, Ghent University Hospital, Ghent, Flanders, Belgium ; and 3 Neural Rehabilitation Engineering Laboratory, Université Catholique de Louvain, Brussels, Brussels, Belgium ) Rationale: The precise mechanism of action of vagus nerve stimulation (VNS) in suppressing seizures remains to be elucidated. This study investigated whether VNS affects the threshold for provoking focal motor seizures using cortical electrical stimulation. Methods: Male Wistar rats (n = 8) were implanted with a custommade cuff‐electrode around the left vagus nerve. Stainless steel electrodes were bilaterally implanted on the rat motor cortex for stimulation and recording of EEG. Motor cortex stimulation (50 Hz, biphasic current pulses) with increasing amplitude (0 to max 2000 μA) was performed. At the time of the first clinical seizure symptoms, the stimulation train was acutely aborted and the threshold intensity was determined by means of continuous video‐EEG monitoring. The vagus nerve was stimulated for one hour with an intensity of 0,75 mA and a maximum duty cycle. For each rat, seizure thresholds (ST) were assessed before and immediately after VNS during 2 or 3 sessions on different days. Results: In 7/8 rats, ST increased after one hour of VNS. In 1/8 rat ST remained unchanged. In total, 21 sessions of VNS were performed in the 8 rats; for each session, ST were assessed at least 3 times before and 3 times after VNS. There was a significant increase (99% CI) of the ST after one hour of VNS (mean: 1424 μA; SD: 315; range: 869–1999 μA) compared to the baseline value (mean: 1077 μA; SD: 244; range: 673–1754 μA). Conclusions: In this study VNS significantly increased ST in a cortical stimulation model used for evaluating efficacy of antiepileptic treatment for motor seizures. These data indicate that VNS is capable of modulating cortical excitability. (Supported by V. De Herdt is supported by a grant (‘aspirant’) from the FWO‐Flanders.) 1 Stefanie M. Dedeurwaerdere, 2 Marie‐Claude Gregoire, 3 Peter Roselt, 3 David Binns, 2 Andrew Katsifis, 3 Rod Hicks, 1 Terrence J. O’Brien, and 1 Damian E. Myers ( 1 The Department of Medicine/Royal Melbourne Hospital, The University of Melbourne, Melbourne, Vic, Australia ; 2 Radiopharmaceuticals Research Institute, Australian Nuclear Science and Technology Organisation, Sydney, NSW, Australia ; and 3 Centre for Molecular Imaging, Peter MacCallum Cancer Centre, Melbourne, Vic, Australia ) Rationale: Changes in GABAergic inhibition are known to be associated with epilepsy. Human positron emission tomography (PET) studies have shown altered GABAA receptor expression using flumazenil (FMZ), a ligand with a high affinity for the central benzodiazepine receptor (cBZR). Small animal PET, using serial scans in the same epileptic animal, is potentially a powerful technique to investigate the evolution and pathophysiolgical basis of these changes during epileptogenesis. The aim of this study was to explore the potential of 2’‐(18F)fluoroflumazenil ((18F)FFMZ) to study the GABAA/cBZR complex in rats utilizing small animal PET. Methods: In five male non‐epileptic control rats (350–375 g), two serial dynamic PET scans were acquired under ketamine/xylazine anaesthesia after bolus injection of (18F)FFMZ (66.6 ± 3.7 MBq). Six scans were acquired after injecting a tracer dose. In addition, a pre‐saturation scan, consisting of a cold FMZ injection followed 10 min later by (18F)FFMZ, and two displacement studies (cold FMZ injection during the dynamic scan) were performed. PET/MRI co‐registration was performed to delineate different regions of interest (whole brain, hippocampus, pons). Results: In the hippocampus, a structure with high GABAA/cBZR density, the signal intensity was higher (10%) than in the whole brain (p < 0.05) with no difference between left and right hippocampus. In the pons, which has low GABAA/cBZR density, the signal intensity was approximately 20% lower than in the whole brain (p < 0.01). Pre‐saturation and displacement studies showed a high non‐specific component in the measured signal, which may be due to recirculation of (18F)‐fluoroethanol, a metabolite of (18F)FFMZ. (figure 1) Conclusions: This study has demonstrated differential uptake of (18F)FFMZ into key structures in the brain, with the highest activity in the hippocampus. However, significant non‐specific activity in the brain will limit the ability to detect changes in GABAA/cBZR density during epileptogenesis and therefore alternative ligands need to be investigated. (Supported by the CRC for Biomedical Imaging Development.) 1,2 Karine Bressand, 1 Corine Roucard, and 2 Antoine Depaulis ( 1 Co, SynapCell, Grenoble, France ; and 2 INSERM, U704, Grenoble, France ) Rationale: The syndrome of mesial temporal lobe epilepsy (MTLE) represents a major challenge in the clinical management of focal seizures mainly due to drug refractoriness. Both morphological and electroclinical features of MTLE can be mimicked in the adult mice by a unilateral injection of kainic acid (KA) in the dorsal hippocampus. Neuronal loss in CA1, CA3 and hilus as well as gliosis and mossy fiber sprouting are observed in the injected side, as well as dispersion of the dentate gyrus, as observed in MTLE patients. In addition, recurrent focal hippocampal discharges occur two weeks after KA which share many similarities with human focal discharges. Whether these discharges are suppressed by antiepileptic drugs (AEDs) was addressed in this study where the effects of different classical AEDS were compared to a new AED: pregabalin. Methods: C57BL/6 adult mice were stereotaxically injected with KA (1 nmol in 50 nl) in the dorsal hippocampus and equipped with cortical and hippocampal electrodes. After 3 weeks, mice were injected with different doses of AEDs or vehicle in a random order and EEG recordings were performed for 20 min pre‐injection and up to 120 min post‐injection. Diazepam (0.5, 1, 2, 3 mg/kg) (n = 6), valproate (100, 200, 400 mg/kg) (n = 6), lamotrigine (30, 60, 90 mg/kg) (n = 7), carbamazepine (25, 50 mg/kg) (n = 12) and pregabalin (n = 8) (10, 50, 100 mg/kg) were injected i.p. Only animals with cell loss and dentate gyrus dispersion were used in the analysis. Cumulated durations of hippocampal discharges during 20 min periods were compared between the different drug conditions and versus vehicle using ANOVA for repeated measures. Results: Hippocampal discharges were suppressed in a dose‐dependent way following diazepam (1, 2 and 3 mg/kg). Suppression of seizures was also observed with high doses of valproate (400 mg/kg) and lamotrigine (90 mg/kg) without dose‐dependency and with occasional behavioral side‐effects. No suppression was obtained following carbamazepine (25 and 50 mg/kg) and aggravation was observed after a low dose of lamotrigine (30 mg/kg). Pregabalin significantly suppressed hippocampal discharges in a dose‐dependent way (50 and 100 mg/kg). Conclusions: Our study shows that besides benzodiazepines, only pregabalin suppresses focal discharges in a dose‐dependent way in the kainate‐mouse model of MTLE. On the contrary, classical AEDs were effective only at high doses (valproate, lamotrigine) or were without effects (carbamazepine). In addition to the histological, electrophysiological and behavioral features of this model, the resistance to classical AEDs suggest a form of drug‐refractoriness often observed in human MTLE. Therefore this model fulfils most criteria for an efficient preclinical development of antiepileptic therapeutical strategies for MTLE. The suppressive effects of pregabalin suggest that this compound could be effective on human MTLE. (Supported by Pfizer France, Inserm and Ligue Française Contre l’Epilepsie.) 1 Taku Doi, 1 Yuto Ueda, 1 Keiko Nagatomo, and 2 James L. Willmore ( 1 Department of Psychiatry, Miyazaki Medical College, Kiyotake, Miyazaki, Japan ; and 2 School of Medicine, Saint Louis University, St. Louis, MO ) Rationale: Kindling is a form of epileptogenesis that can be induced with pentylenetetrazol (PTZ). We undertook this study to evaluate the contribution of glutamate and GABA transporters in the process of PTZ kindling. Methods: Rats were kindled with PTZ (40 mg/kg) injected i.p. three times per week until two consecutive stage five seizures were observed. Experiment 1 Animals underwent pentobarbital anesthesia and decapitation at 24 hours (Group A) or at 30 days (Group B) after their last PTZ induced seizure. Rats in Group A and B were compared with each control rats (Group C and D). Experiment 2 We collected early‐established animals (Group E) and kindling resistant animals (Group F). Early‐established rats had consecutive stage five seizures with less than 5 injections of PTZ. Kindling resistant animals did not have fully kindled seizures by 12 injections of PTZ. Under anesthesia with sodium pentobarbital brain tissue was collected at 30 days after the last PTZ injection. We used western blotting to measure the changes in hippocampal glutamate and GABA transporters in all animals. Results: GLAST, GLT‐1 and EAAC‐1 of Group A were elevated significantly compared with Group C. No change was found in any transporters of Group B compared with Group D. However, at 30 days (Group B) after their last PTZ induced seizure, animals had stage 5 seizure with PTZ. In early‐established rats (Group E), levels of EAAC1 and GAT1 were decreased by 30% when compared to kindling resistant animals (Group F). In these animals, the level of GLT‐1, GLAST and GAT3 equaled control. Conclusions: Activation of NMDA receptors is critical to the kindling process. Since all glutamate transporters were increased at 24 hours after establishing kindling with PTZ we considered this effect a response to seizure induced glutamate turnover. While animals remain kindled at the chronic phase, the lack of sustained changes in glutamate transporters suggest these proteins are not involved in the maintenance of epileptogenesis. We wonder if the glutamate and GABA transporters might be operant in the convulsion threshold set factor or as a pace factor for kindling. (Supported by a Grant‐in‐Aid for Encouragement of Young Scientists (17591223) from the Ministry of Education, Science, Sport and Culture, Japan (to T.D.).) 1,2 Rastislav Druga, 1 Hana Kubova, and 1 Pavel Mares ( 1 Developmental Epileptology, Institute of Physiology, Academy of Sciences, Prague, Czech Republic ; and 2 Anatomy, Charles University, 2nd Medical Faculty, Prague, Czech Republic ) Rationale: Pilocarpine induced status epilepticus (SE) in adult and immature animals results in neurodegenerative changes in many thalamic nuclei In the group of midline thalamic nuclei massive neuronal degeneration was found in immature animals in the reuniens nucleus. The nucleus reuniens (RE) is the largest of the midline nuclei and major source of projections to the hippocampus. Afferent projections originate mainly from limbic structures. Immunocytochemical studies have revealed calretinin (CR) and calbindin (CB) immunoreactive neurons in the RE and 70% of CR‐immunoreactive cells also showed CB‐immunoreactivity. The present study was focused on distribution and dynamics of neuronal degeneration in the RE and on changes in CR‐immunoreactivity within this nucleus in immature rats surviving up to one week after lithium‐pilocarpine SE. Methods: Experiments were performed in Wistar rats 12, 15, 18, 21 and 25 days old. Lithium‐pilocarpine model of SE was used. Only animals exhibiting convulsive SE were included in this study. Rats of all age groups were sacrificed 4, 8, 12, 24, 48 hours and 1 week after SE. Animals were deeply anesthetized with urethane (2.5 g/kg i.p) and perfused with 0.01 PBS (pH 7.4) followed by 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4). The brains were postfixed and cryoprotected. Blocks of the brain were sectioned in a coronal plane (50 μm), mounted on gelatin‐coated slides and processed for FluoroJade B (FJB) histochemistry. Another group of 25‐day‐old animals surviving 1 week after SE was processed for localization of CR. To determine the CR‐expressing cells the standard immunocytochemical techniques were used. Sections were observed under an epifluorescence microscope Olympus Provis AX 70 using FITC filter set. Results: Since P15 massive neuronal degeneration was evident in the RE. Damage was found in all age groups (P15, 18, 21, 25) and number of degenerated (FJB positive) neurons¨increased with age of SE elicitation and with survival interval. Neuronal degeneration was evident in all parts of nucleus, but the majority of FJB‐positive cells was in the rostral half of the RE. A significant decrease in number of CR‐positive neurons was apparent in animals surviving 1 week. CR‐reactive neurons were absent in the lateral portions of the RE, while the medial part of the nucleus exhibited a few of CR‐positive cells. Conclusions: The reuniens nucleus belong among thalamic nuclei where massive neuronal degeneration was evident since P15. The number of degenerated neurons increased with age and survival time after SE. Significant decrease in number of CR‐immunoreactive neurons indicate that calretinin has no neuroprotective effect within the reuniens nucleus. (Supported by a grant No.304/04/0464 of the Grant Agency of the Czech Republic.) 1 Mark J. Dunleavy, 2 Satchiko Shinoda, 2 Clara Schindler, and 1 David C. Henshall ( 1 Department of Physiology and Medical Physics, Royal College of Surgeons In Ireland, Dublin, Ireland ; and 2 Legacy Research, Portland, OR ) Rationale: Prolonged seizures in adults can damage the hippocampus and cause epilepsy. However, the effects of prolonged seizures on the immature brain remain poorly understood. A small number of studies have suggested seizures in rats < P14 cause damage, however the date at which vulnerability emerges is contraversial. If these seizures were able to cause hippocampal sclerosis, a characteristic of temporal lobe epilepsy (TLE), this could have significant implications for seizure treatment and approaches to anti‐epileptogenesis. Methods: We have developed a novel animal model in P10 Sprague‐Dawley rat pups that provides the unique ability to investigate the molecular pathogenesis of hippocampal sclerosis following seizures in the developing brain. The model features early‐life prolonged seizures, induced by intraamygdalar injection of kainic acid (KA), that trigger acute neuronal death in the hippocampus, leading to hippocampal sclerosis at adulthood. EEG was recorded both acutely and over a period of weeks at adulthood to confirm the presence of seizures. Results: Seizures were induced and hippocampal injury was then examined at various time points thereafter. Seizures induced largely unilateral hippocampal damage, affecting CA1 and CA3 regions. Western blot revealed that cell death was associated with activation of apoptotic pathways. By adulthood animals exhibited profound unilateral hippocampal sclerosis. Long‐term EEG confirmed the presence of spontaneous seizures in these animals. Administration of a neuroprotectant FK506 during the initial seizures provided protection from hippocampal sclerosis at adulthood. Conclusions: We have successfully developed a novel model of unilateral hippocampal sclerosis and TLE in P10 rat pups. Using this model valuable insights may be gained into epileptogenesis in the developing brain and provide targets for anti‐epileptogenic treatments. (Supported by Wellcome Trust and NIH.) 1 Tobias Engel, 1 Seiji Hatazaki, 1 Ina Koegel, 1 Carmen Bellver‐Estelles, 1 Jochen H.M. Prehn, and 1 David C. Henshall ( 1 Physiology, Royal College of Surgeons In Ireland, Dublin, Ireland ) Rationale: Epilepsy is one of the most common neurologic disorders, affecting approximately 50 million people worldwide and its characteristics are a predisposition to recurrent unprovoked seizures caused by abnormal hypersynchronous discharges of neurons. Through experimental modelling and clinical neuroimaging it has been shown that seizures are capable of causing neuronal death and this may contribute to epileptogenesis, impairments in cognitive function or the epilepsy phenotype. Neuronal death in this setting is, at least in part, due to the induction of the molecular machinery of apoptosis. Studies using clinical material from patients who suffer recurrent seizures and animal epilepsy models have confirmed changes in expression and activation of pro‐ and anti‐apoptotic proteins of the caspase and Bcl‐2 protein family. Another consistently up‐regulated pro‐apoptotic protein in seizure models is the tumour suppressor protein p53. Recently a new p53 up‐regulated target gene has been identified named PUMA. PUMA, a BH3‐only protein, is one of the most potent pro‐apoptotic proteins of the Bcl‐2 family which is also expressed in the central nervous system. In this study we examine the role of PUMA in seizure‐induced neuronal death using PUMA knock‐out mice. Methods: Seizures were focally evoked in adult C57Bl/6 mice by unilateral stereotaxic microinjection of kainic acid into the basolateral amygdala nucleus. The EEG was continuously monitored until lorazepam was administered to terminate seizures after 40 minutes. For Western blotting, samples were homogenized in lysis buffer and loaded on a SDS‐PAGE gel. For immunohistochemical analysis, coronal brain sections were stained with NeuN and analysis of cells exhibiting DNA fragmentation was performed using fluorescein‐linked TUNEL. Data are presented as mean ± SEM. Data were analyzed using one‐way ANOVA with post hoc Fisher's paired least significant difference test as appropriate. Significance was accepted at p < 0.05. Results: In C57Bl/6 wild‐type mice, p53 and PUMA were up‐regulated in the hippocampus shortly after seizure induction. Assessment of the phenotype of PUMA knock‐out mice revealed normal numbers of hippocampal and amygdala neurons, and normal expression of the kainate receptor in the amygdala and hippocampal site of seizure elicitation. Seizures elicited by intra‐amygdala kainate were of similar frequency and severity in the PUMA knock out mice. However, neuropathological analysis revealed neuron survival was significantly greater in PUMA knock‐out mice than in wild‐type mice and TUNEL counts were significantly lower in PUMA knock‐out mice than in wild‐type mice. Conclusions: The present data suggest that PUMA contributes to the neurodegeneration taking place during seizure‐induced neuronal death and may therefore represent a therapeutic target for adjunctive therapy to prevent hippocampal damage following status epilepticus. (Supported by Health Research Board, Ireland; Science Foundation Ireland.) 1 Dario J. Englot, 1 Ulrich Schridde, and 1,2 Hal Blumenfeld ( 1 Department Neurol., Yale University School Med, New Haven, CT ; and 2 Department Neurobiol., Yale University School Med, New Haven, CT ) Rationale: Partial seizures in patients with temporal lobe epilepsy (TLE) are often classified as either complex, characterized by deficits in consciousness, or simple, involving no loss of consciousness. Prior intracranial EEG studies of complex‐partial seizures in TLE patients have shown large amplitude slow waves in the frontal and parietal cortices ictally, most prominent in the orbitofrontal cortex. This neocortical slowing is associated with decreased cerebral blood flow in these areas measured using single photon emission computed tomography (SPECT). It has been hypothesized that ictal neocortical slowing may reflect abnormal cortical and subcortical network interactions during temporal lobe seizures, responsible for the deficits in consciousness seen in TLE patients. However, an animal model is needed to further study the mechanisms and pathophysiology of neocortical slowing. Methods: Sprague‐Dawley rats were implanted with a bipolar stimulating/recording electrode in the dorsal hippocampus, and a frontal recording electrode in the association (n = 5), cingulate (n = 6), or lateral orbital (n = 5) cortex. After one week of recovery, animals were kindled with a daily 1s train of 1ms, 60 Hz biphasic pulses at the threshold current necessary to produce an afterdischarges of >3s duration. EEG recordings from the hippocampal and frontal electrodes were compared. Behavior was rated using the standard Racine scale. Results: Hippocampal EEG showed seizure activity, consisting of polyspike discharges, during all events. While frontal recordings in the association and cingulate cortices showed rapidly propagating fast (polyspike or sharp) activity within 1–2s of hippocampal stimulation, large amplitude slow waves of approximately 3–4 Hz were seen ictally in the lateral orbital cortex. Behaviorally, these seizures did not involve motor convulsive activity. As daily stimulations proceeded, the slow rhythm in the lateral orbital cortex progressively converted to fast spike activity, suggesting seizure propagation. Conclusions: Our results show neocortical slowing in the lateral orbital cortex during hippocampal seizures in rats. These data suggest that hippocampal kindling may provide a good animal model of complex‐partial seizures in patients with TLE, demonstrating neocortical slow activity outside the region of seizure initiation. Further studies into the fundamental mechanisms of neocortical slowing in this model may provide insight into pathophysiology of altered consciousness in complex‐partial seizures, and may lead to improved treatments of TLE. (Supported by NIH R01 NS049307, the Betsy and Jonathan Blattmachr Family, and NIH Medical Scientist Training Program T32 GM07205.) 1 Ebru Erbayat‐Altay, 1 Kelvin A. Yamada, 1 Michael Wong, and 1 Liu Lin Thio ( 1 Neurology, Washington University, St. Louis, MO ) Rationale: Leptin is a neurohormone that helps regulate body weight and also has complex effects on neuronal excitability. Because its predominant effect on neuronal excitability in vivo is unclear, we examined the susceptibility of leptin deficient ob/ob mice to pentylenetetrazol (PTZ) induced seizures. Methods: We studied PTZ induced seizures in 31 wild type (C57BL/6J) and 30 ob/ob (B6.V‐Lepob/J) male mice at 44–58 days of age using continuous video EEG recordings. After 15 minutes of baseline recording, a 25, 50, 75 or 100 mg/kg dose of PTZ was injected intraperitoneally. The recording continued for 30 minutes after the injection or until death, whichever came first. PTZ induced generalized absence, myoclonic, clonic and clonic‐tonic seizures. Maximal clinical seizure severity, latency to clinical seizure onset, latency to the first epileptiform abnormality and the normalized cumulative epileptiform duration were determined. Clinical seizure severity was rated as: 0 = no seizures, 1 = absence seizures with normal clinical behavior and EEG interictally, 2 = myoclonic seizures with normal clinical behavior and EEG interictally, 3 = generalized clonic or clonic‐tonic seizures with normal clinical behavior and EEG interictally, 4 = repetitive absence or myoclonic seizures without normal interictal behavior or EEG throughout the 30 minute recording, 5 = death. The normalized cumulative epileptiform duration was the sum of the duration of all epileptiform abnormalities divided by the total time of the recording after PTZ injection. Comparisons were made using a t‐test with statistical significance set at p < 0.05. Results: The background EEG for both wild type and ob/ob mice showed 6–7 Hz theta activity without any epileptiform abnormalities. In both groups, PTZ induced the same types of clinical and electrographic seizures in a dose‐dependent manner. The latency to the first clinical seizure and to the first epileptiform abnormality was the same in wild type and ob/ob mice at all PTZ doses. The maximal clinical severity at 25 mg/kg and 100 mg/kg PTZ were the same in wild type and ob/ob mice. However, at 75 mg/kg PTZ the clinical severity was 5.0 ± 0.0 for the ob/ob mice because all mice died whereas it was 3.5 ± 0.8 (p < 0.05) for wild type mice. At 75 mg/kg PTZ, the normalized cumulative epileptiform duration was 9 ± 3% for ob/ob mice and 5 ± 1% for wild type mice (p < 0.05). ob/ob mice also showed a higher maximal clinical severity and normalized cumulative epileptiform duration at 50 mg/kg PTZ. A greater number of generalized clonic and clonic‐tonic seizures accounts for the greater severity of seizures in the ob/ob mice. Conclusions: These results indicate that PTZ induces more severe seizures in ob/ob mice. These results are consistent with leptin acting as an endogenous anticonvulsant and support a role for leptin in the treatment of epilepsy. (Supported by Washington University McDonnell Center for Cellular and Molecular Neurobiology, NS 42744 and JDRF 1–2004–594.) 1,3 Chris J. Feeney, 1,3 Miron Derchansky, 1 Shanthini Mylvaganam, 1 Yana Adamchik, and 1,2,3,4 Peter L. Carlen ( 1 Division of Cellular and Molecular Biology, Toronto Western Research Institute, Toronto, ON, Canada ; 2 Department of Medicine (Neurology), University of Toronto, Toronto, ON, Canada ; 3 Department of Physiology, University of Toronto, Toronto, ON, Canada ; and 4 The Epilepsy Research Program, University of Toronto, Toronto, ON, Canada ) Rationale: Fragile X syndrome is the most common inherited form of mental retardation, affecting 1 in 1200 males and 1 in 2500 females. The syndrome is characterized primarily by intellectual impairment, ranging from mild learning disability to profound mental retardation. We are interested in studying the cellular electrophysiological factors underlying the pathophysiology of seizure activity in the brain tissue of Fragile X mental retardation protein (FMRP) gene knockout mice (KO). Methods: We have characterized the induction of seizure‐like events (SLE) in young adult mouse hippocampal slices exposed to low Mg‐ACSF and tetanic stimulation bouts, as with our previous KO mice. All recording were performed on transverse hippocampal slices (400 μm) from mice aged 6–9 weeks, in oxygenated ACSF, at 33°C. SLEs and extracellular field responses were recorded from the CA1 pyramidal layer of the hippocampal slices, in response (were appropriate) to Schaffer collateral stimulation. Results: The KO mice hippocampal slices were more easily induced into SLEs with either low Mg‐ACSF, or by repetitive tetanizations. KO slices showed a shorter lag time to SLE of 15.4 min. vs 26.5 min. in wt slices, required fewer tetanic stimuli to produce SLEs. A brief application of the mGluR5 antagonist MPEP (at 1 or 10 μM delayed the onset of low Mg‐induced, but not tetanic stimulation‐induced SLEs in KO hippocampal slices by more than 3‐fold. However, we found that longer application of this same drug leads to diminished (with low Mg‐ACSF exposure) or increased susceptibility (with tetanic stimuli) to SLEs in the KO hippocampus. Also, we have found that inhibitory network that regulates paired‐pulse inhibition in CA1 of the hippocampus, is dramatically reduced in the KO hippocampus, leading to a complete lack of PPI that can be elicited in the wt slices. Conversely, we have found that paired‐pulse facilitation is significantly enhanced in the KO hippocampus as compared to the wt, and that this may point to alterations in presynaptic calcium handling in these FMRP knockout mice. Conclusions: We have found that hippocampi from FMRP knockout mice are more easily induced into seizure‐like events in vitro. We are continuing work to further elucidate the nature of mGluR regulation of SLEs in this model, as well as the influence of inhibitory inputs in the KO hippocampus on seizure susceptibility in these mice. (Supported by the Fragile X Research Foundation of Canada (CJF), The Savoy Foundation (MD), and CIHR (PLC).) 1 Neil M. Fournier, 1 Devon R. Andersen, 1 Sumeer A. Mann, 2 Hector J. Caruncho, 3 G. Campbell Teskey, 4 Dan C. McIntyre, and 1 Lisa E. Kalynchuk ( 1 Psychology, University of Saskatchewan, Saskatoon, SK, Canada ; 2 Cell Biology, University of Santiago de Compostela, Santiago, Spain ; 3 Psychology, University of Calgary, Calgary, AB, Canada ; and 4 Psychology, Carleton University, Ottawa, ON, Canada ) Rationale: Reelin is a large secreted glycoprotein that has important roles in the developing nervous system as well as in the adult nervous system. It is believed that extracellular reelin may act as either a stop or as a chemoattracting signal for radially migrating neurons. Deficits in reelin signaling have been found in a variety of neurodevelopment and neurodegenerative disorders, including epilepsy. However, in the adult nervous system, the specific role of reelin is not fully understood. Reelin is synthesized and released by GABAergic interneurons and recent studies have suggested that reelin might modulate and even enhance long‐term potentiation in the adult brain. Importantly, enhancement in synaptic transmission is thought to play a crucial role in the intensification of afterdischarges and progressive development of convulsions in the amygdala kindling model of epilepsy. Methods: To determine if amygdala kindling was associated with changes in reelin expression, rats were surgically implanted with an electrode into the basolateral amygdala and subjected to 99 kindled or sham stimulations. An additional group of non‐kindled genetically prone FAST kindled rats was also processed in order to determine if basal reelin expression was altered in rats that were genetically more susceptible to enhanced amygdaloid excitability and the effects of kindling. Brains were processed for immunohistochemical determination of reelin, NeuN, and doublecortin. Results: The total number of reelin positive cells in the hilus was not substantially different between all groups; however, there was a slight decrease in the number of these cells within the subgranular zone for the kindled and non‐kindled FAST rats. A decrease in reelin expression within the stratum lacunosum‐moleculare was also found for kindled and non‐kindled FAST rats. Doublecortin immunohistochemistry revealed that some newly born neurons migrated ectopically into the hilus for kindled rats but not for sham‐stimulated controls or for non‐kindled FAST rats. Interestingly, we often observed several abnormally located reelin positive cells scattered throughout the dentate granule cell layer for non‐kindled FAST rats suggesting a possible pre‐existing GABAergic circuit malformation in this group of rats. Conclusions: These findings suggest that changes in reelin expression within the hippocampus during development might influence network formation that may influence susceptibility to epilepsy. (Supported by NSERC.) 1 Patrick D. Fox, 1 Hyokwon Chung, and 1 Sookyong Koh ( 1 Neurology, Children's Memorial Hospital, Northwestern Medical School, Chicago, IL ) Rationale: Chronic epilepsy often starts with an isolated early life seizure, a period of remission, and then the re‐emergence of seizures later in life. To mimic this event in rats, we developed a “two‐hit”‐kainic acid (KA) model and demonstrated that an early life seizure increases susceptibility to seizures and to seizure‐induced neuronal injury in adulthood. We have found that activation of resident microglia and subsequent increases in mRNA of inflammatory cytokines and complements appear to be the key initiating events for seizure‐induced inflammatory responses. We hypothesize that the inflammatory reaction provoked by early‐life seizures primes the developing brain and microglia, leading to rapid reactivation by a second seizure in adulthood. Blocking of this microglia activation through pharmaceuticals such as minocycline may prevent some neuronal injury typically seen following these later life seizures. Methods: Postnatal day 21–25 LE rats were injected with KA (10 mg/kg, I.p.) or saline. The first dose of minocycline (10 mg/kg, i.p.) or vehicle was administered immediately after KA injection, and 6 more doses were given 24 hours apart over the next 7 days. Two weeks after KA/PBS injections, all remaining rats were given KA. Rats were perfused two days later and all brain sections were processed for immunocytochemistry (ICC) using IbA1/Aif antibody to visualize activated microglia. CA1 subregions of the hippocampal sections were analyzed using MetaMorph (Universal Imaging Corp.). An average percent of area above the set threshold (threshold area containing immunoreactive cells captured at 20x) and the number of stained cells were calculated per animal. Results: Cellular hypertrophy of microglia suggestive of activated state was observed within 24 h after KA‐induced seizures. The group that experienced early‐life seizures (K/K) had a marked increase in the number of activated microglia compared to control littermates who received single KA in adulthood (S/K). These microglia were larger and formed a complex, web‐like network localized within the hippocampus especially over the CA1 region. Over two fold increase in immunoreactive areas was noted in K/K compared to S/K (KA: 9.0 ± 1.3 vs. Saline: 4.0 ± .1.0, p < 0.04, n = 8). The microglia count was also significantly greater (KA: 233.1 ± 12.1 vs. Saline 153.5 ± 9.2, p < .039, n = 8). This hypertrophic, hyperplastic, reactivating effect of an early‐life seizure on microglia was abolished in minocycline treated rats. Conclusions: Our results indicate that the experience of early life seizures primes the microglia and elicits a much more robust activation in response to later life seizures. Whether microglial reactivation is causally related to the epileptogenic effect of early‐life seizures to increase seizure susceptibility and cell death will be determined. (Supported by Child Neurology Foundation and K02NS048237.) 1 Brita Fritsch, 1 Maciej Gasior, 1 Rafal M. Kaminski, and 1 Michael A. Rogawski ( 1 Epilepsy Research Section, NINDS/NIH, Bethesda, MD ) Rationale: Kainate (KA) receptors are formed from homomeric or heteromeric combinations of subunits, including GluR5 and GluR6. Studies with transgenic mice indicate that KA receptors containing GluR6 subunits play a role in KA‐induced seizures (1). Here we used kainate receptor knockout (KO) mice to determine the clinical seizure and EEG correlates of GluR5 kainate receptor activation, and the roles of kainate receptor subunits in other seizure models. Methods: Video‐EEG monitoring was carried out during slow tail vein infusion of the selective GluR5 agonist ATPA in wild type (WT), GluR5‐/‐, and GluR6‐/‐ mice. We also assessed seizure thresholds using the i.v. PTZ and 6 Hz models. Results: Slight myoclonic jerks/twitches associated with single sharp waves occurred at lower doses of ATPA in WT than in GluR5‐/‐, but generalized seizures occurred at similar doses. Myoclonic jerks occurred at comparable doses in WT and GluR6‐/‐ mice; paradoxically, EEG seizure patterns began with lower doses in GluR6‐/‐ mice. There were no significant differences between any of the genotypes in their thresholds for clonic seizures induced by i.v. PTZ or in the 6 Hz seizure model. Conclusions: GluR5 kainate receptors contribute to ATPA‐induced myoclonic jerks, but not to generalized seizures. The increased sensitivity of GluR6 KO mice to ATPA was unexpected. In the KO animals, the existence of GluR5 homomers instead of GluR5/GluR6 heteromers may account for the phenomenon since GluR5 homomers are about 2.4‐fold more sensitive to ATPA than are GluR5/GluR6 heteromers (2). GluR5 and GluR6 kainate receptors do not appear to play a role in seizure activity in the PTZ and 6 Hz models. 1 Mulle C, Sailer A, Perez‐Otano I, et al. Nature 1998;392:601–605. 2 A. Alt, B. Weiss, A.M. Ogden, et al. Neuropharmacology 2004; 46:793–806. 1 Michael A. Galic, 2 James G. Heida, 3 Campbell G. Teskey, and 2 Quentin J. Pittman ( 1 Neuroscience, University of Calgary, Calgary, AB, Canada ; 2 Neurology, Albert Einstein College of Medicine, Bronx, NY ; and 3 Psychology, University of Calgary, Calgary, AB, Canada ) Rationale: There is increasing evidence that immune and inflammatory processes in early life are capable of producing lasting effects on physiology and CNS function and raise the possibility that a single inflammatory episode during development could evoke changes in seizure susceptibility later in adulthood. Methods: To explore this hypothesis, postnatal day 14 male rats received either the bacterial endotoxin lipopolysaccharide (LPS) (25–250 μg/kg), or the viral mimetic Polyinosinic:Polycitydylic acid (POLY I:C) (1 mg/kg) to induce a mild inflammatory process. Two months later, rats received lithium and pilocarpine (LI‐PILO), kainic acid (KA) or were subjected to amygdala kindling to discern any changes to seizure susceptibility. Following KA seizures, subsets of animals were killed and brains evaluated for neurodegeneration using Fluoro jade histochemistry. Results: Neonatally LPS‐treated, but not POLY I:C‐treated rats, showed significantly faster seizure onset times (SOTs), by about 25% compared to controls. Adult rats given LPS and then LI‐PILO two months later, did not show the same reduction in SOT as found with the neonatal treatment. Separate experiments determined a comparable susceptibility of neonatally LPS‐treated rats to adult convulsions evoked with KA, but not to amygdala kindling. Behavioural sensitivities to KA were also reflected in the amount of neuronal degeneration in the CA3 and CA1 hippocampus 24 h after SOT. Conclusions: Taken together, we have observed that long‐lasting changes in seizure susceptibility can be evoked following a mild inflammatory episode during development, possibly through a mechanism involving, or specific to, the Toll‐like receptor (TLR)‐4 (LPS‐activated), but not TLR‐3 (POLY I:C‐activated) pathways. (Supported by Canadian Institutes of Health Research (CIHR), the Natural Sciences and Engineering Council of Canada (NSERC), and the Alberta Heritage Foundation for Medical Research (AHFMR).) 1,2 Levi B. Good, 3 Shivkumar Sabesan, 2 Trevor D. Boone, 2 Leon D. Iasemidis, and 1,2 David M. Treiman ( 1 Neurology Research, Barrow Neurological Institute, Phoenix, AZ ; 2 Harrington Department of Bioengineering, Arizona State University ; and 3 Department of Electrical Engineering, Arizona State University, Tempe, AZ ) Rationale: In the last decade, substantial progress has been made in the study of the human epileptic brain by utilizing concepts and measures from nonlinear dynamics. The hallmark of this research is the ability to predict seizures prior to their clinical or electrographic onset (IEEE TBME 2003; 50:616–627). As the ability to predict leads to the possibility of control, research in controlling seizures with closed‐loop systems is expected to flourish in the near future and will most likely include studies performed in animal models of epilepsy. We have thus applied the concepts from nonlinear dynamics, namely the Largest Short‐Term Lyapunov exponent (STLmax), to evaluate the effectiveness for real‐time seizure prediction in the lithium pilocarpine rat model of chronic epilepsy as a precursor for a closed‐loop seizure control system that utilizes deep brain stimulation. Methods: Three male Sprague‐Dawley rats (300–350g) were stereotaxcially implanted with a customized array of Tungsten depth wire electrodes which included four cortical, two hippocampal, and two thalamic contacts. Chronic epilepsy was established several weeks after an episode of prolonged status epilepticus using the lithium pilocarpine model (3 mmol/kg LiCl followed by 30 mg/kg pilocarpine 24 hrs later). Continuous EEG/video recordings were made during the entire experiment and all data analyzed in real‐time. STLmax values were calculated for each electrode and entrainment of STLmax was evaluated utilizing a pair t‐statistic (T‐index). True and false predictions were noted with a seizure prediction horizon set at 180 minutes. Results: Continuous EEG data from Rat1 (55 seizures in 166 hours), Rat2 (25 seizures in 384 hours), Rat3 (16 seizures in 739 hours) resulted in a total of 96 seizures in 1289 hours of recording. The seizure prediction results from each rat were: Rat1, prediction rate 35/55 = 63.4%, 11 false positives (0.0663/hr), mean prediction time = 111 minutes; Rat2, prediction rate 22/25 = 88.0%, 50 false positives (0.1302/hr), mean prediction time = 149 minutes; Rat3, prediction rate 10/16 = 62.5%, 114 false positives (0.1543/hr), mean prediction time = 114 minutes. Overall prediction rate‐ 67/96 = 69.8%, 175 false positives (0.1358/hr), mean prediction time = 125 minutes. Conclusions: While the overall prediction rate of ∼70% is less than the one reported in human data (∼80%), similar false positive rates (∼1 every 8 hours of recording), and an extended prediction time were noted. Given these results, prediction of seizures in the lithium pilocarpine rat model of chronic epilepsy appears feasible as a model of chronic epilepsy for testing real‐time feedback control systems utilizing thalamic deep brain stimulation. (Supported by Epilepsy Research Foundation of America and Ali Paris Fund for LKS Research, and Barrow Neurological Foundation.) 1 Heidi L. Grabenstatter, 2 Suzanne Clark, and 3 F.E. Dudek ( 1 Biomedical Sciences, Colorado State University, Fort Collins, CO ; 2 School of Pharmacy, University of Wyoming, Laramie, WY ; and 3 Physiology, University of Utah, Salt Lake City, UT ) Rationale: Animal models of chronic epilepsy, such as the kainate‐treated rat, may be useful in the discovery of new antiepileptic drugs (AEDs). The present study evaluated the effectiveness of carbamazepine on the frequency of spontaneous motor seizures in rats with kainate‐induced epilepsy. Another aim of the study was to of develop a paradigm for continuous, long‐term oral administration of potential AEDs. Methods: Single intraperitoneal (IP) injections of carbamazepine (10–100 mg/kg) were compared to vehicle (i.e., 20% (2‐Hydroxypropyl)‐β‐cyclodextrin) injections via six AED‐versus‐vehicle tests using a repeated‐measures cross‐over protocol. To establish an effective dose and the time‐course of the anticonvulsant effect, two trials (8–9 rats per trial) evaluated single per os(i.e., oral‐ by mouth (PO)) doses of 30 mg/kg and 100 mg/kg carbamazepine. Each PO trial included 3 AED‐versus‐vehicle tests comprised of carbamazepine‐containing or control pellet feedings on alternate days with a recovery day between each treatment day. Based on the single‐feeding studies, carbamazepine at 100 mg/kg was administered three times per day (TID) for 5 days. Results: Carbamazepine significantly reduced seizure frequency at 10, 30, and 100 mg/kg after single IP injections, and caused complete seizure cessation during the 6‐h post‐drug epoch in 25% and 70% of animals at 30 and 100 mg/kg. Single IP injections and single oral feedings of 30 mg/kg and 100 mg/kg carbamazepine significantly (p < 0.0001) reduced seizure frequency relative to control treatments during 6‐h and 22‐h epochs following drug administration. A single feeding of carbamazepine (100 mg/kg PO) was significantly effective for 20 h and full recovery was complete within 23 h. Continuously administered carbamazepine (100 mg/kg PO, TID) significantly reduced seizure frequency by >50% for 24 h and completely suppressed motor seizures in 50% of the animals tested. Conclusions: High doses of carbamazepine (30 and 100 mg/kg) were highly effective at suppressing spontaneous seizures, and were associated with a high percentage of motor seizure freedom. Administration of higher doses of carbamazepine did not further suppress spontaneous motor seizures. Oral administration of 30 mg/kg and 100 mg/kg carbamazepine was as effective as IP injections. Significant anticonvulsant effects of 100‐ mg/kg PO carbamazepine were sustained for the duration of the 5‐day trial. (Supported by NIH (HLG) and (FED).) 1 Jan Grashof, 1 Julika Pitsch, 1 Sandra Bauerkämper, 1 Albert Becker, and 1 Susanne Schoch ( 1 Neuropathology, University of Bonn Medical Center, Bonn, NRW, Germany ) Rationale: The pathogenesis of temporal lobe epilepsy (TLE) is characterized by activity‐dependent changes in synaptic plasticity. Certain forms of synaptic plasticity have been shown to depend on modifications of the presynaptic release machinery. However, our knowledge about the underlying molecular mechanisms and the role of the presynaptic release machinery in functional changes that underlie epileptic activity at the level of the synapse is very limited. Therefore, we have analyzed whether the expression patterns of components of the presynaptic release machinery are altered in different stages of an experimental model of temporal lobe epilepsy. We have furthermore characterized knock‐out mice deficient for three proteins that are integral components of the release machinery, RIM1α, RIM2α or Synaptotagmin 10 (Syt 10), in an experimental model of TLE. Methods: Status epilepticus (SE) was induced by systemic application of pilocarpine to rats as well as to three mutant mouse lines. Changes in mRNA levels were analyzed by Real‐time PCR and in situ hybridization (ISH). Seizure susceptibility was determined after induction of SE. The frequency and severity of chronic recurrent seizures was analyzed with a telemetric EEG system (DSI) and parallel video analysis. To study hippocampal damage, the amount of segmental hippocampal loss of neurons and gliosis were analyzed in wild‐type and knock‐out mice after SE. Results: To investigate changes in the mRNA expression level of various components of the presynaptic release machinery hippocampal subregions (dentate gyrus (DG), CA3 and CA1) were isolated from rats 4, 6 and 8 hours and 5 and 28 days after SE induced by pilocarpine. Analysis of these samples by Real‐time PCR, ISH and Affymetrix‐Micro‐Arrays showed distinct changes in the expression patterns of several constituents of the release machinery; e.g. Syt 10 could not be detected in the CA1 region of control rats/mice, neither by single cell PCR nor by ISH, however, its mRNA was strongly expressed in CA1 cells 6 h after SE. To study if ablation of individual components of the presynaptic release machinery affects epileptogenesis we are analyzing three mutant mouse lines in the pilocarpine model of TLE: knock‐out mice for the active zone proteins RIM1α and RIM2α, which exhibit defects in short‐ and long‐term synaptic plasticity, and Syt 10, whose mRNA expression is induced by seizures. So far the results obtained by Video‐ and EEG‐Monitoring of these animals suggest that RIM1α knock‐out mice have a higher susceptibility to seizures than wild‐type littermates (n = 5). Furthermore, the latency to the first seizue and to SE seem to be reduced. Conclusions: Our results suggest that components of the presynaptic release machinery are modified during epileptogenesis, may play a pathogenetic role and constitute targets for new therapeutic approaches in TLE. (Supported by DFG (SFB TR3, Emmy‐Noether‐Program), BONFOR.) 1 Alberto Lazarowski, 2 Veronica Trida, 3 Christian Höcht, 2 Nelida Gonzalez, 3 Javier Opezzo, 3 Guillermo Bramuglia, 3 Carlos Taira, and 2 Elena Girardi ( 1 Clinical Biochemistry, Faculty of Pharmacy and Biochemistry, UBA., Buenos Aires, Argentina ; 2 Celular Biology Institute, University of Buenos Aires, Buenos Aires, Argentina ; and 3 Pharmacology, Faculty of Pharmacy and Biochemistry, UBA., Buenos Aires, Argentina ) Rationale: The present work addressed possible participation of the efflux transporter P‐glycoprotein (P‐gp) in the alterations of the central pharmacokinetics of phenytoin in an experimental model of epilepsy induced by the administration of mercaptopropionic acid. Methods: Male Wistar rats were administered for 7 days with physiological solution (SF) or mercaptopropionic (MP) acid. The day of the experiment, a shunt micrordialysis probe or a concentric probe was inserted into the carotid artery or the hippocampus respectively, in order to monitor extracellular phenytoin (PHT) levels. In SF and MP treated rats, PHT (30 mg/kg, i.v.) was administered 30 after an intraperitoneal administration of SF or nimodipine (NIMO, 2 mg/kg). Pharmacokinetic analysis of the data was made using TOPFIT program. Results: No differences were found in PHT plasmatic levels comparing all experimental groups. In rats pretreated with SF, hipocampal PHT concentrations were signicantly lower in MP treated rats (maximal concentration (Cmax): 2.7 ± 0.3 μg/ml, p < 0.05 vs SF rats) compared to SF animals (Cmax: 5.3 ± 0.9 μg/ml). Whilst pretreatment with NIMO did not modify central pharmacokinetics of PHT in SF treated rats (4.5 ± 0.8 μg/ml), PHT levels were significantly higher in MP rats pretreated with NIMO (Cmax: 6.8 ± 1.0 μg/ml, p < 0.05 vs MP rats pretreated with SF) compared to MP animals with previous administration of SF. Conclusions: Thus, our results suggest that central pharmacokinetics of PHT was altered in epileptic rats by MP administration. The effect of NIMO on hipocampal concentrations of PHT suggested that P‐gp is implicated in the reduced central bioavailability of PHT in MP epileptic rats. (Supported by University of Buenos Aires.) 1 Susan L. Campbell, and 1 John J. Hablitz ( 1 Department of Neurobiology, University of Alabama at Birmingham, Birmingham, AL ) Rationale: Elevated glutamate levels have been reported in several epilepsy models. The electrophysiological changes that occur when transporter function is disrupted in experimental models of epilepsy have not been examined. The freeze‐induced model of cortical dysplasia results in abnormal lamination of cortical structures associated with hyperexcitability in the region adjacent to the microsulcus. Here we used this model to test the role of glutamate transporters in synaptic transmission in the hyperexcitable region. Methods: Focal freeze lesions were induced in postnatal day (PN) 0 Sprague‐Dawley rats using methods described previously. In freeze‐lesioned cortex, recordings of layer II/III pyramidal cells were obtained 0.3–1.5 mm lateral to the microsulcus. Synaptic responses were evoked with a bipolar stimulating electrode (twisted pair of 25 mm Formvar insulated nichrome wires) positioned 150–200 μm below the recording pipette. Stimuli were current pulses 10–100 μA in amplitude and 50–100 ms in duration. A stimulation frequency of 0.05 Hz was used. The glutamate uptake antagonists DL‐threo‐β‐benzylozyaspartic acid (TBOA) and dihydokainate (DHK) were bath applied. Results: TBOA prolonged postsynaptic currents (PSCs) and decreased the threshold for evoking spontaneous epileptiform activity in lesioned cortex. Uptake inhibition also induced an increase in holding current in the dysplastic cortex. In contrast, responses in sham operated and control animals were not affected. The effect of blocking uptake was mediated primarily by the glia glutamate transporter (GLT‐1) since the selective GLT‐1 blocker DHK mimicked the effects of TBOA. The increases in the holding current suggested that glutamate was accumulating extracellularly and activating ionotropic glutamate receptors. The increase in excitability following uptake inhibition appeared to be mediated by N‐methyl‐D‐aspartate (NMDA) receptors since prior application of D‐(‐)‐2‐amino‐5‐phosphonovaleric acid (APV) prevented TBOA‐induced effects and TBOA‐induced changes were reversed by APV. In the freeze‐lesioned cortex spontaneous epileptiform activity does not occur at a high rate. Under control conditions, the probability for the occurrence of spontaneous discharge was 0.09 ± 0.06 (n = 22) which increased significantly to 0.59 ± 0.1 (n = 22) when TBOA was applied. Conclusions: These results indicate that glutamate transporter function is altered in rat focal cortical dysplasia and these changes contribute to the observed hyperexcitability. Alteration in transporter function could result in increasing extracellular glutamate concentrations and over activation of NMDA receptors during neuronal activity. (Supported by NS22373.) 1 Kiyotaka Hashizume, 1 Seiji Takebayashi, and 1 Tatsuya Tanaka ( 1 Neurosurgery, Asahikawa Medical College, Asahikawa, Hokkaido, Japan ) Rationale: The aim of this study was a comparison of the antiepileptic effect between deep brain stimulation (DBS) of the subthalamic nucleus (STN) and of the anterior thalamic nucleus (ATN), using the same epilepsy model. Methods: A guide cannula was stereotactically implanted into the left sensorimotor cortex of male Wistar rats. Focal cortical seizures were induced by a microinjection of kainic acid (KA) into the left sensorimotor cortex. Using this focal seizure model, High‐frequency electrical stimulation of STN or ATN was performed intermittently when the animal developed into the seizure status. The seizure frequency of DBS‐on and DBS‐off was measured from a video‐EEG system. The data was analyzed statistically. Results: An intracortical injection of KA was induced focal and secondarily generalized seizures in all animals. The ipsilateral (the side of KA injection) STN‐DBS suppressed the seizure frequency to 50–60% of the DBS‐off phase. The effect of bilateral STN‐DBS showed no significance than the ipsilateral DBS. When the electrode was implanted into bilateral ATN, no seizure was induced by KA injection. The implantation of the ipsilateral ATN lead to seizure suppression, and the ipsilateral ATN‐DBS suppressed seizures more strongly. Conclusions: Both DBS of STN and ATN suppressed KA‐induced focal seizures. The antiepileptic effect of ATN‐DBS was stronger than STN‐DBS in this animal model. The seizure suppression of the electrode implantation alone may be caused by a focal destruction of ATN. 1 Henry J. Hasson, and 1,2 Solomon L. Moshe ( 1 Neurology, Albert Einstein College of Medicine, Bronx, NY ; and 2 Neuroscience and Pediatrics, Albert Einstein College of Medicine, Bronx, NY ) Rationale: Status epilepticus (SE) is a neurological emergency and a common problem in neonates. To date, effective treatments to stop ongoing status epilepticus in this age group have not been identified. To assess efficacy and morbidity from the treatment of prolonged seizures in the immature brain, it is necessary to develop screening assays. Methods: SE was induced in postnatal day 15(P15) and 21(P21) rats by systemic administration of kainic acid (KA) 3.5 mg/kg (P15) and 10 mg/kg (P21) or lithium‐pilocarpine (PI) 60 mg/kg (P15) and 30 mg/kg (P30). Rats were monitored by EEG with an electrode in the CA1 region of the hippocampus. In both models behavioral seizures continue for several hours and electrographic seizures continue for more than 24 hours if left untreated. After one hour of electroclinical SE, rats were treated with pentobarbital (PB), diazepam (DZ), or saline. Rats were then intermittently monitored for cessation of behavioral and electrographic seizures, and for mortality for 3 days following treatment to confirm that the seizures stopped and did not recur. Results: Both drugs can stop the ongoing seizures both behaviorally and electrographically. PB (50 mg/kg) was XX% effective in stopping KA and PI induced SE in both P15 and P21 rats. DZ (20 mg/kg) was effective in stopping SE in 50–64% of rats in 3 of the 4 groups. In the KA P21 group, it was not effective. Higher doses of diazepam did not improve the outcome in P15 pups in KA‐SE. When effective, DZ stopped SE much quicker than PB (p < .0001). In most groups, mortality from SE did not change irrespective of whether SE was stopped or not. In the P15 PI group, treatment with either drug had increased mortality compared to untreated SE. Conclusions: This study shows that it is possible to abort seizures in the developing rat even after 1 hour of SE using drugs at appropriate doses. The drugs, when effective, stop both the behavioral and the EEG seizures. The action of DZ is rapid, suggesting that subsequent treatments can be quickly initiated if SE does not stop within 1–2 minutes. On the other hand, PB takes significantly longer to work, and more time is needed to determine if the therapy will stop the seizures. Age and cause of SE may contribute to drug‐induced morbidity. Our protocol can be used to assess the efficacy and morbidity of new agents. (Supported by NIH NINDS grants K12‐NS048856 (NSADA) (HH) and R01‐NS20253 (SLM).) 1 Seiji Hatazaki, 1 Carmen Beller‐Estelles, and 1 David C. Henshall ( 1 Physiology, Royal College of Surgeon In Ireland, Dublin, Dublin, Ireland ) Rationale: Sublethal brain insults can confer protection against cell death induced by subsequent (prolonged) damaging insults. While best described for ischemia, there is some evidence brief seizures may precondition the brain against damage following prolonged seizures (e.g. status epilepticus). To study this so‐called “epileptic tolerance”, we examined neuronal damage following status epilepticus in mice that received various doses of sublethal preconditioning seizures and at various intervals between the conditioning and injurious stimuli. Methods: Seizure preconditioning activity was induced by intra‐peritoneal (i.p.) injection of kainic acid (KA) at various doses (5.0, 7.5 and 15.0 mg/kg; n = 5, plus vehicle controls, n = 9) and subsequent status epilepticus was induced by intra‐amygdala (i.a.) injection of KA (1.0 ug) at two time‐points (24 or 72 hours) after the first insult. Terminal deoxynucleotidyl transferase‐mediated dUTP nick end labeling (TUNEL) and NeuN immunostaining was used to detect neuronal damage within the hippocampus. Results: Preconditioning seizures did not induce significant hippocampal damage at any dose tested. Mice that received preconditioning seizures induced by KA i.p. at 15.0 mg/kg given 24 hours before i.a. KA showed significant reductions in TUNEL counts and significantly more surviving neurons as assessed by NeuN in the ipsilateral hippocampus, particularly the CA3 subfield (p < 0.01). There were no significant differences in the duration of high amplitude polyspike seizures evoked by i.a. KA in mice that received preconditioning. Surprisingly, increasing the time interval between preconditioning and prolonged seizures to 3 days resulted in raised mortality rates (e.g. 3/5) when mice underwent the second prolonged (i.a. KA) seizures: evidence of preconditioning‐induced sensitization to seizures. Conclusions: This study provides a new model with which to investigate the molecular mechanisms underlying endogenous neuroprotection programmes in brain. (Supported by Science Foundation Ireland grant B466.) 1 James G. Heida, 1 Jana Veliskova, 1 Libor Velisek, 1 Ondrej Cudomel, 1 Solomon L. Moshe, and 1 Aristea S. Galanopoulou ( 1 Neurology and Neuroscience, Albert Einstein College of Medicine, Bronx, NY ) Rationale: It is widely accepted that status epilepticus (SE) in early life does not result in major neuronal death as observed in adults. There is increasing evidence however that more subtle aspects of brain physiology are permanently affected. Of particular interest is the potential effect of repetitive early life seizures on the physiology of endogenous seizure controlling circuits, such as those involving the substantia nigra reticulata (SNR). The aim of the current study was to determine the effect of 3 episodes of SE, during early postnatal development, on the seizure controlling function of the SNR in rats. Methods: Three episodes of SE were induced in Sprague Dawely rat pups of both sexes on postnatal days 4 (P4), P5, and P6 by systemic injection of kainic acid ((KA) 1.25, 1.5, and 2 mg/kg i.p. respectively). Animals were left until P28 when they were implanted with bilateral cannulae aimed at the anterior SNR. At P30, rats received bilateral infusions of the GABAA receptor agonist muscimol (100ng/0.25 μl per side), which has been shown previously to have an anticonvulsant effect when infused into the anterior SNR at P30 (Veliskova & Moshé 2001). Control rats were infused with 0.25μl/SNR of saline. Flurothyl seizure threshold for the first clonic seizure was determined 30 minutes later as the amount of flurothyl necessary to induce the seizure. Subsequently, cannulae placements were confirmed histologically. Only results from animals with cannulae placement in the anterior SNR are reported. Results: Systemic injection of KA resulted in behavioral SE in all animals. Within 5 min of the injections, pups manifested scratching‐like behavior and hindlimb clonic movements, hyperactivity, tonic or tonic clonic seizures followed by swimming‐like movements intermixed with tonic or tonic clonic seizures. These seizures lasted for the entire observation period of 6 hrs. At P30 flurothyl thresholds were 179 ± 15.7 μl (saline n = 4) and 169 ± 10.9μl (muscimol n = 7) in males, and 174 ± 25.9μl (saline n = 5) and 181 ± 15.7μl (muscimol n = 8) in females. There were no statistically significant differences between any of the groups (two way ANOVA, p > 0.05, NS). Conclusions: Repetitive episodes of SE during early postnatal development alter the GABAA receptor mediated seizure controlling function of the SNR. The results suggest that early life SE produces long‐term alterations in the SNR GABAA receptor system, which leads to loss of the well‐known anticonvulsant effects of GABAA receptor activation in the SNR of rats without prior seizures. Such alterations in seizure controlling circuits may predispose the brain to future seizures and epileptogenesis. (Supported by NIH NINDS grants NS 20253, NS 045243 and a grant from the Rett Syndrome Research Foundation.) 1 William Gomes, 2 Fred Lado, 3 Nihal De Lanerolle, and 1 Hoby Hetherington ( 1 Radiology, Albert Einstein College of Medicine, Bronx, NY ; 2 Neurology, Albert Einstein College of Medicine, Bronx, NY ; and 3 Neurosurgery, Yale University, New Haven, CT ) Rationale: Reductions in N‐acetyl aspartate (NAA) have proven useful for the lateralization and localization of seizure foci in patients with intractable epilepsy. Since most patients to date have been studied only after years of intractable seizures, it is unclear, if the deficits in NAA precede the onset of overt seizures or, alternatively, are solely a consequence of chronic, intractable seizures. To determine when the reductions in NAA occur, we used a pilocarpine rat model and magnetic resonance spectroscopic imaging (MRSI). Methods: Status epilepticus (SE) was induced in male rats with pilocarpine and terminated after 1 hour with diazepam. Control rats received sham‐injections of saline (n = 10 control rats). The severity of the SE was graded according to Racine, and rats with severe generalized seizures (stages 3, 4 and 5) were included for analysis (n = 11). Two and seven days post‐SE, MRSI data were acquired with 2ul resolution. Three voxels within each hippocampus (6 voxels per rat) were selected for analysis. Following the day 7 MRSI study, the rats were maintained under anesthesia and transcardially perfused with 4% paraformaldehyde. Neuronal loss in the regions spanned by the spectroscopic images was assessed using stereologic methods as described by West. Results: Displayed in Fig. 1 are hippocampal spectra from a sham injected animal (control) and 2 and 7 days post‐SE in a pilocarpine treated rat. NAA is significantly reduced at 2 days post‐SE (27.5 ± 6.9% decrease, p < 0.001) and 7 days post‐SE (17.3 ± 6.9% decrease, p < 0.001) in comparison to sham‐injected controls (Fig. 1). Despite the substantial reductions in NAA 7 days post SE, neuronal counts averaged over the entire hippocampus showed a non‐significant reduction of 2% in comparison to sham treated animals. Conclusions: Hippocampal NAA content in this rat model of epilepsy is significantly reduced during the latent period prior to the development of overt spontaneous seizures. The reduction in NAA, 17%, is not due to neuronal loss, since neuronal numbers are not significantly decreased over the volume sampled. The presence of decreased NAA during the latent period suggests that spectroscopic studies of NAA provide an early marker of the processes underlying the development of epilepsy, prior to the manifestation of overt seizures. If validated in patients, this would provide a non‐invasive method for evaluating and identifying those patients at risk to develop epilepsy following a first provoked seizure. (figure 1) (Supported by National Institutes of Health R21 EB‐001748.) 1 Lan B. Hoang‐Minh, 1 Hector Sepulveda, 2 Mansi Parekh, 3 Angela Hadlock, 3 Wendy Norman, 3 Justin C. Sanchez, 1 William L. Ditto, 3 Michael A. King, 3 Paul R. Carney, 5 Zhao Liu, and 4 Thomas H. Mareci ( 1 Biomedical Engineering, University of Florida, Gainesville, FL ; 2 Neuroscience, University of Florida, Gainesville, FL ; 3 Pediatrics, University of Florida, Gainesville, FL ; 4 Biochemistry and Molecular Biology, University of Florida, Gainesville, FL ; and 5 Neurology, University of Florida, Gainesville, FL ) Rationale: Using MRI at 17.6 Tesla, we have studied structural changes that occur during the latent period of epileptogenesis in an animal model of mesial temporal lobe epilepsy (MTLE) and correlated these changes with histological analysis. Methods: Five 50‐day‐old adult Sprague‐Dawley male rats were anesthesized and implanted with stimulating electrodes to induce self‐sustained status epilepticus in 4 out of 5 rats. Subsequently, rats were video monitored to capture spontaneous seizures. After eight weeks, the animals were sacrificed, transcardially perfused, and the intact fixed brains imaged in vitro at 17.6 Tesla. After imaging, brains were processed for histology and stained with Fluoro‐Jade C, Timm, GFAP, Black Gold and Perl iron stains to correlate structural changes with those observed with MRI. Results: Three of 5 rats presented with observable spontaneous seizures during the epileptogenic period. These rats showed significantly decreased T2 in the pyramidal cell layer CA1 of the hippocampus and the dorsal thalamic nuclei bilaterally (Fig. 1). One presented with a cavity and increased T2 in the parahippocampal gyrus on the contralateral side of stimulation. No T2 changes were observed in the frontal cortices of these seizing rats. The two rats with no observable seizures either hadn't been stimulated or had stimulating electrode placement in the dorsal thalamus instead of the hippocampus. These rats showed no significant T2 abnormalities in any brain region. Histological analysis has been completed for one of the seizing rats and revealed mossy fiber sprouting around the dentate gyrus bilaterally, iron deposition in the dorsal thalamus and demyelination and astrocytic proliferation in the CA1 and CA3 hippocampal regions. Conclusions: These results suggest that there may be a correlation between T2 changes in the parahippocampal gyrus and the hippocampus (related to mossy fiber sprouting and gliosis on histology) and the seizure activity of animal models of MTLE. Future work will correlate these changes with in vivo MRI at different time points after electrical stimulation and structural changes using specific histological methods. (figure 1) (Supported by NIH grant R01 EB004752, the Wilder Epilepsy Research Center and University of Florida Alumni Foundation.) 1 Riikka J. Immonen, 1 Asla Pitkänen, and 1 Olli H.J. Gröhn ( 1 Department of Neurobiology, A.I.Virtanen Institute for Molecular Sciences, University of Kuopio, Kuopio, Finland ) Rationale: Mn2+‐enhanced magnetic resonance imaging (MEMRI) can reveal functional, structural, and connectional alterations in the brain with a high spatial resolution in vivo. We showed that Mn2+ injected in the entorhinal cortex is transported via perforant pathway to the dentate granule cells resulting in labeling that correlates with the density of mossy fiber sprouting (Nairismägi et al. Neuroimage 2006). Now we hypothesized that systemic Mn injection and MEMRI can be used for detection of mossy fiber sprouting. Methods: MnCl2 (45 mg/kg) was injected intraperitoneally to 10 adult Wistar rats. After 12 hours, status epilepticus (SE) was induced with kainic acid (KA) in 6 rats (MnSE group). T1 weighted (gradient echo, TE = 2.7ms, TR = 120ms) and quantitative T1 mapping (IR‐ fast spin echo, TI = 10, 400, 1000, 1600 ms) MRI was performed under 1% halothane anaesthesia at 4.7T, 3 and 25 hours after KA injection. Two months later, follow‐up MRI was performed before and 24 hours after injection of Mn. In the end, rats were perfused for histology, and brains were stained to detect neurodegeneration, gliosis, and mossy fiber sprouting. Results: At 15 h after KA injection, T1 relaxation times did not differ between the groups. This indicates that Mn accumulation during generalized seizures is not activity dependent. Two months after KA administration, Mn‐enhanced MRI indicated thickening of the dentate gyrus (DG) and proximal CA3 in T1‐weighted images in the MnSE group (260 ± 40% increase compared to controls, p < 0.05). The CA1/adjacent corpus callosum signal intensity was 1.37 ± 0.03 in the MnSE and 1.22 ± 0.03 in the control group (Fig. 1, p < 0.05), confirming the increased accumulation of Mn in the CA1 region of the KA‐injected animals. The initial histological analysis shows both mossy fiber sprouting (DG) and neuronal loss related gliosis (CA1) in the signal enhancement areas. (figure 1)Fig.1 Manganase enhanced T1‐wt images 25h (A,C) and 2 months after KA injection from two representative animals. Conclusions: MEMRI contrast after systemic Mn injection is not associated with brain activation during and after SE. Alterations in the thickness and shape of the DG and CA3 or CA1 regions of hippocampus proper can be attributed either to mossy fiber sprouting or gliosis, respectively. According to our data it is evident that MEMRI after systemic administration is able to reveal cellular level changes in the hippocampus. It remains to be studied if the accumulation of Mn into mossy fibers and/or glia can be differentiated. (Supported by Finnish Cultural Foundation of Northern Savo.) 1 John R. Ives, 2 Krista L. Gilby, and 2 Dan C. McIntyre ( 1 Department of Neuroscience, University of Western Ontario, London, ON, Canada ; and 2 Department of Psychology, Carlton University, Ottawa, ON, Canada ) Rationale: We tested the effects of TMS using a model MagPro X100 (Medtronic Canada Inc., Toronto, ON) on well‐established kindled seizures in rats as there seems to be no studies in this area. First we determined whether 60 Hz TMS could evoked seizures in kindled rats. Second, we determined whether low frequency (1 Hz) TMS (LFTMS) had any effect on the threshold or performance of kindled seizures. Methods: Since kindling was originally established in 21 rats using 2 sec of low intensity 60 Hz sinusoidal electrical stimulation, we set the TMS parameters to 2 sec of 60 Hz, biphasic, magnetic stimulation at 100%. The 21 rats, 7 amygdala and 14 ventral hippocampal kindled, previously received more than 10 generalized seizures and their thresholds were stabile. These 21 rats were also from 3 selectively bred groups, 6 “Fast kindler”, 5 “Slow kinders “ and 10 “F1 hybrids”. Two types of TMS coils were used to stimulate the rats, the round and the figure 8. The round coil had a central‐hole that was positioned over the rats' implanted headplug connector, when the kindling leads to the rat were not attached. The figure 8 coil was used when the leads were attached. In 12 ventral hippocampal kindled rats, we tested the efficacy of LFTMS against the kindled seizures in two experiments, A) 1 Hz, 70%, for 30sec, wait 1min and determine the seizure threshold and behavioral profile, and B) 1 Hz, 70%, for 5min, wait 24hrs before testing. Results: No TMS protocol alone triggered a convulsive seizure or focal after‐discharge in any rat. Further the focal after‐discharge threshold to the kindling stimulus was not affected by LFTMS. However, the latency to forelimb clonus of the kindled seizure was significantly lengthened by both LFTMS protocols, indicating a delay in the recruitment of the convulsive kindled seizure. Surprisingly, TMS did not block or even marginally obscure the online EEG from the implanted bipolar Nichrome electrodes. Conclusions: TMS alone does not trigger seizures in seizure‐prone or ‐resistant rats with a well‐established kindled focus in either the amygdala or hippocampus. 1 Hz TMS does not effect the AD threshold, but significantly affects the recruitment of the convulsive seizure, which is an anti‐epileptic effect. (Supported by Medtronic of Canada for loan of TMS equipment.) 1 Kenneth A. Jenrow, 2 Tavarekere N. Nagaraja, 3 Robert A. Knight, 1 Nancy W. Lemke, 2 Joseph D. Fenstermacher, and 1 Kost V. Elisevich ( 1 Neurosurgery, Henry Ford Health System, Detroit, MI ; 2 Anesthesiology, Henry Ford Health System, Detroit, MI ; and 3 Neurology, Henry Ford Health System, Detroit, MI ) Rationale: Epileptogenesis precipitated by status epilepticus (SE) is believed to evolve during a latent period which intervenes between SE and the emergence of spontaneous seizures. Vascular pathologies have been observed in several limbic structures in chronically epileptic animals following SE, including upregulated vascular endothelial growth factor (VEGF) among reactive astrocytes, angiogenesis, and increases in neovascular permeability. Here we have investigated the temporal evolution of these vascular pathologies following SE in the rat. Methods: SE was induced in male wistar rats (250 g – 350 g) by systemic kainic acid injection (10 mg/kg, i.v.), and was terminated four hours after onset by injection of pentobarbital (20 mg/kg, i.p.). The control group received vehicle injections and did not develop SE. Bromodeoxyuridine (BrdU) injections (50 mg/kg, i.p.) were administered daily for seven consecutive days to separate cohorts beginning 1, 7, 14, 21, 28, 35, 42, or 49 days post‐SE. Animals in each cohort were sacrificed 14 days after completing the BrdU injection series, and angiogenesis was subsequently assayed by double‐label immunohistochemistry for BrdU and von Willebrand factor. Prior to sacrifice, vascular permeability was assayed serially at 7, 14, 21, 24 and 56 days post‐SE, using gadolinium‐diethylenetriaminepentaacetic acid (Gd‐DTPA)‐enhanced magnetic resonance imaging (MRI). Gd‐DTPA was administered using a stepped down infusion protocol that maintained a relatively constant blood Gd‐DTPA level during 20 min of imaging. Pre‐contrast T1‐weighted images (T1WI) were subtracted from post‐contrast T1WI to identify enhancing areas. Results: Endothelial proliferation was significantly increased (p < 0.05) during the first two weeks (day 1 and day 7 cohorts) post‐SE, spanning the duration of the average latent period. At sacrifice, coincidently labeled cells were commonly found in abnormal neovessels within regions damaged by SE. MRI image subtraction indicated that vascular permeability increased toward the end of the latent period, coinciding with the formation of abnormal neovascular complexes within regions of epileptogenesis. Conclusions: The evolution of vascular pathologies following kainic acid‐induced SE occurred primarily during the latent period which preceded the onset of spontaneous seizures. These were observed exclusively in regions of the brain damaged by SE and consist of abnormally permeable neovascular complexes formed via angiogenesis. The timing of events in relation to SE suggests that the resultant increase in vascular permeability may bias the adjoining neuropil toward hyperexcitability and promote the emergence of spontaneous seizures. (Supported by Henry Ford Health Sciences Center: Mentored Scientist Grant #A10222) 1 Nigel C. Jones, 1 Gaurav Kumar, 4 Michael R. Salzberg, 3 Margaret J. Morris, 4 Sandra M. Rees, and 1 Terence J. O’Brien ( 1 Department of Medicine – Royal Melbourne Hospital, University of Melbourne, Melbourne, Vic, Australia ; 2 Department of Anatomy and Cell Biology, University of Melbourne, Melbourne, Vic, Australia ; 3 Department of Pharmacology and Physiology, University of New South Wales, Sydney, NSW, Australia ; and 4 Department of Psychiatry, University of Melbourne, Melbourne, Vic, Australia ) Rationale: Epilepsy patients commonly suffer from affective co‐morbidities, particularly anxiety and depression. This has widely been thought to be a consequence of the epileptic process, but recent evidence suggests there could be a common causal link. This study investigates whether the GAERS (Generalised Absence Epileptic Rats from Strasbourg) strain of rats, a phenotypic animal model of generalised absence epilepsy, also display behavioral abnormalities, and attempts to determine whether these behaviors are present prior to the development of absence seizures. Methods: Mixed gender GAERS rats were compared with Non‐Epileptic Controls in neurobehavioral tests for anxiety (elevated plus maze) and depression (sucrose‐preference test). Comparisons were made at 7 or 13 weeks, the equivalent neurodevelopmental age of adolescence in a rat, and also before and after the development of seizures. Results: GAERS rats showed increased anxiety in the elevated plus maze at both 7 and 13 weeks of age compared with NEC rats, as evidenced by a significantly lower number of open arm entries (at 7 weeks: 3.14 ± 0.53 entries (n = 14) vs 7.90 ± 1.01 (n = 20); at 13 weeks: 4.67 ± 1.03 entries (n = 15) vs 6.09 ± 0.73 (n = 11) – two‐way ANOVA, p = 0.0019). This strain of rats also displayed a significantly reduced preference for sucrose drinking water (at 7 weeks: 3.18 ± 0.66ml (n = 11) vs 6.67 ± 0.88ml (n = 3); at 13 weeks: 3.00 ± 0.84ml (n = 5) vs 4.83 ± 0.63ml (n = 12) – two‐way ANOVA, p = 0.0054), indicating an anhedonic state and suggestive of a depressive phenotype. Conclusions: This study provides strong evidence that GAERS rats, when compared with non‐epileptic controls, display anxious and depressive behaviors which are present prior to the onset of absence seizures. This suggests that affective disorders may not be a consequence of seizures, and also raises the possibility that there may be a common neurodevelopmental causality which predisposes humans to both affective disturbance and epilepsy. (Supported by an NHMRC project grant (400088), and a NARSAD Independent Investigator Award (TOB).) 1 Peter I. Jukkola, 2 Roger Strong, 2 Jaroslaw Aronowski, and 1,3 Kevin M. Kelly ( 1 Department of Neurology, Center for Neuroscience Research, Allegheny‐Singer Research Institute, Pittsburgh, PA ; 2 Department of Neurology, Stroke Program, The Medical Center at the University of Texas at Houston, Houston, TX ; and 3 Department of Neurology, Drexel University College of Medicine, Philadelphia, PA ) Rationale: Several rat strains have been studied extensively and are well‐known as genetic models of absence epilepsy. Less well‐known is the fact that many common laboratory rat strains frequently express spike‐wave discharges (SWDs) in a similar fashion. These SWDs are usually 6–9 Hz, and are associated with motor arrest of the animal. Generally, SWDs tend to be more frequent and longer in duration as the animals grow older. To our knowledge, however, no study in Long Evans rats has included a detailed observation of SWDs and how SWD expression changes throughout their lifetime. Previously, we reported that cortical infarction by middle cerebral artery/common carotid artery occlusion results in a significant reduction in the frequency and duration of SWDs compared to sham‐operated control animals (Kelly et al, 2006). In this study, we sought to extend these observations to gain a better understanding of the timecourse of SWDs and the effect of lesion‐associated changes on their expression in these animals. Methods: Five Long Evans rats lesioned by MCA/CCAO and four Long Evans sham‐operated controls were intermittently monitored by video‐EEG over a six‐month period as previously reported (Kelly et al, 2006). Fifty‐hour samples were then chosen from each animal's EEG record and visually reviewed for the presence of SWDs. The time and date of each SWD observed was documented, as well as the duration of the event. Preliminary results were obtained by plotting SWD data on a timeline for analysis by linear regression. Results: Preliminary results in control rats indicate that in 4 of 4 animals (100%) SWDs tend to increase in frequency of occurrence as the animal grows older. The average duration of SWDs tended to increase with age in 2 of 4 animals (50%), while decreasing in the other 2 animals (50%). Preliminary results in lesioned animals show a trend toward both greater frequency of occurrence and longer duration of SWDs in 2 of 5 (40%) of animals, and a decrease in both frequency of occurrence and duration in 2 of 5 animals (40%). One animal (20%) showed an increase in the frequency of occurrence, but a decrease in the average duration of SWDs. Conclusions: Since 7 of the 9 animals included in this study showed a trend toward an increase in the frequency of occurrence of SWDs over the six‐month monitoring period, these studies appear to be consistent with reports of an aging‐related increase in SWD expression in other rat strains. Changes in average duration of SWDs over time appear to be variable animal to animal. No reported changes in SWD expression are clearly related to the lesion. (Supported by RFA 01–07–26 to KMK.) 1 Tohru Kamida, 1 Hiroshi Ooba, 1 Mitsuhiro Anan, 1 Minoru Fujiki, and 1 Hidenori Kobayashi ( 1 Neurosurgery, Oita University Faculty of Medicine, Yufu, Oita, Japan ) Rationale: Nitric oxide (NO) is a free radical synthesized from L‐arginine by a Ca2+‐calmodulin‐dependent NO synthase (NOS). The functional meaning of NO and NOS in epilepsy remains controversial despite previous studies from animal models. In present study, we investigated NOS expression in the hippocampus of rat with pilocarpine‐induced seizures. Methods: Adult male Wistar rats (200–250g) were pretreated with atropine methylbromide (1 mg/kg s.c.) and 30 min later with pilocarpine hydrochloride (330 mg/kg i.p.). After 40 min, the Status epilepticus (SE) was terminated by injection of diazepam (4 mg/kg i.p.). Animals were sacrificed 1 day, 3days, 2 wk (silent phase) or 1 month (chronic phase) after status. Neuronal NOS (nNOS), inducible NOS (iNOS) and epithelial NOS (eNOS) expressions and the density of mossy fiber sprouting (MFS) in the hippocampus were evaluated using immunohistochemistry and neo‐Timm's histological procedures. Results: iNOS and eNOS expressions were more strongly upregulated than nNOS expression at both phases. nNOS expression gradually increased until 3 days and decreased after 2 wk. iNOS expression had no alteration at both phases. eNOS expression tended to decrease after 2 wk. The density of MFS gradually increased and was strongest at 1 month. Conclusions: Our data indicate that nNOS and eNOS with the strongest upregulation in the silent phase may have a correlation with epileptogenesis while iNOS with the durability of upregulation in both phases may have something of influence on the hippocampus also after the epileptogenesis was acquired. 1 Rafal M. Kaminski, 1 Hamid Sheikhi, and 1 Michael A. Rogawski ( 1 Epilepsy Research Section, NINDS/NIH, Bethesda, MD ) Rationale: 11‐Beta hydroxylase is involved in final steps of steroid hormone synthesis. Inhibition of this enzyme leads to accumulation of hormone precursors, which may eventually enter an alternative metabolic pathway and become neurosteroids. Neurosteroids are potent anticonvulsants acting via positive modulation of GABAA receptors. We have previously shown that they are particularly effective against 6 Hz seizures in mice. In the present study we investigated whether inhibition of 11‐beta hydroxylase protects against 6 Hz seizures and whether this effect is associated with increased neurosteroid synthesis. Methods: Male NIH Swiss mice were the experimental subjects. The animals were pretreated with either metyrapone or etomidate, which both are potent inhibitors of 11‐beta hydroxylase, and challenged at various time points with corneal 6 Hz electrical stimulation (32 mA, 3 sec.) to induce seizures. Separate groups of animals were additionally treated with finasteride to test whether seizure protection following administration of metyrapone or etomidate is attenuated by inhibition of neurosteroid synthesis. Results: Metyrapone afforded dose‐dependent protection against 6 Hz seizures 30 min. after injection (ED50 = 190.6 mg/kg). However, the potency of metyrapone was much higher (6‐fold increase) when the animals were tested 6 hrs. following injection (ED50 = 30 mg/kg). Etomidate displayed very similar pattern of protection against 6 Hz seizures with respective ED50 values of 4.5 and 1.7 mg/kg. Finasteride partially attenuated the anticonvulsant effects of both 11‐beta hydroxylase inhibitors. Conclusions: Inhibition of 11‐beta hydroxylase produces prolonged anticonvulsant actions in mice. This effect may be associated with an increased neurosteroid production, which could be responsible for protection from seizures. 1 Irina Kharatishvili, 1 Riikka Immonen, 1 Olli Grohn, and 1 Asla Pitkanen ( 1 A.I. Virtanen Institute, Kuopio University, Kuopio, Finland ) Rationale: The risk of posttraumatic epilepsy positively correlates with the severity of brain injury. Up to 50% of patients with penetrating head injury can develop epilepsy later in life, whereas for moderate injury the incidence of epilepsy is only 4.2%. Our previous study with lateral fluid percussion brain injury model (FPBI) demonstrated that severe impact strength resulting in 31–33% mortality, causes development of spontaneous recurrent seizures in 43–50% of experimental animals (Kharatishvili et al., Neuroscience 2006;140:685–97). In this study, we aimed to examine if in FPBI model, similarly to humans, probability of developing epilepsy is associated with injury severity. Methods: Head trauma was induced by lateral FPBI in adult Sprague‐Dawley rats (n = 15, 5 shams). The pressure of the applied impact, as recorded by a transducer, varied from 2.3 to 3.3 atm. The low mortality rate (23%) indicated moderate severity of the injury. Quantitative MRI was used to monitor the progression of the lesion starting at 3 hrs post‐injury and repeated at 3 d, 9 d, 2 month‐, 3 month‐ and 6 month post‐injury time‐points. Morris water maze test (MWM) was used to assess the long‐term behavioral outcome (spatial learning and memory) 7 months after FPBI. After that, 24h/day continuous video‐EEG monitoring was performed to assess the development of seizure activity; animals were implanted with depth electrode in the ventral hippocampus ipsilateral to the injury site and contralateral cortical electrode, and followed by continuous video‐EEG for 2 + 2 weeks with 3 week interval for up to 12 months. Results: The mean volume of the cortical damage at 6 months post‐injury was 40.8 ± 38.1mm3 (range 2.7–130.0 mm3). MWM test demonstrated no difference in spatial learning in injured rats compared to shams (p>0.05). EEG monitoring revealed 2–20 sec runs of 6–8 Hz monomorphic rhythmic activity with a spike‐like negative phase in 70% rats both in the FPBI and control groups. This rhythm was registered in drowsiness over the injury side and was responsive to movement. No epileptiform discharges/seizures were recorded during the entire monitoring period. Conclusions: Rats with moderate FPBI had substantial cortical lesion. However, the hippocampus‐dependent spatial learning and memory deficit was not as pronounced as previously found in animals which developed epilepsy after severe FPBI (p < 0.05). Also, none of the rats had electrographic seizures. In lateral FPBI, the moderate force of the impact results in progressive brain damage but does not increase the likelihood of posttraumatic epileptogenesis. (Supported by Academy of Finland, Sigrid Juselius Foundation, CIMO, EUCARE, Finnish Neurological Foundation.) 1 Sookyong Koh, 2 Carlo Condello, 2 Jaime Grutzendler, and 1 Hyokwon Chung ( 1 Neurology, Children's Memorial Hospital, Northwestern University School of Medicine, Chicago, IL ; and 2 Neurology and Physiology, Northwestern University School of Medicine, Chicago, IL ) Rationale: Microglia are the resident immune cells in the central nervous system (CNS) and constitute the first line of defense against pathological changes within the CNS microenvironment. Growing evidence suggests that pro‐ and anti‐inflammatory molecules are synthesized during seizures in the brain at the sites of seizure initiation and propagation, and that rapid induction of these inflammatory mediators may play a role in the pathophysiology of epilepsy. We hypothesize that activation of microglia and subsequent increases in cytokines, chemokines, and complements are the key initiating events for seizure‐induced inflammatory responses. Methods: In a total of 40 heterozygous Cx3cr1GFP/+ adult mice, Cx3cr1 encoding fractalkine chemokine receptor has been replaced by a green fluorescent protein (GFP) reporter gene by targeted deletion via homologous recombination in embryonic stem cells. Seizures were induced by intraperitoneal injection of kainic acid (KA, 20 mg/kg); control with phosphate buffered saline (PBS). The mice were sacrificed 30 min, 1day, 2 days, 7days, or 14 days after the injections. Three 50‐μm‐thick horizontal sections (6 hippocampal sections per brain) were selected for quantification of microglia. An average percent of area above the set threshold (containing green fluorescent cells/hippocampal hilar region capture at 20X magnification) and number of fluorescent element were calculated per animal, and comparison was made between KA and Control animals. Results: Enhanced fluorescent signaling indicative of microglial activation was already apparent in the hippocampus 30 minutes after KA‐induced seizures. Within 2 days, microglia in the hippocampus appeared enlarged and more numerous. A nearly 2 fold increase in the area of fluorescent cells was noted in KA treated mice compared to control littermates (PBS:1.67 ± 0.15 vs KA:3.08 ± 0.28, p < 0.005). Also, significant microglial proliferation occurred 2 days after KA‐induced seizures (113 ± 5 vs 189 ± 9, p = 0.0002). These increases in the area and number of fluorescent cells persisted to 7 days after KA (area%: 1.64 ± 0.10 vs 2.35 ± 0.21, p < 0.01; spot counts: 102 ± 3 vs 141 ± 5, p < 0.0001). Conclusions: Taking advantage of transgenic mice in which all microglia are fluorescently labeled, we have unequivocally demonstrated an acute and persistent microgliosis after KA‐induced seizures. Neuroinflammation provoked by seizures thus appears to be mediated via activation of resident CNS immune cells, microglia. Our data implicates microglia in seizure‐induced neuronal excitability and degeneration, and suggests an inhibition of microglia activation as a potential target for therapeutic intervention. (Supported by Child Neurology Foundation and K02NS048237.) 1 Seogkyoung Kong, 2 Ariana Lorenzana, 2 Jon E. Vanleeuwen, and 2 Paula E. Schauwecker ( 1 Biochemistry and Molecular Biology, University of Southern California, Los Angeles, CA ; and 2 Cell and Neurobiology, University of Southern California, Los Angeles, CA ) Rationale: Susceptibility to seizure‐induced excitotoxic cell death varies among inbred mouse strains due to variation in genetic background. Our QTL mapping studies have shown that distal chromosome 18 (Sicd1) contains a gene(s) conferring different susceptibility to seizure‐induced excitotoxic cell death between C57BL/6J (B6) and FVB/NJ (FVB) inbred mouse strains. We have verified that the causal gene(s) reside in the Sicd1 QTL utilizing congenic strains. We found the congenic strains having the Sicd1 QTL‐resistant locus from a cell‐death resistant strain (B6) and the remainder of their genome from a cell‐death susceptible strain (FVB) are less susceptible to seizure‐induced excitotoxic cell death as compared to the cell‐death susceptible strain (FVB). We hypothesized that a strain‐specific variation in one or more genes is responsible for the Sicd1 QTL effect. Methods: We have focused on two potential candidate genes, Galr1 and Mbp, based on their putative role in modulating seizure susceptibility and excitatory amino acid‐induced cell death. To test this hypothesis we sought to identify sequence variants in the Galr1 and Mbp genes between B6 and FVB strains using DNA sequencing. As a complimentary method to identify the Sicd1 QTL gene(s), we have examined changes in gene expression at the mRNA level of Galr1 and Mbp in hippocampal homogenates from both the parental inbred and the reciprocal congenic strains. Studies are underway to detect aberrant splicing based on cDNA sequencing and differential expression of Galr1 and Mbp at the protein level between our parental strains, B6 and FVB, as well as our reciprocal congenic strains. Results: Based on preliminary results, no coding sequence variants have been detected in the Galr1 and Mbp genes. We have found genotype‐dependent differences in Galr1 mRNA expression between B6 and FVB parental strains using qRT‐PCR (p = 0.023). Differential expression of hippocampal mRNA for Galr1 was also found by qRT‐PCR between the FVB.B6‐Chr18 congenic strain and the parental FVB strain. In contrast, no statistically significant differences in Mbp mRNA expression were observed either between the parental strains or recombinant congenic strains (p = 0.816). Conclusions: Based on the expression variation and its putative role in modulating seizure susceptibility and excitatory amino acid‐induced cell death we consider Galr1 as a strong candidate gene responsible for the Sicd1 QTL effect. Follow‐up studies whether there is functional expression differences in Galr1 gene between B6 and FVB would support Galr1 as a Sicd1 QTL gene. Results from the proposed studies will add to our understanding of the molecular determinants critical for the pathogenesis of seizure‐induced excitotoxic cell death. (Supported by NS038696 to P.E.S.) 1 Gaurav Kumar, 1 Nigel C. Jones, 2 Michael R. Salzberg, 4 Sandra M. Rees, 1 Abbie J. Coupar, 1 Terence J. O’Brien, and 3 Margaret J. Morris ( 1 Department of Medicine – Royal Melbourne Hospital, University of Melbourne, Melbourne, Vic, Australia ; 2 Department of Psychiatry, University of Melbourne, Melbourne, Vic, Australia ; 3 Department of Physiology and Pharmacology, University of New South Wales, Sydney, NSW, Australia ; and 4 Department of Anatomy and Cell Biology, University of Melbourne, Melbourne, Vic, Australia ) Rationale: We have previously demonstrated that low‐dose corticosterone (CS) supplementation, used as a model of the effect of chronic stress/depression, accelerates epileptogenesis in the amygdala kindling rat model of temporal lobe epilepsy (TLE). This study examined the contributions to this effect of the subtypes of glucocorticoid receptors: mineralocorticoid (MR, high‐affinity for CS) and glucocorticoid (GR, low‐affinity for CS). Methods: Female Non‐Epileptic Control rats 10–13 weeks of age were ovariectomized and implanted with a bipolar electrode into the left amygdala. Five experimental groups were studied: 1. Water‐control (n = 6), 2. CS treated (n = 6), 3. CS + Mifepristone (Mif – GR‐antagonist, 25 mg/kg sc, n = 7), 4. CS + Spironolactone (Spir – MR antagonist, 50 mg/kg, n = 5), and 5. CS + Mif (25 mg/kg) + Spir (50 mg/kg) (n = 6). Corticosterone was dissolved in drinking water (6 mg/100 ml) and administered ad libitum (groups 2–5) throughout the kindling period. Rats were injected subcutaneously with vehicle (Groups 1 and 2) or the relevant antagonist in vehicle (Groups 3–5) twice daily for the entire kindling period. Rapid Amygdala Kindling (RAK), which began two days after initial treatment, consisted of 10 sec bursts of 400μA every 15–20 mins, with up to 24 stimulations per day. Results: Groups differed significantly in the number of stimulations required to reach the ‘fully kindled state’ defined as five Class V seizures on the Racine scale (p = 0.04, Kruskal‐Wallis ANOVA). The CS‐treated group kindled faster than either the water‐treated control group (p < 0.05) or either of the three CS+antagonist groups (p < 0.10). No significant differences were found between the any of the CS+antagonist groups and the water‐treated controls. Conclusions: The data implicate both high affinity (MR) and low affinity (GR) glucocorticoid receptors in the mechanism by which low‐dose CS accelerates kindling epileptogenesis. This may have implications for understanding the potential effects of chronic stress and depression in initiating and/or exacerbating TLE. (Supported by an NHMRC project grant (400088), and a NARSAD Independent Investigator Award (TOB).) 1 Qizhi Zhang, 2,3 Zhao Liu, 2,3 Paul R. Carney, and 1 Huabei Jiang ( 1 The J. Crayton Pruitt Family Department of Biomedical Engineering, University of Florida, Gainesville, FL ; 2 Division of Pediatric Neurology, University of Florida, Gainesville, FL ; and 3 McKnight Brain Institute, University of Florida, Gainesville, FL ) Rationale: Photoacoustic tomography (PAT) is an emerging imaging modality that has the potential to image the dynamic function of the brain due to its unique ability of imaging biological tissues with high optical contrast and ultrasound resolution. Here we report the first application of our finite element based PAT method for imaging of seizure‐onset zone in an animal model of focal seizures. Methods: Male Harlen Sprague‐Dawley rats, weighing 50∼60g, were used in this study. The rats were anesthetized with urethane 1 mg/g in all the experiments. Focal seizures were induced by microinjection of a GABAA antagonist bicuculline methiodide (BMI) into the left parietal neocortex. A mechanical scanning photoacoustic system with single acoustic transducer was employed for collecting the light‐induced acoustic signals. A reconstruction algorithm was used to generate PAT images. Results: Following microinjection of BMI into the left parietal cortex, focal seizures were induced and confirmed with electroencephalogram recordings which showed high amplitude spike and wave discharges. These seizures were accurately localized tomographically by PAT. The dynamic changes in PAT imagings were comparable to EEG changes during seizures. Conclusions: Our result showed that PAT is an emerging non‐invasive imaging modelity to localize the seizure‐onset zone. Compared to other neuroimaging modalities, PAT has not only the advantage of high spatial and temporal resolution, but also the ease of use. It has the potential to bring brain imaging to the bedside, clinic and daily life, and coincide a scan with a seizure such that an epileptic focus and seizure occurrence can be visualized in real‐time. (Supported by: This work was supported in part by the National Institutes of Health (R01 AR048122) and Wilder Epilepsy Research Center of the University of Florida.) 1 Ariana Lorenzana, 1 Ruth I. Wood, and 1 Paula E. Schauwecker ( 1 Cell and Neurobiology, University of Southern California, Los Angeles, CA ) Rationale: Studies have shown a correlation between estrous cycle and seizure susceptibility in humans and most mammals. This effect is also seen during menopause, or reproductive senescence. It has been hypothesized that the hormonal fluctuations throughout the various stages of estrous exert some type of influence on seizure occurrence, prevalence, and severity. The goal of this study was to analyze the effect of ovarian hormone cyclicity at different stages of estrous, the loss of ovarian hormones and the relationship to kainate‐induced seizure susceptibility and seizure induced cell death. Methods: Control female mice of the FVB/NJ (FVB) strain were administered kainic acid (KA) at different stages of the estrous cycle. Estrous cycle was determined by vaginal cytology and only regularly cycling animals were chosen for the experiments. KA or saline was administered systemically to female FVB mice at various stages of the estrous cycle, to ovariectomized (OVX) females, OVX females treated with estradiol (OVX + E), and VCD‐treated female mice. Mice were monitored for 4 hours for seizure onset, severity, and activity and scored for latency to onset of first severe seizure and duration of severe seizures. After 7 days of recovery, neuronal damage was evaluated throughout the hippocampus proper. Results: Among ovary‐intact female FVB mice, estrous cyclicity was without effect on seizure susceptibility. In particular, no significant differences in latency to onset or duration of severe seizures were noted. In contrast, in the OVX + E female mice, seizure duration was significantly increased, and latency to onset of seizures was reduced, as compared to OVX and ovary‐intact female mice. The extent of hippocampal cell death among ovary‐intact mice was not affected based on estrous cycle staging. In particular, no significant differences in the extent of cell death were observed between ovary‐intact female FVB mice injected with KA during estrus, when gonadal hormone levels peak, or during diestrus, the nadir of gonadal hormone levels. Conclusions: Our results suggest that estradiol may lower the threshold for seizure activity. While we found no modulatory effects of estradiol on seizure susceptibility or cell death based on estrous cyclicity, it is important to note that two different models for the modulation of ovarian steroid levels were used. In one model, ovariectomy (OVX) causes a complete disruption of gonadal hormones. In contrast, ovarian steroid levels vary across the estrous cycle in ovary‐intact female mice. In addition, OVX + E mice have none of the protective effects of progesterone present. Studies are ongoing to compare two models of reproductive senescence: the OVX model and the chemically‐induced VCD model. Insight into the mechanisms of brain related‐changes as a result of the loss of ovarian steroids will help identify risk factors for seizure susceptibility and cell death in aging women. 1 Jesus Machado‐Salas, 3 Patricia Guevara, 1 Dongsheng Bai, 1 Reyna Duron, 2 K. Amano, 2 T. Suzuki, 2 Kazuhiro Yamakawa, 1 Eaian Cornford, and 1 Antonio Delgado‐Escueta ( 1 Epilepsy Center of Excellence, UCLA David Geffen School of Medicine, Los Angeles, CA ; 2 Laboratory for Neurogenetics, Riken Brain Science Institute, Saitama, Japan ; and 3 Neurociencias. Neurobiologia, Instituto Nacional de Neurologia y Neurocirugia, Mexico, DF, Mexico ) Rationale: We have identified two genes associated to Lafora disease.In our null mutant mice (EPM2A), we have studied the immunocytochemical architecture of the neurocytoskeleton and the Lafora bodies. Methods: Our Laforin‐deficient knockout mice were produced by deleting the Laforin domain coding region of the EPM2A gene (Hum Mol Genet 11:1252–1262.2002). Similar groups of control and homozygous null mutant mice were sacrificed at fixed intervals and their encephalons were fixed with glutaraldehyhde and processed with PAS and immunocytochemical techniques. Monoclonal antibodies were used against neurofilaments M and L (NFM/L) . Results: Our experimental animals showed neuronal death at 2 months of age, in the absence of apoptotic changes or accumulation of Lafora bodies (LfB). At 9 months of age the number of LfB was very prominent in large nerve cells. They were positive to PAS, Ubiquitin and AGEP. Large neurons from the brainstem, hippocampus and cerebellum showed abnormal immunoreaction for NFM/L, which, in some cases, very much resembled neurofibrillary degeneration. Clusters of fragmented and distorted neurites were seen, very much like “senile” (neurite) plaques. Analysis of LfB has revealed some unexpected structural components that appear to establish an anatomofunctional link between them and the neurocytoskeleton . Conclusions: Our observations apppear to indicate that our model of Lafora disease presents morphological evidence of a neurodegenerative process in absence of apoptotic cell death. Lafora bodies do not primarily participate in neuronal death. With age, they increase in size and number, accumulating poliglucosans and, probably, also misfolded proteins. The presence of a complex structure in the ring of Lafora bodies, allows us to postulate an active role for these still enigmatic inclusions. (Supported by NINDS 5RO1NS042376–03.) 1 Pavel Mares, 1 Denisa Lojkova, and 1 Anna Mikulecka ( 1 Department of Developmental Epileptology, Institute of Physiology, Academy of Sciences, Prague, Czech Republic ) Rationale: Many antagonists of NMDA receptors exhibit a powerful anticonvulsant action but also strong unwanted side effects. To utilize potential of these drugs attention is now focused on specific NMDA antagonists. We decided to study effects of representants of the two above mentioned classes low‐affinity noncompetitive antagonist memantine and NR2B‐specific antagonist ifenprodil in immature rats. Methods: Two models of epileptic seizures were used – pentetrazol(PTZ)‐induced motor seizures and cortical epileptic afterdischarges (ADs) induced by stimulation of the sensorimotor cortex. Male rats of the Wistar strain 12, 18, and 25 days old were tested. The same age groups were studied behaviorally to find possible side effects. Spontaneous behavior in the open field was registered and motor performance was checked by means of age‐specific tests. Memantine was administered intraperitoneally in doses of 10, 20 and 40 mg/kg, ifenprodil in doses of 10, 20 and 40 mg/kg. Each age and dose group in each experiment was formed by 8 to 10 animals. Results: Memantine suppressed generalized tonic‐clonic seizures (GTCS) in a dose‐dependent manner in all age groups. In contrast, it potentiated minimal clonic seizures. Ifenprodil suppressed only the incidence of the tonic phase of GTCS in 12‐ and 18‐day‐old rats. Memantine increased threshold intensities for elicitation of spike‐and‐wave type of AD in 12‐ and 18‐day‐old rats as well as threshold for the other (mixed) type of ADs (due to a transition of epileptic activity into the limbic system) in 25‐day‐old rats. Duration of ADs was shortened in all age groups. Ifenprodil pretreatment decreased threshold intensities for elicitation of mixed type of ADs in 18‐ and 25‐day‐old rats and increased duration of these ADs in 25‐day‐old rats. If administered when the animals allready passed repeated ADs, it significantly shortened ADs in 12‐ and 18‐ but not 25‐day‐old animals.All doses of memantine increased locomotor activity in 12‐ and 18‐day‐old rats, the change in the oldest group was not significant. The 20‐ and 40‐ mg/kg doses impaired motor performance in all age groups. Ifenprodil increased locomotor activity only in 12‐day‐old rats and motor performance was changed only with the highest dose in 12‐ and 25‐day‐old rats. Duration of the impairment of motor performance by either drug was shorter than duration of an anticonvulsant effect. Conclusions: Memantine exhibits anticonvulsant action as well as side effects described for common NMDA atagonists. Ifenprodil has an action mostly in the youngest group what is in agreement with developmenal decrease of NR2B subunits in the NMDA receptors. (Supported by research projects No. LC554 and AV0Z 50110509.) 1 Steven T. Marsh, 1,2 Levi B. Good, 1 Kevin J. Garvey, and 1,2 David M. Treiman ( 1 Neurology Research, Barrow Neurological Institute, Phoenix, AZ ; and 2 Harrington Department of Bioengineering, Arizona State University, Tempe, AZ ) Rationale: Status epilepticus (SE) is a dynamic state. There are progressive changes in clinical behavior, response to treatment, extent of neuronal damage, and EEG pattern during prolonged clinical and experimental status epilepticus. SE also causes deficits in learning and memory, but progressive impairment with increasing severity of SE has been little studied. We tested the hypothesis that the EEG stage during SE predicts the extent of subsequent cognitive impairment. Methods: Forty‐three 250–280 gm male Sprague‐Dawley rats were implanted with 4 stainless steel epidural electrodes for EEG recording. Status epilepticus was induced by 3 mmol/kg IP LiCl followed 24 hrs later by 30 mg/kg pilocarpine. EEG was continuously recorded before and after injection, and SE was stopped at EEG stage I (discrete seizures), III (continuous ictal activity), IV (continuous ictal activity punctuated by flat periods), or V (periodic epileptiform discharges) (Treiman et al., Epilepsy Res5:49–60, 1990) with IP diazepam (10 mg/kg) plus phenobarbital (25 mg/kg). Control rats were given PB and DZM prior to SE induction and thus did not develop SE. Spatial learning and memory were assessed one week prior to SE induction and one week after SE using a modification of the Morris Water Maze (Morris, J Neurosci Methods11:47–60, 1984). Learning impairment was tested in six trials on each of four sequential days. Effect of the SE stage on the mean escape times to a submerged platform were determined by a multiple analysis of variance. Pair‐wise multiple comparisons of groups were done by Fisher's least significant difference. The mean escape time for the trials on test day 1 post‐SE was compared with the mean of the trials on test day 4 pre‐SE to assess the effect of SE on spatial memory. Results: Mean time to stop SE from pilocarpine injection was: Stage I 25 min, Stage III 65 min, Stage IV 100 min, and Stage V 210 min. There were no significant differences between these groups in pre‐SE water maze performance. However, post‐SE day 1 mean escape times were significantly worse compared with pre‐SE day 4 (test of memory) as well as the mean escape times for all test days (test of learning) in Stage IV and Stage V rats when compared to controls, Stage I, and Stage III rats (p < 0.001). No significant differences were found between control, Stage I and Stage III rats, nor between Stage IV and Stage V rats. Conclusions: SE results in a marked impairment of both visual‐spatial memory and learning in the late stages of SE. Stage IV, and V rats demonstrated a greater impairment than controls, or Stage I‐III. Thus progressive cognitive impairment, in addition to other dynamic changes cited above, is also an effect of prolonged SE. These data provide further evidence that the EEG stage during prolonged SE is a marker of its severity. (Supported by the Barrow Neurological Foundation.) 1 Stella M. Valiensi, 1 Oscar A. Martinez, 1 Ricardo C. Reisin, 1 Rubens Granillo, 1 Silvia Christiansen, and 1 Federico J. Bottaro ( 1 Neurology, Hospital Italiano, Buenos Aires, Buenos Aires, Argentina; Neurology, Hospital Britanico, Buenos Aires, Buenos Aires, Argentina; Neurology, Hospital Britanico, Buenos Aires, Buenos Aires, Argentina; Neurology, Hospital Italiano, Buenos Aires, Buenos Aires, Argentina; Pathology, Hospital Italiano, Buenos Aires, Buenos Aires, Argentina; and Internal Medicine, Hospital Britanico, Buenos Aires, Buenos Aires, Argentina ) Rationale: S.E. can produce brain damage. Propofol is an anaesthetic with anticonvulsant effect, but its use is limited by complications such as Rhabdomyolisis® There are no experimental studies that can demonstrate neuroprotective activity of Propofol in S.E, neither R. attributed to this drug . Objectives: 1 – To evaluate if Propofol in an animal model of Pilocarpine – induced S.E can prevent neuronal damage in hippocampal structures. 2 – To evaluate if Propofol in S.E. in this experimental model can cause R. Methods: Adult male Wistar rats, (200 to 300 g) were used. Biosecurity rules defined by the Institute of Basic Sciences and Experimental Medicine of our Hospital were followed. 1.CONTROL GROUP : n = 9 rats: 4 cm3 saline solution intraperitoneal (i.p) 2.PILOCARPINE GROUP : n = 10: 1 mg/kg N‐Hyoscine butylbromide+ 350 mg/kg of i.p Pilocarpine. 3.PROPOFOL GROUP : n =10: 1 mg/kg N‐Hyoscine butylbromide+ 350 mg/kg of i.p Pilocarpine. After 30 minutes of S.E., 20 mg/kg Propofol. Behaviour and Scalp EEG monitoring were evaluated. Brains were removed after 2 hours,and stained with hematoxylin‐eosin (H.E). The hippocampus was assessed. 4 histologic groups with signs of neuronal damage(dark and shrunken) were identified . 1) No damage 2) Mild damage: (+) 2–3 neurons per field (400 X) 3) Moderate damage (++) 3–5 neurons per field 4) Severe damage (+++) 5–7 neurons per field The muscle quadriceps femoralis was removed and stained with H.E. The presence of macrophage in and around the fibres was evaluated . The pathologist performed analyses blinded to the treatment received by the rats. Results: Brain damage in every area was analyzed with a non‐parametric test (Wilcoxon sum rank).A p < 0.05 was considered a significant difference. Between Group1 and Group 2, greater difference was found in temporal neocortex and subiculum in Group 2. Significant damage decrease was observed in Group 3, with less damage in the CA3–4 regions (p = 0.0019). BEHAVIOUR All the rats had S.E. EEG: Group1: normal EEG. Group 2: continuous epileptiform activity . Group 3: continuous epileptiform activity during S.E. Propofol produced suppression of the epileptiform activity. R.: 2 rats in each group(Group 2 and Group 3) had signs of R.There were no statiscally significant differences between treatment groups. Conclusions: Neuronal damage was less severe in the rats treated with Propofol, with decrease neuronal damage regarding the CA region, specifically in the CA3–4 regions, which suggests its neuroprotective effect. R. was not considered a statiscally significant finding. 1 Brona M. Murphy, 2 Clara K. Schindler, 2 Sachiko Shinoda, 1 Carmen Bellver‐Estelles, 1 Mark Dunleavy, 2 Robert Meller, 1 Jochen H.M. Prehn, 2 Roger P. Simon, and 1 David C. Henshall ( 1 Physiology and Medical Physics, Royal College of Surgeons In Ireland, Dublin, Ireland ; and 2 Robert S. Dow Neurobiology Laboratories, Legacy Research, Portland, OR ) Rationale: Seizures whether brief or prolonged are capable of causing neuronal death. Apoptosis signaling pathways play a role is such neuronal loss. Previous studies have demonstrated that experimental seizures can activate the apoptotic intrinsic mitochondrial cell death pathway and that components of this pathway are altered in involved brain structures in patients with temporal lobe epilepsy. We have developed a mouse model of seizure‐induced neuronal death with features of programmed cell death. Presently, we examined the activation of the mitochondrial apoptotic pathway in this model and contrasted this to events within human epilepsy brain. Methods: The experimental methods used in this study included, a mouse seizure model: intraamygdala kainic acid injection in adult mice, terminated after 40 min by diazepam, human temporal lobe specimens: hippocampal resections obtained following surgery for intractable epilepsy compared to matched autopsy controls, Western blotting and Immunofuorescence. Results: Seizures evoked by intraamygdala kainic acid in C57BL/6 mice caused ipsilateral death of CA1 and CA3 neurons within the hippocampus. Western blotting revealed seizures induced overexpression of Bax, cytochrome c release and activation of caspase‐9 and −7. Analysis of hippocampi from patients with intractable epilepsy revealed cleaved caspases 9 and 7 were present. In searching for relevant anti‐apoptotic proteins we found no changes to expression of Bcl‐2 and Bcl‐xl but a rapid decline in Bcl‐w levels at both 0.5 (p < 0.01) and 4 h following seizures. Reasoning that loss of Bcl‐w may promote cell death after seizures we investigated hippocampal injury following seizures in mice lacking Bcl‐w. Bcl‐w knockout mice exhibited no obvious neuroanatomical differences to their wildtype littermates and expressed similar hippocampal levels of a range of apoptosis‐associated proteins including Bcl‐2 and similar amygdala levels of the kainic acid (KA) receptor GluR5–7. Seizures evoked by intraamygdala KA induced significantly more hippocampal neuronal loss and DNA fragmentation (terminal deoxynucleotidyl dUTP nick end labeling) in Bcl‐w knockout mice compared to wildtype mice. Extending these data, we found higher cytoplasmic levels of Bcl‐w in hippocampus from patients with TLE compared to autopsy controls. Conclusions: These data identify Bcl‐w as an endogenous neuroprotective gene that may be a therapeutic target for the treatment of seizure‐induced brain injury and temporal lobe epilepsy. (Supported by the Health Research Board, Ireland, Wellcome Trust, UK and NIH/NINDS, USA.) 1 Anna Tostevin, 1 David Liu, 1 Terence O’Brien, 2 Rod Hicks, 3 David Williams, and 1 Damian Myers ( 1 Department of Medicine, Royal Melbourne Hospital, University of Melbourne, Melbourne, Victoria, Australia ; 2 Centre for Molecular Imaging, Peter MacCallum Cancer Institute, Melbourne, Victoria, Australia ; and 3 Department of Physiology, The University of Melbourne, Melbourne, Victoria, Australia ) Rationale: Reduced GABAA/central benzodiazepine receptor (GABAA/cBZR) expression, mossy fibre sprouting (MFS) and hippocampal sclerosis are common pathological changes seen in both human and animal models of temporal lobe epilepsy (TLE). This study was designed to investigate immediate (24 hours) and long‐term changes (2, 4 and 6 weeks) in GABAA/cBZR expression, neuronal loss and MFS in the hippocampus in the kainic acid (KA)‐induced status epilepticus (SE) model of TLE. Methods: Age‐matched non‐epileptic control male rats were randomized into control and KA groups (at least n = 6/group). MFS and GABAA/cBZR expression, neuronal loss and MFS were measured in hippocampal subregions, dentate gyrus (DG), CA3 and CA1 in both the control and epilepsy groups. GABAA/cBZR expression (Bmax value) was measured by saturation‐binding analysis using (3H)‐Flumazenil, a cBZR antagonist. Stereological counting was performed for cell counting and volume analysis of the CA3 SPc subregion of the hippocampus. MFS into the dentate molecular layer was quantified from Timm‐stained brain slices. Results: At 24 hours post‐SE, Bmax was significantly increased compared to controls in the whole hippocampus (+29.4%± 5.7; p < 0.001) and parietal cortex (Par Ctx) (+30.6%± 12.8; p < 0.05) as well as in all subregions of the DG, CA1, and CA3 with the exception of CA3 SPc (Repeated measure ANOVA, p < 0.02). Calculated from absolute values, no changes in receptor expression were seen over 2–6 weeks. There was no difference in MFS 24 hours post‐SE compared with controls but MFS increased in the dentate molecular layer at 2 weeks post SE (20% above control; p < 0.05) and showed a progressive increase over time reaching 70% above controls (p < 0.001) at 6 weeks. At 24 hours, statistically significant neuronal loss occurred (−24 ± 9.3%) in the CA3 SPc region (control = 6, KA group = 6, p < 0.05) associated with a significant increase in GABAA/cBZR expression per surviving neuron. No recovery of neurons was apparent through to the end of 6 weeks post‐SE. Conclusions: Changes in GABAA/cBZR expression, cell loss and synaptic reorganization occur during epileptogenesis and may influence the severity of spontaneous recurrent seizures (SRS). Elevated receptor expression observed early after status epilepticus (within 24 hours) may be a protective response by the brain to dampen excitability and increase resistance to future seizures. MFS, a form of neuronal reorganization, is a prominent feature in the KA model of TLE progressively increasing during the latent epileptogenic period up to onset of SRS. GABAA/cBZR expression and MFS were not correlated until 6 weeks implying that MFS may affect seizure activity in late stages of epilepsy onset. 1 Sandeep P. Nair, 4 Wendy M. Norman, 1 Linda K. Dance, 1,3 Panos M. Pardalos, 1,2 Jose C. Principe, and 1,4 Paul R. Carney ( 1 Biomedical Engineering, University of Florida ; 2 Electrical and Computer Engineering ; 3 Industrial and Systems Engineering ; and 4 Pediatrics, University of Florida, Gainesville, FL ) Rationale: Gradual, progressive preictal dynamical convergence among brain regions has been reported in human epilepsy (IEEE TBME 2003; 50; 616–27) and in an animal model of chronic limbic epilepsy (CLE) (Epilepsia 45 (S7): 211–212). We have used a spontaneously seizing animal model of limbic epilepsy to investigate the occurrence of spatiotemporal transitions involving such dynamical convergence among different regions in an epileptogenic brain. We hypothesize that the frequency of these transitions increases progressively from the time of initial insult to the subsequent manifestation of epilepsy. Methods: Young adult male Sprague Dawley rats (n = 4) were used in the study. The model was developed using chronic hippocampal stimulation (CHS). The animals animals, after allowing sufficient time for recovery from the stimulation procedure,were connected to an automated system that monitored the T‐index calculated from short‐term maximum Lyapunov exponents estimated from multiple brain regions. A (dynamical convergence (transition) was defined as a drop in T‐index value from a preset upper threshold (UT= 10) to a lower threshold (LT= 2.662). The time leading to the first seizure was divided into blocks of 12 hours and the frequency of transitions in each block was calculated. The null hypothesis is rejected if the regression slope is found to be significantly positive (p < 0.05). Results: Two out of the four animals developed spontaneous seizures after CHS. The first showed a progressive increase in frequency of spatiotemporal transitions leading up to the first seizure (recording time before seizure∼232.9 hrs, slope = 0.09, p = 0.0005), the second animal showed an abrupt increase in incidence immediately preceding the seizure (recording time before seizure∼142.5 hrs, slope = 0.07, p = 0.0228). No dynamical transitions were observed in the two animals that did not develop spontaneous seizures. Conclusions: These results suggest that spatiotemporal transitions involving convergence of dynamical values among brain regions may be a marker for epileptogenesis. (Supported by NIH grant R01EB002089, Children s Miracle Network, University of Florida Division of Sponsored Research and Development and Department of Veterans Affairs.) 1 Jari Nissinen, and 1,2 Asla Pitkanen ( 1 Department of Neurobiology, A.I.Virtanen Institute/University of Kuopio, Kuopio, Finland ; and 2 Department of Neurology, Kuopio University Hospital, Kuopio, Finland ) Rationale: Prevention of epileptogenesis in patients with acute brain damaging insults like status epilepticus (SE) is a major challenge. We investigated whether carbamazepine (CBZ), levetiracetam (LEV), and valproic acid (VPA) as monotherapy are antiepileptogenic or disease modifying. To mimic clinical study design, treatment was started 24 h after the beginning of electrically induced SE. Methods: SE was induced by stimulating the lateral nucleus of the amygdala of adult Sprague‐Dawley rats (n = 98) electrically for 20–40 min. One group of vehicle treated non‐stimulated animals served as controls for histology and behavioral analysis (SHAM, n = 37). SE was stopped with diazepam (20 mg/kg, intraperitoneally, i.p.) at 4 h after SE induction (additional injection of DZP 10 mg/kg was given at 3h after the first injection). Administration of compounds (CBZ 120 mg/kg/d, n = 17; LEV 300 mg/kg/d, n = 18; VPA 600 mg/kg/d, n = 16; vehicle VEH, n = 47) was started 24 h after the beginning of SE. On the first day, AEDs were administered intraperitoneally, and thereafter i.g. All compounds were administered at 8 h intervals for 7 d. SE and the development of spontaneous seizures were monitored with video‐EEG. The occurrence of epilepsy and seizure frequency were assessed at 10 weeks after SE induction with a continuous video‐EEG monitoring (24h/day) for 2 weeks. To confirm antiepileptogenesis or disease modification, the second 2‐weeks continuous video‐EEG was started 14 weeks after SE. Thereafter, animals underwent behavioural testing (Morris water‐maze and fear conditioning) and were perfused for histology. Results: Altogether, 62% of VEH, 89% of LEV, 53% of CBZ, and 88% of VPA treated animals developed epilepsy. The mean seizure frequency in the VEH group was 6.3 ± 12.8 sz/d (range 0.04–50, median 0.39), in the LEV group 4.8 ± 7.6 sz/d (range 0.04–24.2, median 0.88), and in the VPA group 7.6 ± 18.5 sz/d (range 0.07–70). In CBZ group the mean seizure frequency was 0.6 ± 0.6 sz/day (range 0.04–1.63, median 0.46). However it did not differ from that in the VEH group. Mean behavioural seizure severity did not differ between groups. The mean seizure duration was longer in the CBZ group compared to that in the VEH group (77 ± 16 sec vs. 68 ± 19 sec, p < 0.05). Conclusions: The three AEDs with different mechanisms of action, LEV, CBZ, and VPA were not antiepileptogenic when started at 24 h after SE induction and continued for 7 days. (Supported by NIH/NINDS R21 NS 049525.) 1 Joseph A. Oommen, 1 Alli Kraus, and 1 Robert S. Fisher ( 1 Neurology, Stanford, Stanford, CA ) Rationale: Since CSF drug delivery is clinically useful for antibiotics and antineoplastics, direct delivery of anti‐epileptic medication to the CSF may also be useful and improve the tolerability and efficacy of these drugs. We performed a proof‐in‐principle study of intraventricular gabapentin (GPN) in the rat. Methods: Anesthetized Sprague‐Dawley rats were implanted with two Alzet model 2001 osmotic pumps, each delivering 1.0 μL/hour for one week. Pumps were connected to left and right intraventricular catheters, with tips in relation to Bregma at: 1.4 mm lateral and 0.8 mm posterior on the right, 2 mm lateral and 1.3 mm posterior on the left, and 2.5 mm below the top of cortex. GPN 80 μg/μL was installed into each pump and delivered at 3.8 mg per day. Half of the animals had pumps filled with isotonic normal saline (control group), and half with GBP. Seizure theshold was tested after five days, by dripping the convulsant liquid anesthetic flurothyl onto tissue paper at 10 μL/minute in a sealed 3.2 L container. A single observer, blinded as to treatment group, recorded the time to first myoclonic twitch, first partial seizure and first tonic‐clonic seizure. Seizures were judged behaviorally; no EEG was used as part of this protocol. Results: Of 60 rats used, 6 were lost due to mortality or pump/catheter malfunction. For the surviving 54, time to first tonic‐clonic seizure was 295.8 ± 58.8 s (n = 28) for control group, versus 338.0 ± 89.9 s (n = 26) for the rats with GBP in the pump (p = 0.049 by two‐tailed t‐test). The mean time to onset of first myoclonic jerk was 158.7 ± 20.8 vs. 164.6 ± 33.5 s (p = n.s.). Effects of flurothyl can be influenced by animal weight; however, regression of time to tonic‐clonic seizure versus weight yielded no significant relationship. Serum gabapentin levels obtained immediately after testing the seizure threshold were undetectable (<1 μg/ml) in every animal. We did not perform quantitative studies of GPN distribution through brain, but methylene blue dye in the pumps distributed into periventricular white matter and also over cortex, especially ipsilaterally. Conclusions: Intraventricular delivery of gabapentin elevates the threshold for generalized tonic‐clonic seizures in the rat flurothyl model, but not time to the first myoclonic jerk or first partial seizure. Interpretation is limited, as the degree of penetration and distribution of the GPN into brain cannot be detailed. Unmeasurable serum levels suggest the effect is not systemic. Intraventricular delivery of antiepileptic drugs would be useful for drugs with poor oral absorption or which do not penetrate the blood‐brain barrier. Intraventricular administration of drugs that do penetrate into brain after oral ingestion (e.g. GPN) may or may not be useful, depending upon achievable improvements in the therapeutic/toxic ratio and in steady brain levels of drug. (Supported by Medtronic, and by the Maslah Saul MD Chair, the James and Carrie Anderson Epilepsy Laboratory, and the Susan Horngren Fund.) 1 Mansi B. Parekh, 2 Lan B. Hoang‐Minh, 1 Hector Sepulveda, 3 Angela Handlock, 1 Wendy Norman, 1 Justin C. Sanchez, 1 William L. Ditto, 1 Paul R. Carney, and 4 Thomas H. Mareci ( 1 Neuroscience, University of Florida, Gainesville, FL ; 2 Biomedical Engineering, University of Florida, Gainesville, FL ; 3 Pediatrics, University of Florida, Gainesville, FL ; and 4 Biochemistry and Molecular Biology, University of Florida, Gainesville, FL ) Rationale: In this study, we used MR diffusion tensor imaging (DTI) to determine the temporal hippocampal and parahippocampal structural changes in vivo at 11.1 Tesla (T) during the latent period of epileptogenesis and to examine these same regions in the excised intact brain at 17.6 T. The mesial temporal lobe model was also used to determine if chronic stimulating/recording microelectrodes distort MR images collected at 11.1 T. Methods: Fifty micron gold plated tungsten wires (2) were implanted in the ventral hippocampus in Sprague Dawley rats (n = 6). Spontaneously seizing rats were obtained following experimentally inducted status epilepticus (SE). Rats were video recorded to capture spontaneous seizures. Injured rat brains were examined in vivo at 11.1 T pre/post wire implantation and after SE at 3, 5, 7, 10, 20, 40, 60 days. The intact excised brain of 2 rats were imaged at 17.6 T. Diffusion weighted images were acquired using gradient directions specified by the tessellations of an icosahedron on the hemisphere. The diffusion weighted images were fit to a rank‐2 tensor model of diffusion. Results: Five of the 6 rats developed seizures. The DT images of the rat, without seizures, looked identical to controls. In the other 5 rats, diffusivity increased in the hippocampus and fimbria in vivo. In 2 of these rats, diffusivity also increased in the entorhinal cortex. Increased diffusivity was seen as early as day 7 post‐stimulation. We examined the excised brains of a seizing rat and one that did not. In the seizing rat, the diffusivity increased bilaterally in the hippocampal region, identified with mossy fiber sprouting. The hippocampus showed atrophy ipsilateral to the simulation. Decreased fractional anisotropy indicated a loss of structure in the dentate gyrus and CA2/3 region of the pyramidal cells (Fig.1). Conclusions: The induced SE produces bilateral structural changes, visible in detail with DTI, in the hippocampal and parahippocampal structures. Stereotaxic images allowed repeated measures of the structural changes during epileptogenesis and the implanted wires produced minimal distortions. Future work will relate the physiological and structural changes, during epileptogenesis, using electrophysiology and DTI. (figure 1) (Supported by NIH grants, R01 EB004752 and R01 NS42075, Wilder Epilepsy RC, and the UF Alumni Foundation.) 1 Alexia Paucard, 2 Akitaka Yamamoto, and 1 David C. Henshall ( 1 Physiology and Medical Physics, Royal College of Surgeons, Dublin, Ireland ; and 2 RS Dow Neurobiology Laboratories, Legacy Research, Portland, OR ) Rationale: Hippocampal sclerosis is the commonest lesion identified in temporal lobe epilepsy and has been proposed as both cause and effect of seizures. Additionally, an emerging concept in epilepsy research is that the propagation of paroxysmal activity is necessary and sufficient to transform a naïve structure into one that is capable of generating spontaneous (epileptic) seizures. This transformation underlies the formation of an epileptogenic focus that can entrain the generation of seizures in interconnected structures. However, whether seizure‐injury occurs within such communicating fibre tracts is largely unknown but recent reports document the presence of lesions within white matter tracts in epilepsy patients. According, we evaluated white matter tract injury within the corpus callosum in a novel seizure model in mice. Methods: Adult male C57Bl/6 mice underwent focally‐evoked seizures induced by intraamygdala kainic acid microinjection with continuous electroencephalography. Seizures were terminated by intravenous lorazepam 40 (n = 10) or 50 minutes (n = 7) following kainate/vehicle (n = 3) injection and brains obtained 24 hours later. Coronal sections were processed for detection of DNA fragmentation using the TUNEL technique. To identify the cell populations of the corpus callosum, the slides were labelled for oligodendrocytes with antibodies against 2,3‐cyclic‐nucleotide 3‐ phosphodiesterase (CNP) or myelin basic protein (MBP) and for astrocytes with anti‐glial fibrillary acidic protein (GFAP). Results: Seizures in mice terminated after 40 min resulted predominantly in unilateral TUNEL staining within ipsilateral CA3 and CA1. In contrast, when seizure activity was extended to 50 min contralateral CA1 and CA3 hippocampal damage was more frequently observed. Examination of the ipsilateral and contralateral corpus callosum revealed the presence of cell death. Our data also revealed significant correlations between numbers of degenerating cells in contralateral (r = 0.460, p < 0.05) corpus callosum and the severity of neuronal cell death in CA3 hippocampus ipsilateral. Seizures induced oligodendrocyte death in the corpus callosum as revealed by the immunofluorescence colocalization of TUNEL/MBP or CNP. In contrast, the presence of a gliogenesis is shown around the corpus callosum. Conclusions: The present data establish that prolonged focal limbic seizures in mice induce damage to non‐neuronal elements within the corpus callosum and show the degree of this injury is a function of hippocampal damage. Our data indicate seizures injure white matter tracts and offer a potential mechanism for seizure generation and spread to contralateral and extrahippocampal structures. (Supported by Marie Curie Grant.) 1 Challon R. Perry, 1 Minah Suh, 1 Hongtao Ma, 1 Mingrui Zhao, and 1 Theodore H. Schwartz ( 1 Department of Neurological Surgery, Weill Medical College of Cornell University, New York, NY ) Rationale: Cortical stimulation is playing an increasing role in clinical neuroscience in the treatment of epilepsy, stroke and pain. The neuromodulatory effects of cortical stimulation include both excitation and inhibition, the balance of which varies with the stimulation parameters. As a result, cortical stimulation also influences cerebral hemodynamics, including perfusion and oxygenation, which may also vary depending on the stimulation parameters, although little data exists on this subject. In this study, we investigate the effects of stimulation frequency on the hemodynamic response using optical recording of intrinsic signals (ORIS). Methods: Using bipolar tungsten electrodes, the rat sensory neocortex was stimulated at variable frequencies in a pseudo‐random fashion at 5 Hz, 10 Hz, 50 Hz, 100 Hz, and 250 Hz at a constant charge (CC). The amplitude remained constant at 1 mA. CC was maintained by delivering stimulation at 100 spikes at each frequency and varying the duration of stimulation. Simultaneous field potential (f.p.) recording and ORIS was performed at 570 nm and 610 nm to quantitatively measure total hemoglobin (Hbt), deoxygenated hemoglobin (Hbr) and oxygenated hemoglobin (HbO2) using a pathlength correction. Optical data was averaged over 10 trials at each frequency and each trial lasted 100 sec so the optical signal could return to baseline. Trials with afterdischarges were eliminated. Results: A large increase in Hbr was recorded at all stimulation frequencies but the maximum occurred at 10 Hz (2.12 ± 0.20%, n = 3). The relationship between the increase in Hbr and stimulation frequency (in log scale) could be characterized by the equation y =−1.1164 × 2 + 3.1836x – 0.2979 (R2 = 0.74). There was a significant correlation between the duration of stimulation and the duration of the increase in Hbr (r = .86, p < 0.05). The largest increase in Hbt occurred at 50 Hz (4.04 ± 1.08%, n = 3) and the relationship between the increase in Hbt and stimulation frequency (in log scale) could be characterized by the equation y =−1.3816 × 2 + 5.3414x – 1.2291 (R2 = 0.98, p < 0.05). The duration of the increase in Hbt was generally longer than the duration of stimulation by 80 seconds, regardless of frequency. Conclusions: These results demonstrate that the duration of stimulation is more important than the frequency in determining the duration of the dip and that the relationship between frequency and both Hbt and Hbr is not linear, with a plateau occurring between 10–50 Hz. (Supported by NIH.) 1 David Revesz, 1 Mikael Tjernstrom, 1 Thorleif Thorlin, and 1 Elinor Ben‐Menachem ( 1 Department of Clinical Neuroscience and Rehabilitation, Institute of Neuroscience and Physiology, the Sahlgrenska Academy at Gothenburg University, Gothenburg, Sweden ) Rationale: The birth of new neurons from neuronal stem cells, the process called neurogenesis, has been seen in adult brains from both rodents and humans. The process is especially prominent in the hippocampus, which is involved in mood and memory functions and epilepsy. A new hypothesis for the development of depression in humans has been formed over the last few years, which includes neurogenesis as a factor of importance in the depressive disease: Reduction of neurogenesis in the hippocampus is a causality factor in the generation of depression and stimulated neurogenesis is part of the recovery process from the depressive state. The same applies for epilepsy. VNS therapy, which has been used successfully for many years in the treatment of epilepsy, was early on in its development noted to have mood elevation effects in VNS treated patients, with effects on depressive symptoms. This anti‐depressive effect from VNS has been proven in clinical studies, and VNS is now approved by the FDA for the treatment of severe depression in humans as well as epilepsy. The mechanisms behind the antiepileptic and anti‐depressive effects from VNS are, however, not fully understood. We hypothesize that a VNS‐induced stimulation of the neurogenesis is part of the mechanism behind the observed clinical effects from VNS. Methods: The degree of stem cell proliferation within the dentate gyrus of the hippocampus was studied. The number of new cells were measured by BrdU (bromodeoxy‐uridine) incorporation. Immunohistochemistry with antibodies against BrdU were used for the detection of the newly divided cells. Eight rats implanted with VNS‐leads and generators were stimulated for two days with the following parameters: 0.75 mA output current, 250 microseconds pulse width, 20 Hz frequency, 30 seconds on and 5 minutes off. These rats were compared with eight control rats that had VNS surgery for implantation of the VNS equipment, but with devices constantly in the off position. Results: There was a statistically significant difference between the two groups with a 50% increase of stem cell proliferation at the 0.03 significance level in the VNS‐stimulated rats as compared to sham operated animals. Conclusions: These results indicate that VNS treatment in the rat affects the stem cell proliferation in the hippocampus. In the light of the prevailing theory of a link between hippocampal neurogenesis, epilepsy and depression, this finding might represent an interesting path in the search for mechanisms behind the actions of VNS. (Supported by Cyberonics, Houston, TX, USA; Vastra Gotalandsregionen, Sweden.) 1 Alexander Rotenberg, 1 Khanh Hoang, 2 Alvaro Pascual‐Leone, 1 James J. Riviello, and 1 Frances E. Jensen ( 1 Neurology, Children's Hospital, Boston, MA ; and 2 Neurology, Beth Israel Deaconess Medical Center, Boston, MA ) Rationale: rTMS is a noninvasive method for cortical stimulation that is emerging as a new therapeutic tool in epilepsy. The technique is based on the principle of electromagnetic induction where small intracranial electrical currents are induced by a strong rapidly‐changing extracranial magnetic field. Limited human trials and animal experiments in our laboratory show acute suppression and reduction in frequency of partial seizures with EEG‐guided single pulse and 0.5 Hz rTMS. However, the capacity of rTMS to suppress generalized (absence) seizures has not been thoroughly examined. Accordingly, we sought to determine whether whether 0.5 Hz rTMS can suppress frequent PTZ‐induced generalized seizures in a rat absence epilepsy model. Methods: Generalized (absence) seizures were triggered in adult male Long Evans rats (n = 10) with PTZ (50 mg/kg, i.p.). Immediately after PTZ injection, and thereafter at 30‐minute intervals, one group (n = 4) was treated with 0.5 Hz repetitive TMS for 10 minutes (300 stimuli). A second group (n = 6) remained as untreated controls. Continuous EEG was recorded with subdermal wires electrodes for 120 minutes after PTZ injection, and then analyzed for seizure frequency during periods between TMS sessions. Results: (1) Frequent brief absence seizures accompanied by generalized spike‐wave discharges on surface EEG were triggered by PTZ injection in all rats. (2) 3/6 untreated rats died within the monitoring period – 0/4 rats died in the 0.5 Hz rTMS group. (3) Average cumulative seizure frequency was lower in the treated group after 60 min (t = 2.65; p < 0.033), 90 min (t = 2.98; p < 0.021) and 120 min (t = 2.85; p < 0.030) of monitoring. Conclusions: 0.5 Hz rTMS is effective in suppression of generalized seizures in the PTZ absence epilepsy rat model. Or finding may be a step toward eventual treatment of generalized epilepsy with low frequency rTMS. (Supported by Citizens United for Research in Epilepsy (CURE).) 1,2 Raman Sankar, 1,3 Stephane Auvin, 1 JoAnne Nakagawa, and 1 Don Shin ( 1 Pediatrics, David Geffen School of Medicine at UCLA, Los Angeles, CA ; 2 Neurology, David Geffen School of Medicine at UCLA, Los Angeles, CA ; and 3 Pediatric Neurology, EA 1046 School of Medicine, Lille, France ) Rationale: Lithium pilocarpine status epilepticus (LiPC‐SE) causes prolonged seizures and subsequent widespread neuronal injury. In the developing rat brain, untreated LiPC‐SE results in age dependent hippocampal injury as well as widespread extrahippocampal neuronal damage. We examined whether brief durations of SE are sufficient to induce injury in the immature brain. Methods: EEG data was collected using acquisition software obtained from wireless transponders implanted s.c. in two week old wistar pups. Rats were pretreated with LiCl (3 mEq) on the day of surgery and then subjected to SE 16–20 hr later with pilocarpine (60 mg/kg) at P14. Diazepam (5 mg/kg) and phenobarbital (25 mg/kg) were given after varying durations of SE using EEG recordings to determine both the delay of onset after PC injection, as well as the total duration of SE. 24 hr after the onset, rats were given pentobarbital (100 mg/kg) followed by perfusion/fixation with 4% paraformaldehyde. Brains were embedded in paraffin and cut at 8 μ intervals. Sections were stained with Fluoro‐Jade B (F‐J B) as a marker of neuronal injury, and analyzed using a semiquantative scale (0 = no injury, 1 = trace, 2 = < 10%, 3 = 10–25%, 4 = 26–50%, 5 = > 50% F‐J B labeled neurons) bilaterally in three sections containing the dorsal hippocampus. Results: SE was established in over 90% of the P14 rats. Behavioral manifestations of seizures appeared almost immediately after PC injection, but synchronized video monitoring revealed a delay of 9.8 ± 1.6 min before the appearance of continuous, rhythmic, high frequency electrographic activity used to mark the onset of SE. Prolonged recordings showed that the combination of diazepam/phenobarbital was sufficient to prevent the return of seizures over the following 24 hr period. While 30 min of SE was not sufficient to induce hippocampal injury with the exception of trace dentate granule cells, injury was seen in the amygdala (2.3 ± 0.1) and cortex (1.8 ± 0.1). No brightly labeled neurons were seen in the thalamus, but some lightly stained, irregularly shaped neurons were detected. Conclusions: Previous studies have shown that P14 rats subjected to LiPC experience several hours of SE resulting in extensive neuronal injury, particularly in the CA1 hippocampal subfield. While early pharmacological intervention was sufficient to protect those neurons from injury, extrahippocampal structures remain vulnerable after only 30 min of SE. Studies in adult rats suggest protection of the Ammon's horn alone is not sufficient to prevent epilepsy. 25% of untreated P14 rats undergoing LiPC‐SE go on to demonstrate spontaneous seizures. Whether the neuronal injury that results after only 30 min is sufficient to induce epilepsy is under investigation. (Supported by NS046516 & DAPA Foundation (RS), AEAC Association (SA).) 1 Matthew R. Sarkisian, 1 Christopher M. Bartley, 2 Hongbo Chi, 3 Fumihiko Nakamura, 1 Kazue Hashimoto‐Torii, 1 Masaaki Torii, 2 Richard A. Flavell, and 1 Pasko Rakic ( 1 Neurobiology, Yale University School of Medicine, New Haven, CT ; 2 Immunobiology, Yale University School of Medicine, New Haven, CT ; and 3 Medicine, Harvard Medical School, Boston, MA ) Rationale: Periventricular nodular heterotopia (PVNH) is a congenital malformation of human cerebral cortex frequently associated with mutations in the Filamin‐A (FLN‐A) gene but the pathogenetic mechanisms remain unclear. Both loss and gain of FLN‐A function have been proposed to disrupt the migration initiation of cortical neurons. Here we show that MEKK4 (a MAP3K member of the mitogen‐activated protein kinase pathway) is involved in Fln‐A regulation and PVNH formation. Methods: We performed histological analyses on the brains of MEKK4‐deficient mice. Additionally, we designed short interference RNA (siRNA) constructs targeting MEKK4, incorporated these constructs into the developing cerebral cortex by in utero electroporation and analyzed the fates of transfected cells. Immunocytochemistry, in situ hybridization, electron microscopy, GST‐pulldown and western blot analyses were also performed Results: MEKK4−/− mice develop bilateral PVNH in the cerebral cortex that arise from breaches in ventricular surface lining. Using neuronal markers, we show that cells within PVNH have differentiated and failed to migrate into the cortical plate. In utero electoporation of MEKK4 siRNA at E14.5 and analysis at P0 revealed dramatic disruption of migration‐initiation of cortical neurons compared to control. Analysis of Fln‐A expression in MEKK4−/− forebrain revealed elevated and ectopic Fln‐A, while MEKK4 RNAi increased Fln‐A phosphorylation in vitro. Finally, recombinant‐MKK4/SEK1 protein precipitated a complex containing MEKK4 and Fln‐A suggesting that MKK4/SEK1 bridges these molecules during development. Conclusions: Disrupted migration initiation in MEKK4‐deficient cells may be related to both defects of the ventricular surface lining and dysregulated Fln‐A. Our data demonstrate an essential role for MEKK4 in the migration of developing cortical neurons and provide insight into the pathogenesis of human PVNH. (Supported by the Epilepsy Foundation of America postdoctoral fellowship (to M.R.S), an under‐represented minority fellowship (to C.M.B) on U.S. Public Health Service grants (to P.R.), a Child Health Research Grant from the Charles H. Hood Foundation, Inc. (Boston) (to H.C.). R.A.F. is an Investigator of the Howard Hughes Medical Institute.) 1 Steven C. Schachter, 2 H. Steve White, 3,4 Lauren Murphree, and 4 James Stables ( 1 Osher Institute, Harvard Medical School, Boston, MA ; 2 Pharmacology and Toxicology, University of Utah, Salt Lake City, UT ; 3 SRA International, Fairfax, VA ; and 4 Anticonvulsant Screening Project, NINDS, Bethesda, MD ) Rationale: Huperzine A (HupA), a sesquiterpene alkaloid derived from Chinese club moss (Huperzia serrata), has traditionally been used in China for swelling, fever and inflammation. HupA reversibly inhibits acetylcholinesterase and reduces glutamate‐induced neuronal death as an NMDA receptor antagonist. We previously reported that huperzine potently protects mice against clonic seizures induced by s.c. pentylenetetrazol at 1 mg/kg (White, 2005) with a TD50 of 0.83 mg/kg. We now report results from the 6‐Hz model of psychomotor seizures. Methods: Anesthetic/electrolyte solution (0.5% tetracaine hydrochloride in 0.9% saline) was applied to the eyes of male, 18–25 g CF1 mice (Charles River Laboratories, Wilmington, MA) prior to placement of corneal electrodes. Seizures induced by 6‐Hz corneal stimulation (Barton, 2001) using three levels of stimulation – 22, 32 and 44 mA – (3 sec duration) were assessed at various times after i.p. administration of 0.09 – 1.1 mg/kg HupA. Median effective doses (ED50) and 95% confidence intervals were calculated for each the three current stimulations evaluated. Results: ED50 values for i.p. HupA were 0.28, 0.34 and 0.78 mg/kg for 22, 32, and 44 mA, respectively. The following results were observed at the highest doses tested at each stimulation level: at 22 mA, 7/8 animals were protected with a dosage of 0.7 mg/kg; at 32 mA, 8/8 animals were protected with 0.83 mg/kg; and at 44 mA, 6/8 animals were protected with 1.1 mg/kg. Conclusions: HupA produced a potent anticonvulsant effect in the 6‐Hz model of psychomotor seizures at all stimulation levels tested and with ED50 values less than the TD50, suggesting a possible advantage over phenytoin, carbamazepine, lamotrigine and topiramate, each of which display limited efficacy in this model at doses devoid of behavioral toxicity. Further, the less than 2‐fold ratio of dosages effective across the range of stimulations suggests a possible advantage over other drugs active in this model such as levetiracetam. These findings and the known action of huperzine as an NMDA receptor antagonist warrant further pre‐clinical and clinical evaluation. References Barton ME et al. Epilepsy Res 2001;47:217–27. White HS et al. Epilepsia 2005;46(suppl 8):220. (Supported by: The investigators gratefully acknowledge the support of the NINDS Anticonvulsant Screening Project and a grant from The Epilepsy Project (S.S.).) 1 Melodie R. Winawer, 2 Nicholas Makarenko, 2 Tana M. Hintz, 2 Sameh M. Kamel, 2 Daniel P. McCloskey, 2 Nandini Nair, 3 Abraham A. Palmer, and 2,4 Helen E. Scharfman ( 1 Neurol. & Sergievsky Center, Columbia University, New York, NY ; 2 CNRRR, Helen Hayes Hospital, West Haverstraw, NY ; 3 Human Genetics, University of Chicago, Chicago, IL ; and 4 Pharmacol. & Neurol., Columbia University, New York, NY ) Rationale: To identify new genes involved in temporal lobe epilepsy (TLE) using quantitative trait locus (QTL) mapping, initial studies were conducted to compare the pilocarpine model of TLE in A/J vs. DBA/2J mouse strains, two strains that are fully sequenced. The incidence and latency to status epilepticus were compared using pilocarpine doses that spanned the dose‐response curve. Animals that had status were also compared after recurrent, spontaneous seizures developed, to examine structural changes that often accompany chronic seizures: hippocampal neuronal loss, mossy fiber sprouting, mossy fiber neuropeptide Y (NPY) expression, and hilar ectopic granule cells (EGCs). Methods: Male mice (10 wks old) were administered atropine methylbromide (5 mg/kg, s.c.) 30’ before pilocarpine hydrochloride (200, 220, 250, or 300 mg/kg, i.p.), and 1 hr after the onset of status, diazepam was administered (5 mg/kg, i.p.). Mice were transcardially‐perfused > 4 wks after status, and immunocytochemistry was conducted using antibodies to NeuN, a neuronal marker (1:5,000, Chemicon), NPY (1:30,000, Peninsula), or PROX1 to label granule cells (1:60,000, Covance). PROX1‐labeled EGCs were quantified using stereology (StereoInvestigator software). Results: A/J mice had a higher incidence of status, if animals which died after a tonic‐clonic seizure, prior to clear signs of status, were excluded. However, the incidence of these tonic‐clonic seizures was higher for A/J mice than DBA mice, and the tonic‐clonic event could have been the start of status, so an incidence comparison is complex. When status occurred, the mean (±sem) latency to status was longer for A/J mice at all doses (A/J: 132.6 ± 6.5min, n = 26; DBA: 30.4 ± 1.7min, n = 24; p < 0.05). Recordings in A/J mice confirmed that the onset of behavioral status reflected the onset of electrographic status (n = 4). All DBA mice demonstrated neuronal loss in CA3 (n = 6/6), but all A/J mice did not (n = 3/6). Neuronal loss was detected in area CA1 in DBA (n = 3/6) but not A/J mice (n = 0/6). Both strains demonstrated NPY expression in mossy fibers (A/J: n = 5/6; DBA: n = 4/5). There were more EGCs in DBA mice (A/J: 3029.9 ± 1053.4 cells/hippocampus, n = 4; DBA: 9590.2 ± 1979.2, n = 6; p < 0.05). Conclusions: A/J and DBA/2J mice show differences in acute behavioral and long‐term anatomical changes with the pilocarpine model. Differences suggest that use of one variable to define susceptibility can be misleading. The results suggest guidelines for future studies of strain‐dependent seizure susceptibility, and characterize an infrequently‐studied strain for QTL mapping of seizure susceptibility. (Supported by NINDS R01 41490, K02 NS050429, K23 NS02211.) 1 Irene Schlifke, 2 Biljana Georgievska, 2 Deniz Kirik, and 1 Merab Kokaia ( 1 Experimental Epilepsy Group, Wallenberg Neuroscience Center, BMC A‐11, Lund, Sweden ; and 2 Neurobiology, Wallenberg Neuroscience Center, BMC A‐11, Lund, Sweden ) Rationale: The glial cell line derived neurotrophic factor (GDNF) has recently been indicated as an endogeneous anti‐epileptic agent, as well as it, due to its trophic action, could rescue neurons from excocytotic cell death and promote hippocampal neurogenesis. In the present study, we aimed to evaluate the ability of viral vector‐mediated hippocampal overexpression of GDNF to supress seizures and alter seizure‐induced cell death and neurogenesis. Methods: A recombinant adeno‐associated viral vector (rAAV) encoding for GDNF or GFP (control) was injected in multiple sites of the hippocampal formation of rats. A recording/stimulating electrode was placed in the ventral hippocampus, and animals were stimulated either for 1 s once daily, inducing kindling epileptogenesis, or for 1 h continously to induce self‐sustained status epilepticus (SE). Kindled animals were re‐stimulated (i.e. re‐kindled) four weeks following termination of kindling to assess the sustainability of hyperexcitability. Behavioural seizure activity (graded according to the scale of Racine) and EEG were recorded throughout the experiments. All experiments were conducted according to international ethical guidelines for animal care and were approved by the local ethical comittee. Results: Using the kindling model, we found that transfection with rAAV‐GDNF decreased the number of generalized seizures, prevented an increase of the duration of behavioural convulsions seen in re‐kindled control animals, and increased the threshold for induction of re‐kindled seizures. In the SE‐model of epilepsy, animals overexpressing GDNF developed fewer generalized stage 5 seizures and exhibited a significantly better survival rate following SE. We further found that in GDNF‐overexpressing animals, a higher proportion of new‐born (BrdU+) cells acquired a neuronal fate, as determined by staining for DCX and Neun, compared to control animals. The SE‐induced damage in hippocampal CA1/CA3 and dentate gyrus, as assessed by fluorojade staining, was unchanged by rAAV‐GDNF transfection. Conclusions: Taken together, these data suggest that rAAV‐mediated gene transfer is a feasible method for long‐term delivery of GDNF to the seizure focus, and that such GDNF overexpression is able to suppress seizures in different animal models of epilepsy. (Supported by Swedish Research Council (Vetenskapsradet), the Segerfalk Foundation, the Craaford Foundation and the Kock's Foundation.) 1 Hector Sepulveda, 1 Lan Hoang‐Minh, 2 Mansi B. Parekh, 3 Angela Hadlock, 3 Wendy Norman, 3 Justin C. Sanchez, 1 William L. Ditto, 3 Paul R. Carney, and 4 Thomas H. Mareci ( 1 Biomedical Engineering, University of Florida, Gainesville, FL ; 2 Neuroscience, University of Florida, Gainesville, FL ; 3 Pediatrics, University of Florida, Gainesville, FL ; and 4 Biochemistry and Molecular Biology, University of Florida, Gainesville, FL ) Rationale: We used an 11.1 Tesla (T) magnet to determine in vivo temporal hippocampal and parahippocampal structural changes in the latent period of epileptogenesis in a rat model of mesial temporal lobe epilepsy. We also determined if chronic microelectrodes distort MR images to assess the feasibility of MRI with electrophysiological arrays. Methods: Fifty micron gold plated tungsten wires (2) were implanted in the ventral hippocampus in Sprague Dawley rats (n = 4). Spontaneously seizing rats were obtained following experimentally inducted status epilepticus (SE). Rats were video recorded to capture spontaneous seizures. An MRI compatible stereotaxic frame was developed, to allow repeatable positioning under isoflurane in the 11.1T/40 cm bore magnet, along with a 3 cm, 150 degree arc, linear surface MR coil. Rats were imaged pre/post wire implantation and after SE at 3, 5, 7, 10, 20, 40, 60 days. T2‐weighted images were collected and quantified. Results: Brain images and wire locations were obtained with the stereotaxic frame, which provided a repeatable platform. Significantly increased T2 values were observed in the amygdala, entorhinal and piriform cortices, particularly on the contralateral side to stimulation, in three out of four rats. T2‐weighted images of the first rat showed a contralaterally growing volume of hyper‐intense tissue in the hippocampus and amygdala (Fig. 1). The brain of the second rat did not change; the wires were medial to the hippocampus near the alveus. The third rat showed hyper‐intensity in the contralateral amygdala, with wires in the ventral hippocampus. The fourth rat was hyper‐intense in both sides of the hippocampus and amygdala by day 3, but slightly decreased in time, with wires in the hippocampus near the alveus. Conclusions: Unilateral electrically induced SE produces bilateral structural changes in the hippocampal and parahippocampal structures. Stereotaxic images allowed repeated measures of the structural changes during epileptogenesis. The implanted wires produced minimal distortions. Their positioning in the ventral hippocampus may play a role in which structures show changes. Future work will relate the physiological and structural changes during epileptogenesis using electrophysiology and high resolution MRI. (figure 1) (Supported by NIH grant R01 EB004752, Wilder ERC, UF Alumni Foundation.) 1 Margaret N. Shouse, 1 John C. Scordato, and 1 Paul R. Farber ( 1 Sleep Disturbance Research, VA GLAHS, North Hills, CA ) Rationale: To describe some similarities and dissimilarities between feline epilepsy models, Landau‐Kleffner Syndrome (LKS) and Electrical Status Epilepticus in Slow‐sleep (ESES) with respect to sleep state modulation, neurobehavioral consequences and early intervention. Methods: The sleep‐wake state distribution of seizures and neurobehavioral consequences were quantified in relation to monoamine concentrations (microdialysis) as well as before and after lesions, systemic and localized drug administration and/or photic stimulation in feline epilepsy models (n = 86), mostly in the amygdala kindling epilepsy model. 1 Similarities and Dissimilarities of age‐related amygdala kindling in kittens to Electrical status epilepticus (ESES) and/or Landau‐Kleffner Syndrome (LKS) SIMILARITIES DISSIMILARITIES AGE DEPENDENCY Critical period for onset GTCs can last to adulthood in kittens SPONTANEOUS SEIZURE‐RELATED SYMPTOMATOLOGY: Seizure types (GTCs, partial motor, atypical absence, complex‐partial); IID types & cortical “foci;” Foci change over time ESES rare in kittens;Clinical seizures start 1st kittens & ESES vs. LKS; Fronto‐central foci in ESES vs. Centro‐ temporal foci in LKS and kittens SLEEP PATTERNS: NREM onset seizures; ↑ IIDs & spread in NREM & ↓ IIDs & spread in REM. Occasional ESES symptoms in kittens; Sleep deficits in kittens DEVELOPMENTAL/NEUROBEHAVIORAL DEFICITS: ↑ sensitivity & IIDs to sound; Social isolation; Mood change, Stereotypies, Hyperactive, atonic/dystonic aggressive/placid Type/extent of residual deficits depend on early onset, site and duration of CSWS in humans vs.GTCs in kittens PROGNOSIS & AED RESPONSE Good for seizures; ESES self‐limited; Residual neurobehavioral symptoms Same AEDs untested ENCEPHALOPATHY; PATHOPHYSIOLOGY Localization‐related with subtle/minor cell loss/pathology at foci ? GTCs = generalized tonic‐clonic convulsions; ↑= increased ↓= decreased; CSWS = Continuous slow spike wave during sleep (Supported by the Department of Veterans Affairs.) Results: 1) Table 1 compares the kindling model to LKS and ESES with respect to sleep state modulation, spontaneous neurobehavioral consequences and early intervention. 2) Neural generators of synchronous EEG oscillations (tonic background slow waves and sleep EEG transients), can combine to promote electrographic seizure propagation during NREM; antigravity muscle tone permits seizure‐related movement. 3) Neural generators of asynchronous neuronal discharge patterns can reduce electrographic seizures during REM sleep; skeletal motor paralysis blocks seizure‐related movement during REM. 4) Neurobehavioral consequences depend on the site of the focus, regardless of overt histopathological change in the young. 5) The prognosis is best with early intervention. Conclusions: The kindling process is thought to have trans‐synaptic effects on cells distal to the kindled focus. Age‐related changes could underlie the progression of epileptic and non‐epileptic symptoms in kindled kittens, LKS and ESES. 1 Ajay K. Srivastava, and 1 H. Steve White ( 1 Pharmacology and Toxicology, University of Utah, Salt Lake City, UT ) Rationale: Recently we have characterized the pharmacology of antiepileptic drugs (AEDs) in the lamotrigine (LTG)‐resistant pentylenetetrazlole kindled (Epilepsia 2003; 44,42) and LTG‐resistant amygdala kindled rat (Epilepsia 2004; 45(7):12). In this model of pharmacoresistant epilepsy, CBZ, PHT, and TPM were all inactive at non‐toxic doses. In contrast, VPA and retigabine (Epilepsia 2005;46(8):217) were both effective. The broad‐spectrum AED felbamate (FBM) displays a favorable preclinical profile in various models of epilepsy including the phenytoin‐resistant amygdala kindled rat (Neuropharmacology 2000;39(10):1893–903). The present study aimed to evaluate the efficacy of FBM in the LTG‐resistant amygdala kindled rat model of refractory epilepsy. Methods: Two groups of male Sprague‐Dawley rats were kindled via basolateral amygdala stimulation according to the method described by Srivastava et al., (Epilepsia 2004; 45(7):12). One hour before each kindling stimulation, rats in the control group received 0.5% methylcellulose and rats in the experimental group received LTG (5 mg/kg, i.p.). Seizure severity and afterdischarge duration was recorded after each kindling stimulation. Treatment was continued until all rats displayed 4 consecutive Stage 5 seizures (Clin Neurophysiol 1972; 32:281–294). Two days later, both groups were challenged with 15 mg/kg, i.p LTG to confirm LTG‐sensitivity (control) and LTG‐resistance (experimental). The efficacy of FBM (100, 150 and 200, mg/kg administered 90 min prior to kindling stimulation) was then evaluated. The ED50 and 95% confidence intervals were calulated using Probit analysis. Results: Felbamate dose‐dependently reduced the seizure severity and decreased the ADD in the fully kindled LTG‐resistant rat. At the highest dose tested (200 mg/kg), FBM decreased the seizure score from 5 ± 0 to 2.25 ± 0.6 and the ADD from 57 ± 5 sec. to 22 ± 30 sec. The calculated ED50 and (95% confidence intervals) for FBM against Stage 4 and 5 seizures was 169 and (131 – 206) mg/kg, respectively. Conclusions: In an animal model of pharmacoresistant epilepsy, the broad‐spectrum AED FBM was found to effectively block the fully kindled secondarily generalized kindled seizure. In this particular model, FBM is as effective as VPA and the investigational AED retigabine (albeit less potent than retigabine). These findings demonstrated that the LTG‐resistant kindled rat can be utilized to differentiate the AED profile of novel AEDs and confirm that FBM possesses a unique profile relative to CBZ, PHT, LTG, and TPM. Ongoing studies continue to evaluate the mechanism underlying the development of pharmaco‐resistance in this model. (Supported by NINDS grant 1 R21‐NS049624–01.) 1,2 Kimberly Statler, 1 Seth Swank, and 2 H. Steve White ( 1 Pediatrics, University of Utah, Salt Lake City, UT ; and 2 Pharmacology and Toxicology, University of Utah, Salt Lake City, UT ) Rationale: Pediatric traumatic brain injury (TBI) is a common cause of childhood death and disability. Post‐traumatic epilepsy (PTE) and cognitive disabilities are common sequelae that may manifest years after TBI. PTE afflicts 10% of children after severe TBI but remains poorly understood. To test the hypothesis that TBI during brain development increases seizure susceptibility later in life, we used standard electrical stimulation techniques to assess seizure thresholds in a rat TBI model. Methods: Immature, male Sprague‐Dawley rats (n = 20/grp) underwent TBI by controlled cortical impact (6‐mm rounded tip, 4 m/sec velocity, 2‐mm deformation, 100 msec duration) to left parietal cortex using isoflurane anesthesia (1%) on post‐natal day (PND) 16–18. Results were compared to those obtained from age‐matched sham (craniotomy only using isoflurane) and naive rats (n = 20/grp). Seizure thresholds were assessed during adolescence and adulthood for tonic (PND 34 and 60), clonic (PND 37 and 63) and limbic (PND 40) seizures. Full convulsive current (CC) curves were generated by staircase procedure and the median CC (CC50) was calculated using Probit analysis. A p‐value < 0.05 was considered significant. Results: Tonic and clonic seizure thresholds increased with age in naive rats (p < 0.05). As shown in Table 1, seizure responses during adolescence (PND 34–40) were similar among groups for tonic and clonic seizures; however, limbic seizure responses showed a trend toward lower thresholds after TBI. In adults (PND 60–63), tonic seizure thresholds were decreased in both sham and TBI (vs. naive) groups. Conversely, clonic seizure thresholds were similar among sham and naive rats but decreased after TBI (vs. either naive or sham). Conclusions: TBI to left parietal cortex during brain development attenuates normal maturational increases in clonic seizure thresholds and may increase susceptibility to limbic seizures. Given the involvement of frontal cortex and limbic system in learning and decision‐making functions, this finding may have implications not only for PTE, but also for post‐injury learning and behavioral deficits. Delineation of the incidence of PTE and cognitive deficits, as well as further characterization of longitudinal maturational changes in seizure thresholds post‐TBI are ongoing. 1 Seizure Thresholds Tonic Clonic Limbic adolescent naive 27.7 (25.1, 30.5) 12.3 (10.4, 13.9) 89.0 (87.6, 90.5) sham 30.4 (29.0, 33.0) 13.7 (12.0, 15.5) 87.8 (86.4, 89.2) TBI 26.9 (23.7, 29.6) 11.6 (9.7, 13.5) 87.2 (85.4, 88.9)# adult naive 36.5 (35.1, 37.8) 15.0 (13.7, 16.3) sham 34.5 (33.8, 35.8)* 14.8 (13.5, 16.1) TBI 33.7 (32.1, 35.3)* 12.0 (10.3, 13.7)*∧ Data presented as CC50 (95% CI) and compared by Probit analysis; *p < 0.05 vs. naive, ∧p < 0.05 vs. sham; #p < 0.1 vs. naive (Supported by NIH K12‐HD 01410; PCMC Foundation; University of Utah Child Health Research Center.) 1 Kerry‐Ann A. Stewart, 2 Robert S. Fujinami, 1 Karen S. Wilcox, and 1 H. Steve White ( 1 Pharmacology & Toxicology, University of Utah, Salt Lake City, UT ; and 2 Neurology, University of Utah School of Medicine, Salt Lake City, UT ) Rationale: Viral infections of the CNS are associated with a significantly increased risk for seizures and epileptogenesis. Up to 50% of adult‐onset epilepsy cases in developing countries are associated with CNS infection. To our knowledge, no available animal model accurately reproduces the pathology of human encephalitis‐induced epilepsy. We have established an animal model where infection of C57BL/6 (B6) mice with Theiler's murine encephalomyelitis virus (TMEV) leads to the development of seizures during the acute encephalomyelitis phase of infection. Stage 4/5 seizures (Racine, Electroenceph. Clin. Neurophys. 32: 281–94, 1972) occur in approximately 50% the mice between 3–10 days post infection and appear to remit as the infection clears. The present study tested the hypothesis that seized mice have altered seizure thresholds post inoculation, which increases their risk for the development of spontaneous seizures. Methods: B6 mice were infected intracerebrally with the Daniel's strain of TMEV (TMEV‐DA). Animals were monitored for the development of seizures and separated into two groups: seized versus non‐seized. Two months after infection with 3 × 105 PFU TMEV‐DA, minimal clonic and maximal tonic hindlimb extension (THE) seizure thresholds of seized and non‐seized mice were assessed by electroconvulsive seizure testing (EST). Additionally, the rate of corneal kindling acquisition (number of stimulations required to reach fully kindled state, i.e., 3 consecutive Stage 5 seizures) was assessed to determine whether TMEV inoculation leads to an increased rate of kindling. Results: At two months post infection, seized mice had a significantly lower minimal clonic seizure threshold compared to the non‐seized mice (seized CC50= 6.7 mA, non‐seized CC50= 7.8 mA, p < 0.01). In contrast, there was no significant difference in maximal THE seizure thresholds between the seized and non‐seized mice (seized and non‐seized CC50= 13.5 mA, p < 0.99). In a preliminary study to assess acquisition rate of corneal kindling, 8/8 seized mice displayed fully generalized Stage 5 seizures, versus 5/8 non‐seized mice. Furthermore, seized mice had much shorter latencies to generalized seizures than non‐seized mice. Conclusions: The results obtained in the present study demonstrate altered seizure susceptibility following recovery from seizures associated with TMEV‐ induced encephalitis. Furthermore, they suggest that these animals may be more susceptible to epileptogenesis following subsequent challenges. Although preliminary, these results further support this as a model of encephalitis‐ induced hyperexcitability. (Supported by Robert and Joyce Rice Epilepsy Fellowship (KAS); Max Abrams Memorial Fellowship from Epilepsy Foundation of America (KAS).) 1 Nikolaus Sucher, 2 Steven C. Schachter, 3 Nam‐In Baek, 1 Bombi Lee, 2 Jongbae Park, 1 Frances Jensen, 4 H. Steve White, 5,6 Lauren Murphree, and 6 James Stables ( 1 Department of Neurology, Children's Hospital, Boston, MA ; 2 Osher Institute, Harvard Medical School, Boston, MA ; 3 Graduate School of Biotechnology & Plant Metabolism Research Center, Kyung Hee University, Seoul, Korea ; 4 Pharmacology and Toxicology, University of Utah, Salt Lake City, UT ; 5 SRA International, Fairfax, VA ; and 6 Anticonvulsant Screening Project, NINDS, Bethesda, MD ) Rationale: Herbal medicines for epilepsy have a long tradition and continue to be used widely. An international team has been assembled to identify Asian herbal medicines used for epilepsy, isolate their constituent compounds, and test these compounds in established animal models of epilepsy and in vitro assays to determine their potential mechanisms of action. We report the preliminary findings from the first herbal medicine to be evaluated. Methods: Five pure compounds were isolated from an Asian herbal extract commonly used for epilepsy and were designated ADD 377008, 377009, 377010, 377011, and 377012 by the NIH/NINDS Anticonvulsant Screening Project. Compounds were assessed in the mouse maximal electroshock and s.c. Metrazol (pentylenetetrazol; s.c. Met) models of generalized tonic‐clonic and myoclonic seizures, respectively, following p.o. administration to Swiss‐Webster mice. Minimal motor impairment was evaluated in the rotarod test. Four of the compounds were tested for effects against NMDA receptor‐mediated excitotoxicity in primary neuronal cultures derived from rat cortex (Sun et al., NeuroSignals 2003;12:31–8). Results: ADD 377010 protected 5/5 animals after a dose of 300 mg/kg (i.p.) in the scMET model and had no effect in the MES model. No other compound was active up to 300 mg/kg in either model, and none caused motor impairment. All four compounds tested prevented NMDA receptor‐mediated cytotoxicity when co‐applied with NMDA. Conclusions: Compounds isolated from an Asian herbal medicine demonstrate promising activity in the NMDA assay, and one also effectively blocked seizures in the scMET model, suggesting it may be responsible for the anticonvulsant properties of its parent extract. Further evaluations of compounds isolated from herbal medicines used for epilepsy are warranted. (Supported by an American Epilepsy Society Research Infrastructure Award. The investigators gratefully acknowledge the support of the NINDS Anticonvulsant Screening Project.) 1,2 Lucie Suchomelova, 1 Roger Baldwin, and 1,2 Claude G. Wasterlain ( 1 Epilepsy Research, VA GLAHS, Los Angeles, CA ; and 2 Neurology, UCLA, Los Angeles, CA ) Rationale: The role of febrile seizures (FS) in epileptogenesis and the contribution of fever to the long‐term consequences of prolonged FS and febrile status epilepticus (SE) are poorly understood. Controversy remains regarding any causal relationship between prolonged FS and temporal lobe epilepsy. This study investigated the contribution of hyperthermia per se to SE‐induced epileptogenesis, using the lithium‐pilocarpine model of SE in immature rats, and carefully controlling for brain temperature and seizure severity. Methods: SE was induced in rats at postnatal day 10 (P10) by administration of lithium (3mEq/kg, i.p.), and 24 hours later by pilocarpine (60 mg/kg, s.c.). Body temperature was maintained during 30 min of SE at 39 ± 1°C (hyperthermic animals) or at 35 ± 1°C (normothermic animals). SE was terminated by diazepam (0.5 mg/kg, i.p.). Control animals received lithium and diazepam in presence of hyperthermia or normothermia. The course of SE was recorded by video/EEG for 24 hours. The severity of SE was assessed by measuring the following parameters: duration of SE (time of last seizure), and total seizure time (time spent in seizures subtracting interictal time) using Harmonie Software. Separate groups of animals, subjected to SE at P10 without surgery, were monitored four months later for the occurrence of spontaneous recurrent seizures (SRS) by telemetry/videotape continuously for one month. Results: We were able to control brain and body temperature during and after SE in both study groups: hyperthermic (39 ± 1°C) and normothermic (35 ± 1°C). Terminating SE by diazepam was effective in both groups (10.2 ± 1.5 min in hyperthermic animals and 12.3 ± 2.2 min in normothermic animals). Total seizure time (54.5 ± 15.2 min after SE in hyperthermic animals and 56.7 ± 12.1 min after SE in normothermic animals) and duration of SE (87.2 ± 23.2 min after SE in hyperthermic animals vs. 95.2 ± 28.2 min after SE in normothermic animals) were similar between groups. Only hyperthermic animals developed convulsive seizures (50%). Both groups of animals that experienced SE at P10 developed electrographic seizures 4 months later (83 vs 55%). However, the frequency and seizure severity were significantly higher in hyperthermic animals (12.5 ± 3.5 vs 4.2 ± 2.0 SRS/day). Conclusions: Since the precise role of hyperthermia during SE in long‐term consequences such as epileptogenesis and neuronal injury has never been adequately defined, this study tightly controlled seizure duration and seizure severity in order to isolate hyperthermia as the main variable and to study its consequences. In this model, hyperthermia greatly increased the epileptogenicity of SE, leading to subsequent behavioral seizures not seen in the normothermic group, and increased the severity and incidence of spontaneous electrographic seizures. (Supported by Grant NS13515 from NINDS, National Institutes of Health, and by the Research Service of the Veterans Health Administration.) 1 Jessica R. Sudbury, and 1,2 Massimo Avoli ( 1 Neurology and Neurosurgery, Montreal Neurological Institute, McGill University, Montreal, QC, Canada ; and 2 Human Physiology and Pharmacology, La Sapienza University, Rome, Italy ) Rationale: The participation of the insular cortex (IC) in temporal lobe epilepsy (TLE) has long been suggested from studies of human epileptic patients; however, its network interactions with the perirhinal cortex (PC) and hippocampus remain uninvestigated. Here, we used a hippocampus‐parahippocampus‐IC rat brain slice to provide the first characterization of network epileptiform interactions among these structures during application of the convulsant 4‐aminopyridine (4‐AP). Methods: Brain slices (450 μm) were prepared from adult male Sprague‐Dawley rats. Field potential recordings were acquired using ACSF‐filled glass electrodes positioned within area CA3 of the hippocampus, the PC and the IC. 4‐AP (50 μM) was bath applied to induce epileptiform synchronous network activity. Results: 4‐AP induced three types of epileptiform activity: (i) fast interictal events (duration = 0.1–0.2 s; interval of occurrence = 0.7–1.2 s) that were mainly observed in CA3 as well as (ii) slow interictal (duration = 0.5–2.0 s; interval of occurrence = 4.0–21.7 s) and (iii) prolonged ictal‐like (duration = 11.3–48.7 s; interval of occurrence = 17.9–332 s) that occurred synchronously among the PC and the IC and occasionally entered the hippocampus. Only slow interictal discharges were present in PC and IC in 15/30 slices. Moreover, slow interictal and ictal events were generated by the IC following surgical isolation. Ictal discharges were abolished by NMDA receptor antagonism while interictal activity persisted. Further antagonism of non‐NMDA glutamatergic receptors abolished interictal discharges and disclosed slow field potentials that recurred at intervals of 11.4–28.0 s in all structures. Bath application of the μ‐opioid receptor agonist (D‐Ala2,N‐Me‐Phe4‐,Gly‐ol5)enkephalin (DAGO, 10μM) significantly reduced the occurrence of these slow potentials in all structures (n = 5); this effect was reversible with naloxone methiodide (20μM). Application of the GABAA receptor antagonist picrotoxin (50μM) abolished these field potentials. Conclusions: Our results demonstrate that epileptiform activity occurs synchronously among PC and IC networks, while analogous activity can be generated independently within the isolated IC. These observations suggest that the network characteristics necessary and sufficient to initiate epileptiform network events are present within the IC. (Supported by: Funded by CIHR (Grant 8109), CURE and the Savoy Foundation.) 1 T. Sutula, 1 J. Ockuly, 1 C. Stafstrom, and 1 A. Roopra ( 1 Department of Neurology, University of Wisconsin, Madison, WI ) Rationale: During investigation of mechanisms of the ketogenic diet, the glycolysis inhibitor 2DG had unexpected anticonvulsant and antiepileptic effects in hippocampal slices and against seizures evoked by kindling of the perforant path. Anticonvulsant and antiepileptic properties of 2DG were further characterized by kindling of olfactory bulb and in screening models in cooperation with the NINDS Antiepileptic Screening Program (ASP). Cognitive and systemic toxicity was examined in rats receiving 2DG for as long as 6 months. Methods: Rats received olfactory bulb or perforant path kindling stimulation twice daily with 1‐sec trains of 60‐Hz 1‐ms pulses at the lowest intensity evoking ADs by a standardized protocol. Interictal and ictal burst frequencies were examined in rat hippocampal slices in 7.5 mM (K+)o. 2DG was evaluated in NINDS ASP screening models in rats and mice including maximal electroshock, subcutaneous metrazol, 6‐Hz stimulation, and Frings audiogenic mice. Cognitive effects were evaluated in rats in the Morris water maze. Results: Application of 2DG (10mM) to hippocampal slices in 7.5 mM (K+)o reduced CA3 interictal bursts by ∼40% and ictal bursts by 60%. 2DG (250 mg/kg) 30 min before kindling of perforant path in rats induced a 2‐fold increase in AD threshold (an anticonvulsant effect) and a 2‐fold increase in the number of ADs required to evoke Class V seizures (an antiepileptic effect). In contrast, 2DG prior to olfactory bulb kindling also resulted in 2‐fold slowing of kindling progression, but had no effect on AD threshold. Initial evidence of anticonvulsant activity was observed in preliminary screening against seizures evoked by subcutaneous metrazol in rats, by 6‐Hz stimulation in mice, and in Fring's audiogenic mice, but not against seizures evoked by maximal electroshock. Additional testing is underway to further characterize the activity found in these models. 2DG at 2 gm/kg daily for 2 weeks in rats had no effect on Morris water maze performance, and body weights of rats receiving 0.5 gm/kg daily for 6 months did not differ from controls. Conclusions: 2DG has novel anticonvulsant and antiepileptic activity against experimental seizures compared to currently available anticonvulsants and a favorable preclinical toxicity profile. Acute anticonvulsant action was observed in vitro against both interictal and ictal discharges. Robust chronic antiepileptic action against kindling progression was independent of the seizure initiation site, but anticonvulsant action against AD threshold was stimulation site specific. The in vivo chronic effects on kindled seizures have been associated with a novel mechanism of antiglycolytic metabolic suppression of seizure‐induced BDNF and trkB expression by the NADH sensor CtBP and the transcription factor NRSF, which regulates neuronal gene expression. (Supported by NINDS R01 25020, the Wisconsin Alumni Research Foundation, and the NINDS Antiepileptic Screening Program.) 1,2 Janos Szabadics, 2 Gabor Tamas, and 1 Ivan Soltesz ( 1 Department of Anatomy & Neurobiology, University of California, Irvine ; and 2 Department of Comparative Physiology, University of Szeged, Hungary ) Rationale: A distinguishing feature of neurogliaform cells is their extremly dense axonal arborization. Moreover, neurogliaform cells are the only known cortical cell type that can reliably activate postsynaptic GABAA and GABAB receptors after discharging a single action potential. Using three‐dimensional light microscopic reconstructions and correlated electron microscopic investigations, we found that neurogliaform cells formed synapses (62 ± 28 μm) closer to the soma of layer 2/3 pyramidal cells than Martinotti cells (102 ± 48 μm) that evoked faster inhibitory responses. Thus, the slow kinetic properties of GABAA responses evoked by neurogliaform cells can not be explained exclusively by the dendritic localization of the synapses. However, there is a twofold difference in unitary axon densities between neurogliaform cells and other interneurons (0.954 ± 0.299 and 0.453 ± 0.015 μm axon in 1000 μm3).We propose that neurogliaform cells release GABA all over their dense axonal field to activate GABA receptors on dendrites present in the axonal cloud. Methods: To test this hypothesis, we studied the physiological properties of responses between neurogliaform cells, other identified interneurons and layer 2/3 pyramidal cells using whole cell patch clamp recording in neocortical slices with the AMPA, NMDA and GABAB receptors blocked. Results: Neurogliaform cells evoked slower GABAA currents (10–90% rise time: 4.5 ± 1.6 ms, weigthed decay time constant: 42 ± 9.7 ms, amplitude: 97 ± 69 pA) than basket cells (10–90% rise time: 0.7 ± 0.2 ms, weigthed decay time constant: 8.1 ± 1.9 ms, amplitude: 130 ± 109 pA). The variability in amplitudes of neurogliaform cell‐evoked IPSCs (0.14 ± 0.04, S.D./mean) was much less than in the basket cell‐evoked responses (0.29 ± 0.1). Furthermore, gradual changing of extracellular calcium resulted in continuous increases in the amplitude of the responses. Blocking of GABA transporter 1 by NO‐711 dramatically increased the decay time constant of neurogliaform cell‐evoked IPSCs (by 203 ± 110%) and did not affect basket cell‐evoked IPSCs. The low affinity, competitive antagonist of GABAA receptors, TPMPA had a larger effect (58 ± 17%) on the amplitude the neurogliaform cells effect compared to basket cells (35 ± 17%). Alfa1 subunit specific GABAA agonists (zolpidem, Cl218,872) had significant effets on the neurogliaform cell‐evoked responses. However, the properties of this modulatory action was different than on basket cell‐evoked IPSCs. Conclusions: These results indicate that a spill‐over‐like mechanism of GABA release takes place following single action potentials in neurogliaform cells, in contrast to the more conventional basket cell synapses. These data suggest that neurogliaform play highly specialized role in regulating neuronal activity in the cortical networks. (Supported by NIH NS35915 to IS & GT; OTKA T049535 and HHMI 55005625 to GT; George E. Hewitt Foundation and Boehringer Fonds to JS.) 1 C. Akos Szabo, 2 Shalini Narayana, 2 Peter V. Kochunov, 2 Crystal Franklin, 1 Koyle Knape, 2 M. Duff Davis, 2 Peter T. Fox, 3 M. Michelle Leland, and 3,4 Jeff T. Williams ( 1 South Texas Comprehensive Epilepsy Center, UTHSCSA, San Antonio, TX ; 2 Research Imaging Center, UTHSCSA, San Antonio, TX ; 3 Southwest National Primate Research Center, San Antonio, TX ; and 4 Department of Genetics, Southwest Foundation for Biomedical Research, San Antonio, TX ) Rationale: The baboon (Papio hamadryas spp) offers a natural primate animal model of photosensitive generalized epilepsy. This study compared changes in cerebral blood flow (CBF) during intermittent light stimulation (ILS) between photosensitive and asymptomatic baboons. Methods: Six photosensitive and four asymptomatic baboons, matched for age, gender and weight, were selected based upon previous scalp EEG evaluation. Continuous intravenous ketamine (5–13 mg/kg) was used for sedation. Subjects underwent five sequential blood flow PET studies within 60 minutes using 20 mCi 15O‐labelled water. Images were acquired in 3D mode (CTI/Siemens HR+ scanner, 63 contiguous slices, 2.4mm thickness). Three resting scans were alternated with two activation scans. ILS was performed at 25 Hz for 60 seconds before to 60 seconds after the start of an activation scan. PET images were coregistered with MRI (3T Siemens Trio, T1‐weighted 3D sequence, TE = 6, TR = 14, TI = 800msec, flip angle 19°, NEX = 6). PET scans were reviewed and corrected for motion artifact. Resting scans were contrasted with activation scans and averaged independently for both groups. Results: The controls showed greatest ILS‐induced activation in the left middle frontal and inferior temporal gyri, left brainstem structures and right postcentral gyrus, and bilaterally in the occipital lobes, posterior cingulate gyrus and cerebellum. In contrast, the photosensitive animals showed strongest ILS activation in the right anterior cingulate and medial orbital gyri, amygdala, globus pallidum, left inferior and superior temporal gyri, and the pineal gland. A striking finding was the absence of occipital activation during ILS in the photosensitive animals. Deactivations were noted in the right orbitofrontal and anterior cingulate cortices in the controls, and in the posterior cingulate gyrus, brainstem and cerebellum of the photosensitive animals. Conclusions: The patterns of ILS‐induced CBF changes differed between control and photosensitive groups. These differences of activations and inhibitions suggest involvement of specific cortical‐subcortical networks in photosensitivity. (Supported by National Institutes of Health (P51 RR13986 through the SFBR to CAS, 1 R01 NS047755 to JTW, and by facilities constructed with support from Research Facilities Improvement Program Grants C06 RR15456 and C06 RR014578 from the National Center for Research Resources).) 1 Patricia Szot, 1 Sylvia S. White, 2 Krista Gilby, and 2 Dan C. McIntyre ( 1 MIRECC, VA Puget Sound Health Care System, Seattle, WA ; and 2 Department of Psychology, Carleton University, Ottawa, ON, Canada ) Rationale: Kindling is a progressive increase in seizure activity following the daily application of focal, low‐intensity electrical stimulation of different forebrain regions. Two strains of rats were developed to exhibit different rates of amygdala kindling, and were called Fast and Slow kindlers. These animals also show different response profiles to α2‐adrenoreceptor (AR) agonists. Therefore, the objective of this work was to examine α2‐AR binding sites and subtype mRNA expression in the CNS of naive Fast and Slow rats. Methods:α2‐AR binding sites were measured in naive Fast (n = 8) and Slow (n = 8) kindlers with the α2‐AR antagonist (3H)RX821002. α2‐AR binding sites were measured in the locus coeruleus (LC), neocortex, bed nucleus of the stria terminalis, thalamus, septum, amygdala, habenula and hippocampus. α2‐AR is composed of three different subtypes. α2A‐ and α2C‐AR mRNA were determined in the CNS of Fast and Slow kindlers by in situ hybridization with oligonucleotides. α2A‐AR mRNA was measured in the LC, cortex, septum, hypothalamus and amygdala, and α2C‐AR mRNA was measured in the LC, cortex, hippocampus, amygdala, cerebellum, caudate and nucleus accumbens. Results:α2‐AR binding sites, as determined by (3H)RX821002, were not statistically different between Fast and Slow kindlers in all regions of the CNS. α2A‐AR mRNA expression, the subtype responsible for (3H)RX821002 binding, therefore, was not different. However, α2C‐AR mRNA expression was different between Fast and Slow kindlers in many regions of the CNS, including nucleus accumbens, dentate gyrus granule cell layer, cingulate gyrus, cerebellum and neocortex layer II: in these cases, α2C‐AR mRNA expression was significantly elevated in the Slow compared to Fast kindlers. Conclusions: These data indicate a signifcant increase in several brain structures in the α2C‐AR mRNA expression in Slow versus the Fast kindlers. These differences in α2C‐AR mRNA may explain the different α2 adrenergic pharmacological profiles between the two strains, and speak additionally to adrenergic involvement in epileptogenesis. (Supported by VA Puget Sound health Care Center and CIHR.) 1 Mayuko Takaki, 2 Yuto Ueda, 3 Akira Nakajima, 4 Yoshiya Murashima, 2 Taku Doi, 2 Keiko Nagatomo, and 5 James Willmore ( 1 Section of Integrative Physiology, Department of Medical Sciences, Faculty of Medicine, University of Miyazaki, Miyazaki, Japan ; 2 Department of Psychiatry, Faculty of Medicine, University of Miyazaki, Miyazaki, Japan ; 3 Section of Chemistry, Department of Medical Science, Faculty of Medicine, University of Miyazaki, Miyazaki, Japan ; 4 Department of Neural Plasticity, Tokyo Institute of Psychiatry, Tokyo, Japan ; and 5 Department of Pharmacology and Physiology, Saint Louis University School of Medicine, St. Louis, MO ) Rationale: The EL mouse is an inbred mutant strain that is used as an animal model of secondarily generalized seizures. Epileptogenesis in EL mice is thought to be caused by collapse of redox state induced by either decreased antioxidant protection or excessive free radical formation. However, measurement of reductants in brain tissue extracts is problematic because synergistic actions of antioxidants such as ascorbic acid, alpha tocopherol or glutathione are effective in scavenging free radicals. Using in vivo methods we developed, we evaluated hippocampal antioxidant ability in EL mice with X‐band Electron Paramagnetic Resonance (EPR) spectroscopy and microdialysis. Methods: Using sodium pentobarbital anesthesia, each mouse underwent stereotactic implantation of guide cannula into the ventral hippocampus. EL mice were injected i.p. with the nitroxide radical 3‐methoxycarbonyl‐2,2,5,5‐tetramethylpyrrolidine‐1‐oxyl (PCAM) at 677 micromol/kg (3.3 ml saline/kg). Following PCAM injection sequential measurements with EPR were used to monitor levels of nitroxide radicals in the extracellular fluid of the hippocampus. In vivo antioxidant efficacy was estimated based on the principle that PCAM loses paramagnetism and that the decay rate of the EPR signal expressed as half life will reflect tissue antioxidant efficacy. Animals used included experimental groups of 6 five week old and 6 eight week old EL mice; all animals were interictal. Mice from the originating strain of EL mice, DDY mice of similar age, were used as controls. Results: EPR half‐life in the five week old mice was equal to control. However, by eight weeks there was significant prolongation in EPR half‐life in EL mice when compared to both control groups of DDY mice. None of the EL mice used in this experiment were subjected to seizure‐inducing stimulation and they remained in the interictal state. Conclusions: Since five week EL mice do not seize in response to tossing stimulus while by eight weeks EL mice are subject to seizures, we concluded that the decreased in vivo antioxidant ability is a fundamental change related to development of seizures. (Supported by Grant‐in‐Aid for Scientific Research (C) (2) (18591297) from the Ministry of Education, Science, Sport and Culture, Japan (to Y.U.).) 1 Nobuhiro Tanaka, 1 Hirochika Imoto, 1 Masami Fujii, 2 Jouji Uchiyama, 1 Hirosuke Fujisawa, 2 Kimihiko Nakano, 3 Ichiro Kunitsugu, 2 Takashi Saito, and 1 Michiyasu Suzuki ( 1 Neurosurgery, School of Medicine Yamaguchi University, Ube, Yamaguchi, Japan ; 2 Applied Medical Engineering Science, Graduate School of Medicine Yamaguchi University, Ube, Yamaguchi, Japan ; and 3 Public Health, School of Medicine Yamaguchi University, Ube, Yamaguchi, Japan ) Rationale: Local cortical cooling for the elimination of epileptic discharges (EDs) has recently become a focus of research. However, the effect of focal cooling on seizures arising from the temporal lobe has yet to be thoroughly investigated. We devised an apparatus for the cooling of the hippocampus in order to examine the effect of focal cooling on experimental hippocampal seizures. Methods: We developed a cooling needle (measuring 1 mm diameter and 6mm length) made of copper which was attached to a thermoelectric (Peltier) chip. Next, experiments were performed on adult male Sprague‐Dawley rats under halothane anesthesia. The cooling needle, a thermocouple and a needle electrode for the EEG recording were then inserted into the right hippocampus after performing a craniotomy. Kainic acid (KA) (1 mg/ml, 1 μl) was injected into the left hippocampus to provoke epileptiform discharges (EDs). Thereafter, focal cooling was started and it was continued for 5 minutes after confirming the appearance of EDs in the right hippocampus which had been provoked by the KA in the contralateral hippocampus. Results: The temperatures of the cooled site in the hippocampus were 9°C in the center, while they were under 25°C within 2.4 mm in radius from the cooling center in phantom experiments using the cooling device.The EDs began to significantly decrease in amplitude from 90 seconds after the start of cooling (p < 0.01). The mean amplitudes of the EDs were suppressed to 68.1 ± 4.8% at the end of the cooling and then stayed low values during the subsequent period. The temperatures of the cooling site were maintained at 33.1 ± 0.7°C before cooling and then they decreased to 14.5 ± 1.3°C finally. Conclusions: We confirmed that hippocampal cooling suppressed the EDs provoked by KA injection in the contralateral hippocampus. Our results therefore seem to suggest that hippocampal cooling may thus be a potentially useful minimally invasive therapy, which might thus be able to replace the need to perform a temporal lobectomy. 1 Kristina Surdova, 1 Erik Taubøll, 2 Anette Krogenæs, 2 Ellen Dahl, 2 Jan L. Lyche, 2 Irma C. Oskam, and 2 Erik Ropstad ( 1 Department of Neurology, Division of Clinical Neuroscience, Rikshospitalet‐Radiumhospitalet, University of Oslo, Oslo, Norway ; and 2 Department of Production Animal Sciences, Norwegian School of Veterinary Science, University of Oslo, Oslo, Norway ) Rationale: It is well known that valproate (VPA) affects endocrine function. There are, however, contradictory reports in the literature as to whether or not VPA affects the hypothalamic‐pituitary‐adrenal axis. Such effects, if present, are important since this hormonal axis is crucial in the homeostasis of fundamental bodily functions and in regulation of stress responses. In the present study the effects of VPA exposure on basal and stress‐induced cortisol was investigated in male adolescent goats. Methods: Twenty‐three bucks aged 2 months at the start of experiment were given VPA mixture (n = 16) or control solution (n = 7) perorally twice daily for 8 months. VPA mixture was given in single doses up to 62.5 mg/kg to sustain therapeutic plasma concentrations (300–600 μmol/l). In Experiment 1, blood samples were analyzed for cortisol weekly from weeks 35 to 50 of the year. In experiment 2, bucks aged 9 months were exposed to a period of mild stress associated with blood sampling from indwelling jugular vein catheters every 15 minutes for 7 hours. The effects of VPA on plasma cortisol concentration were studied in repeated measured mixed model analyses. Results: VPA treated bucks in experiment 1 had significantly lower basal cortisol concentrations than control animals during the study period (F1,14= 6.22, p = 0.026). A significant time effect (F7,98 = 5.92, p < 0.0001) was found with a decrease in basal cortisol plasma concentrations towards the end of the study period. A significant time/treatment interaction (F7,98= 2.63, p = 0.016) was found with a more pronounced decrease in plasma cortisol in the control group than in the treated group. In experiment 2 the cortisol response to mild stress was different in the VPA group compared to control. Overall, there was a lower cortisol response in the VPA group compared with controls. Mean basal plasma cortisol concentration was 2 ng/ml in both groups. During the first period of mild stress plasma cortisol concentration in the control group was 4–8 ng/ml compared to 3–6 ng/ml in the VPA group. In both groups cortisol decreased after 1.5 hours of blood sampling. The group difference was significant when only main effects were included in the model (F1,14= 47.2, p < 0.0001). Conclusions: VPA exposure was associated with a significant reduction in both basal and stress‐induced plasma cortisol concentration in male adolescent goats. The observed effects can be a potential problem with long‐term VPA treatment. On the other hand, it may be speculated that the reduction in cortisol levels and reduced response to mild stress could be related to the mood stabilizing effect of VPA. 1 Michael R. Taylor, 1 Matthew T. Dinday, 2 Sally Chege, and 1,2 Scott C. Baraban ( 1 Neurological Surgery, UCSF, San Francisco, CA ; and 2 Graduate Program In Neuroscience, UCSF, San Francisco, CA ) Rationale: Studies from humans, rodents, and fruit flies have identified dozens of single gene mutations that result in heritable forms of epilepsy or increased seizure susceptibility. Although this genetic basis for epilepsy and seizure susceptibility is well established, there have been no efforts to identify gene mutations conferring seizure resistance. Large‐scale mutagenesis screening in simple vertebrates offers a powerful approach to further studies of the genetic basis of epilepsy. Here we describe the initial results of the first forward‐genetic screen to identify seizure resistant zebrafish mutants. Methods: Approximately 5,000 F1 adult fish were generated from matings between ENU‐treated males and wild‐type females. From these founders 1,896 F2 families were generated and ∼500,000 F3 larvae were used for screening. Using a behavioral assay, we isolated several seizure‐resistant (SR) familes based on their ability to survive prolonged exposure to the convulsant pentylenetetrazol (PTZ). All SR mutants were further characterized by examining resistance to another convulsant (pilocarpine), tracking behavioral responses, and recording electrical activity. In addition, mutants were genetically mapped by bulked‐segregant analysis, and recombination analysis with microsatellite markers was used for fine‐mapping. Results: Six seizure‐resistant mutant families were identified in a forward‐genetic screen over a four‐year period. One mutant, s198, was also shown to be resistant in a pilocarpine survival assay, suggesting multi‐convulsant resistance. Three of the mutants exhibited reduced behavioral activity upon acute PTZ exposure (s139, s333 & s387), and one of these also showed decreased c‐Fos expression by RT‐PCR (s333). Electrophysiology data on mutant s334 revealed an inability to generate long‐duration ictal‐like discharge in response to PTZ or 4‐aminopyridine. Mutant s198 was randomly chosen for mapping studies. Analysis of 192 microsatellite markers, evenly distributed throughout the zebrafish genome, placed the mutated gene on linkage group 24. Recombination analysis with 658 meioses localized the mutation within a 1.1‐centimorgan region (approximately 800 kb). Conclusions: In conclusion, SR zebrafish mutants were successfully identified in a large‐scale forward‐genetic screen. Mutants were isolated with multi‐convulsant resistance, reduced seizure behaviors, and no ictal‐like discharge. Linkage studies on one mutant, s198, have been completed. Further identification and characterization of the defective genes in these mutants will present a more complete understanding of the genetics of seizures, and provide a unique opportunity for developing a genetically‐based cure for epilepsy. (Supported by Epilepsy Foundation of America/Milken Family (M.R.T.) and Klingenstein Fund (S.C.B.).) 1 Michael A. Treiman, 1 Jie Wu, 1 Steven T. Marsh, 1,2 Levi B. Good, 1 Norman Wang, 1 Ryoichi Kimura, 1 Kevin Ellsworth, and 1,2 David M. Treiman ( 1 Neurology Research, Barrow Neurological Institute, Phoenix, AZ ; and 2 Harrington Department of Bioengineering, Arizona State University, Tempe, AZ ) Rationale: Treiman et al. (Epilepsy Res 5:49–60, 1990) described five stages of progressive electroencephalographic (EEG) changes during clinical and experimental status epilepticus (SE), which reflect the underlying severity of the episode of SE. Longer duration of SE and later EEG stages translate to increased problematic cessation of seizures, perhaps due to progressive attenuation of GABA‐mediated inhibition. We hypothesized that marked electrophysiologic changes in GABAA receptor function correlate with a specific EEG stage during experimental SE in rats. Methods: Ten‐week‐old Sprague‐Dawley rats were injected (i.p.) with 3 mM LiCl followed by 30 mg/kg pilocarpine 24 h later to induce SE. Animals were sacrificed before initiation of SE (control), or during EEG Stages I, III, or V, and brains were rapidly removed, placed in cold (4°C) artificial cerebrospinal fluid (ACSF) for 1 min, sliced using a vibratome, and then oxygenated for 1 h in ACSF at room temperature. Paired‐pulse responses with interpulse intervals (IPI) of 30, 50, 100 and 150 msec were induced by stimulation of Schaffer collaterals and recorded from the cell‐body layer of the CA1 region of the hippocampus. γ oscillations (30–80 Hz) were then elicited by tetanic stimulation at a voltage twice that of threshold and were also recorded from CA1. Some Stage V hippocampal slices were perfused with 100 μM GABA or 1 μM diazepam following initial electrophysiological recordings. Results: Paired‐pulse stimulation resulted in increased inhibition in Stage I hippocampal slices compared to control slices (p < 0.05, IPI = 50 msec) but decreased inhibition in Stage V slices compared to Stage I slices (p < 0.05, IPI = 50 msec). γ oscillations were progressively attenuated in slices from later EEG stages. Amplitude (p < 0.01), duration (p < 0.05) and spike count (p < 0.05) were all reduced in Stage V slices compared to control slices. Perfusion with GABA, and to a lesser extent with diazepam, resulted in recovery of γ oscillations in Stage V slices. Conclusions: These results are consistent with other reports of attenuation of GABA‐mediated inhibition during prolonged SE, which may explain the increased refractoriness of later stages of SE to treatment. In addition, our results indicate a possible impairment of presynaptic release of GABA occurred in this model. This observation suggests GABA reuptake inhibitors could be useful for the treatment of refractory status epilepticus. We also report the new finding that paired‐pulse inhibition is greater during Stage I compared to control slices and speculate that this may represent an initial compensatory response to excessive excitation before GABA‐mediated inhibition begins to deteriorate as SE progresses. (Supported by Barrow Neurological Foundation.) 1,2 David M. Treiman, 1 Norman Wang, 1,2 Levi B. Good, 1 Steven T. Marsh, and 1 Kevin J. Garvey ( 1 Neurology Research, Barrow Neurological Institute, Phoenix, AZ ; and 2 Harrington Department of Bioengineering, Arizona State University, Tempe, AZ ) Rationale: There is a need for more effective drugs for the treatment of status epilepticus (SE) but testing potential SE drugs in human patients is difficult. The cobalt‐homocysteine model of generalized convulsive status epilepticus (GCSE) (Epilepsy Res 2:79–86, 1988) can be used to predict the median effective dose (ED50) and median effective concentration (EC50) of drugs against SE. We used this experimental model to study the potential efficacy of levetiracetam in the treatment of GCSE. Methods: 135 male 200–300 gm male Sprague‐Dawley rats were implanted with 4 stainless steel epidural electrodes for EEG recording. 2.5 mg powered cobalt was placed under the left frontal electrode. When the rat exhibited interictal polyspikes and/or right UE twitching, SE was induced by injecting 8 mM homocysteine thiolactone IP. Immediately after the 2nd GTC seizure, levetiracetam (LTA) was administered IP and was considered effective if no further seizures occurred 10–30 minutes after injection. Animals were euthanized at 30 minutes after onset of SE and blood and brains collected for LTA determination by HPLC. Because LTA was not effective alone, its ED50 was determined after administration of 0.1 mg/kg diazepam (ED50 of DZM alone =∼ 5 mg/kg), using doses from 0 to 1000 mg/kg. The neuroprotective potential of LTA was determined by comparing visual‐spatial performance in a Morris Water maze at various EEG stages of SE with and without LTA (500 mg/kg). Results: LTA alone at doses of 10 to 6000 mg/kg was not effective at stopping GCSE in the cobalt‐homocysteine model. The ED50 of LTA, when given in combination with 0.1 mg/kg DZM, was 550 mg/kg (CI 385–800 mg/kg), based on doses from 0–800 mg/kg. Paradoxically, at 1000 mg/kg only 14% of the rats stopped GCSE. 500 mg/kg LTA had no neuroprotective effect on water maze performance in rats in which SE was stopped at EEG stage V, whether given during SE (Stage III), or when SE was stopped at Stage V. Serum and brain LTA concentrations 30 minutes after administration will be shown. Conclusions: LTA is not effective alone in the cobalt‐homocysteine model of GCSE. Its efficacy is enhanced by prior‐ or co‐administration of a subtherapeutic dose of diazepam, but the ED50 is still ∼ 550 mg/kg. Whether or not LTA should be considered for use in human SE will at least partially depend on the serum concentration that corresponds to the ED50 of 550 mg/kg in this model. If the serum concentration is in a range safe for human administration, it can serve as a target concentration in further study of the potential of LTA in the treatment of human SE. It may be that LTA will be an effective second treatment that should be used to provide long‐term protection against relapse in human GCSE after initial use of a benzodiazepine such as diazepam or lorazepam. (Supported by UCB and the Barrow Neurological Foundation.) 1 Andrew K. Tryba, and 2 Charles J. Marcuccilli ( 1 Physiology, Medical College of Wisconsin, Milwaukee, WI ; and 2 Pediatric Neurology, Children's Hospital of Wisconsin, Milwaukee, WI ) Rationale: The neurotrophin, brain derived neurotrophic factor (BDNF), plays a critical role in neural development, migration, plasticity, recovery from injury and learning. In epileptogenic brain areas, such as the temporal lobe, BDNF is up‐regulated during repeated seizures, whereby it is proposed to enhance glutamatergic synaptic transmission and decrease GABAergic inhibition, enhancing the overall excitability of the network (Neuroscientist. 2005 11(4):282–7). BDNF additionally enhances intrinsic neuronal membrane excitability (Nature 1999;401 (6756):918–21). Thus, BDNF has been proposed to promote epileptiform activity (Trends Neurosci. 2001; 24(1):47–53). Although, BDNF over‐expression and exogenous application lead to reduced seizure threshold in experimental models where inhibition is reduced, no one has before shown that BDNF alone induces cortical epileptiform activity. Here, we test the hypothesis that BDNF application can trigger epileptiform activity. Methods: We used brain slice preparations containing temporal lobe cortex of CD‐1 mice (age P10‐P15), where we simultaneously recorded network population activity and from individual cortical neurons using whole cell current clamp techniques. Results: BDNF application (50ng/mL) triggered paroxysmal depolarization shift (PDS) bursting (n = 9/10); PDS bursts were NMDA‐dependent and blocked by the NMDA antagonist, CPP. BDNF enhanced bursting properties in intrinsic bursting (IB) neurons (n = 3), but not in non‐rhythmogenic neurons (n = 7). Application of BDNF also induced seizure‐like population activity in some (n = 8) preparations. Conclusions: Our data support our working hypothesis that BDNF can preferentially enhance bursting properties of rhythmogenic neurons that may trigger SLA. Additionally, BDNF can trigger PDS bursting, one of the hallmarks of epilepsy. Thus, our results suggest that BDNF signaling may be involved in the genesis of epilepsy. (Supported by NIH RO1 HL079294–02 (AKT).) 1,2,3 Pamela A. Valentine, 1 Elizabeth J. Jensen, and 2 Lana J. Ozen ( 1 Department of Clinical Neurosciences, University of Calgary, Calgary, AB, Canada ; 2 Department of Psychology, University of Calgary, Calgary, AB, Canada ; and 3 Neuroconnections Program, Hotchkiss Brain Institute, Calgary, AB, Canada ) Rationale: Deep brain stimulation (DBS) has a long history of therapeutic use in the treatment of a number of neurological disorders, perhaps most effectively for movement disorders such as Parkinson's disease. Neural stimulation may also provide a promising innovative technique for the treatment of pharmacoresistant seizure disorders. While there is experimental and clinical evidence that direct electrical stimulation of deep brain structures can prevent or decrease seizure activity, fundamental questions remain to be resolved including the identification of the most effective brain sites and stimulation parameters, long‐term consequences and potential side effects. The mechanism underlying deep brain stimulation is as yet unknown, although it is thought that the application of high‐frequency stimulation results in a net inhibitory effect that serves to desynchronize activity, effectively abolishing pathologically synchronized epileptiform activity. This may explain why stimulation of a number of structures have been shown to have anticonvulsant properties. Methods: In this study we examined the effect of high frequency stimulation (applied to a number of subcortical sites) on seizure thresholds of electrically kindled temporal lobe seizures. The DBS consisted of unipolar, biphasic square wave pulses, with pulse durations of 300 μs, delivered at a frequency of 100 Hz and an intensity of 200 μA. The DBS was applied 1 second prior to the application of the kindling stimulation to ensure effective activation prior to the seizure inducing stimulus. The seizure activity was scored with respect to the afterdischarge (AD) duration and seizure severity. Results: Preliminary data revealed DBS applied to the entopeduncular nucleus resulted in a strong anticonvulsant effect on the seizure threshold and AD characteristics. We observed the increase in seizure threshold to occur over time which was observed to commence at about the 12th kindling session. The nominal progression in AD characteristics typically observed during kindling were largely absent, with the AD duration becoming shorter, the spike frequency lower and the spike complexity more simplistic. Conclusions: We conclude there may be strong anticonvulsant effects of DBS on temporal cortex seizures that are dependent on the site of stimulation. (Supported by Natural Sciences and Engineering Council of Canada (NSERC), Faculty of Medicine, Hotchkiss Brain Institute, University of Calgary.) 1,2 Erwin A. Van Vliet, 1 Silvana da Costa Araujo, 3 Sandra Redeker, 1 Rosalinde Van Schaik, 3 Eleonora Aronica, and 1,2 Jan A. Gorter ( 1 Center for Neuroscience, Swammerdam Institute for Life Sciences, Amsterdam, Netherlands ; 2 Epilepsy Institute of the Netherlands, Heemstede, Netherlands ; and 3 Department of (Neuro) Pathology, Academic Medical Center, Amsterdam, Netherlands ) Rationale: Disruption of the blood‐brain barrier (BBB) has been found to be associated with various neurological disorders, including the acquired epilepsies such as temporal lobe epilepsy (TLE). However, it is not known to what extent the BBB integrity changes during epileptogenesis and whether alterations in BBB permeability may contribute to the development of epilepsy. To get more insight into this, we determined BBB permeability in epileptic rat and human and studied the possible consequences of BBB opening for the subsequent progression of TLE. Methods: BBB permeability was investigated in the rat at different time points after electrically induced SE using various tracers that bind to albumin. The cellular localization of albumin was further investigated using immunocytochemistry in rat and human epileptic brain specimens. In addition, we investigated whether BBB opening was related to seizure progression. Results: In a rat model for TLE, in which epilepsy develops after an electrically induced status epilepticus (SE), the BBB was most affected 1 day after SE in limbic brain regions. Although a partial recovery took place during the latent period, the BBB was still permeable in the chronic epileptic phase. Tracer deposits (albumin) were found in microglia, astrocytes and neurons. Similarly, albumin extravasation in human was most prominent directly after a fatal SE in astrocytes and neurons, and to lesser extent in hippocampi of temporal lobe epilepsy patients. BBB permeability in chronic epileptic rats was positively correlated to seizure frequency. Artificial opening of the BBB by mannitol in the chronic epileptic phase induced a persistent increase in the number of seizures in the majority of rats. Conclusions: These findings indicate that BBB leakage occurs during the latent and chronic epileptic phase and suggest that this permanent disruption can contribute to the progression of epilepsy. (Supported by the Epilepsy Institute of The Netherlands.) 1 Lia C. Liefaard, 2 Bart A. Ploeger, 3 Adriaan A. Lammertsma, 1 Meindert Danhof, and 1,4 Rob A. Voskuyl ( 1 Division of Pharmacology, LACDR, Leiden, Netherlands ; 2 LAP&P Consultants BV, Leiden, Netherlands ; 3 Department of Nuclear Medicine & PET Research, VU Medical Centre, Amsterdam, Netherlands ; and 4 Epilepsy Institute of the Netherlands (SEIN), Heemstede, Netherlands ) Rationale: The GABAA‐receptor plays an important role in epileptogenesis. A key question is whether downregulation or changes in receptor properties are involved in reduced efficacy of antiepileptic drugs, in particular in pharmacoresistance. Recently, a full saturation approach was developed, in which the whole range of receptor occupancies, obtained from a single PET‐experiment per animal, was used to calculate both GABAA‐receptor density (Bmax) and affinity (KD). This method was used to examine the influence of kindling on Bmax and KD. Methods: Sprague Dawley rats were amygdala kindled up to 5 consecutive stage V seizures (n = 12). Implanted but unstimulated rats served as controls (n = 22). The PET experiment was performed 10–14 days after the last seizure. After injection of an excess amount of flumazenil (FMZ) to fully saturate the receptors, the concentration‐time curves of FMZ were measured in blood with HPLC‐UV and in brain with PET. Next, Bmax and KD were estimated using population pharmacokinetic (PK) modelling. The model consists of a blood, a tissue and 2 brain (“free” and “bound”) compartments. The total FMZ concentration in brain, as measured by PET, reflects the sum of “free” and “bound” concentrations. In the model, exchange between blood, tissue and “brain free” compartments is described by first order rate constants, and specific binding depends on both the concentration of free ligand and the concentration of receptors available for binding. Population PK modelling allows for simultaneous analysis of data from all animals, taking interindividual parameter variability into account. Results: In control rats, all model parameters, including specific binding of FMZ in the brain, as characterized by Bmax (14.5 ± 3.7 ng/ml) and KD (4.68 ± 1.5 ng/ml), could be estimated adequately. Kindling did not affect KD, but Bmax decreased to 64 ± 16% of control. This finding corresponds with a reduction in maximal effect of the allosteric modulator midazolam to 72% of control, as reported by Cleton et al. In addition, the volume of distribution of the brain was increased to 180 ± 27% of control, indicating that transport from the brain was decreased. Conclusions: This study shows that the full saturation technique, using a single injection of FMZ, allows detection of moderate reductions in GABAA‐receptor density and changes in brain PK. The decreased Bmax may underlie the reduced efficacy of midazolam in kindled rats. In addition, the study shows the feasibility to conduct (longitudinal) studies on epileptogenesis and pharmacoresistance development in chronic epilepsy models. (Supported by Christelijke Vereniging voor de Verpleging van Lijders aan Epilepsie and National Epilepsy Fund, #02–06.) 1 Yue Wang, 3 Hitten P. Zaveri, 1 Arko Ghosh, 2 Henning Beckstrom, 1 Nihal C. De Lanerolle, and 1 Tore Eid ( 1 Department of Neurosurgery, Yale University School of Medicine, New Haven, CT ; 2 Department of Anatomy, University of Oslo, Oslo, Norway ; and 3 Department of Neurology, Yale University School of Medicine, New Haven, CT ) Rationale: Glutamine synthetase (GS), which converts glutamate to glutamine, is deficient in specific areas of the sclerotic hippocampus in patients with mesial temporal lobe epilepsy (MTLE). A key question is whether hippocampal GS deficiency is causally related to the generation of seizures in MTLE. Methods: The GS inhibitor methionine sulfoximine (MSO) was infused chronically into the hippocampus of rats at the following doses: 1.0 mg/ml (n = 4), 2.5 mg/ml (n = 18), 5.0 mg/ml (n = 11), and 10 mg/ml (n = 39). Saline infused animals (n = 30) were used as controls. The animals underwent long‐term intracranial EEG and video monitoring for up to 83 days; the activity of GS was assessed biochemically; and the brain pathology was explored by Nissl‐ and silver‐staining of histological sections. Results: Infusion of MSO, but not saline, caused significant (p < 0.01) inhibition of brain glutamine synthetase. A dose of ≥ 2.5 mg/ml MSO consistently produced spontaneously recurrent seizures (stages I – IV) after a clinically silent interval of 5–9 days, whereas no seizures were observed in saline treated animals. Hippocampal gliosis and neurodegeneration were present in all MSO‐treated animals in a dose‐dependant manner (i.e. increasing severity in pathology change with higher dose of MSO). Conclusions: These studies suggest that the deficiency of hippocampal GS is a causative factor in the generation of spontaneously recurring seizures in MTLE, and that GS may represent a novel therapeutic target for this disease. 1 Andrew M. White, 1 Philip A. Williams, 1 Jennifer L. Hellier, 2 F. Edward Dudek, and 1 Kevin J. Staley ( 1 Department of Pediatrics, University of Colorado, Denver, CO ; and 2 Department of Physiology, University of Utah, Salt Lake City, UT ) Rationale: Following severe head trauma, 10 – 50% of survivors will develop epilepsy. Identifying the subset of individuals that will go on to have chronic seizures would greatly improve their clinical care. We used the repeated low‐dose kainate model of temporal lobe epilepsy to investigate the spectral content of the EEG before and after treatment (pre‐ictal) to determine if there are EEG predictors of future epilepsy. Methods: We used implantable radiotelemetry units to continuously monitor the EEG in freely behaving rats. Three sets of rats were observed: (1) rats treated until motor status (MSE) (n = 5), (2) rats treated until electrical status (ESE) (n = 5), and (3) saline treated rats (n = 3). An average Fast Fourier Transform was calculated using the Welch method and a rectangular window. Each window contained 65,536 points (262 sec). The results were averaged over an entire day (329 windows). Three power sectra were compared: (1) prior to kainate, (2) 3 days after kainate, and (3) 8 days after kainate. Results: All MSE rats developed chronic epilepsy. Their power spectrum did change after kainate administration. In 4/5 of the rats considered, the total power increased. The increases corresponded mostly to the lower frequencies, below 5 Hz. These increases persisted in 3/5 of the rats 8 days after treatment. By comparing the spectral density when frequent spikes are present (this was present only with MSE rats) with that when they were absent, we note that there is an increase in both high frequency (due to the sharpness of the spike) and low frequency (due to the periodicity of the spiking) power. We note that the results were invariant to the type of window used (Hamming, Hann, rectangular). None of the ESE rats developed chronic seizures. In these rats, we note that the total power decreased after kainate treatment in 4/5 of the rats. The fact that the power does not increase is not surprising because visually there was no change in the signal (no spikes). We also do not see the low frequency increase in the power spectrum noted in the MSE rats. For control rats the average EEG total power changed by only 10% from before to after kainate treatment. Conclusions: Kainate‐injured rats that develop epilepsy exhibit spectral changes (increases in low‐frequency power) in the EEG that appear prior to the onset of frequent motor seizures. Some of these spectral changes are due to the appearance of spikes (either individually or in clusters). If similar changes are found to be present in humans, it could be possible to predict whether an individual is likely to go on to develop chronic epilepsy following a brain injury. This will have a large positive influence on the care of patients with brain injury and could focus treatment trials on those at highest risk. (Supported by Funding was provided by the NIH.) 1 Dawn N. Wilson, 1 Isabel Chung, and 1 Sookyong Koh ( 1 Neurology, Children's Memorial Hospital, Chicago, IL ) Rationale: Prolonged seizures during childhood can lead to neuropsychiatric consequences including memory impairment, learning disability, and behavioral problems. No effective post‐seizure treatment exists to prevent this neurological insult. Using kainic acid (KA)‐induced seizures, a rat model of temporal lobe epilepsy, we have demonstrated a reversal of seizure‐induced decrease in exploratory behavior in open field testing of young rats by exposure to an enriched environment for 7–10 days after KA. These behavioral changes were accompanied by transcriptional alterations of genes involved in proliferation, long‐term potentiation, and memory consolidation including Arc, Homer1a, and Egr1. The expression of these genes were decreased after KA‐induced seizure but significantly increased after post‐seizure exposure to an enriched environment. To demonstrate a causal role for these genes in alleviating seizure‐induced behavioral deficits, we have delivered recombinant adeno‐associated virus (rAAV2) containing Arc into the hippocampus of young rats after KA‐induced seizure and assessed the subsequent behavioral effects. Methods: Arc, Homer1a, and Egr1 were cloned into AAV2 shuttle vectors and high‐titer rAAV2 vectors were produced by the CMRC Viral Vector Core. Long‐Evans male rats were injected either with PBS or KA at postnatal day 21 (P21) and divided into 5 groups (KA‐ARC, PBS‐ARC, KA‐GFP, PBS‐GFP, PBS‐No Surgery). One day later, rAAV2‐Arc or rAAV2‐GFP was stereotactically injected into the hippocampus of anesthetized rats in the surgical groups. Rats underwent the open field test at P28, 35, and 42. Animals were sacrificed at P44 and brains were immediately flash‐frozen. 40 μm cryostat sections were cut and processed for quantitative real‐time RT‐PCR (qRT‐PCR), in situ hybridization (ISH), immunocytochemistry (ICC), and western blot. Results: In the open field test, KA‐ARC animals explored significantly more than KA surgical controls (two‐way ANOVA, p = 0.0196, n = 8). By the third week of open field testing, a 1.6‐fold increase in exploratory behavior of KA‐ARC rats compared to KA controls was paralleled by a >2.5‐fold increase in Arc mRNA (qRT‐PCR, p = 0.0013, n = 8) and a 2‐fold increase of Arc protein in western blot with Arc antibody. Shown by ISH and ICC, Arc mRNA and protein were primarily localized to the dentate granule cells of the hippocampus. Exploratory behavior significantly correlated with Arc mRNA expression seen in qRT‐PCR (r = 0.701, p = 0.0012, n = 18). Conclusions: Expression of Arc mRNA and protein was increased by injection of rAAV2‐Arc into rat brain, which led to an improvement in exploratory behavior after KA‐induced seizure. These results demonstrate that Arc expression can reverse seizure‐induced behavioral impairments. Future in vivo studies injecting viral vectors containing Egr1 and Homer1a into rat brain will elucidate whether these genes will also afford protection against behavioral deficits after seizure. (Supported by Child Neurology Foundation, K02NS048237.) 1 Melodie Winawer, 3 Rachel Kuperman, 1 Martin Niethammer, 4 Steven Sherman, 5 Daniel Rabinowitz, 5 Irene Plana‐Guell, 7 Christine Ponder, 6 Helen Scharfman, and 2 Abraham Palmer ( 1 Neurology & G.H. Sergievsky Center, Columbia University, New York, NY ; 2 Human Genetics, University of Chicago, Chicago, IL ; 3 Pediatric Neurology, Columbia University, New York, NY ; 4 Epidemiology, Columbia University, New York, NY ; 5 Statistics, Columbia University, New York, NY ; 6 Pharmacology and Neurology Columbia University, Helen Hayes Hospital, West Haverstraw, NY ; and 7 Genetics and Development, Columbia University, New York, NY ) Rationale: Inbred mouse strains are valuable to identify genes contributing to complex genetic diseases such as epilepsy. Chromosome substitution strains (CSS) are strains in which a single chromosome from one inbred strain (donor) has been transferred onto a second strain (host) by repeated backcrossing. A panel of CSS in which each chromosome is represented may be used to identify quantitative trait loci (QTLs) that contribute to a phenotype. Pilocarpine hydrochloride is a muscarinic cholinergic agonist that causes limbic seizures in mice, modeling human temporal lobe epilepsy. QTLs for susceptibility to pilocarpine‐induced seizures have not been reported. C57BL/6J (B6) mice have been shown to be resistant to seizures; the A/J strain has been less studied, but appears more susceptible to seizures. Here, we report QTLs identified using a B6 (host) X A/J (donor) CSS panel to localize genes involved in susceptibility to pilocarpine‐induced seizures. Methods: 340 adult male mice–at least 10 of each CSS, 50 B6, and 39 A/J–were tested for susceptibility to pilocarpine‐induced seizures between 10 and 12 weeks of age. Mice were pretreated with atropine (5 mg/kg), followed in 30 minutes by pilocarpine (250 mg/kg), then continuously observed for 3 hours. Using a seizure staging system adapted from established (Racine) rodent seizure scales, highest stage reached, latency to each stage, and latency between stages were recorded for all mice. Significance of results in the CSS panel screen was determined by nonparametric comparison of mean highest stage reached (Mann‐Whitney), and cox proportional hazard regression was used to generate hazard ratios and p‐values for latency variables. Results: B6 mice were more resistant to seizures and were slower to reach each stage compared to A/J mice (p < .01). The CSS for chromosomes 10 and 18 progressed to the most severe seizure stages, diverging dramatically from the B6 seizure‐resistant phenotype (p < .01). Latency to all seizure stages and between stages was also significantly shorter for CH10 and CH18 mice, with hazard rates as high as 25 times that of B6 mice. Conclusions: Results of CSS mapping suggest the presence of seizure susceptibility loci on mouse chromosomes 10 and 18. This approach can provide a framework for identifying potentially novel homologous candidate genes for human temporal lobe epilepsy. (Supported by NINDS K23 NS02211, K02 NS050429, R01 NS36319, R0141490.) 1 Yan Yang, 1 Connie L. Mahaffey, 1 Terry P. Maddatu, 1 Gregory A. Cox, 1 Joel H. Graber, and 1 Wayne N. Frankel ( 1 Neurogenetics, The Jackson Laboratory, Bar Harbor, ME ) Rationale: BRUNOL4 (Bruno like 4) is a brain specific RNA binding protein, the loss of which leads to both convulsive and absence seizures in mice. The expression of Brunol4 is concentrated in neurons of the brain regions controlling synchronization and oscillation such as cerebral cortex, hippocampus and thalamus, suggesting that BRUNOL4 may play important roles in the maintenance of normal neuronal rhythmic activities. We hypothesize that BRUNOL4 is involved in the processing of mRNAs encoding proteins critical to neuronal excitation. Thus, BRUNOL4 deficiency would lead to altered expression of several molecules which in turn would result in seizures. This study aims to identify differentially expressed genes in the Brunol4 mutant brains and to determine how BRUNOL4 regulates gene expression. Methods: RNAs were extracted from mutant and control brains before the onset of overt seizures. The transcriptomes between mutants and controls were compared by microarrays, followed by northern and western analysis to confirm the expression difference. A potential BRUNOL4 recognition site in the target RNA transcripts was identified by multi‐species sequence comparison. Direct binding between BRUNOL4 and its RNA targets was demonstrated by RNA‐immunoprecipitation in neurons. The consequences of the BRUNOL4‐RNA interactions were investigated through reporter assays and RNA decay studies in cultured neurons. Results: Multiple potential BRUNOL4 target RNAs with altered expression were identified in mutants. Subsequent work confirmed the reduced expression of four RNAs encoding proteins critical to neuronal excitation in the mutant brain. The RNAs encode N‐ethylmaleimide‐sensitive factor, serotonin receptor 2c, synapsin II and α‐synuclein, some of which have been implicated in seizures while others may represent novel pathways in epilepsy. Sequence analysis revealed a highly conserved U/G rich motif in the 3′ untranslated region (UTR) of all four genes. RNA‐ immunoprecipitation showed that BRUNOL4 interacts with the 3′ UTRs of the four RNA transcripts through binding to the conserved U/G rich region. In luciferase reporter assays, the expression of chimeric reporters containing 3′ UTRs of the target RNAs was down‐regulated in the Brunol4 mutant neurons compared with wildtype neurons, a defect that can be rescued by adding back BRUNOL4 through transfection. Preliminary studies suggest that the mRNA decay rates of the four RNA transcripts are shorter in mutant neurons. Conclusions: BRUNOL4 appears to modulate neuronal excitability at the RNA level through fine‐tuning the stability of RNA transcripts encoding proteins critical to neuronal excitation. Brunol4 mutant mice represent the first animal model where compromised mRNA processing leads to epilepsy, thus providing a novel disease mechanism for seizure disorders. (Supported by a research award from Citizens United for Research in Epilepsy (YY) and a grant from NINDS (NS31348 to WNF).) 1 Guo Yin, 1 Peter I. Jukkola, 1 Hong Xu, and 1,2 Kevin M. Kelly ( 1 Neurology, Center for Neuroscience Research, Allegheny‐Singer Research Institute, Pittsburgh, PA ; and 2 Neurology, Drexel University College of Medicine, Philadelphia, PA ) Rationale: Macroscopic properties of the EEG are frequently correlated with abnormal behaviors. We observed an apparent correlation of increased generalized spike wave discharge (SWD, absence seizure) amplitude to abnormal behaviors in aged F344 rats during long‐term video‐EEG recordings. We sought to validate quantitatively whether SWD amplitude could be used as a reliable marker to predict myoclonic arousals and limbic seizures in these animals. Methods: SWD amplitude recorded from six EEG channels during the initial 5 hours of a 24 hour recording was observed, and 7 schema were established to predict two major types of abnormal behaviors (myoclonic arousal and limbic seizures) in the following 19 hours of recording. The 7 schema consisted of: 1) event occurrence frequency; 2) myoclonic arousal; 3) head nodding (class II); 4) forelimb or hindlimb clonus (class III); 5) running and rearing (class IV); 6) falling and jumping (class V); and 7) fatality. Five episodes of SWDs were chosen randomly in the first 5 hours of recording and an average amplitude was generated from 6 EEG derivations: F3‐C3, C3‐P3, F3‐P3, C4‐P4, F3‐F4, and C3‐C4. Quantitative prediction for different abnormal behaviors were determined based on the amplitude variation of the 6 channels. For example, it would be predicted that there would be no myoclonic arousal when the amplitude of F3‐C3 was greater than that of F3‐P3 by more than 5%, whereas there would be 1 to 2 events of myoclonic arousal in the subsequent 19 hours when the F3‐P3 amplitude was over 10% greater than F3‐C3 while the amplitude of C3‐P3 was less than or equal to C4‐P4. For purposes of simplicity, a standard scheme of quantitative comparison was designed: 1 class III event = 2 class II events, 1 class IV event = 4 class II events. Accuracy was judged by comparing the number of observed events to the predicted results. For example, based on SWD amplitude, the range of actual class II events is predicted between 5 and 8 times. Prediction failure occurs when the actual class II event number is 3 or 10. Results: Twenty two EEG files of 20 month old F344 rats (n = 9) were analyzed and 3 prediction categories included: 1) frequency of occurrence of myoclonic arousal; 2) event frequency; and 3) limbic event class and fatality. Analysis indicated prediction accuracy of 80% for myoclonic arousals, 70% for event frequency, and 75% for limbic events ± fatalities. Taken together, prediction accuracy was ∼70%. Conclusions: These findings suggest that macroscopic properties of electrocerebral activity in aged F344 rats may be used as predictors of abnormal ictal behaviors in these animals. These findings also suggest the possibility that specific differences in EEG spectral power may be quantifiable and equally predictive of these behaviors. (Supported by NIH R01NS046015 to KMK.) 1 Junli Zhou, and 1 Gregory L. Holmes ( 1 Neuroloscience Center at Dartmouth, Dartmouth Medical School, Hanover, NH ) Rationale: Patients with epilepsy suffer from varying degrees of postictal impairment including confusion and amnesia. This impairment adds substantially to the disease burden of epilepsy. However, the mechanism responsible for postictal cognitive impairment is unclear. In this study we monitored place cells in the CA1 region of the hippocampus before and after spontaneous seizures in epileptic rats. Place cells fire action potentials when the animal is in a specific location in space, the so‐called place field. Place cell function correlates well with performance in tasks of visual‐spatial memory and are an excellent measure of spatial memory. Methods: Male Sprague‐Dawley rats at P70 were given lithium chloride (127 mg/kg) given intraperitoneally (IP) 18 hours before an IP injection of pilocarpine (34 mg/kg) to induce SE. One month later the rats were implanted with eight tetrodes and place cells and interneurons recorded. We compared place cell and interneuron firing patterns before and after spontaneous seizures. For the pyramidal neurons we measured coherence, information content, AP firing, and total number of APs. Results: A total of 12 spontaneous seizures were recorded from six rats. Of 87 cells recorded in these six rats, 20 met the requirements for a place cell and were detected during all four recording sessions (two before and two after the seizure). In these cells there were significant decreases in coherence, information content, firing rate and total spikes following the seizure (Figure). In addition, eight cells that did not have place firing prior to the seizure developed firing fields following the seizure. Nine cells that had firing fields on both recording sessions prior to the seizure could not be recorded following the seizure. Seven interneurons were recorded from four rats were recorded. The seizures had no effect on firing frequency of the interneurons. Conclusions: Following seizures there was a dramatic decrease in firing rate of action potentials. In addition, firing fields were aberrant with reduced coherence and information content. In addition, to postictal suppression of firing patterns, seizures lead to the emergence of previously silent cells. Interneuron firing was not affected by seizures. The concept that postical impairment is due to simple neuronal exhaustion is no longer tenable. Our findings demonstrate that the effects of seizures are complex, affecting neuronal function in a cell specific manner. (figure 1) (Supported by Western Massachusetts Epilepsy Awareness Fund, Friends of Shannon McDermott, the Sara fund, and grants from NINDS (Grants: NS27984 and NS44295).) 1 Jacqueline J. Ardesch, 1 Laetitia J.J.C. Wagener‐Schimmel, 2,3 Hans E. Van Der Aa, 1 Gerard Hageman, and 3 Hendrik P.J. Buschman ( 1 Neurology, Medisch Spectrum Twente, Enschede, Netherlands ; 2 Neurosurgery, Medisch Spectrum Twente, Enschede, Netherlands ; and 3 Twente Institute for Neuromodulation, Medisch Spectrum Twente, Enschede, Netherlands ) Rationale: Few adverse events on heart rate have been reported with vagus nerve stimulation (VNS) for refractory epilepsy. This effect may be caused by electrical stimulation of the efferent nerve fibers innervating the heart. We describe three cases with intraoperative bradycardia during device testing. Methods: From 1999 until 2006 approximately 90 patients with medically refractory epilepsy have been implanted with a VNS system in our centre. We review 3 patients who showed a bradycardia during intraoperative device testing (Lead Test) and their VNS‐therapy follow‐up outcome. Results: Patient 1, a 32 year old female with complex partial epilepsy, experienced a bradycardia of 43 beats/min during the first Lead Test, and 25 beats/min during a second test. Ten days after the operation the VNS system was activated under ECG‐monitoring. No cardiac events occurred. Patient 2, a 52 year old male with partial epilepsy, experienced bradycardia of 40 beats/min. Ten days after implantation no cardiac effects were recorded when VNS therapy started. Patient 3, a 59 year old female with tonic clonic seizures, had a change in heart rate from 63 beats/min to 54 beats/min. No change in cardiac rhythm occurred when VNS therapy started. All three have had 4–6 years of chronic stimulation without cardiac side effects. The effect on epilepsy varied between 30%– > 50% reduction in seizure frequency. They also had less severe seizures and improved postictal recovery. Patient 1 and 2 have received a new pulse generator when battery was depleted. During the subsequent intraoperative Lead Tests no cardiac changes on ECG occurred. Conclusions: Our results suggest that bradycardia during intraoperative device testing does not have to be a reason for aborting the operation. However, postoperative VNS‐therapy onset should be done carefully and under ECG monitoring. Furthermore, long‐term VNS in these patients seems safe in respect to heart rate. 1 Edward Bromfield, 1 Barbara Dworetzky, 2 Shelley Hurwitz, 1 Zina Eluri, 1 Lara O’Brien, 1 Sonia Replansky, and 3 David Mostofsky ( 1 Neurology, Brigham and Women's Hospital, Boston, MA ; 2 Medicine, Brigham and Women's Hospital, Boston, MA ; and 3 Psychology, Boston University, Boston, MA ) Rationale: Fatty acids reduce seizures in animal models, but there is little human data. In a recent randomized trial (Epilepsy and Behavior 2005;7:253–8), patients with intractable epilepsy showed only a transient response. We are reporting the second such trial. Methods: Adults with uncontrolled epilepsy (at least 4 seizures/mo) were randomized 1:1, double‐blinded, to PUFA (eicosapentanoic acid (EPA) plus docosahexanoic acid (DHA), 2.2 mg/d in a 3:2 ratio) or placebo (mineral oil). Following a 4‐wk prospective baseline and 1‐wk titration phase, subjects entered a 12‐wk treatment period. Quality of life was assessed (QOLIE‐31) before and after treatment. Upon completion, subjects had the opportunity to participate in a 1‐mo open‐label PUFA trial. Nonparametric statistics were used. Results: 27 subjects were randomized; 4 decided not to participate before entering baseline (3 had been randomized to placebo and 1 to PUFA). 2 other subjects, both randomized to placebo, were dropped during baseline, 1 because of noncompliance and the other because of increased seizures. Completers included 12 PUFA subjects (7F, 5M; ages 25–55 (mean 37)), vs 9 on placebo (5F, 4M; ages 22–62 (mean 39)). 10 on PUFA and 6 on placebo had focal epilepsy, while 2 and 3 had generalized epilepsy (cryptogenic/symptomatic in 2 and 1, idiopathic in 2 on placebo). Subjects took from 1 to 3 antiepileptic drugs, with no significant group differences. 2 on placebo vs. 0 on PUFA had >50% decrease in seizure frequency from baseline during the 12‐week treatment (p = 0.17). Seizure frequency increased 6% on PUFA and decreased 12% on placebo (p = 0.21). On the other hand, of 19 subjects completing the 4‐wk open‐label PUFA trial, 15 experienced fewer seizures than during baseline (p = 0.02), 5 by at least 50%; 4 of these 5 had previously been on placebo. QOLIE‐31 scores increased an average of 1 point among PUFA subjects, and decreased 6 points among placebo subjects (p = 0.23). PUFA subjects during the randomized trial and all subjects during the open‐label phase showed no differences in serum drug concentrations other than lamotrigine, which declined by a mean of 17% among 6 patients. Conclusions: In this trial, PUFA was not superior to placebo as a treatment for intractable epilepsy. Given PUFA's absence of adverse effects and efficacy in animal models, additional trials of different doses or EPA:DHA ratios could be justified, particularly a longer, larger trial including children as well as adults. Adequate controls will be needed to account for what can be a strong placebo effect. (Supported by CURE Foundation; The General Clinical Research Center of Brigham and Women's Hospital, supported by the National Center for Research Resources (5M01 RR‐02635); Carlson Laboratories, suppliers of the treatment and placebo capsules.) 1,2 Hendrik P.J. Buschman, 2 Jasper R. Sikken, 3 Willem Kersing, 2 Peter H. Veltink, and 1,4 Hans E. Van Der Aa ( 1 Institute for Neuromodulation (TWIN), Medisch Spectrum Twente, Enschede, Netherlands ; 2 Biomedical Signals & Systems Group, Department Electrical Engineering, University of Twente, Enschede, Netherlands ; 3 Otolaryngology and Phoniatrics, Medisch Spectrum Twente, Enschede, Netherlands ; and 4 Neurosurgery, Medisch Spectrum Twente, Enschede, Netherlands ) Rationale: VNS is an effective treatment for patients with medically refractory epilepsy. It is generally accepted that VNS owes its antiepileptic effect from afferent nerve fiber stimulation. Co activation of efferent fibers can cause side‐effects. One of the most frequent reported side‐effects is hoarseness due to stimulation of the recurrent laryngeal nerve. At high stimulation amplitudes even spastic contractions of ipsilateral intrinsic laryngeal muscles can occur. For a group of VNS‐therapy users this side‐effect interferes severely with their daily activities. We investigated the recruitment properties of the recurrent laryngeal nerve and analyzed electromyographic and morphometric alterations on the vocal folds during VNS. Methods: Vocal fold EMG experiments were conducted intra‐operatively during the implantation of a VNS system. When the patient was anesthetized the endotracheal EMG tube (XOmed) was inserted, and the recording wires connected to the Monitor. During surgery the VNS‐therapy stimulation lead and pulse generator were implanted following normal procedure. The pulse generator was then programmed to stimulate in magnet mode for 14 seconds at 2 Hz, and one of the following pulse durations: 130, 250, 500, 750, or 1000 micros. At each pulse width the EMG‐threshold current was determined by electrical stimulation of the VN with increasing stimulation currents. When EMG signal amplitude saturated upon stimulation a second pulse duration was randomly selected and the above procedure repeated. Laryngostroboscopic examination was performed 1 and 6 months after surgery. Special attention was given to the effects of spontaneous stimulation on the larynx. If no effect was noticed a higher intensity stimulation cycle was given by means of ‘magnet‐activation’. Results: The vocal fold EMG and morphodynamic changes in the larynx have been analyzed in 5 patients. In all patients left vocal fold EMG threshold was between 0.25 and 0.50 mA. Pulse duration had little influence on the EMG threshold level. Vocal fold EMG saturation levels were reached between 0.75 and 1.00 mA. Videostroboscopic monitoring at therapeutic levels (1.25 – 2.25 mA) showed that VNS is well tolerated. ‘Magnet’ stimulation induced an adductory spasm of either the ipsilateral vocal fold or the vestibular fold, and was present remarkably irrespective of the presence of hoarseness. Conclusions: VNS causes pronounced effects on the vocal folds even at low stimulation amplitudes. At therapeutic levels the effect on the vocal folds is maximal even at the lowest stimulation pulse durations. The vocal fold contractions, however, do not necessarily give audible effects (hoarseness). 1 Robert S. Fisher, and SANTE Study Group ( 1 Neurology, Stanford Medical Center, Stanford, CA; and See Listing on Epilepsycontrol.com ) Rationale: We present an interim report for the SANTE (Stimulation of the Anterior Nucleus of the Thalamus for Epilepsy) trial, to illustrate the structure of the trial. Methods: The subject pool consists of patients 18–65 years old, inclusive, with partial onset seizures with or without generalization, and with observable clinical seizure manifestations. Seizures must be documented by prior video‐EEG monitoring, and they must average at least six per month. Patients must have failed at least three AEDs, with dosing stable for 30 days prior to enrollment. Women must be using a reliable method of birth control. Excluded are patients with IQ < 70 or inability to complete neuropsychological testing, progressive lesions or diseases, or psychiatric hospitalization, suicide attempt or psychosis (excluding postictal psychosis) within five years. If present, a VNS is to be explanted at time of thalamic implantation. After a three‐month baseline, eligible subjects receive bilateral anterior thalamic nuclei implantation of a dual‐channel Medtronic Model 7428 Kinetra Neurostimulation System with Model 3387 DBS Brain Leads. One month later, stimulation is initiated at either 0 (placebo) or 5 (active treatment) V, with pulse width 90 us, 145 pulses per second, on for 1 minute and off for 5. The double‐blind period continued for 3 months, after which there is 9 months of systematic variation of certain stimulation parameters. Results: As of 3/27/06, 98 patients were enrolled and 58 implanted at 15 US medical centers. Of these, 25 discontinued prior to implantation. Reasons included withdrawal of consent, inability to maintain stable antiepileptic drugs, status epilepticus, insufficient numbers of seizures in the baseline, and miscellaneous other reasons. Among 85 patients completing the baseline visit, age (mean ± SD, range) was 35.4 ± 10.9 (18.5–60.9) years, 45 male, 40 female. Epilepsy duration was 22.2 ± 13.1 (2–60) years. AEDs numbered 1 in 9 patients (11.3%), 2 in 37 (46.3%), 3 in 32 (40.0%), and 4 in 2 (2.5%). A vagus nerve stimulator previously had been implanted in 36 (42.4%); 17 (20.0%) had prior epilepsy surgery and 12 had both epilepsy surgery and VNS implantation. Conclusions: The SANTE trial is proceeding satisfactorily. Efficacy and safety data will be analyzed and presented after 102 patients have completed the blinded phase. (Supported by Medtronic.) 1 Eric H. Kossoff, 2 Hannah M. Rowley, 3 Saurabh R. Sinha, and 1 Eileen P. Vining ( 1 Epilepsy, Johns Hopkins Medical Institutions, Baltimore, MD ; 2 Nutrition, Johns Hopkins Medical Institutions, Baltimore, MD ; and 3 Neurology, Sinai Hospital, Baltimore, MD ) Rationale: The ketogenic diet is not typically offered to adults with epilepsy due to the significant lifestyle alterations needed for its use. The modified Atkins diet has been recently demonstrated to be therapeutic for children with intractable epilepsy without the need for admission, weighing foods, and fluid, calorie, and protein restrictions. We hypothesized the modified Atkins diet would be similarly well tolerated and effective for adults. Methods: Study inclusion criteria included age over 18, and at least weekly seizures and 2 attempted anticonvulsants. Carbohydrates were initially restricted to 15 grams/day, fats encouraged, and fluids, protein, and calories were ad lib. Medications were unchanged for the first 3 months. Patients were examined and laboratory results obtained at baseline, 1, 3, and 6 months on the diet. Results: Thirty patients, median age 36 years (range, 18 to 53), 19 (63%) female, consented and 29 started the diet. The median number of prior medications was 8 (range, 2–14); seizures were occurring 10 times per week (range, 1–140). Twelve patients had vagus nerve stimulators, 8 had failed resective surgery. Using an intent‐to‐treat analysis, approximately 40% had a >50% seizure reduction after 1 and 3 months on the diet. In those with seizure reduction, the median time to improvement was 2 weeks (range, 1 week to 2 months). Improvement was slightly higher in those with moderate to large urinary ketosis at the 1‐month visit, p = 0.08. The mean weight loss was 6.9 kg; mean BMI decreased from 29.0 to 26.5, p = 0.001. Cholesterol increased from a mean of 182 to 201 mg/dl after 1–3 months, p = 0.02. Compliance was problematic; 11 (37%) discontinued the diet prior to 3 months. Conclusions: A modified Atkins diet appears to be an effective therapy for adults with intractable epilepsy and can also induce weight loss. Considering the high dropout rate, a 2‐month trial appears to be adequate. (Supported by General Clinical Research Center, NIH/National Center for Research Resources grant M01‐RR00052.) 1 Kevin Novak, 2 JoAnn Hoeppner, 1 Ruzica K. Ristanovic, 1 Lawrence P. Bernstein, 1 Jesse Taber, and 3 Jeffrey Cozzens ( 1 Neurology, Evanston Northwestern Healthcare; Northwestern University, Evanston, IL ; 2 Pediatrics ; and 3 Neurosurgery ) Rationale: Patients receiving VNS therapy for refractory epilepsy often appear on follow‐up to be more alert, attentive, and energetic, independent of the effects of VNS on seizure frequency. This effect has not been fully demonstrated, and potential mechanisms of action have not been explored. We hypothesized that the apparent improvement in attention and energy results from a reduction in IEDs, which may interfere with cognitive processes. Methods: A heterogeneous group of 12 subjects with medically refractory, localization related epilepsy was evaluated at baseline, and at 6 months after VNS implantation surgery. Subjects had different epilepsy syndromes, baseline seizure and IED rates, and baseline IQs. During a 6 hour EEG recording session, an extensive battery of neuropsychological tests was administered to test for global cognitive function in several domains, with particular emphasis on attention. EEG spikes were detected with Persyst software, while custom software was used for EEG noise reduction and analysis of power spectrum, coherence, and statistics. Quality of life was examined with the QOLIE‐31 questionnaire. Seizure frequency was recorded by subject and caregiver report. The paired Student's t test or Wilcoxon signed rank test was used for statistical analysis (p < 0.05 considered significant). Results: The group had nonsignificant reductions in seizure frequency (mean 10.0 to 2.4/month, p = 0.07) and IED rate (64.1 to 7.7 spikes/hr, p = 0.08). There were no significant changes in relative EEG power acutely (VNS‐on vs ‐off) or chronically (baseline vs VNS‐off), although there was a trend of decreased delta and increased alpha power after VNS. Coherence did not change acutely with stimulation, but 9 of 12 subjects had significant chronic changes. Quality of life trended toward improvement (QOLIE‐31 raw score 52.5 to 61.9, p = 0.17). No changes were seen in any of the cognitive domains examined in standard neuropsychological testing, but the longer lasting of the two computerized tests of sustained attention/vigilance trended toward improvement (Steer clear driving: 7.02 vs. 4.95 cows/min hit, p = 0.11). The longer duration sustained attention driving test performance was correlated with baseline IED rate (r2 = 0.27, p = 0.05), and improved performance correlated with decreased IED rate after VNS (r2 = 0.31, p = 0.04). Conclusions: Although nonsignificant, 6 months of VNS in a small number of subjects resulted in a trend of reduced seizure frequency and IED rate, but had no measurable effect on cognition. The trend of reduced IED rates significantly correlated with improved ability to sustain attention in a longer duration task. Perhaps a larger study on a more homogeneous group of subjects selected for high IED rates and low to moderate baseline cognitive function would show larger changes in cognition. (Supported by Cyberonics, Inc.) 1,2 Marianna V. Spanaki, 3 David Greene, 1,2 Brien J. Smith, 1 David Burdette, 3 Martha Morrell, and 1,2 Gregory L. Barkley ( 1 Neurology, Henry Ford Hospital, Detroit, MI ; 2 Neurology, Wayne State University, Detroit, MI ; and 3 NeuroPace, Inc., Mountain View, CA ) Rationale: A significant percentage of women with epilepsy (WWE) have reported increased seizure frequency during the perimenstrual period (catamenial epilepsy). We first documented catamenial exacerbation of epileptiform activity using the Responsive Neurostimulator System (RNS) in 2005 (M V Spanaki, et al. Epilepsia 2005;46(Suppl 8):222). In this study we continuously measured epileptiform activity in two WWE over a total of 17 menstrual cycles. Methods: As part of a FDA‐approved multicenter feasibility study investigating the safety and potential efficacy of the RNS for epilepsy, we have implanted two WWE who were not good surgical candidates. The patients kept diaries of their menstrual periods. After seizure onset patterns were identified for both patients, detectors were programmed to identify electrocorticogram (ECOG) patterns that occurred prior to clinical seizures. These patterns appear epileptiform and occur frequently, but the vast majority of these bursts do not result in clinical seizures. Detection data from the RNS were then analyzed in multi‐day windows around the date of menses onset and compared to detection rates on days exclusive of these perimenstrual windows to determine if the daily detection rate differed between the menstrual and non‐menstrual intervals. Results: Data was gathered for 9 and 8 menstrual cycles for patients H and M, respectively. The detection rates for the different windows are shown in the graph. (figure 1)The detection data show statistical difference for both patients at window sizes of ± 3 days relative to the onset of the patient's menstrual cycle at the alpha = 0.05 level (95% confidence). Conclusions: The RNS has detection and diagnostic capabilities that allow objective measurement of epileptiform activity over prolonged periods of time in ambulatory patients. Use of RNS has confirmed the existence of perimenstrually exacerbated seizure activity. (Supported by: The RNS is limited by U.S. law to investigational use only. This study was sponsored by NeuroPace, Inc.) 1 Alan W.C. Yuen, 1 Josemir W.A. Sander, 1 Dominique Fluegel, 1 Philip N. Patsalos, 2 Lucy Browning, 1 Gail S. Bell, 3 Tony L. Johnson, and 1 Matthias M. Koepp ( 1 Department of Clinical and Experimental Epilepsy, UCL Institute of Neurology, London, United Kingdom ; 2 MRC Human Nutrition Research, Cambridge, Cambridge, United Kingdom ; and 3 MRC Biostatistics Unit, Cambridge, United Kingdom ) Rationale: Fatty acids (FA) have important roles in determining optimal structural and functional properties of neuronal membranes and may have an important impact on epilepsy. Pharmacology studies have shown that the FA, eicosapentaenoic acid (EPA, an omega‐3 FA), docosahexaenoic acid (DHA, an omega‐3 FA) and arachidonic acid (AA, an omega‐6 FA) can increase seizure thresholds. One study showed that the ketogenic diet increased plasma concentrations of EPA, DHA and AA. This increase may be a factor in the diet's anti‐seizure effects. In a double blind study examining effects of omega‐3 FA supplementation in people with epilepsy, we also examined FA profiles prior to and following supplementation. Methods: Patients with chronic epilepsy were randomised to receive capsules providing EPA (1g) and DHA (0.7g) daily or placebo (mixed vegetable oils) in a 12‐week double blind study. Seizure counts, adverse events, AEDs and 28 different red blood cell (RBC) FA concentrations were assessed during the study. The effect of concomitant AEDs on baseline FA concentrations was assessed by comparing those on AEDs with those not on the AEDs, using the Wilcoxon rank sum test. This study was approved by the local ethics committee. Results: The RBC FA profile obtained in 56 subjects prior to the start of the study showed mean% of AA, EPA and DHA of 13.1, 0.9 and 3.9. Carbamazepine (CBZ) and oxcarbazepine (OXC) appear to affect baseline FA concentrations, so analysis was done comparing those on CBZ (26) and those on OXC (20) with those taking neither CBZ nor OXC (10). Those taking CBZ showed 22% lower mean DHA (p < 0.05), 20% lower long chain omega‐3 FA (p < 0.05) and 5% lower monosaturated FA (p < 0.01). Those taking OXC showed non‐significantly higher EPA (23%), DHA (14%) and AA (6%) levels, which together contributed to a higher total polyunsaturated FA of 8% (p < 0.05). Like CBZ, OXC was also associated with lower (8%) monosaturated FA (p < 0.001). Following 12 weeks of supplementation, the 29 subjects in supplement group showed a 13% reduction in AA (p < 0.001), 216% increase in EPA (p < 0.001) and a 55% increase in DHA (p < 0.001). Caution needs to be taken in interpreting these results as no correction factor was applied for the multiple analyses undertaken. Conclusions: CBZ appears to be associated with lower DHA and long chain omega‐3 FA whilst OXC appears to be associated with higher total polyunsaturated FA. Hence OXC appears to be associated with a potentially more favourable FA profile. These findings need to be confirmed in subjects starting on these drugs as monotherapy. This study also showed that EPA and DHA concentrations are increased with omega‐3 FA supplementation. (Supported by Seven Seas Ltd. UK provided study medication.) 1 Ann Bergin, 2 Karen Costas, 2 Vanessa Ludlow, 1 Francie Mandel, and 3 Virginia Kimonis ( 1 Neurology, Childrens Hospital ; 2 GI/Nutrition ; and 3 Genetics, Childrens Hospital, Boston, MA ) Rationale: Angelman syndrome (AS) is associated with deletion of the maternal chromosome 15q11.2–15q13 in >70% of cases. Uniparental disomy, mutations of the UBE3A gene or of an imprinting center, cause an additional 15% of cases. Refractory epilepsy is particularly common in deletion‐associated cases. Multiple seizure types occur including atypical absences and prolonged non‐convulsive status epilepticus (NCSE). Anticonvulsant drugs (AEDs), often in combination, are incompletely effective. Each of the molecular mechanisms known to be associated with AS causes dysfunction of the UBE3A gene product, the E3 ubiquitin‐ligase protein. In deletion cases other genes, including the gene encoding g‐aminobutyric acid (GABAA) β3 subunit is deleted. A model for AS, mice lacking the β3 subunit, show myoclonic jerks, abnormal EEG and seizures with motor and learning defects. The β3 subunit deficit may result in abnormal thalamocortical functioning, favoring atypical absence seizures and NCSE. The ketogenic diet (KD) is an effective treatment for refractory epilepsy. The mechanisms by which the KD exerts its anticonvulsant effect are not known. Proposed molecular targets include mitochondrial uncoupling proteins, potassium channels and, of interest for deletion‐positive AS patients, GABA‐ergic neurotransmission favoring inhibitory effect. In this context we present 3 patients with Angelman syndrome and refractory epilepsy who were successfully treated with the ketogenic diet. Methods: A modified ketogenic diet (brief fast, full hydration, acid‐base correction) was initiated in hospital after neurological, nutritional and neurometabolic testing revealed no contraindication to the KD. Final KD ratio was determined based on efficacy and level of ketosis. Results: A: 3 year old girl, deletion positive AS and refractory NCSE, on 4 AEDs. Seizure free without complications on KD. Improved alertness, motor control, respiratory health, wellbeing. B: 2 year old boy, deletion positive AS, on 2 AEDs. Seizure free off AEDs on KD. More alert, more progress. C: 14 year old boy, one of 2 affected sibs with linkage to 15q11–13 but no mutation, with refractory status epilepticus, on 5 AEDs. Decrease in admissions for status epilepticus from 5/year to 1/3 years. Continued rarer seizures. Improved alertness, wellbeing. No adverse effects. Conclusions: Epilepsy in 2 deletion‐positive AS cases was completely controlled, and a useful response occurred in an undeleted case. Based on these initial observations we recommend consideration of KD treatment early in the course of refractory epilepsy in AS individuals. Different molecular pathology may underlie the differential response seen in these cases. There are now 2 animal models of AS available with different molecular characteristics. Further clinical and in vivo studies of the KD in AS may clarify the role of the KD and offer insights into the pathophysiology of epilepsy in AS and into KD mechanisms. 1 A.G. Christina Bergqvist, 2 Joan I. Schall, 1 Erica L. Richard, 3 Paul R. Gallagher, and 2 Virginia A. Stallings ( 1 Division of Neurology, The Children's Hospital of Philadelphia, Philadelphia, PA ; 2 Division of Gastroenterology, Hepatology, and Nutrition, The Children's Hospital of Philadelphia, Philadelphia, PA ; and 3 Division of Biostatistics, The Children's Hospital of Philadelphia, Philadelphia, PA ) Rationale: The purpose of this study was to determine if hypoglycemia or hyperglycemia predicts clinical ketogenic diet (KD) response by evaluating first morning blood glucose during the initial 21 days after initiation of the KD in a cohort of children with intractable epilepsy. In addition, the correlation between change in weight status and blood glucose was explored. Methods: First morning fasting whole blood glucose was measured daily for the first 21 days after KD initiation. Moderate hypoglycemia was defined as a glucose < 60 mg/dL at any time during the 21 days; severe hypoglycemia as < 45 mg/dL; and hyperglycemia as > 100 mg/dL. Z‐scores for weight (WAZ) and height (HAZ) were obtained at baseline, the day of discharge, 0.5 and 1 month of full KD therapy. Associations between clinical response to the KD, responder defined as > 50% reduction of seizure frequency at 3 months, hypoglycemia, hyperglycemia, style of KD initiation (fasting or gradual) and weight status were evaluated using Fisher's exact test and longitudinal mixed effects models. Results: Forty‐five subjects ages 1 to 12 years (mean age 5.3 ± 2.7 y) participated. Growth status was suboptimal before KD initiation (WAZ −0.3 ± 1.6 and HAZ −0.4 ± 2.0). Weight status declined over 1 month of KD therapy (to −0.6 ± 1.7 for WAZ). KD responder status was not associated with low or elevated blood glucose. A fasting or gradual KD initiation was also not associated. Children who had declining weight status during KD initiation were more likely to be hypoglycemic during full KD therapy, but this was unrelated to response to KD. Conclusions: Low blood glucose during KD therapy was not necessary for response to KD with clinically significant seizure reduction. Hypoglycemia was related to declining weight status. An effective KD can be provided in a manner to minimize side‐effects and maximize efficacy. Low and elevated blood glucose by ketogenic diet responder status Number Subjects All Responders Non‐responders p‐value Total Number of Subjects 45 30 15 Ever any glucose > 100 13 10 3 0.285 Ever any glucose < 60 28 19 9 0.539 Ever any glucose < 45 4 2 2 0.407 (Supported by: This study was supported in part by RRK‐23 16074 and General Clinical Research Center (MO1RR00240), the Nutrition Center of the Children's Hospital of Philadelphia, and the Catharine Brown Foundation.) 1 Thomas A. Buzzerd ( 1 Clinical, Associates In Behavioral Health, Tunkhannock, PA ) Rationale: Roughly 500,000 children and adolescents across the U.S. suffer from seizure activity. Currently, medication management programs are the primary intervention with only varying success at controlling seizure activity. Increased attention and study needs to be focused on non‐pharmacological intervention strategies like stem cell therapy. This case study evaluated the effectiveness of human, umbilical cord stem cell (hUCSC) therapy over a ten month period of time in an eight year old male with cerebral palsy and epilepsy. Methods: Direct observations and indirect measures (such as standardized assessment tools) were conducted and administered; interviews with support persons were also conducted. Seizure activity, motor skills, communication skills and academic performance baseline were evaluated for change following two stem cell injections (1.8–2.0 and 2.8–3.0 million stem cells) administered six months apart. A single subject research design was used to collect baseline and post‐injection data. Results: A t‐test analysis of data groups presented statistical significance (at p = .05) in seizure activity. The post injection data represented the subject as exhibiting approximately a 90% reduction in both seizure frequency and duration during the period of this study. Conclusions: Although a single case study does not serve as “proof” that stem cell therapy is an effective treatment. Studies into this new area of science must critically evaluate all outcomes that may be discovered. (Supported by Private Funder.) 1 Maria G. Dahlin, 2 Lena Hjelte, and 1 Per E. Amark ( 1 Neuropediatrics, Astrid Lindgrens Childrens Hospital, Karolinska Hospital, Stockholm, Sweden ; and 2 Cystic Fibrosis, Astrid Lindgrens Childrens Hospital, Karolinska Hospital, Stockholm, Sweden ) Rationale: The ketogenic diet (KD) is a high‐fat, low‐protein, very low‐carbohydrate diet used in the treatment of medically refractory epilepsy in children. Its anticonvulsant mechanism is still unclear but alterations of fatty acid (FA) levels has been proposed as a mechanism of action. We examined the influence of the KD on FA levels and seizure control. Methods: The levels of seventeen FAs in plasma phospholipids were determined before and 1, 6, and 12 months after initiation of the KD in 25 children (mean age 6.3 years) with severe epilepsy. The majority were on a KD ratio of 4:1 (fats:proteins and carbohydrates). After 1 month, the KD was supplemented with polyunsaturated omega‐3 FAs by adding fish‐oil (4–8 g/day). Seizures were quantified. Results: Highly significant changes in the levels of several FAs were found in relation to the KD. After 1 month on KD significant changes were seen in the levels of 8 FAs, after 6 months on KD in 10 FAs and after 12 months on KD in 12 FAs. Changes were found in the levels of polyunsaturated, monounsaturated as well as saturated FAs. Among others, increases were seen in linoleic acid (18:2 n6) and EPA (20:5 n3) and decreases in arachidonic acid (20:4 n6). The ratio of omega‐6 to omega‐3 FAs decreased from 7.0 before diet to 4.9 at 12 months after diet start. However, no correlation of FA levels to seizure response was found. Conclusions: The ketogenic diet induced prominent changes in several major plasma fatty acid levels but a correlation to seizure control was not found. A decreased ratio of omega‐6 to omega‐3 fatty acids was observed which may be of benefit to reduce risks of cardiovascular disease. Further studies are needed on how different diet compositions, in terms of type and amount of various fatty acids, affect blood fatty acid profile, in order to define an optimal diet regime for maximal seizure control as well as minimal cardiovascular risks. (Supported by the Margaretahemmet Association Sunnerdahl Handicap Fund.) 1 Eric Dinnerstein, 1 Karen R. Richardson, and 1 Barbara C. Jobst ( 1 Section of Neurology, Dartmouth Hitchcock Medical Center, Lebanon, NH) Methods of terminating seizures Patient Epileptic focus Praying Talking Swallowing Listening Writing Rubbing 1 Right + + + 2 Right + 3 Right + 4 Right + 5 Right + 6 Left + + + 7 Left + + Rationale: Patients often report that they can terminate seizures with certain cognitive tasks. Case reports decribe the ability of cognitive tasks to suppress/interrupt seizure activity. A systematic review of such experiences has not yet been documented in the literature. Is it not known whether cognitive activity arresting seizures is hemispheric specific, and whether it correlates with side of seizure onset. Methods: 20 adult patients with mesial temporal epilepsy were retrospectivley interviewed by telephone or in writing. All patients had unilateral temporal lobe seizure onset and underwent temporal lobectomy. Patients were asked to describe, whether they developed a cognitive method/technique to terminate seizures, when they were preceded by an aura. In addition a list of cognitive activities specific to each hemisphere was presented to all patients. Results: Seven patients reported cognitive acitivities that terminated seizures after an aura occured.Five patients had right mesial temporal seizure onset, two had left mesial temporal seizure onset. Patients with right mesial temporal onset seemed to control their seizures with praying, while patients with a left mesial temporal onset controlled their seizures via rubbing or tapping the contralateral affected limb (see table). The small number of patients precluded statistical analysis. Conclusions: Despite being a limited study, this survey provides an interesting observation. Some patients with right sided mesial temporal seizure onset can terminate their seizures with right hemispheric activities such as praying. Some select patients with a left mesial temporal seizure onset are able to terminate their seizures using left hemispheric tasks, such as talking, writing, tapping and rubbing the right side. Neurocognitive tasks represented in close proximity to the seizure focus could influence electrical seizure activity. Neurocognitive therapy could evolve as treatment option for epilepsy. These findings need to be confirmed in a larger prospective study with objective validation. 2 Celina C. Martinez, 1 Adam L. Hartman, and 1 Eric H. Kossoff ( 1 Pediatric Epilepsy, Johns Hopkins Medical Institutions, Baltimore, MD ; and 2 School of Medicine, Johns Hopkins Medical Institutions, Baltimore, MD ) Rationale: Approximately 10% of children on the ketogenic diet become seizure‐free after its initiation. The diet typically is discontinued after 2 years, but there is no information regarding the likelihood and severity of seizure recurrence. Methods: A retrospective chart review examined children who became seizure‐free on the ketogenic diet at our institution over a 12‐year period; the indication for discontinuing the diet was freedom from seizures. All children had at least 6 months of seizure freedom and were on the diet for that period prior to discontinuation. Results: Sixty‐six children discontinued the diet after a median of 2.1 years due to seizure freedom; 92% were medication‐free. Twelve (18%) had recurrence of their seizures at a median of 3.0 years after the diet was stopped. Seven (58%) became seizure‐free a second time, 2 after reinstitution of the diet. Four patients restarted the diet but remain intractable to date. The presence of recent electrographic epileptiform activity and an abnormal MRI each significantly increased the likelihood of recurrence (45% and 38%, respectively); age, seizure type, seizure frequency, medication freedom, and time to initial diet response did not. Three patients with recurrent seizures had tuberous sclerosis complex. Conclusions: The risk of recurrent seizures subsequent to discontinuing the diet after achieving seizure freedom was approximately 1 in 5, and more than half became seizure‐free again, usually with anticonvulsants alone. The risk of recurrent intractability was three times less; however the diet, when restarted in these patients, was not as successful. Obtaining a recent EEG prior to discontinuing the diet in this situation may be advisable in counseling families. 1 Hrisimir Kostov, 1 Pål G. Larsson, and 2 Geir K. Røste ( 1 Department of Neurodiagnostics, The National Center for Epilepsy, Rikshospitalet University Hospital, Sandvika, Norway ; and 2 Department of Neurosurgery, Rikshospitalet University Hospital, Oslo, Norway ) Rationale: Vagus nerve stimulation (VNS) is accepted as an effective and safe treatment for refractory partial epilepsy. Its value for treating patients with refractory idiopathic generalized epilepsy (IGE) is not well documented. Methods: All patients with medically refractory IGE, with EEG/video documented seizures, implanted with VNS in our epilepsy surgery program between 1995 and 2006 were included. A total of 12 patients (2 M, 10 F) from 11 to 48 years (mean age: 30.7 ± 14.4 years) were identified. Mean follow‐up period was 18 months (±14; range, 3 – 48). Patients were treated with an average of 2.25 AEDs/patient at implantation time. Wilcoxon's paired signed rank test was used for statistical analysis. Results: There was 62% seizure reduction for generalized tonic‐clonic seizures (p = 0.0020), 58% for absence seizures (p = 0.0003) and 40% for myoclonic seizures (p = 0.0156). Eight patients were considered responders (≥ 50% seizure reduction); two of these patients became seizure‐free. Two patients had < 50% seizure reduction, and two patients showed no change in seizure frequency. At the last follow‐up visit, the patients were receiving an average of 1.67 AEDs (−0.58 AEDs/patient; p = 0.0625). Two patients (both responders) are currently being treated with VNS therapy only. Most commonly used stimulations parameters were: pulse width 250 μs, frequency 20 Hz, 30 seconds on and 5 minutes off. Median output current at the last follow‐up was 1.5 mA. Due to lack of response in four patients, parameters were changed to rapid cycling (7 sec. on and 0.2/0.3 min off); two of these patients became then responders. An improvement in alertness (“better” or “much better”) was observed in 6 of the patients. According to 5 patients, the magnet was always or most times effective. Nine patients reported some side effects, which were mild and tended to decrease over time. Conclusions: In this patient population, VNS therapy was effective and safe for the treatment of medically refractory IGE. Rapid cycling seems to be effective for some of the non‐responders. There was an improvement in alertness in half of the patients, and a trend towards a reduction in the number of AEDs/patient. 1,2 Elizabeth A. Leleszi, 1,2 Eishi Asano, and 1,2 Gyula Acsadi ( 1 Neurology, Children's Hospital of Michigan, Detroit, MI ; and 2 Pediatrics, Children's Hospital of Michigan, Detroit, MI ) Rationale: Celiac disease can be associated with a wide spectrum of neurological manifestations including ataxia, epilepsy, dementia, neuropathy, myopathy, cerebral calcification, cerebellar atrophy and leukoencephalopathy, which are mostly related to the metabolic consequences of severe malabsorption. We describe a four year‐old child, presenting with myoclonic epilepsy described as head drops and occasional upper extremity clonus. The seizures continued to happen up to thirty times a day for several months even on multiple antiepileptic medications. An ictal EEG showed generalized high amplitude spike and slow wave activity with duration lasting 1–2 seconds. A brain MRI was normal. As part of an allergy workup, high antigluten antibodies were found and the HLA typing (HLA‐DQB1*0302, 0301) was also consistent with celiac disease. Unfortunately, a jejunal biopsy was not preformed. After the patient was started on a gluten‐free diet, the seizures resolved completely within three days and the EEG was normalized. The anti‐epileptic medications were gradually discontinued without any relapse for at least four more months. Conclusions: Gluten from dietary sources may evoke an immune response of the gastrointestinal cells via gliadin and tissue transglutaminase in susceptible individuals and lead to celiac disease with multiorgan manifestations. However, the clinical and immunological consequences of asymptomatic gluten intolerance are not well established. This is the first case report describing a complete resolution of seizure activity with a gluten‐free diet alone. Our case suggests a possibility that the epilepsy was caused by a so far unknown immunological process directed to the central nervous system (e.g. antibodies to some ion channels.) Further investigations are necessary to determine a relationship and pathomechanism between intractable epilepsy and gluten intolerance. 1 Jullie W. Pan, 2 C.J. Segal‐Issacson, and 2 Brandy Cowell ( 1 Neurosurgery, Yale University School of Medicine, New Haven, CT ; and 2 Epidemiology and Social Medicine, Albert Einstein College of Medicine, Bronx, NY ) Rationale: A number of reports have found that low carbohydrate (e.g., Atkins) or low glycemic index diets may help in reducing seizure frequency. Given the relatively lower ketosis achieved in such diets as compared to the ketogenic diet, it is possible that the effect of seizure control is not via direct ketone (or ketone derivatives) levels in the brain, but through a physiological mechanism. We examined for such a mechanism by studying the effects of 4 weeks of an Atkins diet on cerebral energetics in n = 7 healthy overweight control adult volunteers. Methods: We used high field 31P whole brain MRSI to study healthy overweight (n = 7, mean age 34.7 ± 10.0years, 4F) volunteers, examining them prior to initiating a low carbohydrate diet (Atkins Phase 1) and after 4 weeks of sustaining this diet. For optimum diet control, meals from a licensed caterer were provided to the subjects, and the macronutrient distribution of the diet was 65% fat: 30% protein: 5% carbohydrate. The mean body mass index at baseline was 34.0 ± 2.7, and after 4 weeks of the diet was 31.2 ± 2.2. Weight loss was achieved by all volunteers, averaging 6.7 ± 3.2 kg. 31P MRSI was performed at 4Tesla. Including scout imaging and calibrations, the duration of the 31P study was ∼75min. The 31P acquisition has an effective sampling radius of 1.4cm; previous studies have shown the reproducibility to be 10%. Quantification was performed accounting for tissue volume, coil loading and relaxation. Plasma samples were frozen at –80C until analysis; b‐hydroxybutyrate concentrations were spectrophotometrically assayed after each study using a COBAS‐FARA analyzer. Results: A mild ketosis was achieved in all subjects, with plasma BHB rising from 0.24 ± 0.05 to 1.64 ± 0.93. Example data (position and spectra) are shown in Fig. 1, showing PCr, ATP and inorganic phosphate Pi. Data and loci of spectral analysis are shown in the Table. From all volunteers, the ATP and PCr concentrations in the hippocampus and thalamus rose significantly between D0 to D28, while there was no significant change in occipital cortical tissue. Conclusions: Cerebral energetics changed with use of a low carbohydrate diet, with increased ATP and PCr concentrations in the thalamus and hippocampus. The data are consistent with a view that the change in fuel type and availability is characterized by an alteration in energetics particularly in subcortical brain structures. The increased availability of high energy phosphates in key limbic structures may be an important factor in how a low carbohydrate diet affects seizure control. (figure 1) (Supported by NIH R01NS40550.) 1 Roberto Poma, 2 John Ives, 3 Alexander Rotenberg, and 3 Alvaro Pascual‐Leone ( 1 Clinical Studies, Ontario Veterinary College—University of Guelph, Guelph, ON, Canada ; 2 Clinical Neuroscience, University of Western Ontario, London, ON, Canada ; 3 Division of Epilepsy, Harvard Medical School, Boston, MA ; and 4 Department of Neurology, Beath Israel Deaconess Medical Center, Boston, MA ) Rationale: The purpose of this study was to investigate the role of repetitive transcranial magnetic stimulation (rTMS) as an alternative treatment for 3 dogs suffering from refractory idiopathic epilepsy. Methods: Three client‐owned epileptic dogs refractory to AED treatment were studied. All dogs were diagnosed with idiopathic epilepsy based on the history of generalized seizures, normal neurological examination, unremarkable brain MRI and EEG features. Affected dogs were followed for at least 60 days before and after rTMS. A calendar of the seizures was kept to evaluate the interval between seizures before and after rTMS. Each dog received multiple AED treatment including phenobarbital, potassium bromide and levetiracetam. No dosage change was made in the two months prior to and following the study. A total amount of five treatments were delivered to each dog over five consecutive days (1 treatment per day). The technique of stimulation consisted in a 9 cm focal point coil located over the skull in proximity of the vertex. Different rates of stimulation were used between dog 1 and dogs 2 and 3. In dog 1, each treatment consisted in low‐rate rTMS delivered in 20 consecutive trains of 90 pulses each at 1 Hz frequency and 60% of the maximal coil output. The inter‐train interval (ITI) was 120 seconds. Dog 2 and 3 received a treatment protocol consisting of two consecutive sessions at different rates. The first session consisted of priming the cerebral cortex with high‐rate rTMS characterized by 20 trains (40 seconds ITI) of 60 pulses each at 6 Hz frequency and 55% of the maximal coil output. The second session of stimulation consisted in low‐rate rTMS characterized by 10 trains (60 seconds ITI) of 60 pulses each at 1 Hz frequency and 60% of the maximal coil output. The longest seizure‐free interval during the pre‐rTMS period was compared with the seizure‐free interval immediately post‐rTMS. Results: Dog 1 and 3 had a short‐term improvement of the seizure interval in the immediate post‐rTMS period (27 days and 14 days respectively) as compared to the longest seizure‐free interval of the pre‐rTMS period (20 days and 12 days respectively). Dog 2 had an equal seizure‐free interval between the immediate post‐rTMS period (19 days) and the longest seizure‐free interval of the pre‐rTMS period (19 days). Conclusions: Regardless of the technique of stimulation used in this study (low‐rate rTMS versus high‐rTMS followed by low‐rate rTMS), an equal or prolonged seizure‐free interval in the immediate post‐rTMS period was observed in all dogs suggesting the possible short‐term beneficial effect of rTMS in canine epilepsy. (Supported by Ontario Veterinary College Pet Trust Foundation.) 1 Steven M. Rothman, 1 Xiao‐Feng Yang, and 2 Brigitte F. Schmidt ( 1 Neurology and Hope Center for Neurological Disorders, Washington University School of Medicine, St. Louis, MO ; and 2 Molecular Biosensor and Imaging Center, Carnegie Mellon University, Pittsburgh, PA ) Rationale: Caged compounds are inert prodrugs that are capable of releasing small bioactive molecules when excited by light at the appropriate wavelength. They were initially developed to facilitate rapid, step increases in the local concentration of neurotransmitters and are typically activated with a very brief pulse from a mercury bulb or laser. We wondered whether we could use low levels of light from an ultraviolet diode (UV LED) to release GABA from a new caged analog and modulate seizure‐like activity in cultured neurons. Methods: We used cultured hippocampal neurons after about 2 weeks in vitro and recorded in whole cell bridge or voltage clamp mode. GABA was applied by whole bath perfusion or by uncaging 4‐(((2H‐1‐benzopyran‐2‐one‐7‐amino‐4‐methoxy)carbonyl)amino)butanoic acid (BC204). A UV LED (Nicchia 365 nm/100 mW) was activated by a custom fabricated driver for 4 seconds in the uncaging experiments. Results: We first verified that our UV LED was capable of generating sufficiently high GABA concentrations to activate GABAA receptors in our cultures. We found that driver currents between 25 and 250 mA produced readily detectable currents in the presence of BC204 (30 μM). These currents are well below the maximum capacity of the UV LED (700 mA). The current elicited by 150 mA corresponded to a 10 μM GABA current, when normalized to a GABA dose response curve. The currents produced by uncaging BC204 were reduced by 89% in the presence of picrotoxin (100 μM). BC204 itself at 30 μM is a very weak GABAA blocker, reducing 3 μM GABA currents by about 5%. When cultures were exposed to medium lacking magnesium, they generated increased synaptic activity, paroxysmal depolarization shifts, and often “seizure like” discharges that were abolished by perfusion with GABA as low as 3 μM. In the presence of BC204 (30 μM), a 4 second illumination with 100–200 mA also eliminated spontaneous activity. The figure below shows the effect of a 4 second light pulse on bursting. (figure 1)Illumination in the absence of BC204 produced no detectable current. Conclusions: These results suggest that it may be possible to utilize a compact, low power UV LED in combination with locally applied caged GABA to activate tonic GABAA and GABAB receptors and modulate paroxysmal activity. These receptors can be activated by GABA concentrations in the low micromolar range, which it may be possible to achieve in vivo. (Supported by Alafi Family Foundation, the NIH (R01 NS 42936 to SMR), and NSF (MCB‐8920118 to BFS).) 1 Christina A.G. Bergqvist, 2 Joan I. Schall, and 2 Virginia A. Stallings ( 1 Pediatrics/Neurology, The Children's Hospital of Philadelphia, University of Pennsylvania, Philadelphia, PA ; and 2 Pediatrics/Division of GI, Hepatology and Nutrition, The Children's Hospital of Philadelphia, Philadelphia, PA ) Rationale: The aims of this study were to describe vitamin D status including dietary intake in a contemporary cohort of children with intractable epilepsy prescribed newer anti‐epileptic drugs (AED), and to determine the effects of the ketogenic diet (KD) on serum vitamin D status over 15 months. Methods: Prepubertal children 1–14 years with intractable epilepsy were eligible. Serum vitamin D (25‐OHD and 1,25‐OHD) and parathyroid hormone (PTH) were obtained prior to initiation of KD therapy and every three months thereafter. Three‐day weighed dietary records were obtained at baseline and one month on KD therapy, and included vitamin and mineral supplementation. Results: 45 children with intractable epilepsy (age 5.1 ± 2.7 years) enrolled in the study. At baseline 4% had deficient and 51% had insufficient serum 25‐OHD levels. Vitamin D intake was inadequate; 46% consuming less than the recommended intake. Adequate vitamin D intake, fewer AED and generalized seizures were associated with higher serum 25‐OHD levels (p < 0.01). After 3 months on the KD, vitamin D status improved, 25‐OHD levels increased (p < 0.001), and PTH declined (p < 0.001). Over the next 12 months of KD exposure, the 25‐OHD levels steadily declined (p < 0.001), while 1,25‐OHD and PTH were not significantly changed. Conclusions: Children with intractable epilepsy treated with newer AED had poor vitamin D intake and blood levels. KD therapy with vitamin D supplementation improved vitamin D status; however, over the next 12 months of KD therapy a decline in 25‐OHD was observed. (figure 1) (Supported by: Supported in part by RRK‐23 16074 and General Clinical Research Center (MO1RR00240), the Nutrition Center of the Children's Hospital of Philadelphia, and the Catharine Brown Foundation.) 1 Christie Snively, 1 Paul R. Carney, 1 Zhao Liu, 1 Debbie Ringdahl, 1 Kelly Winger, 1 Crystal Jackson, 2 Thameshwarie Singh, 2 Lauren E. Little, and 2 Peggy R. Borum ( 1 Pediatric Neurology, University of Florida, Gainesville, FL ; and 2 Food Science and Human Nutrition, University of Florida, Gainesville, FL ) Rationale: Ketogenic Therapy (KT) is a treatment for intractable epilepsy that is designed to mimic metabolism during starvation. Although KT has been practiced for many decades, the guidelines for practice have often been communicated orally and based on personal experience. Subjective measurements of seizure increase or decrease and of antiepileptic drug (AED) requirements are usually used. The goal of our project was to develop an evidence based KT program for treatment of epilepsy. Methods: A review of the literature on KT over the past decade was conducted and an evidence grading system was used to document current practice. A retrospective database was developed to evaluate KT provided at our center for the past decade. There are currently 184 current and non‐current patients in the retrospective database of which 95 are males and 89 are females. The average age at initiation was 7.1 ± 4.9 years old and the average time on diet from non‐current patients (n = 133) was 17.7 ± 18.9 months. Patient tracking forms for pre‐initiation, initiation, post‐initiation, and termination of KT were created. Systems for quantifying and tracking seizures and medications were created to evaluate patients' progress on KT. This includes Clinic Seizure Score (CSS), Daily Diary Seizure Score (DDSS), and a Medication Score (MS). The CSS is dependant upon the caregiver's recollection at clinic of seizure activity since the last clinic. The DDSS is based on daily seizure records of the caregiver and used to track the total amount of seizure activity between clinic visits. The MS allows us to evaluate both the total number and dosage of all the AEDs that a patient is taking based on their age and weight. Results: The evidence based guidelines have been created by the University of Florida Pediatric Comprehensive Epilepsy Center. For initiation, pre‐therapy records are collected and the patient comes in after an overnight fast and starts on a 1.5:1 ratio at full calories which is then increased by 0.5 every other meal for oral feeders and every day for tube feeders. During post‐initiation, parents document daily ketones, specific gravity, and seizure activity as well as dietary intake for 24 hours per week. Patients are monitored using the CSS, DDSS, MS as well as the following lab parameters: comprehensive metabolic panel, complete blood count, lipid profile, serum beta‐hydroxybutyrate, lactate, and acylcarnitine profile. Data from the patient tracking forms are entered in a prospective database on a weekly basis. Conclusions: The data obtained from the databases will allow us to evaluate and expand the evidence based KG guidelines for optimal administration and monitoring of patients throughout KT. 1 Jessica A. Yankura, 1 Amy French, 1 Michael A. Rogawski, 2 Adam Hartman, and 1 Maciej Gasior ( 1 Epilepsy Research Section, NINDS, National Institutes of Health, Bethesda, MD ; and 2 Pediatric Epilepsy Center, John Hopkins Hospital, Baltimore, MD ) Rationale: Anticonvulsant effects of the ketogenic diet seem dependent on high fat and low carbohydrate availability that results in the development of ketosis along with a modest reduction in serum glucose levels. However, it remains unclear if low glucose levels alone can decrease neuronal excitability and have anticonvulsant effects. Glucose depletion attenuates seizure activity in in vitro models whereas both proconvulsant and anticonvulsant effects have been reported in in vivo models (1,2). In the present study we sought to clarify the influence of glucose in regulating seizure susceptibility in vivo by simulating glucose deprivation with 2‐deoxyglucose (2‐DG), a structural analog of glucose that does not undergo glycolysis. Methods: Anticonvulsant effects of 2‐DG were assessed in amygdala‐kindled rats and in Swiss mice in the intravenous pentylenetetrazol (PTZ) and kainic acid models, and in the electroconvulsive threshold (ECT) and 6 Hz models. 2‐DG (125 – 500 mg/kg) was administered intraperitoneally 30 min before seizure testing. Results: In amygdala‐kindled rats, 2‐DG had no effect on afterdischarge threshold, afterdischarge duration, behavioral seizure score and behavioral seizure duration. In mice, administration of 2‐DG had no effect on seizure threshold in the PTZ test and was slightly proconvulsant in the kainic acid model. It had no effect on ECT. Only in the 6 Hz seizure test did 2‐DG cause a small, yet significant, dose‐dependent elevation in seizure threshold. Conclusions: In contrast to a prior study using olfactory bulb and perforant path kindling (2), we did not find 2‐DG to have potent anticonvulsant efficacy in the various in vivo seizure models we studied. Our findings suggest that low glucose availability plays a limited role in the seizure protection conferred by the ketogenic diet. 1 Kirchner, A., Veliskova, J. & Velisek, L. Differential effects of low glucose concentrations on seizures and epileptiform activity in vivo and in vitro. Eur. J. Neurosci. 23:1512–1522, 2006. 2 Stafstrom, C., Kriegler, S., Valley, M., Ockuly, J., Roopra, A. & Sutula, T. 2‐Deoxyglucose exerts anticonvulsant and antiepileptic actions in experimental epilepsy models. Epilepsia 46 (Suppl. 8), 268–269, 2005. (Supported by NINDS, NIH.) 1 Mahmoud M. Abu‐Ata, 1 Bassel Abou‐Khalil, 1 Kevin Haas, 1 Patrick Lavin, and 1 Peter Konrad ( 1 Neurology, Vanderbilt University Medical Center, Nashville, TN ) Rationale: To evaluate the visual field defects in patients who underwent SAH (selective amygdalohippocampectomy) for intractable epilepsy using automated static visual field testing. Background: Visual field defects, specifically contralateral upper quadrant homonymous quadrantanopsia, are well known complications after conventional temporal lobectomy, however, few studies have evaluated visual field defects after SAH using automated static visual field perimetry. Methods: We evaluated 6 adult patients with history of intractable temporal lobe epilepsy who had SAH within the last 3 years using neuro‐ophthalmologic evaluation and automated Humphrey visual field analysis post‐operatively. All patients were diagnosed with mesial temporal lobe sclerosis and had their surgery at our institution using the same image guided surgical technique, in which an incision is made in the middle temporal gyrus to locate the temporal horn, then the amygdala and 3 cm of hippocampus are removed using the microscope. After obtaining written consents, all patients were questioned about visual complaints and were examined for visual fields by confrontation using finger counting, visual acuity ocular motility, pupil size and reactivity and fundoscopy. Humphrey automated visual field analysis was then performed using the 24–2 program that tests 54 points that straddle the horizontal and vertical meridians. The visual field for each eye was divided into 4 quadrants. The depth of any identified defect was determined by calculating the algebraic difference between measured value and expected value and was displayed on a grid corresponding to the location of the test points creating a map that showed the location and extent of the visual field defects. The pattern deviation was then averaged in the quadrant contralateral to surgery, separately for the ipsilateral and contralateral eye. Deviations less than −5 were displayed (Figure). Results: 4 patients had left and 2 had right SAH. None had identifiable visual field defects by history or confrontation visual field exam. All had identifiable visual field defects in the contralateral superior quadrant by perimetry. 4 had incomplete contralateral superior quadrantonopsia and 2 had only small peripheral defects in the contralateral superior quadrants. The defect affected mostly the medial sector near the vertical meridian and was steeper in the ipsilateral eye. Conclusions: Visual field defects are an expected complication of SAH. The expected defects are predominantly incomplete contralateral superior quadrantonopsias, similiar to those previously reported with conventional anterior temporal lobectomy. (figure 1) 1 Indrani E. Acosta, 1 William O. Tatum, 1 Fernando L. Vale, and 1 Selim R. Benbadis ( 1 Department of Neurology & NeuroSurgery, Tampa General Hospital, Tampa, FL ) Rationale: Temporal Lobectomy is the most common performed epilepsy surgery. Specialized surgical centers infrequently perform temporal lobectomies on patients 60 years of age or older. We reviewed our experience on patients in this age group who underwent a temporal lobectomy at our Comprehensive Epilepsy Program. Methods: We reviewed our surgery database and identified patients who underwent a temporal lobectomy for intractable seizures at the age of 60 years or older. All patients underwent a comprehensive presurgical evaluation, including EEG‐video monitoring, MRI, SPECT or PET, neuropsychologicla testing, and Wada test. Post surgical outcome was measured using Engel classification. Follow up ranged from 5 months up to 20 months. Results: Over a 5 year period, 7 patients 60 years or older underwent a temporal lobectomy at our Epilepsy Center. This represents less than 5% of resective epilepsy surgeries performed. Patient's ages ranged from 60 to 76 years of age (mean 63.4 years). Duration of epilepsy ranged from 1 year to 53 years (mean 34.8 years). Etiology was mesiotemporal sclerosis in 4 patients and unclear in 3. Invasive EEG was performed in 1 patient. Post‐surgical outcome was class I in 4 patients, class II in 2 patient. No patients fell into Class III or Class IV. One patient developed psychogenic seizures after surgery. Conclusions: Temporal lobectomies are rarely performed after the age of 60, but should be considered and seem so have a comparable outcome to that of younger patients. 1 Katherine Bayless, 1 Jana E. Jones, 2 Lincoln Ramirez, and 1 Bruce P. Hermann ( 1 Neurology, University of Wisconsin, Madison, WI ; and 2 Neurological Surgery, University of Wisconsin, Madison, WI ) Rationale: This study examined the long‐term (10 years+) psychosocial outcomes of a consecutive series of patients who underwent anterior temporal lobectomy (ATL) compared with a group of medically managed patients. These same individuals were initially assessed in 1999 (mean 5 years follow‐up) (Jones et al., 2002). This second follow‐up interview occurred 6 years later in 2005. The primary focus of this study was to examine the changes maintained in actual life performance after surgery in the following areas: employment, independent living, driving and financial independence. Self‐reported quality of life was also assessed. Additionally, during this second follow‐up interview questions pertaining to the presence or absence of mood and anxiety disorders were included. Methods: The sample consisted of 57 adults who underwent ATL and 18 controls that were evaluated for surgery during the same time period as those who had surgery. However, the medically managed group did not have surgery and continued to receive medical management. Participants were interviewed 8–15 years after surgery or presurgical evaluation. A structured clinical interview was conducted to obtain information regarding seizure frequency and psychosocial status. Seizure frequency was evaluated for the number seizures in the year prior to the interview and number of seizures since the initial follow‐up interview in 1999. Results: As indicated in the initial follow‐up evaluation in 1999, significant differences in psychosocial outcomes were maintained between the surgical and medical management groups (p < .001). Favorable psychosocial outcomes were significantly more common in surgical participants (50.9%) than medically managed patients (11.1%). Individuals in the surgery group were more likely to be employed (70.2% v. 44.4%), driving (77.2% v.33.3%), living independently (80.7% v. 44.4%) and financially independent (80.7% v. 44.4%). Interestingly, there were no significant differences between the two groups in terms of self reported quality of life. In terms of seizure frequency, 71.9% of the surgery group was seizure free in the past year compared to 27.8% in the medically managed group. Since 1999, 50.4% of participants in the surgery group have remained seizure free. Among the two groups 15.8% of the surgery participants reported a diagnosis of a mood or anxiety disorder in the past year compared to 17.6% in the medically managed group. There were no significant differences between the two groups. Conclusions: Surgery appears to be associated with the long‐term maintenance of favorable psychosocial outcomes at a very long‐term follow‐up of 8–15 years after surgery. Rates of employment, driving, independent living and financial independence appear to be positively influenced by surgery. Seizure freedom and reduced seizure frequency appear to be a significant long term outcome of surgery. 1 Christine M. Bower, 1 Eric M. Cheng, 2 Susan S. Spencer, 1 Stefanie Vassar, and 1 Barbara G. Vickrey ( 1 Department of Neurology, University of California Los Angeles, Los Angeles, CA ; 2 Department of Neurology, Greater Los Angeles VA Medical Center, Los Angeles, CA ; and 3 Department of Neurology, Yale University School of Medicine, New Haven, CT ) Rationale: Although the clinical goal of resective epilepsy surgery is seizure freedom, patients may have a wide set of expectations for this invasive procedure. This study's goals were to assess the nature, range, and frequency of preoperative expectations for resective epilepsy surgery, and to explore whether expectations vary across patient sociodemographic and clinical characteristics. Methods: 396 adults and adolescents with refractory epilepsy were enrolled in a seven‐center cohort study and underwent resective surgery. During baseline interviews, patients responded to open‐ended questions about expectations for surgery. Investigators reviewed text responses and identified a preliminary set of expectation themes. Each text response was placed on a card, then “sorted” into piles representing distinct themes (expectations) by consensus of three investigators. Differences in the frequencies for which each expectation theme was identified were explored across gender, ethnicity, education, whether temporal or extratemporal lobe epilepsy, and side of surgery, using chi‐square. Results: Among 391 respondents, nine unique expectations themes were identified by >15% of the sample: driving (n = 241, 61.6%); job/school (n = 166, 42.5%); independence (n = 115, 29.4%); seizure cessation (n = 100, 25.6%); social functioning (n = 90, 23.0%); quality of life (n = 82, 21%); medication discontinuance (n = 78, 20.0%); physical activities (n = 71, 18.2%); and cognition (n = 67, 17.1%). Non‐whites and whites differed on 3 of 9 expectation themes: non‐whites (n = 72, 18% of sample) endorsed job/school and cognition more frequently and driving less frequently than did whites (p < 0.03). Respondents with a higher level of education (n = 95, 24% of sample with bachelor's degree or higher) endorsed cognition more frequently than those with less education (p < 0.003). Subjects with a right‐sided resection (n = 190, 49% of sample) endorsed social functioning more frequently than those with a left‐sided resection (p < 0.05). There were no differences in expectations by gender or temporal versus extratemporal lobe epilepsy. Conclusions: Patients awaiting resective epilepsy surgery have definable expectations for surgery, the most predominant of which are driving and employment/educational attainment. While the majority of these expectations were similar across a range of patient characteristics, there were differences on one‐third of the most frequently reported expectations by ethnicity. Future studies are needed to explore these potential sociodemographic differences, and to understand how expectations may aide in the clinical decision‐making process of resective epilepsy surgery. (Supported by Grant #RO1 NS 32375 from the National Institute of Neurological Disorders and Stroke.) 1 Mar Carreño, 1 Juan Luis Becerra, 1 Joaquin Castillo, 1 Antonio Donaire, and 1 Iratxe Maestro ( 1 Neurology, Hospital Clinic, Barcelona, Barcelona, Spain ) Rationale: Little is known about the long term outcome of patients with medically refractory epilepsy who undergo presurgical evaluation but are not considered good surgical candidates or reject surgery. A previous study with a small number of patients found out that up to 20% of those patients may eventually become seizure free (Selwa et al, 2003). This information is essential to counsel patients attended in a surgical programme. Methods: We performed a retrospective chart review and telephone survey on all patients who had undergone presurgical evaluation at the Epilepsy Monitoring Unit (EMU) of Hospital Clinic (Barcelona, Spain) from 1995 until 2004, but did not undergo epilepsy surgery afterwards, either because they were not considered to be good surgical candidates or because the patient declined surgery. We inquired about current seizure frequency, antiepileptic drug treatment, overall impression of change, quality of life, marital and laboral situation. Results: From a total of 167 telephone calls, we finally contacted 80 patients who agreed to answer the survey. Of them, only 7 patients (9.3%) were currently seizure free, with a mean seizure free period of 3 years (2–6). Seizure freedom was associated with use of new AEDs in 2/7 patients. Five patients (6.2%) had died, with death being possibly related to seizures (SUDEP or CRA during prolonged seizures) in all of them. Among those patients who were not seizure free, the vast majority (89%) reported that their seizure frequency was either the same or lower than at the time of presurgical evaluation. No significant differences were found between seizure free and not seizure free patients respect to the type of epilepsy, etiology, presence of a lesion on the MRI and reason not to have surgery. Overall, 54% of all patients interviewed reported feeling better and more satisfied with their lives than when they were admitted to the EMU. Although most patients did not report changes in their marital or laboral status, a higher proportion of patients in the seizure free group (71%) were actively working, compared to 32% of patients in the group with continued seizures. Conclusions: A small percentage (9.3%) of patients with refractory epilepsy who undergo presurgical evaluation but not epilepsy surgery may eventually become seizure free on drugs. Up to 6% of patients may die as a consequence of their epilepsy. In patients who do not become seizure free, seizure frequency will be, in general, similar or lower with time, and up to 50% of patients will report feeling better and more satisfied with their lives than at the moment of presurgical evaluation. This information may be used to counsel patients attending epilepsy surgery programmes. 1 Shiro Chitoku, 1 Yshihiro Sumi, 2 Masahiro Mizobuchi, 1 Toshiaki Osato, 1 Takehiko Sasaki, 1 Jouji Nakagawara, and 1 Hirohiko Nakamura ( 1 Department of Neurosurgery, Nakamura Memorial Hospital, Sapporo, Hokkaido, Japan ; and 2 Department of Neurology, Nakamura Memorial Hospital, Sapporo, Hokkaido, Japan ) Rationale: To analyze characteristics and risk factors of seizures following ruptured aneurysmal subarachnoid hemorrhage (SAH). Methods: We studied a total of 544 patients with acute SAH treated with clipping (444), detachable coil embolization (37), and conservative treatment (67) at Nakamura memorial hospital between 1998 and 2004. The age ranged from 26 to 92 years old (mean 60 years old). We characterized seizures following SAH and compared risk factors of seizures between seizure group and non‐seizure group. Results: Sixty‐seven (12%) patients had seizures (seizure group) 58 (87%) in clipping, 7 (10%) in embolism, and 2(3%) in conservative treatment. Other 477 (88%) patients had no seizure history for 7 years follow‐up (non‐seizure group). Thirty (44%) of 67 patients presented seizure within 24 hours following SAH (acute onset), and 37 patients (56%) after 24 hours (late onset). Sudden generalized convulsions occurred in 27 (90%) patients in the acute onset group. Focal seizures occurred in 31 (84%) patients in the late onset group. Two (7%) patients had following epileptic seizure disorders in the acute onset group and 11 (30%) in the late onset group. Thirty‐one (46%) of 67 patients in seizure group resulted in good prognosis (modified Rankin Scale0–2) and 320 (65%) of 477 patients in non‐seizure group (p < 0.05) Seizure group showed more than three times higher rate of symptomatic hydrocephalus, 34(51%):76(16%); infarction due to vasospasm, 18(27%):41(9%); and other co‐morbidities, 34(51%):48(10%); compared with those of non‐seizure group (p < 0.05). Conclusions: Seizure group revealed worse prognosis and more correlated with hydrocephalus, vasospasms and co‐morbidities than non‐seizure group. Seizures following SAH characterized generalized convulsions at acute onset, and focal seizures at late onset. The late onset seizure group tended to develop epileptic seizure disorders compared with acute onset seizure group. 1,2,3 Luciana L. D´Alessio, 3 Juan José J.J. López‐Costa, 1 Hector H. Konopka, 1 Damián D. Consalvo, 1 Eduardo E. Seoane, 2 Laura L. Guelman, 1,3 Silvia S. Kochen, and 2 Luis María L.M. Zieher ( 1 Epilepsy Center, Ramos Mejía Hospital, Buenos Aires, Capital Federal, Argentina ; 2 Pharmacology Department, Medicine, Buenos Aires University, Buenos Aires, Capital Federal, Argentina ; and 3 CONICET, CONICET, Buenos Aires, Capital Federal, Argentina ) Rationale: Nitric oxide (NO) is a short lived free radical with diverse functions as a biological messenger molecule, and has been implicated in numerous aspects of the physiology and pathology of the CNS including epilepsy. NO can acts as anticonvulsant or proconvulsant depending on the experimental model and the pharmacological parameters used. There is little information about changes in NO levels in human epileptic brains. Neurons containing nicotinamide adenine dinucleotide phosphate‐diaphorase (NADPH‐d) synthesizes NO. The aim of this study is to investigate the anatomical distribution, morphology, optical density and cell sizes of NADPH‐d neurons in the temporal cortex of patients with hippocampal sclerosis who underwent anterior temporal lobectomy for intractable temporal lobe seizures. Methods: Brain samples of 7 × 5 × 5 mm from temporal cortex were obtained from 7 patients was compared with 5 postmortem controls. Samples were fixed and incubated at 37°C in 1 mg/ml of B‐NADPH and 0,2 mg/ml of nitroblue tetrazolio. NADPH d activity was measured as distribution of formazan deposits (endproduct of the histochemical reaction). Neuronal area and neuropil staining optical density was determined. Results: Sprouting (dendritic arborization) and larger NADPH‐d positive neurons, with an increase in staining intensity were found in epileptic´s temporal cortex (p < 0.05). Conclusions: The increased area and the observed morphological changes in NADPH‐d reactive cells found in this study indicate plastic changes and suggest an up‐regulation in NO system in the neocortex. These changes could contribute to abnormal cortical excitability however, could also be involved in adaptative neuroprotective function. 1,2 Alaa Hassan, 1 Heinz Pannek, 1 Friedrich Behne, 1 Reinhard Schulz, 2 Mohammed Nayel, 2 Ahmed Issa, and 1 Alois Ebner ( 1 Preoperative Evaluation, Epilepsy Center Bethel, Bielefeld, NRW, Germany ; and 2 Neurosurgery, Cairo University, Cairo, Cairo, Egypt ) Rationale: To evaluate the long‐term seizure outcome of 15 years of extratemporal neocortical epilepsy surgery of adults in the Bethel Epilepsy Center comparing the outcome between patients operated in three consecutive 5‐year periods. Methods: Between 1991 and 2006, 221 adult patients with extratemporal drug resistant epilepsies were operated on. From 221 patients 28 patients (12.6%) had more than one operation, five patients (2.6%) received VNS. We retrospectively reviewed the outcome of all patients. Patients with Rasmussen's encephalitis and those who only had a biopsy were excluded. Data was gathered by reviewing hospital files, outpatient reports, two‐year in‐patient postoperative follow‐up results, and using mail and telephone questionnaires. Three patients were not available for review, six patients died, and in ten patients data was incomplete. Patients were divided into three groups, depending on the time of operation: group 1 (1991–1995), group 2 (1995–2000), and group 3 (2000–2005). Results: Group 1: (54 patients, 24.4%) mean age of patients at seizure onset in years 10.6 ± 8.8 (range 1–41), mean age at surgery 28.13 ± 10.3 (range 16–57), mean duration of epilepsy 17.4 ± 9.7 (range 1 to 38), mean duration of follow‐up 11.4 ± 2.09 (range 1 to 14); outcome class 1 at the 6‐month follow‐up (47.0%), after 1 year (48.5%), 2 years (45.5%), 5 years (46.6%). Group 2: (90 patients, 40.7%), mean age at seizure onset 13.4 ± 10.4 (range 0.8–52), mean age at surgery 28.02 ± 10.1 (range 16–59), mean duration of epilepsy 15.6 ± 9.2 (range 0–44), mean duration of follow‐up 7.6 ± 1.6 (range 2–10 yrs); outcome class 1 at the follow‐up after 6 months (53.4%), 1 year (50.0%), 2 years (50.0%), 5 years (53.7%). Group 3: (77 patients, 34.4%) mean age at seizure onset 16.8 ± 15.0 (range 0.4–65), mean age at surgery 31.1 ± 13.1 (range 16–69), mean duration of epilepsy 14.2 ± 13.2 (range 0–64), mean duration of follow‐up 2.6 ± 1.5 (range 0.5–5 yrs); outcome class 1 at the follow‐up after 6 months (63.5%), 1 year (57.1%), 2 years (61.9%), 5 years (57.9%). A statistically significant difference was only found in the the 2‐year outcome group (p = 0.002). Conclusions: The comparison of postoperative seizure outcome of extratemporal epilepsy surgery in adults over the last 15 years shows mixed results. A statistically significant improvement was seen only in the two‐year postoperative outcome whereas at other points in time there was a tendency for improvement. 1,3 Frederick A. Schmitt, 2 Curtis A. Given, 1 Meriem K. Bensalem‐Owen, 1 Michelle L. Mattingly, and 1 Toufic A. Fakhoury ( 1 Department of Neurology, University of Kentucky Medical Center, Lexington, KY ; 2 Department of Radiology, University of Kentucky Medical Center, Lexington, KY ; and 3 Sanders Brown Center on Aging, University of Kentucky Medical Center, Lexington, KY ) Rationale: Presurgical evaluation of language and memory using the internal carotid WADA (ICAW) procedure is a standard approach for patients with medically intractable epilepsy. However, ICAW results often provide equivocal memory data, especially in cases with a dominant hemisphere seizure focus, as a result of aphasia with amytal injection. Therefore, selective preoperative intracarotid amytal procedures (IAP) have been developed to target hippocampal functions. Several variations of the selective IAP have been reported and include selective injections of amytal into: (1) anterior choroidal artery (AChA) and (2) posterior cerebral artery (PCA). This report focuses on a clinical series of patients who had selective amytal injections through the internal carotid artery (ICA) with temporary balloon catheter occlusion of the distal ICA beyond the AchA origin. Methods: Twenty‐seven patients with intractable epilepsy were evaluated using the distal ICA occlusion approach. Three patients underwent selective procedures for both hemispheres while the remaining 24 had selective injections targeting the hippocampus in the affected hemisphere along with a standard IAP of the contralateral hemisphere. Of these patients, 7 (26%) had previously ‘failed’ standard IAP injections given generalized anesthetic effects on the dominant hemisphere. Memory and language were evaluated in all patients using a modified McGill protocol that required learning of 3 visual objects and 3 verbal items after selective amytal injection. Subsequent free and cued recall of these items along with recognition was completed after resolution of the patient's homonymous hemianopsia. Results: All patients demonstrated homonymous hemianopsia with the selective injection and 78% showed brief temporal slowing on EEG ipsilateral to the injected hemisphere. Evaluation of memory showed encoding specific effects on memory items for 78% of cases where selective left‐sided injection resulted in proportionally poorer recall of verbally encoded items. In the 3 right‐sided selective IAPs, no memory encoding effects were observed. This ‘selective’ memory effect was independent of the presence of temporal slowing on EEG. Conclusions: Selective WADA procedures targeting the hippocampus through temporary occlusion of the supra‐clinoidal ICA and injection through the AchA can be an effective tool in the presurgical evaluation of patients when evaluating information on memory preservation post surgical resection of a seizure focus. Further, selective WADA procedures generally show memory effects that are consistent with the verbal or visual presentation mode of the memory items used in IAP assessments. 1 Georges A. Ghacibeh, 1 Jeffrey Smith, 1 Keyur Patel, 1 Sandeep Konka, and 1 Stephan Eisenschenk ( 1 Department of Neurology, University of Florida, Gainesville, FL ) Rationale: Patients with mesiotemporal epilepsy have a better chance of achieving seizure freedom after epilepsy surgery than those with neocortical epilepsy. The risk of seizure recurrence depends on several factors. Previous studies revealed that interictal epileptiform activity (IEA) on post‐operative EEG did not predict the risk of seizure recurrence. However, it is not known whether there is a difference based on the type of surgery. The aim of this study was to evaluate whether the relationship between IEA on post‐operative EEG and seizure recurrence after epilepsy surgery is different in patients who underwent neocortical and mesiotemporal resections. Methods: We reviewed retrospectively the electronic medical records of patients who underwent epilepsy surgery at the Comprehensive Epilepsy Program of the University of Florida between January 1997 and December 2002. Post‐operative follow‐up clinic notes were reviewed up to the last clinic note to determine the recurrence of seizures. EEG reports of the post‐operative EEG were reviewed to assess for the presence of IEA. Epileptiform abnormalities were defined as frank spikes or sharp waves or electrographic seizures. Results: A total of 94 patients who had both adequate clinic follow‐up and post‐operative EEGs were identified. EEGs were performed on average 106 days after surgery and patients were followed in clinic on average 32 months after surgery. Seventy patients had mesiotemporal resections. Of those, 16 (23%) had IEA on their EEGs and 22 (31%) had seizure recurrence. A chi‐square test revealed no significant relationship between seizure recurrence and the presence of IEA. Twenty four patients had neocortical resections. Of those, 14 (58%) had IEA and 12 (50%) had seizure recurrence. A chi‐square test revealed a significant relationship between the presence of IEA and seizure recurrence (p = 0.013). However, 14 out of the 22 mesiotemporal and 8 out of the 12 neocortical surgery patients had seizure recurrence within 3 months after surgery, prior to the post‐operative EEG. Conclusions: This study revealed that epileptiform activity on early post‐operative EEG correlate with seizure recurrence in neocortical but not mesiotemporal surgeries. However, since most patients have seizure recurrence within three months after surgery, evaluation of the utility of earlier EEG is needed. 1 Ryder P. Gwinn, 1 John Morgan, 1 Lisa Caylor, 1 Alan Haltiner, 1 Laura Allen, and 1 Michael Doherty ( 1 Epilepsy Center, Seattle Neuroscience Institute, Seattle, WA ) Rationale: Hippocampal depth electrodes are useful in lateralizing and localizing medial temporal lobe seizures. Historically these have been placed either orthogonally or transversely using frame based or frameless stereotaxy. This method can provide quality electrical recordings of hippocampal activity, but is a blind procedure with inherent risks and limitations including bleeding, hippocampal injury, inconsistent hippocampal orientation, and incomplete sampling. We report the placement of ventricular depth electrodes under framless stereotactic and neuro‐endoscopic guidance using direct visualization to assure safe and consistent placement of hippocampal recording electrodes, with quality recordings leading to successful seizure localization. Methods: Frameless stereotactic navigation (Steath Treon Plus, Medtronic) was used to guide occipito‐parietal burr hole placement and endoscope trajectory. A neuro‐endoscope (Channel™ Neuroendoscope, Medtronic) was placed into the atria of the lateral ventricle, then guided down the temporal horn under direct visualization. A 10 contact, 5 mm spacing depth electrode (Ad‐tech, Racine WI) was passed through the endoscope with 1 – 2 contacts placed into the temporal pole. The endoscope was then removed using flouroscopy to verify stable electrode position. Electrode position was verified by co‐registration of post‐operative CT scan to pre‐operative MRI in all patients. Results: Eleven ventricular electrodes were placed in 6 patients undergoing surgery for intracranial localization of partial seizures. Successful navigation of the temporal horn was achieved in all patients, and electrodes were verified to be touching the hippocampus in all patients visually and by post‐operative imaging. No infection, neurologic injury, or post‐operative hemorrhage was seen. Quality recordings were obtained from all electrodes, and seizure onsets were documented in 7/11 electrodes, with reversal potentials seen along the hippocampus using a bipolar montage. Two patients underwent anteromedial temporal resection 6 weeks post‐implantation, with pathology showing gliosis adjacent to the electrode location. Conclusions: Electrocorticography may be performed using depth electrodes placed in the temporal horn of the lateral ventricle. Placement of temporal horn electrodes under direct neuro‐endoscopic visualization can be performed safely, and provide high quality recordings of hippocampal activity without significant injury to the structure. Temporal horn electrodes are situated in a predictable anatomical relationship to hippocampal structures for reproducible recordings from the same hippocampal subfields between patients. Neuro‐endoscopically placed ventricular electrodes may offer a safe and effective alternative to orthogonal and longitudinal hippocampal electrodes for localization of medial temporal lobe seizures. 1 Jeffery A. Hall ( 1 Neurosurgery, Montreal Neurological Hospital, Montreal, QC, Canada ) Rationale: Proponents of tissue‐sparing selective temporal resections for patients with mesial temporal sclerosis and intractable epilepsy begrudgingly acknowledge the lack of neuropsychological data demonstrating any related savings with these procedures. Seizure outcome in well‐selected patients appears to be equivalent between the more “selective” and larger traditional temporal lobectomies. One interpretation of why no obvious savings follows from selective resection is that the tissue is somehow rendered dysfunctional by the operation. Alternatively, it may be that we are not adequately probing the function of the spared temporal tissue. Methods: Ten consecutive patients who underwent transcortical (second temporal gyrus) transventricular selective amygdalohippocampectomy for mesial temporal sclerosis and intractable epilepsy were examined post‐operatively with formal visual field testing. Each patient also had post‐operative MRI to demonstrate the extent of resection and exclude complications. These patients were compared to a reference case of intractable non‐lesional temporal lobe epilepsy operated by the same surgeon. This patient underwent a more radical traditional temporal corticoamygdalohippocampectomy. Pre and post operative visual fields and MRI are available for this patient also. Results: The patient who underwent traditional temporal lobectomy demonstrated the expected homonymous superior quadrantopsia. Patients who underwent the selective procedure demonstrated some evidence of visual field loss, albeit to a lesser extent than that of the reference case. Conclusions: This finding has several important implications the most of salient of which suggests that the selective procedure results in the sparing of viable and functionally important brain tissue. Additionally, these data support the notion from recent studies of MR tractography, that the fibers of Meyer's Loop course further anterior in the temporal lobe than is generally acknowledged. (Supported by Department of Neurology and Neurosurgery McGill University.) 1 Lizbeth Hernandez‐Ronquillo, 1 Jose F. Tellez‐Zenteno, 1 Samuel Wiebe, and 1 Nathalie Jette ( 1 Department of Clinical Neurosciences, University of Calgary, Calgary, AB, Canada ) Rationale: In recent years, advances in surgical techniques have allowed patients without identifiable neuroimaging abnormalities to undergo epilepsy surgery. However, there is a notion that patients with normal MRI or histopathology may have worse surgical outcomes than those with lesional epilepsy. The objective of this study is to provide evidence‐based estimates of the surgical outcome in patients with non‐lesional epilepsy and to identify sources of results variation between published studies Methods: Medline, Index Medicus, bibliographies of reviews, original articles, and book chapters were searched to identify published articles since 1991, describing outcomes in ≥5 patients of any age who underwent resective or non‐resective epilepsy surgery without an identifiable structural lesion by imaging or histopathology. Two reviewers independently assessed study eligibility and extracted the data. Disagreements were resolved through discussion. Results: Twenty‐nine articles fulfilled eligibility criteria and described outcomes in 655 patients with non‐lesional epilepsy surgery. Non‐lesional status was defined by MRI in 56% of studies and by histopathology in 44%. Overall, 24% of patients described were non‐lesional. The median proportion of seizure‐free patients was 44% (95% CI 39–49) in non‐lesional cases vs. 61% (95% CI 58–64) in those with lesions. The corresponding proportion was 46% (95% CI 39–53) if non‐lesional status was defined by MRI, versus 42% (95% CI 35–49) if defined by histopathology. In non‐lesional patients with temporal lobe surgery, 40% (95% CI 33–46) were seizure free. We explore aspects related to the validity of these results. Conclusions: Overall, non‐lesional epilepsy surgery has inferior results compared with lesional cases (44% vs 61% respectively). The response after surgery was the same whether lesional status was defined using MRI or histopathology. Sources of heterogeneity are explored. 1 Hyeon‐Seob Byeon, 1 Seung‐Chul Hong, 1 Jong‐Su Kim, 1 Jung‐Il Lee, 1 Jong‐Hyun Kim, 1 Do‐Hyun Nam, and 2 Dae Won Seo ( 1 Neurosurgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea ; and 2 Neurology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea ) Rationale: To compare seizure outcome in the patients with cavernous angioma treated by lesionectomy alone or further resection and to analyze the factors influencing on the surgical outcome. Methods: Fifty‐nine patients (27 male) with supratentorial cavernous angioma associate with seizure were operated in our institute. Clinical, radiological and electrophysiologic data were collected retrospectively. The mean age was 32 years (ranged 2–67) and the mean follow up duration after surgery was 34.3 months (ranged 12–103). Results: Thirty‐two patients underwent lesionectomy and remaining 27 patients underwent standard temporal lobectomy with tumor removal. Seizure outcome was compared using the Engel classification scheme between two kinds of technique. At the last follow up, 33 patients were involved in Engel class I, 6 patients in class II, 4 patients in class III and 2 patients were class IV. Estimated survival in patients who had shorter seizure histories and fewer preoperative seizures did not differ between the two surgical techniques. The important factors influencing the outcome were duration of seizure, location of tumor, multiple epileptic foci on electroencephalogram and surgical technique. Conclusions: These findings indicated that shorter seizure histories and fewer preoperative seizures can be effectively treated by lesionectomy alone, whereas those with longer histories and more seizures are not effectively treated by this procedure and may require epileptic surgery. And early surgical intervention is important for good treatment outcome. 1 Hariprasad Kunhiveedu, 1 Tobias Loddenkemper, 1 Stephan U. Schüle, and 1 Richard C. Burgess ( 1 Department of Neurology, The Cleveland Clinic, Cleveland, OH ) Rationale: Prediction of ictal onset zone by subdural interictal spike pattern has rarely been analyzed. The objective of our study was to analyze the relationship between interictal spikes and ictal onset zones on subdural EEG recordings. Methods: Fifteen patients (aged 25 to 61 years) undergoing prolonged subdural intracranial EEG recordings were included. On average 92 electrodes (range 42 to 119) were implanted. Representative interictal and ictal EEG recordings were reviewed. Interictal spike populations were identified by voltage from baseline, earliest onset, frequency of the spike population, presence of an aftergoing slow wave and duration. Distribution of the interictal activity was defined as 80% fall‐off from the maximum. The electrodes with the earliest rhythmic evolution determined the ictal onset. The relationship between interictal EEG and ictal onset electrodes was analyzed. Results: Interictal spikes were seen from a discrete single focus in five patients, from two different foci in six patients and were multifocal or diffuse in four patients. Ictal onset was seen from circumscribed areas in 14 patients and was diffuse in one patient. Cases with more than one spiking pattern were not associated with a higher number of implanted electrodes. Single EEG seizure patterns were recorded in eight patients. In five patients two different EEG seizure patterns were recorded, one patient had three different EEG seizure patterns and one additional patient had five different EEG seizure patterns. Interestingly, cases with more than one subdural EEG seizure pattern were associated with a higher number of implanted electrodes (p < 0.05). In eight patients (53%) the most frequent spike focus was concordant with a subdural EEG seizure onset zone. Six out of these eight patients presented with only one subdural EEG seizure type. One of these eight patients presented with very broad spikes, but their maximum was concordant with the ictal onset zone. In three additional patients (20%) the most frequent or second most frequent spike population occurred one or two electrodes (1–2 cm) adjacent to a seizure onset zone. In three patients diffuse spiking was seen whereas the seizure onset was focal. One patient had diffuse seizure onset with focal spikes. Conclusions: Ictal onset zones on subdural EEG recordings were related to the most frequent subdural interictal spike focus in 53% of cases. After inclusion of the second most frequent spike focus and surrounding electrodes, an ictal onset zones could be predicted in 73% by interictal spiking. However, not all ictal onset zones were predicted by the interictal spike focus, in particular if a patient presented with more than one ictal onset zone. Interestingly, a higher number of implanted electrodes and a higher number of recorded ictal onset zones may be related. Analysis of interictal subdural EEG may be helpful in the prediction of the ictal onset zone. 1 Addolorata Mascia, 1 Antonio Sparano, 1 Vincenzo Esposito, 1 Sergio Paolini, 1 Roberta Morace, 1 Giancarlo Di Gennaro, and 1 Pier Paolo Quarato ( 1 Epilepsy Surgery Unit, IRCCS Neuromed, Pozzilli, Isernia, Italy; Epilepsy Surgery Unit, IRCCS Neuromed, Pozzilli, Isernia, Italy; Epilepsy Surgery Unit; Department of Neurosurgery, IRCCS Neuromed; La Sapienza University, Rome, Italy, Pozzilli, Isernia, Italy; Epilepsy Surgery Unit; Department of Neurosurgery, IRCCS Neuromed; La Sapienza University, Rome, Italy, Pozzilli, Isernia, Italy; Epilepsy Surgery Unit; Department of Neurosurgery, IRCCS Neuromed; La Sapienza University, Rome, Italy, Pozzilli, Isernia, Italy; Epilepsy Surgery Unit, IRCCS Neuromed, Pozzilli, Isernia, Italy; and Epilepsy Surgery Unit, IRCCS Neuromed, Pozzilli, Isernia, Italy ) Rationale: To determine the potential effectiveness and tolerability of levitiracetam (Lev) versus oxacarbazepine (Oxc) versus placebo in preventing early postoperative seizures in patients undergoing brain surgery for mass lesions. Methods: 30 patients requiring elective craniotomy for supratentorial brain mass lesions were randomized to three groups of equal size, with a prospective, open‐label, controlled design. One group receveid Lev (13.3–40 mg/kg/die) the other group receveid Oxc (18–20 mg/kg/die), the third group receveid placebo. In all the three groups the drug administration were oral with a titration performed within the 30 days preceding surgery. Early postoperative seizures (seizures occurring in the 7 days following surgery) were considered for the purpose of the study. Patients having chronic epilepsy were not included. On the contrary, heralding seizures were not considered as esclusion criterion Results: Early post‐operative seizures were observed in all three groups. Patients taking Lev or Oxc showed less seizures than patients taking placebo. Conclusions: In our study Lev and Oxc seemed to be similarly effective in preventing early seizures after brain surgery. Both the drugs showed a good profile of tolerability and all patients remaining on each treatment during all the duration of the study. 1 Shearwood McClelland III, 2 Rebeca E. Garcia, 3 Daniel M. Peraza, 4 Tina T. Shih, 5 Lawrence J. Hirsch, 3 Joy Hirsch, and 2 Robert R. Goodman ( 1 Neurosurgery, University of Minnesota, Minneapolis, MN ; 2 Neurological Surgery, Columbia University College of Physicians and Surgeons, New York, NY ; 3 Radiology, Columbia University College of Physicians and Surgeons, New York, NY ; 4 Neurology, University of California at San Francisco, San Francisco, CA ; and 5 Neurology, Columbia University College of Physicians and Surgeons, New York, NY ) Rationale: The nondominant amygdala is crucial for processing facial expression and emotional recognition (ER) of visual stimuli, particularly in facial fear recognition. Patients with temporal lobe epilepsy (TLE) associated with mesial temporal sclerosis (MTS) often incur amygdalar and hippocampal damage. Previous studies have shown that patients with right‐sided (nondominant) MTS experience impaired ER if TLE onset occurred before age six. This finding has resulted in the hypothesis that early right mesiotemporal insult impairs plasticity, resulting in ER deficits, while damage later in life (seizure onset after age five) results in no deficit. However, this hypothesis has not been tested in a uniformly seizure‐free postsurgical population. This study was performed to examine this issue in late‐onset postsurgical patients. Methods: Controls (n = 10) and late‐onset patients (n = 5) were recruited. All patients underwent nondominant anteromedial temporal lobectomy (AMTL), had TLE onset after age five, Wada‐confirmed left‐hemisphere language dominance and memory support, MTS on both pre‐operative MRI and biopsy, and were Engel class I five years after surgery. Using a standardized (Ekman and Friesen) human face series depicting neutrality and the six basic emotions conveyed by facial expression (happiness, sadness, fear, disgust, anger and surprise), subjects were asked to match the affect of one of two faces to that of a simultaneously presented target face. Target faces expressed fear, anger, or happiness. Results: Statistical analysis (t‐test) revealed that the late‐onset group had ER (as measured by percentage of faces correct and reaction time) for fear (p = 0.871), anger (p = 0.256), and happiness (p = 0.608) comparable to controls. All 95% confidence intervals included zero. Conclusions: Following curative AMTL in nondominant MTS patients, TLE onset after age five does not predict significant ER impairment for facial expressions. Future studies are needed to examine whether early TLE onset continues to impair ER following AMTL. 1 Pradeep Modur, and 2 Warren Milteer ( 1 Neurology, University of Louisville, Louisville, KY ; and 2 Hazelwood Center, Louisville, KY ) Rationale: To investigate the impact of vagus nerve stimulation (VNS) therapy in institutionalized patients with Lennox‐Gastaut Syndrome (LGS) with respect to emergency room (ER) utilization for seizures. Methods: We identified patients with LGS from a retrospective chart review of 98 institutionalized patients with epilepsy. LGS was defined when all of the following criteria were met: mental retardation; multiple seizure types; frequent seizures refractory to antiepileptic drugs (AEDs); and prolonged video‐EEG monitoring showing 2 of these 5 findings – slow spike‐wave discharges, generalized fast activity, multifocal (≥3 sites) epileptiform discharges, tonic seizures and atypical absence seizures. The institution's protocol specified administration of intrarectal diazepam (Diastat) for seizures or seizure clusters lasting >3 minutes, and transfer to an ER if seizures persisted >10 minutes. Because of frequent seizure clusters in this population, the number of days with documented seizures was counted instead of the actual number of seizures. The comparator group consisted of patients without VNS (non‐VNS group) followed over a period of 18 months. The intervention group consisted of those with VNS, and were analyzed in two periods: 6 months before and 6 months after implantation. The primary outcome measure was the number of seizure‐related ER visits in the VNS group before and after implantation. Secondary outcome measures were the number of seizure days and the number of Diastat administrations in the VNS group during the same periods. Fixed effects poisson model and poisson regression were used for statistical analysis. Results: There were 9 patients with LGS, and 4 of them had been implanted with VNS. In the non‐VNS group, there were 5 patients (3 males), aged 24–53 years. In the VNS group of 4 patients (2 males) aged 25–56 years, the age at implantation was 22–53 years. Compared to non‐VNS patients, the VNS patients before implantation had significantly higher Diastat administrations (p < 0.0005) and ER visits (p < 0.0005) but not seizure days (p = 0.06). In the VNS group, during 6 months of follow‐up after implantation, the ER visits reduced significantly (10 versus 2, p = 0.04), the number of seizure days reduced minimally (107 versus 99, p = 0.58) while the number of Diastat administrations increased slightly (36 versus 50, p = 0.13). Conclusions: In a group of institutionalized LGS patients more prone to ER visits, VNS therapy significantly reduced the number of ER visits. The observed trend towards an increase in the administration of Diastat after VNS implantation suggests that the favorable outcome with respect to ER visits may be related to the combined effect of VNS and Diastat in reducing the duration of seizures or seizure clusters. Prospective studies are needed to confirm this and evaluate other potential factors such as VNS parameters and concomitant AEDs influencing this outcome. (Supported by Grant from the state of Kentucky.) 1 Paula Pergami, 1 Christian A. Keller, 1 Adriana E. Palade, 1 John F. Brick, 1 Warren Boling, 1 Alexander Torres‐Trejo, and 1 Daunice Lohr ( 1 Neurology, West Virginia University, Morgantown, WV; and Neurosurgery, West Virginia University, Morgantown, WV ) Rationale: The last few years have seen an increased interest in mechanisms regulating body weight. At hippocampal level, obesity can be viewed as the result of an imbalance between the tendency to seek a reward (food) and learned inhibitory processes. Our objective was to investigate the relation between removal of different brain structures and the development of postoperative eating disorder resulting in weight gain. Methods: In order to identify significant differences in weight changes, two groups of patients with intractable epilepsy were studied: patients with medically refractory partial seizures status post amygdalohippocampectomy (n = 25) and patients with extratemporal epilepsy who underwent surgery for the same condition (n = 18). BMI was calculated at baseline and at 6, 12, 18 and 24 months following surgery. Differences between the two groups were analyzed. The two groups were comparable for control of postoperative seizure, medications, age, gender and co‐morbidities. Results: Postoperatively, a significant difference in weight, consisting of weight gain was observed in patients with temporal lobe epilepsy following the amygdalectomy compared to patients undergoing extratemporal surgery. Conclusions: Our results further strengthen the hypothesis that the amygdala plays a role in eating behavior. This might involve complex neuronal circuits controlling behaviors like rewards, learning and memory formation. This finding supports a possible role for amygdala in eating disorders and challenges the traditional idea of eating disorders associated with hypothalamic disturbances. 1 Dimitris G. Placantonakis, 1 Saadat Shariff, 2 Douglas Labar, 2 Cynthia Harden, 2 Syed Hosain, 3 Neil Schaul, 3 Dimitrius Kolesnik, and 1 Theodore H. Schwartz ( 1 Neurological Surgery, Weill Cornell Medical College, New York, NY ; 2 Neurology, Weill Cornell Medical College, New York, NY ; and 3 Neurology, New York Hospital of Queens, Flushing, NY ) Rationale: Medically refractory epilepsy is amenable to neurosurgical intervention if the epileptogenic focus can be accurately localized. In some situations, the scalp EEG and MRI are discordant and/or non‐lateralizing, yet the clinician is still highly suspicious that a single focus exists, either based on a dominant interictal focus or stereotypical semiology. In these situations, a bilateral intracranial survey can be helpful to lateralize the ictal onset zone. The risks and success of such bilateral surveys have not been well described. Methods: We retrospectively reviewed 20 patients with refractory seizures treated over a 4‐year period. Patients underwent bilateral placement of subdural strip electrodes and, in 50% of cases, stereotactic implantation of depth electrodes into the mesial temporal lobes. In all cases the MRI scans and scalp EEG were not sufficiently concordant to lateralize seizure onsets. The mean age was 35.1 years and 75% of patients were male. Results: Patients were electrophysiologically monitored for an average of 7.4 days. Thirty‐five percent (n = 7) subsequently underwent unilateral implantation of electrodes to more accurately localize the seizure focus. Sixty percent (n = 12) eventually underwent a therapeutic procedure. Of those, 58.3% underwent resection of epileptogenic foci without unilateral invasive mapping. Reasons for no further surgery (n = 8) included failing the WADA (37.5%), multifocal onsets (25%), refusal of further treatment (12.5%) and negative intraoperative ECoG (12.5%). There was 1 complication, involving a retained electrode fragment that was removed in a separate minor procedure. Of the surgically treated patients, 58.3% have no seizures currently, 16.7% have improved and 25% remain unchanged. Conclusions: Bilateral placement of subdural strip and depth electrodes to survey the cortex in patients with non‐lateralizing scalp video‐EEG and MRI but clinical suspicion for localized epilepsy is both safe and effective. Of those patients who go on to a therapeutic procedure, 75% are either improved or seizure‐free. 1 Fabio Sebastiano, 1 Pier Paolo Quarato, 1 Vincenzo Esposito, 1 Angelo Picardi, 1 Antonio Sparano, 1 Addolorata Mascia, 1 Mario Manfredi, and 1 Giancarlo Di Gennaro ( 1 Epilepsy Surgery Unit, IRCCS Neuromed, Pozzilli, Isernia, Italy; Epilepsy Surgery Unit, IRCCS Neuromed, Pozzilli, Isernia, Italy; Epilepsy Surgery Unit; Department of Neurosurgery, IRCCS Neuromed; La Sapienza University, Pozzilli, Isernia, Rome, Italy; Center of Epidemiology and Health Surveillance and Promotion, Italian National Institute of Health, Rome, Italy, Rome, Lazio, Italy; Epilepsy Surgery Unit, IRCCS Neuromed, Pozzilli, Isernia, Italy; Epilepsy Surgery Unit, IRCCS Neuromed, Pozzilli, Isernia, Italy; Epilepsy Surgery Unit; Department of Neurology, IRCCS Neuromed; La Sapienza University, Pozzilli, Isernia, Rome, Italy; and Epilepsy Surgery Unit, IRCCS Neuromed, Pozzilli, Isernia, Italy ) Rationale: To evaluate a novel method for localization of subdural electrodes in presurgical assessment of patients with drug‐resistant focal epilepsy. Basic demographic, clinical, and neuroradiological characteristics of patients Patient N. Sex Age (years) Age at epilepsy onset (years) Epilepsy duration (years) Neurological examination Number of seizures per month MRI 1 f 37 19 18 Normal 20 r‐T dysplasia 2 m 21 11 10 Normal 20 r‐ and l‐TPO dysplasia 3 m 36 9 27 Normal 60 r‐P dysplasia 4 m 47 14 33 l‐upper quadranopia 16 r‐ and l‐TPO glyosis 5 m 28 7 21 Normal 20 Normal 6 f 12 2 10 l‐hemisoma sensory‐motor deficit 60 Normal 7 m 21 18 3 l‐upper quadranopia 10 Normal 8 f 31 31 19 Normal 20 r‐T tumour Methods: We studied eight consecutive patients with posterior epilepsy in whom subdural electrodes were implanted for presurgical evaluation. Electrodes were detected on post‐implantation brain CT scans through a semiautomated procedure based on a MATLAB routine. Then, post‐implantation CT scans were fused with pre‐implantation MRI to localize the electrodes in relation to the underlying cortical structures. Two independent raters tested the reliability of this procedure by comparing 3D‐rendered MR images of the electrodes with electrode position as determined by intraoperative digital photography. Results: In each patient, all electrodes could be correctly localized and visualized in a stereotactic space, thus allowing optimal surgery planning. For both raters, the agreement between the procedure‐generated images and the digital photographs was good in all patients. Conclusions: While our findings need confirmation on larger samples including patients with anterior epilepsy, this procedure allowed to localize subdural electrodes and to establish the spatial relationship of each electrode to the underlying brain structure, either normal or damaged, on brain convessity, basal and medial cortex. (figure 1) 1,2 Marianna V. Spanaki, 3 Kost V. Elisevich, and 1 Brien J. Smith ( 1 Neurology, Henry Ford Hospital, Detroit, MI ; 2 Neurology, Wayne State University, Detroit, MI ; and 3 Neurosurgery, Henry Ford Hospital, Detroit, MI ) Rationale: Epilepsy surgery is considered to be the most effective treatment in refractory epilepsy with seizure free outcomes from 45% to 67.9% (Engel J. Surgery for seizures. N Engl J Med 1996;334:647–652). However, little data exists on the number of older patients who are treated surgically, the post surgical complications and outcome rates in this patient population. Methods: We identified all patients who were 50 years and older and had epilepsy surgery for refractory epilepsy at Henry Ford Hospital (HFH) Epilepsy Program from 1994 to 2005. Data on age at surgery, duration of epilepsy, risk factors, noninvasive and invasive video/EEG monitoring, surgery site, pathology, outcome, and medical conditions at surgery as well as complications were collected through search of the Epilepsy Monitoring Unit admission database and the HFH electronic database (CarePlus). Patients who had tumors that presented with seizures and surgery was performed in the context of tumor treatment were not included. Descriptive statistics were applied. Results: Thirty five patients from 50 to 70 years old (median age = 55; 19 females) underwent epilepsy surgery at HFH. Duration of epilepsy was from 6 months to 50 years (median = 35 years). Nineteen patients (54%) with inconsistent localization during scalp continuous video EEG monitoring had invasive monitoring with grids, strips and/or depth electrodes. Thirty one patients had temporal resections (16 had left anterior temporal resection), 3 extratemporal and one subpial transection. Pathology revealed mesial temporal sclerosis in 14 patients, heterotopia in 2, dual pathology in 2, cavernous angioma in 2, dysembryoplastic neuroepithelial tumor in 1. In the remainder pathology was consistent with various degrees of gliosis or non diagnostic. Duration of follow up ranged from 1 month to 13 years (median follow up = 2 years). One patient was lost to follow up. Twenty five patients (25/34; 73.5%) were seizure free after surgery, 4 of which remained seizure free off antiepileptic medication. Medical history prior to surgery was consistent with hypertension (HTN) in 6 patients, hypercholesterenemia and HTN in 3, diabetes and HTN in 2 and coronary artery disease in 3 (2 of which had previous angioplasty). One patient had right MCA infarct one day after surgery that resulted in mild left hemiparesis. Following surgery, 1 patient developed transient angina and another one pulmonary embolism that was successfully treated. Conclusions: Surgical treatment of refractory epilepsy in older patient population is a safe treatment and yields seizure freedom outcomes that are comparable to those reported in young patients. Co‐morbid conditions and advanced age should not prevent physicians from offering epilepsy surgery to appropriate candidates. 1 K. Meng Tan, 1 Elson L. So, 1 Gregory D. Cascino, 2 Fredric B. Meyer, 2 W. Richard Marsh, and 1 Gregory A. Worrell ( 1 Department of Neurology, Mayo Clinic ; and 2 Department of Neurosurgery, Mayo Clinic, Rochester, MN ) Rationale: Patients with medically refractory temporal lobe epilepsy may be excellent candidates for epilepsy surgery. However, lateralization of the ictal onset zone is not always clear from non‐invasive studies. In such cases, bilateral temporal depth electrode implantation may be undertaken prior to considering temporal lobe resection. Although intraoperative pharmacologic activation is known to increase interictal epileptiform discharges on electrocorticography, the clinical relevance of this procedure has not been established. Methods: We performed a retrospective review of all epilepsy patients who underwent bilateral temporal depth electrode implantation during the period 1997–2005 and received intraoperative alfentanil activation. We identified ten patients and characterized the data provided by EEG (interictal and ictal), MRI, intraoperative electrocorticography (before and after pharmacologic activation) and subsequent prolonged intracranial EEG monitoring, with respect to localization of the ictal onset zone. We focused on the correlation between pharmacologic activation and lateralization of the epileptogenic temporal lobe as determined by intracranial EEG monitoring. Results: 6 females and 4 males meeting inclusion criteria were identified. Administration of alfentanil intraoperatively caused increased activation of epileptiform discharges bilaterally in 5 patients, and no definite activation in another 3 patients. In the remaining 2 patients, the side of pharmacologic activation did not correspond with the epileptogenic temporal lobe determined on subsequent prolonged intracranial monitoring. Conclusions: Intraoperative pharmacologic activation does not appear to be useful in patients with temporal lobe epilepsy of unclear lateralization. 1 Denise S. Taylor, 1 Heber L. Varela, 1 William O. Tatum, 1 Fernando L. Vale, and 1 Selim R. Benbadis ( 1 Neurology & Neurosurgery, University of South Florida & Tampa General Hospital, Tampa, FL ) Rationale: Lateralization of temporal lobe epilepsy (TLE) is sometimes unclear based on the non‐invasive presurgical evaluation. In this situation, invasive EEG‐video monitoring is performed, using either depth or subdural electrodes. We reviewed our experience with bitemporal subdural strips for lateralization of TLE. Methods: We reviewed the data on all patients who received bitemporal subdural strips for lateralization of TLE at our center from December 2000 to March 2006. We only analyzed patients whose sole question was lateralization of TLE; if non‐temporal coverage was performed, patients were excluded. Outcome used Engel's classification. Results: Over a 5‐year period, 21 patients underwent bitemporal subdural strips. Out of 1,356 patients monitored, this represents 1.5% of patients. The results for the 21 patients with bitemporal subdural strips evaluation was as follows: • 11 patients (52%) had clear lateralization (100% of seizures on one side) resulting in a temporal lobectomy: 8 class I, 2 class II, and 1 class III. • 5 patients had a predominance but less than 100% lateralization. These were offered a temporal lobectomy and presented before surgery as an unlikely seizure free outcome but a probable improvement (“palliative” interventions). 4 patients underwent a lobectomy (2 class III, 2 class IV), and 1 declined. • 1 patient underwent resection of a large arachnoid cyst that was contralateral to seizure onset (outcome class II). • 4 patients (19%) had true bitemporal epilepsy without predominance and no resection was performed. Conclusions: Subdural strips achieved lateralization of TLE in about half of cases where the non‐invasive evaluation failed to do so. When lateralization was 100% to one side, postoperative outcome was excellent. About 19% had true bitemporal epilepsy. 3 Fuyuki Hirashima, 1 Eric Dinnerstein, 1 Karen L. Gilbert, 2 David W. Roberts, 1 Vijay M. Thadani, and 1 Peter D. Williamson ( 1 Section of Neurology, Dartmouth Hitchcock Medical Center, Lebanon, NH ; 2 Section of Neurosurgery, Dartmouth Hitchcock Medical Center, Lebanon, NH ; and 3 Dartmouth Medical School, Hanover, NH ) Rationale: We wished to study variables that influence outcome, in terms of seizure control, among patients with mesial temporal sclerosis who undergo surgery. We compared two surgical techniques. Some patients had a standard temporal lobectomy, and others had a selective amygdalo‐hippocampectomy. In this retrospective study, we compared the outcomes of these surgical techniques with respect to seizure control. Methods: 61 patients with medically refractory epilepsy and mesial temporal sclerosis (seen on MRI and subsequent pathological examination) were evaluated. The two groups were very similar with regard to age, age at surgery, duration of epilepsy, handedness, risk factors such as febrile seizures, and side of resection. The number of anti‐epileptic medications, before and after surgery, was the same in the two groups. 30 patients had a standard temporal lobectomy, and 31 had the more selective procedure. The choice of procedure was arbitrary. Follow‐up information was collected at 6, 12, 18 and 24 months, and yearly thereafter. The mean duration of follow‐up was 4.38 ± 1.02 years. Outcome was assessed using the Engel classification. Patients were classified as Class I outcomes if they were seizure‐free for the previous year. To simplify the statistics, patients in Classes II, III, and IV were pooled. Neuropsychological testing was done before surgery, and 6–12 months after surgery. Results: Class I outcomes ranged from 63% to 96%. A trend towards a difference between the two surgical groups, with more Class I outcomes among those who had the standard procedure, was detected 4 years after surgery (89% vs. 67%), and a statistically significant difference was apparent 5 years after surgery (96% vs. 63%, p < .01). However, if the data were analyzed with Class I defined as seizure‐free since surgery, there was no difference between the two groups. Neuropsychological testing did not show a significant difference between the two groups before or after surgery. Conclusions: These data suggest that in patients with mesial temporal sclerosis a better outcome with respect to seizure control may be seen with a wider resection of the anterior‐lateral temporal lobe. The difference, if there is one, depends on how a Class I outcome is defined. Roughly the same% of patients in both groups has post‐operative seizures; but after several years, a higher% of those who had the wider resection has been seizure‐free for at least a year. The clinical significance of this difference is uncertain. 1 Alejandro Torres‐Trejo, 1 Adriana Palade, 1 Warren Boling, and 1 John F. Brick ( 1 Neurology, West Virginia Univerisity, Morgantown, WV ) Rationale: After epilepsy surgery for cortical dysplasia, a high percentage of patients may remain seizure free. The clinician for various medical reasons may decide to discontinue one or more antiepileptic drugs. It is reasonable to consider discontinuation of the antiepileptic drugs (AED) to perhaps decrease some of the side effects associated with these medications. We would like to identify whether patients will continue to remain seizure free after discontinuation of AED's. Methods: We are reporting five patients who underwent epilepsy surgery for resection of epileptic focus with pathology reported as cortical dysplasia. The seizure focus was localized electrographically. Structures resected include right posterior temporal region, right central sulcus, amygdalohippocampectomy, right frontal lobectomy, and a left parietal region resection. For various medical resons all of the patients had their AED's discontinued between six months to three years. Results: All of these patients remained seizure‐free between six months to three years after surgery, but had recurrent seizures after withdrawal of medications was attempted. Conclusions: Patients that had successful epilepsy surgery resecting the epileptic focus due to cortical dysplasia remained seizure free while on AED's but had recurrence of seizures when the medications were discontinued. This suggests that patients with cortical dysplasia may fail AED discontinuation despite having good seizure control postoperatively if the AED's are discontinued before three years. 1 K. Upchurch, 2 J. Stern, 2 R. Staba, 2 C.L. Wilson, 2 J. Engel Jr., and 1 I. Fried ( 1 Neurosurgery, David Geffen School of Medicine, UCLA, Los Angeles, CA ; and 2 Neurology, David Geffen School of Medicine, UCLA, Los Angeles, CA ) Rationale: For cases of pharmacologically refractory epilepsy in which data from noninvasive diagnostic studies (Phase I) are discordant with respect to the epileptogenic zone, a hypothesis concerning the location of ictal onset is formulated. On the basis of this hypothesis, patients may undergo invasive tests such as ictal EEG recordings via intracerebral depth electrodes (Phase II). The purpose of this study was to assess clinical outcomes (diagnostic and therapeutic) of invasive chronic depth electrode recording for patients with medically intractable epilepsy in a series of 100 patients who underwent operation by a single surgeon at a single institution. Methods: Using the UCLA Seizure Disorders Center surgical database, the history, operative course, and clinical outcomes of 100 patients with medically intractable epilepsy who underwent implantation of intracerebral depth electrodes at UCLA Medical Center between January 1993 and August 2005 were retrospectively reviewed. Results: Of the 100 patients reviewed, 54 were males; mean age was 33.2 years; and age range was 12–50 years. All surgeries were performed stereotactically by the same surgeon (I.F.) with pre‐operative targeting based on stereotactic cerebral angiography and magnetic resonance imaging. The average number of depth electrodes implanted for each patient was 10 (range 6 to 12). In 60/100 patients, a single epileptogenic zone was localized via the first Phase II study; 10/100 underwent a second Phase II study with subdural grids/strips; and 59/100 underwent Phase III resection. Morbidity secondary to intracerebral depth electrode implantation consisted of: (1) small intracerebral hematoma incidentally revealed on post‐implant or post‐deplant imaging in three asymptomatic patients; (2) contaminant bacteria on either Gram stain or culture on deplanted electrodes in two asymptomatic patients who were clinically without other evidence of infection; and one episode of fever and cerebrospinal fluid pleiocytosis during implantation in one patient whose deplanted electrodes and cerebrospinal fluid showed negative bacterial cultures. Mortality was zero. Conclusions: These data suggest that chronically implanted intracerebral depth electrodes for electrophysiologic recording are safe and continue to be important in patient selection for Phase III epilepsy surgery, despite the emergence of sophisticated noninvasive diagnostic modalities. (Supported by Epilepsy Foundation Post‐Doctoral Research Fellowship.) 1,3 Hans E. Van Der Aa, 2 Jacqueline J. Ardesch, 2 Gerard Hageman, and 3 Hendrik P.J. Buschman ( 1 Neurosurgery, Medisch Spectrum Twente, Enschede, Netherlands ; 2 Neurology, Medisch Spectrum Twente, Enschede, Netherlands ; and 3 Institute for Neuromodulation (TWIN), Medisch Spectrum Twente, Enschede, Netherlands ) Rationale: VNS‐therapy for treatment of patients with medically refractory epilepsy in the Netherlands is part of a continuous quality improvement system. This encompasses the organisatory structure, responsibilities, processes and facilities for carrying out quality care. One of the basic and core elements of a quality system is the systematic monitoring of the process of care via the quality circle. The first outcome results of VNS‐therapy in patients with epilepsy within this setting in the Netherlands are presented. Methods: The quality system was developed in a close collaboration with medical specialists from 5 neurosurgery and 5 epilepsy centers. Monitoring of the VNS‐quality system is performed in three phases (intake, implant, and follow‐up). The main indicators are reduction of epileptic seizures and changes in seizure severity, quality of life (QOLIE‐89) and adverse events. The effect of VNS during follow‐up was recorded after 1, 3, 6, 9 and 12 months. Results: Between 2002 and 2006, 122 patients were included following the national protocol for VNS‐therapy. During the first 2 years patient referral was low (∼12 patients p.y.). Thereafter ∼30 patients per year were included. Of these patients 9 were excluded. Of the remaining patients, at the time of analysis, seventy had received an implant, 22 were on the waiting list for implantation and 21 were in the intake phase. The mean age at intake was 40 y (SE 1.5 y). The onset of epilepsy in these patients was at 12.7 y (SE 1.1 y), and had a history with medically refractory epilepsy of 27.7 y (SE 1.5 y). They had 60.0 (SE 11.1) epileptic seizures per month. Most patients had localized partial epilepsy (59% symptomatic and 31% cryptogenic). A medically significant (>40%) decrease on seizure frequency during VNS was observed in 35% of patients after 1 month, and 49%, 61%, 42% and 64%, after respectively 3, 6, 9 and 12 months. Seizure severity was markedly reduced in 46%, 46%, 52%, 62% and 55% of patients after respectively 1, 3, 6, 9 and 12 months. About 80% of patients used ‘magnet activation’ to prevent or halt a seizure. 52% of patients experience a better or much better post‐ictal recovery period with VNS. Side‐effects were mainly voice related (hoarseness: 68%) and throat pain (32%). Quality of life showed statistically significant improvement in “change in health” and “general health”. Conclusions: Monitoring the effect of VNS‐therapy in a national quality‐system setting guaranties high quality care. Moreover it provides an effective means to review the outcome results of the participating centers. This national data‐base may eventually help identify response predictors to further improve the effectiveness of VNS‐therapy, which will subsequently be implemented in the national VNS‐therapy protocol. 1 Kenou Van Rijckevorsel, 2 Alexandra Volckaert, 1 Marianne De Tourtchaninoff, 1 Geraldo Vaz, and 1 Christian Raftopoulos ( 1 Reference Center for Refractory Epilepsy, Université Catholique de Louvain, Brussels, Belgium ; and 2 Psychology Faculty, Université Catholique de Louvain, Brussels, Belgium ) Rationale: Epilepsy surgery is an effective treatment for seizure control, especially in temporal lobe epilepsy (TLE). Morbidity is low, often limited to visual field defect and memory problems. However, long‐term cognitive and emotional effects are poorly documented, especially in selective amygdalo‐hippocampectomy (Sel‐AH). This paper studied memory (hippocampus,) emotions (amygdala) and quality of life (QOL) in 21 TLE patients after Sel‐AH. Methods: Out of 31 Sel‐AH patients operated between 2001 and 2005, 21 adults (5M, mean age 35 (19–53) years, 15 left) completed the tests; 4 children or teenagers were excluded, 4 refused or did not answer, 2 died of possible SUDEP more than 2 years after AH. Memory (immediate, work and general) was tested subjectively (QAM) and objectively (MEM‐III Wechsler 3rd ed). QOL was appreciated with QOLIE‐31‐P questionary and emotions were analyzed by presenting computed pictures of fear, sadness and joy (Mann‐Witney and Greenhouse‐Geisser tests). Results were compared to matched controls or control population. Results: 15 (71.43%) are seizure free since surgery. 4 complained spontaneously from memory and 4 were chronically depressed, already before AH. For memory, patients' performances are inferior to controls' in QAM for reading and movie, especially in left AH. With MEM‐III test, patients are less competent than controls, but work memory is the most preserved or the best recovered without side inference of AH. There is no correlation between QAM and MEM‐III results. For QOL, patients are less anxious than unoperated epilepsy controls regarding seizures and medications. However, patients are more anxious than controls in general regarding the ongoing (test) situation. In emotional testing, there are significant differences between controls and patients, for control situations (human‐animal and male‐female) and in emotional testings: patients made more errors and more omissions with a slightly increased reaction time (RT). However, these differences were minimal without side inference of AH. Conclusions: Memory, QOL and emotions have been tested in 21 adult Sel‐AH patients, compared to matched controls or control population. Memory is less performant in patients compared to controls, but not deteriorated by surgery. QOL is improved in surgery patients compared to unoperated epilepsy patients regarding seizures and medications. For emotions, patients differ significantly from controls by making more errors and omissions with a slightly increased RT. However, these results are the same for test and emotional situations and are probably unrelated to surgery. 1 Marcos Velasco, 1 Ana L. Velasco, and 1 Francisco Velasco ( 1 Stereotactic and Functional Neurosurgery, Hospital General de México, Mexico City, Mexico ) Rationale: Resective surgery of the epileptic focus yields very good results, nevertheless there are cases that escape this surgical possibility, i.e. patients with bilateral hippocampal foci or patients with epileptic focus located nearby eloquent areas for speech and memory (usually the left side). These latter patients cannot be operated on because it would mean having severe neurological impairment particularly related to short term memory. These patients are candidates for neuromodulating procedures. The present study evaluates the long term efficacy of ESH in the treatment of complex partial seizures in patients with intractable mesial temporal lobe epilepsy. Methods: Nine patients were studied. They had at least a 6 month basal period during which a careful seizure record was obtained. They all underwent bilateral hippocampal electrode implantation to establish focus was laterality and location. 3 patients had bilateral and 6 had unilateral foci. Diagnostic electrodes were explanted and definitive Medtronic electrodes were implanted directed to the hippocampal foci. Position was confirmed with MRI and afterwards the DBS system internalized. Patients signed the informed consent approved by the Hospital's Ethics Committee and started a double blind stimulation protocol. Patients attended every 3 months for seizure count and neuropsychological tests. Results: Follow up went from at least 18 months to 5 years. Patients were divided in two groups: 5 had normal MRIs and 4 had hippocampal sclerosis. Patients with normal MRI had a seizure reduction of >95%. Patients with hippocampal sclerosis had a seizure reduction of 30–50%. None of the patients showed deterioration of neuropsychological function, patients who had >95% seizure reduction, showed improvement in the neurophsychological tests. No patient showed adverse effects. 1 patient had to be explanted after 2 years due to skin erosion in the trajectory of the system. 1 patient had skin erosion which was corrected with plastic surgery. Conclusions: ESH provides a non lesional method that improves seizure outcome without deterioration of memory in patients with unilateral or bilateral hippocampal epileptic focus. 1 Clarissa L. Yasuda, 1 Leonardo Bonilha, 1 Chris Rorden, 1 Helder Tedeschi, 1 Evandro Oliveira, 1 Li M. Li, and 1 Fernando Cendes ( 1 Neurology, State University of Campinas (UNICAMP), Brazil; Neuropsychiatry, University of South Carolina; and Communication Sciences and Disorders, University of South Carolina ) Rationale: Resection of hippocampus (HIP)has been the standard procedure for patients with drug resistant (DR) medial temporal lobe epilepsy (MTLE) (1), with a success rate of approximately 75%(1,2). The reasons for the failure of almost 25% of the operated patients remain unknown. We aimed to investigate the influence of resection of medial temporal lobe (MTL) structures in addition to HIP in the surgical outcome(SO) of these patients. Methods: We studied 43 patients with DR, unilateral MTLE who underwent anteromedial temporal resection in the Epilepsy Service at UNICAMP, between 1998 and 2004. We selected patients according to our standard surgical investigation protocol (2) and included only patients with hippocampal atrophy supporting both clinical and electrophysiological findings. This study was approved by the ethics committee of our institution. All patients underwent a routine MRI scan at least six months after the surgery, with the following parameters: T1weighted with 1 mm isotropic voxels acquired on a 2 Tesla scanner. We used the Engel's postoperative outcome scale (1) and conducted a voxel lesion symptom mapping (VLSM) (3) of their post‐operative MRI to discriminate the contribution of distinct segments of the MTL cortex to the SO (4). We computed linear regressions and applied a cost function transformation of the resection maps for each patient to a common set of spatial coordinates. Results: Patients underwent surgery with the mean age of 37 (range 17–56) years, and the mean follow up period was 40 (range 11–99) months. The SO classification was: 76% as Engel I,14% as Engel II,5% as Engel III and 5% as Engel IV. We observed that the extent of hippocampal removal was associated with better outcomes (t = 2.371, p = 0.023). However, when the resection of the HIP was associated with resection of the entorhinal cortex there was a greater likelihood of seizure control after surgery (t = 3.286, p = 0.002) (Fig. 1). Figure 1. VLSM maps are overlaid in a normal T1 template. Colors that are associated with low p‐values (green, yellow and red) are the most significantly associated with a good SO and correspond to the entorhinal cortex and HIP. (figure 1) Conclusions: Based upon this finding, the efficiency of the surgical treatment of MTLE can be improved by including the entorhinal cortex to the hippocampal resection. References 1. N Engl J Med 2001;345:311–318. 2. Seizure 2006;15:35–40. 3. Nat Neurosci 2003;6:448–450. 4. Hum Brain Mapp 2004;22:145–154. (Supported by FAPESP:05/59258.) 1,2 Arthur Cukiert, 1,2 Jose A. Burattini, 1,2 Pedro P. Mariani, 1,2 Lauro Seda, 1,2 Cristine M. Baldauf, 1,2 Meire Argentoni‐Baldochi, 1,2 Carla Baise‐Zung, 1,2 Cassio R. Forster, and 1 Valeria A. Mello ( 1 Neurology & Neurosurgery, Hospital Brigadeiro, Sao Paulo, Sao Paulo, Brazil ; and 2 Neurology & Neurosurgery, Clinica de Epilepsia de Sao Paulo, Sao Paulo, Sao Paulo, Brazil ) Rationale: Children with refractory epilepsy often present with delayed growth. We studied the effect of epilepsy surgery on linear growth of kids with refractory focal or generalized epilepsy. Methods: Eleven consecutive children with refractory epilepsy were evaluated as for their linear growth after epilepsy surgery. Six kids were submitted to focal cortical resection and five to extensive callosal section (90%). Age ranged from 3 months to 8 years (mean = 4.1 years). All patients had daily seizures. Four children had cortical dysplasia on MRI, two had gliotic lesions and five diffuse atrophy (those submitted to callosal section). Antiepileptic medications were kept at the same regimen as preoperatively, at least for one postoperative year. Follow up time ranged from 1.1 to 5 years (mean = 1.8 years). Results: Preoperatively, one patient was growing under the 2.5 growth percentile, 8 within the 2.5 and 10 percentile and one within the 10 and 25 percentile intervals. All patients submitted to cortical resection have been rendered seizure‐free after surgery; all kids submitted to callosal section had a 90% decrease of the generalized seizure frequency, but none was seizure‐free. All patients improved their linear growth after surgery and all of them were growing at least two percentile intervals above the preoperative level. Although not statistically significant, this “catch‐up” phenomenon was clinically more noticeable in those kids submitted to callosal section. Conclusions: Children with refractory epilepsy that underwent successful epilepsy surgery caught‐up with linear growth after surgery. Although this could be seen as recovery from a chronic illness in the kids with focal resections that were seizure‐free postoperatively, it is unlikely that an isolated response to improved health would be responsible for the additional growth in patients submitted to callosal section, who were not seizure‐free postoperatively. In this pediatric population, an inhibitory effect on growth from the epileptic activity might explain this postoperative catch‐up phenomenon, especially in children with generalized epilepsy. (Supported by Sao Paulo Secretary of Health.) 1 Jason S. Doescher, 1 Frank J. Ritter, 1 Patricia E. Penovich, 1 Deanna L. Dickens, 2 Mary Beth Dunn, and 1 Michael D. Frost ( 1 Minnesota Epilepsy Group, PA, Minnesota Epilepsy Group, PA, St. Paul, MN ; and 2 United Neurosurgery Associates, United Neurosurgery Associates, St. Paul, MN ) Rationale: Many clinical variables are considered when selecting pediatric epilepsy surgery candidates with the objective of predicting beneficial outcome. Methods: We analyzed all pediatric patients who underwent initial resective surgery at Minnesota Epilepsy Group from Jan 2000‐Dec 2002. Beneficial outcome was defined as Engel class I or II. Beneficial outcomes were compared to many diagnostic variables: attainment of early childhood milestones, presence of nocturnal seizures, history of febrile seizures, history of status epilepticus, symptomatic versus cryptogenic etiology, temporal vs. extra‐temporal lobe seizures, complete vs. partial resection, and presence of multiple conflicting diagnostic variables. Outcome measurements were done at 24‐months. Statistical method utilized was Fisher's Exact Test. Results: 46 consecutive subjects were identified and reviewed; 23 (50%) underwent temporal lobe only resection; extra‐temporal resections included 10 (22%) frontal, 1 (2%) parietal, and 12 (26%) multi‐lobe. In subjects with available follow‐up data, 19/33 (58%) were seizure free at 24‐months. Achievement of early childhood milestones was not associated with beneficial seizure outcome. The clinical history of nocturnal seizures, febrile seizures, or status epilepticus did not statistically separate beneficial outcomes groups. Etiology was not a statistically significant influence on beneficial outcome at 24‐months (symptomatic 23/30: 77%; cryptogenic 1/3: 33%; p < 0.174); however, this is may be influenced by low numbers within the cryptogenic group. Temporal lobe surgery was statistically more successful in achieving a beneficial outcome of Engel I‐II in 15/17 (88%) subjects compared to 9/16 (56%) of the extra‐temporal group (p < 0.046). Complete resection of the epileptogenic zone was associated with beneficial outcome in 20/24 (83%) subjects compared to partial resections 4/9 subjects (44%) (p < 0.039). The presence of multiple conflicting diagnostic variables was the strongest predictor of beneficial outcome as 24/27 (89%) subjects achieved Engel I‐II compared to none of the 6 subjects with >1 contralateral diagnostic variable (Fisher's Exact test p < 0.001). Conclusions: Favorable diagnostic variables associated with an Engel class I‐II outcome following epilepsy surgery include temporal lobe resections, complete resections, and absence of multiple contralateral diagnostic variables. In this retrospective review, the absence of multiple contralateral variables was the strongest predictor. 1 Georg Dorfmüller, 1 Emidio Procaccini, 1 Martine Fohlen, 1 Christine Bulteau, 1 Claude Jalin, and 1 Olivier Delalande ( 1 Pediatric Neurosurgery, Fondation Rothschild, Paris, France ) Rationale: To present our single‐center experience with a frameless robot‐guided endoscopic system for the transventricular disconnection of hypothalamic hamartomas (HH) associated with drug‐resistant epilepsy. Methods: Fourty‐three patients (9 m to 34 y, median age: 10 y) were operated on either through a pterional approach (8) or by means of transventricular endoscopy (21) or subsequently by both techniques (14), in order to disconnect the HH below or at the wall of the third ventricle. For the endoscopic approach we used a frameless stereotactic robot‐guided navigation (Neuromate®, Schaerer‐Mayfield). The trajectory of the endoscopic approach was defined by the hamartomas plane of attachment on the hypothalamus, for which we have proposed an anatomo‐topographic classification. Results: In terms of seizure outcome, we achieved the best results in HH with a vertical broad plane of attachment, extending into the third ventricle, enabling us to disconnect it through the endoscopic route. In this subtype (n = 12), 83% of the patients became seizure‐free, as compared to 43% in the other three types (horizontal attachment, attachment in multiple planes and giant HH). Other parameters, such as the patients age, seizure type and frequency or the duration of illness, were not significantly related with the outcome. In all, of 42 patients followed up, 21 (50%) became seizure‐free and 2 had rare events, while 17 (41%) had a worthwhile reduction in their seizure frequency, and 2 patients remained unchanged. Furthermore, improvement in several patients with disease‐related cognitive and behavioural disturbances could be objectified postoperatively. In terms of surgery‐related morbidity, the endoscopic approach had a significantly lower complication rate. Conclusions: Our results confirm the safety and efficacy of the disconnection of HH extending into the third ventricle with our robot‐guided endoscopic approach. 1 Eliana Garzon, 2 Ajay Gupta, 2 Jennifer Haut, 2 Deepak Lachhwani, 2 Prakash Kotagal, 2 Elaine Wyllie, 3 William Bingaman, 1 Americo Sakamoto, and 2 Hans Luders ( 1 Neurology and Neurosurgery, Universidade Federal de Sao Paulo, Sao Paulo, Sao Paulo, Brazil ; 2 Neurology, Cleveland Clinic Foudation, Cleveland, OH ; and 3 Neurosurgery, Cleveland Clinic Foundation, Cleveland, OH ) Rationale: In infants and young children with congenital or early onset acquired lesions, interictal scalp EEG may show generalized or multiregional abnormalities but ictal EEG concordant with semiology and brain MRI usually clarifies the epileptogenic zone for surgical option. We studied 6 children with catastrophic epilepsy whose ictal scalp EEG was classified as maximum over the unaffect hemisphere contralateral to the brain lesion. Factors leading to paradoxical lateralization of ictal scalp EEG over the unaffected hemisphere were evaluated. Methods: Of 209 patients in the pediatric epilepsy surgery database of Cleveland Clinic Foundation, clinical, EEG, brain imaging, surgery and postoperative records of 6 patients, whose EEG was classified as maximum in the unaffected hemisphere were reviewed. Digital records were restored and analyzed independently. Paradoxical lateralization was defined as scalp EEG seizure pattern that developed eiher exclusively or bilaterally but maximally contralateral to the to the side of the lesion. Results: Six patients, 4 females, 3–14 years in age at the time of the pre‐surgical evaluation, were identified. Despite the paradoxical lateralization of ictal scalp EEG during recorded seizures, lateralizing signs (>1 sign in 3 patients) of focal tonic (4 patients), head versive (2 patients), hemiclonic (1 patient) and asymmetric epileptic spasm (1 patient) were noted contralateral (concordant) to the side of lesion. Interictal EEG showed generalized or lateralized and multiregional epileptiform discharges with predominance (> 65%) over the affected hemisphere in 3/6, generalized but maximum over the unaffected hemisphre in 2/6, and generalized with no focal predominance in 1 patient. Brain MRI showed extensive destructive (encephaloclastic) lesions such as stroke in 3 and encephalomalacia (post status epilepticus and infectious) in 3 patients. All patients underwent hemispherectomy and 5/6 patients were seizure free after 5 months to 3 years of follow‐up. Conclusions: Extensive congenital or early onset acquired encephaloclastic lesions may produce an ictal scalp EEG pattern that appear maximum over the unaffected hemisphere. Interictal EEG in these patients may also show bilateral and multiregional abnormalities with or without any predominant region of abnormality. These scalp EEG patterns should be recognized and they do not preclude benefit from epilepsy surgery if the neurological deficits, semiology, and imaging implicate the abnormal hemisphere. Early spread to the normal hemisphere, the presence of encephaloclastic lesion modifying scalp EEG patterns, and secondary epileptogenesis may explain these findings. (Supported by CAPES (Brazil).) 1 Tiziana Granata, 1 Elena Freri, 2 Carlo Marras, 3 Alessandro Pincherle, 4 Marina Casazza, 1 Francesca Ragona, 2 Paolo Ferroli, 3 Flavio Villani, 3 Roberto Spreafico, and 2 Giovanni Broggi ( 1 Department of Child Neurology, Neurological Institute C.Besta, Milan, Italy ; 2 Department of Neurosurgery, Neurological Institute C.Besta, Milan, Italy ; 3 Department of Clinical Epileptology, Neurological Institute C.Besta, Italy ; and 4 Department of Clinical Neurophisiology, Neurological Institute C.Besta, Italy ) Rationale: We report on the outcome of 13 patients who underwent hemispherotomy, for the relief of drug resistant seizures, at the Neurological Institute of Milan, between 2000 and 2005. Methods: The 13 patients were affected by static or progressive encephalopathies, involving the right hemisphere in 7 cases, and the left in 6. Diagnosis was made by MRI and confirmed by pathological study. Motor, mental, and language evaluation have been performed before surgery and at least yearly during a follow‐up lasting 6–72 months (median 35.5 months). All the patient underwent peri‐insular hemispherotomy. Results: Four patients had malformation of cortical development (2 hemimegalencepahly, 2 cortical dysplasia), 3 patients suffered of acquired perinatal brain damage and 6 patients were affected by Rasmussen encephalitis (RE). Age at the first seizure was 1 month – 11 years (median 4.4 years), age at surgery was 6 months‐18 years (median 7.3 years), the mean delay between the onset of seizures and surgery being 3.3 years (range 4 months‐9 years). In all the patients motor deficit contralateral to the affected hemisphere, mental retardation or cognitive decline, of different severity, were present at the time of treatment. Post‐op course was complicated in 2 cases (chemical meningitis, subdural hematoma). There were no life‐threatening complications. Among the 10 patients followed‐up for at least 12 months, surgery was judged as successful in 8 cases. Six patients became seizure‐free (3/4 cases with RE, 2/2 cases with acquired brain damage, 1/2 cases with diffuse cortical dysplasia), whereas 2 children (1 with RE, 1 with cortical dysplasia) showed >50% reduction of seizures. Antiepileptic treatment was withdrawn in 3 patients, and simplified in 4. In all the 8 patients neurological condition improved: motor deficits remained unchanged in 3 patients and improved in 5; a significant cognitive and behavioural improvement was observed in 5 patients, no patient with left hemispherotomy experienced worsening in language. No improvement in seizure frequency, nor in motor and cognitive skills was, by contrast, obtained in the 2 children affected by hemimegalencephaly. Conclusions: Our data support the efficacy of peri‐insular hemispherotomy in unilateral encephalopathies. As already reported, the seizure outcome is related to the underlying pathology, with the best results in acquired brain damage and in RE, and the worst in hemimegalencephaly. Our data also confirm that control of seizures is often associated with improvement of motor and mental abilities when surgery is performed in childhood. 1 Jin Sook Lee, 1 Jee Yoon Park, 1 Jin Hwa Moon, 1 Hee Hwang, 1 Ki Joong Kim, 1 Yong Seung Hwang, 2 Seung Ki Kim, 2 Byung Kyu Cho, and 3 Ho Jin Park ( 1 Pediatrics, Seoul National University Hospital, Seoul, Republic of Korea ; 2 Pediatric Neurosurgery, Seoul National University Hospital, Seoul, Republic of Korea ; and 3 Pediatrics, Eulji University Hospital, Daejeon, Republic of Korea ) Rationale: Dysembryoplastic neuroepithelial tumor (DNT) is a benign brain tumor and often occurs in association with medically intractable partial seizures. The authors report our experience of DNT patients treated with epilepsy surgery. Methods: A retrospective review and analysis of medical records were performed on brain MRI findings, interictal and ictal EEG features pathological findings, other clinical characteristics, and surgical outcome of 14 patients with refractory epilepsy secondary to DNT. Results: All patients showed cortical or subcortical lesions with T1 low and T2 high signal intensity, and only one case showed focal small enhancing portion within the tumor. Other MRI findings included cystic formation (6), calcification (2) and hippocampal atrophy (1). Temporal lobe was involved in 9 cases (64.3%), while frontal, parietal or occipital lobes were involved in 2, 3, and 1 respectively. Mean age at seizure onset was 7.2 years (range 0.3 – 13.8 years). All patients presented with complex partial seizures with or without secondary generalization, however, one patient had previous history of infantile spasms. Ten of the 12 patients revealed abnormal findings on interictal EEG. Long‐term video‐EEG monitoring (VEM) was done in 5 patients. The epileptogenic lobe was correctly localized by VEM in 4 patients who had lesions in the temporal lobe. Intracranial recording was performed in one patient, in whom lesion was located in the parietal lobe but scalp VEM failed to localize ictal onset zone. Lesionectomy was done in 8 cases (57%) and temporal lobectomy with amygdalohippocampectomy was done in 6 cases (43%). Pathology revealed associated cortical dysplasia (CD) in 4 patients (28.6%), and one of them showed oligodendroglioma component as well as CD. Mean follow‐up period after surgery was 2.8 years (range 0.2 – 5.9 years). All had an Engel Class IA outcome (3 patients still have been on antiepileptic drugs) without recurrence of tumor. Second operation was performed in one patient with temporal lobe involvement. Postoperative morbidity (visual field defect and hemiparesis) was observed in 2 cases who had lesions in the temporal lobe (21.4%). Follow‐up MRI could not exclude the possibility of residual tumor in 3 cases, however, no further seizure was reported so far. Six of the 9 patients showed normal postoperative EEG. Conclusions: DNT most frequently involved the temporal lobe and was often associated with CD. Gross total removal is thought to ensure favorable prognosis in most cases regardless of type of surgery. However, careful resection is necessary to reduce morbidity especially in case of temporal lobe involvement. 1 Nicoletta Longoni, 1 Warren W. Boling, 3 Adriana Palade, 4 Angel Wabulya, 2 Frederick Andermann, 1 Priscah Mujuru, 4 Warf Benjamin, 6 Donald Gross, 5 Richard Bittar, 7 Gavin Fabinyi, and 8 Richard Byrne ( 1 Neurosurgery, West Virginia University, Morgantown, WV ; 2 Neurology and Neurosurgery, Montreal Neurological Institute, Montrea, QC, Canada ; 3 Neurology, West Virginia University, Morgantown, WV ; 4 Community Medicine, West Virginia University, Morgantown, WV ; 5 CURE Childrens Hospital of Uganda, Mbale, Uganda ; 6 Neurosurgery, Monash University, Melbourne, Victoria, Australia ; 7 Neurology, University of Alberta, Edmonton, AB, Canada ; 8 Neurosurgery, University of Melbourne, Melbourne, Victoria, Australia ; and 9 Neurosurgery, Rush University Medical Center, Chicago, IL ) Rationale: Since May 2005 an Epilepsy Monitoring Unit (EMU) has been in place at the CURE Children's Hospital of Uganda (CHU) in Mbale, Uganda, East Africa. Children are pre screened in regionalized clinics prior to evaluation at CHU comprehensive epilepsy program, the first such program in East Africa. The program is dedicated to the evaluation and treatment of pharmacoresistant temporal lobe epilepsy. Methods: Prescreening relies on identifying the stereotypical signs and symptoms of temporal lobe epilepsy. Treatment candidates must have intractable epilepsy, seizure frequency over 1 per month, and a history of two of the following: ‐ febrile convulsions ‐ typical aura ‐ typical semiology of stare and/or automatisms Patients less than 6 years of age or multiple extra‐temporal intracranial lesions were excluded. EEG and CTimaging are done at CHU. The data are transferred for analysis to WVU. Epilepsy surgery is performed at CHU. Results: 30 patients have been evaluated at CHU. Seizure frequency ranged between 30/day to 2/month with a mean of 2.2 seizures/day. 25 patients presented an aura. Essentially all patients had a history of malaria. 27 had febrile convulsions. To date, 19 patients have completed an evaluation in the comprehensive program. EEG recording has shown: Left temporal (5), right temporal (2), frontal (2), bitemporal (2), multifocal (3), generalized (4), and non‐epileptic (1). CT showed scar in 4 cases. Four patients had convincing temporal lobe epilepsy and underwent surgery at CHU. Conclusions: Epilepsy prevalence has been estimated to be 20 per 1000 in one region of Eastern Africa, 10 times the prevalence in developed countries making intractable epilepsy a significant public health problem in the region. Malaria was essentially universal in the population screened and we suspect cerebral malaria as an important cause of intractable epilepsy. One case of “non‐epileptic spell” was identified (5% in the population studied). (Supported by USAID, American Schools and Hospitals Abroad.) 1 Angel W. Hernandez, 1 Saleem I. Malik, and 1 David Donahue ( 1 Comprehensive Epilepsy Program, Cook Children's Medical Center, Fort Worth, TX ) Rationale: Patients undergoing brain surgeries are at risk of developing bacterial infections post‐operatively, but herpes simplex virus (HSV) reactivation is a rare complication. Few reports are available on patients that develop reactivation herpes virus encephalitis following epilepsy surgery. We report a case of HSV reactivation post epilepsy surgery to raise awareness of this risk and to consider treatment pre and post surgery on patient with a history of prior HSV infection. Methods: We report one case of reactivation of herpes simplex virus after epilepsy surgery on a 12 year old with pharmaco‐resistant localization‐related epilepsy of left temporo‐parietal onset. This patient had a history of negative cerebrospinal fluid (CSF) polymerase chain reaction (PCR) for HSV as an infant. Four days after resection of the epileptogenic focus the patient developed acute repetitive seizures, meningeal signs, confusion, lethargy and high fever. PCR in the CSF was positive for HSV. Results: The patient's seizures stopped 48 hours after starting treatment with Acyclovir. The confusion and lethargy resolved 48–72 hours after initiating treatment. There were no sequelae to the infection. The patient is now 1.5 years seizure free and off anti‐convulsants. Conclusions: Though early recognition and treatment of HSV encephalitis is essential, we believe that prophylactic treatment with acyclovir pre and post brain surgery is warranted in patients with prior HSV exposure to decrease the chances of this complication. 1 Manuel Marrufo, 1 Yong D. Park, 1 Ki Hyeong Lee, 2 Mark R. Lee, 2 Joseph R. Smith, and 2 Scott Y. Rahimi ( 1 Neurology, Medical College of Georgia, Augusta, GA ; and 2 Neurosurgery, Medical College of Georgia, Augusta, GA ) Rationale: Corpus callosotomy is a treatment option for epilepsy patients who are not resective surgery candidate. EEG as predictor or outcome for corpus callosotomy cases has been studied exclusively among adult population. Methods: Patient selection: A retrospective anlysis included all patients under 18 yrs who underwent corpus callosotomy from July 02 to April 06. Data collection: The demographic data, seizure characteristics and seizure outcome was obtained from the medical records, as well as direct phone interviews at least 6 month following the surgery. The EEG findings were reviewed by two authors (YDP and MM) blinded to the outcome. Seizure and EEG patterns classification: Seizure type was classified and recorded in to two main categories: Category A for drop attacks and tonic seizures; Category B for GTCS, atypical absence, adverse postures, CPS and myoclonus. Ictal EEG was classified into two different patterns: Type A for generalized desynchronization, spike and wave, and diffuse beta activity; and Type B for focal onset confined to one hemisphere or with secondary generalization. Outcome evaluation: Two seizure outcomes were defined by the reduction of seizure: Class I ≥ 90% reduction, and Class II < 90% or less. Statistical Methods: Chi‐Square test run with SPSS 13 package. Results: There were a total of 35 patients followed for 11mo (6–12mo). The average age at surgery and the average age of seizure onset were 7.4 years (range 1 – 17 yrs) and 2.36 yrs (range 0–10 yrs) respectively. Sex ratio was 20:15 (M:F). In terms of seizure characteristic, 24 out of 35 (68%) patients belonged to category A while 11/35 (32%) to category B. When correlated with outcome of surgery, patients with Category A seizures showed a tendency to do better than Category B (Category A 85% (12/14) vs. Category B 15% (2/14), p‐value 0.077). Ictal EEG pattern did not predict the outcome: Type A 47.6% (10/21) vs Type B 30.8% (4/13), p‐value 0.272. Other demographic or pathology characteristics were not different between the two outcome groups. Conclusions: Our data suggest that ictal EEG pattern does not predict outcome following corpus callosotomy among the pediatric population unlike the published data on adult. However, consistent with previous reports, “drop” seizures tended to respond better to this surgical procedure than other seizure types in our population. Careful selection of candidate regarding seizure types, not necessarily ictal EEG patterns, are recommended for corpus callosotomy. 1 Bruno Maton, 1 Prasanna Jayakar, 1 Trevor Resnick, 1 Glenn Morrison, 1 John Ragheb, 1 Catalina Dunoyer, 1 Pat Dean, and 1 Michael Duchowny ( 1 Comprehensive Epilepsy Center, Brain Institute, Miami Children's Hospital, Miami, FL ) Rationale: Temporal lobe epilepsy (TLE) that begins in early life is often a catastrophic disorder with pharmaco‐resistant seizures and secondary neurological deterioration. Few data are available regarding epilepsy surgery performed during infancy and early childhood and no prior study has focused on TLE. Methods: We analyzed the results of temporal resection, performed for epilepsy as the primary indication, between 1979 and 2003, in children less than age 5 years at time of surgery who had at least 2 years of follow‐up. Results: 20 children (14 males) were identified with a mean age at surgery of 26 months and a mean age at seizure onset of 12 months. Clinical presentation was age‐dependent. Typical psychomotor seizures (n = 4; mean age at surgery 37 months) may be followed by prominent motor changes (n = 7; 30 months) that can be isolated (n = 3; 23 months). Epileptic spasms were noted in 6 patients and were frequently associated with partial seizures. The interictal EEG was lateralizing in 15 patients and ictal EEG was lateralizing in 18 patients. Brain MRI had good predictive value in 16 patients; ictal SPECT was concordant in 4/8 patients. Invasive EEG was employed in 6 cases. At mean follow‐up of 5.5 years, 65% of the children were seizure‐free and 15% had > 90% seizure reduction. Morbidity included infection and hydrocephalus in one case and stroke‐related hemiparesis in two cases. Cortical dysplasia was identified in 8 children, tumors in 8 including 2 DNET, 2 ganglioglioma, and 4 malignant tumors. Hippocampal sclerosis was present in 4 cases as dual pathology. Conclusions: Surgery for TLE can be performed during early life with similar results as in older children and in adults. 2 Ann‐Christine Duhaime, 3 Barbara C. Jobst, and 1 Richard P. Morse ( 1 Pediatrics and Neurology, Dartmouth‐Hitchcock Medical Center, Lebanon, NH ; 2 Pediatric Neurosurgery, Dartmouth‐Hitchcock Medical Center, Lebanon, NH ; and 3 Neurology, Dartmouth‐Hitchcock Medical Center, Lebanon, NH ) Rationale: Supplementary Motor Area (SMA) seizures are relatively common in children but there are small numbers reported in terms of surgical approach and outcome. Despite the fact that SMA epilepsy typically begins in childhood and is often refractory to AED treatment, surgical approaches to its management most often are delayed until adulthood. We report on the clinical features, evaluative test results (SPECT, invasive EEG) and surgical outcome in a series of 5 non‐lesional children and adolescents, 3 of whom have had resective epilepsy surgery. Methods: Five children were characterized by means of seizure description, V‐EEG, SISCOM, fMRI, and invasive intracranial recording with brain mapping, and seizures were localized to the SMA. Results: Three of the five patients have had resective surgery, with mixed outcomes. None were initially seizure‐free: one has had a few seizures due to non‐compliance but otherwise has gone months at a time without seizures, one required a second resection in the primary motor cortex and is now seizure‐free, and one has had a >50% reduction in seizures. Of the non‐operated patients, one declined surgery and one has been moderately well‐controlled with AED therapy. All patients were non‐lesional on MRI imaging; pathology showed cortical dysplasia in one and non‐specific changes in the other two patients. The patient who required a second operation has a mild L hemiparesis, but had no deficits after the initial resection limited to the SMA, and the other two patients have no neurologic deficits. Conclusions: SMA seizures are not uncommon in children, with characteristic clinical features and seizure semiology like those described in adults. The seizures can be approached surgically with acceptable results and without deficits. Identification and characterization of SMA in children may allow for earlier successful surgical intervention, especially in cases involving focal cortical dysplasia as the underlying cause of the epilepsy. Use of adjunctive methods to localize, including SISCOM and invasive monitoring with stimulation and mapping, is essential for non‐lesional cases. A common confounder appears to be the rapid propagation to the SMA from adjacent cortical areas (primary motor) or simultaneous onset, making surgery more challenging. 1 Joo Hee Seo, 2 Su Jeong You, 2 Hoon Chul Kang, 1 Heung Dong Kim, 3 Tae‐Sung Ko, 4 Yong Soon Hwang, 5 Dong Suk Kim, 6 Jung‐Kyo Lee, and 4 Sang Keun Park ( 1 Department of Pediatrics, Severance Hospital, Handicapped Children's Research Institute, Brain Research Institute, Yonsei University College of Medicine, Seoul, Korea ; 2 Department of Pediatrics, Epilepsy Center, Inje University College of Medicine, Sang‐gye Paik Hospital, Seoul, Korea ; 3 Department of Pediatrics, Asan Medical Center, Ulsan University College of Medicine, Seoul, Korea ; 4 Department of Neurosurgery, Epilepsy Center, Inje University College of Medicine, Sang‐gye Paik Hospital, Seoul, Korea ; 5 Department of Neurosurgery, Severance Hospital, Handicapped Children's Research Institute, Brain Research Institute, Yonsei University College of Medicine, Seoul, Korea ; and 6 Department of Neurosurgery, Ulsan University College of Medicine, Asan Medical Center, Seoul, Korea ) Rationale: We sought to evaluate the long‐term efficacy of vagus nerve stimulation (VNS) in refractory pediatric epilepsy. Methods: Twenty eight patients could be followed over 12 months in three tertiary care referral epilepsy centers. Data including seizure frequencies, quality of life measures and adverse events had been prospectively filed up through 5 years. Results: VNS resulted in a >50% reduction in seizure frequency in 53.6% (15/28) of children with 28.6% (8/28) patients achieving a >90% reduction. When compared a seizure reduction rate according to seizure types (generalized vs paritial seizure), etiologies (symptomatic vs cryptogenic), and seizure duration before VNS implantation, we could not find any significant difference of it. However, a seizure reduction rate had a tendency to be in inverse proportion to seizure duration. In addition, of 11 patients who had partial seizure disorders, 7 patients (63.6%) obtained a reduction of seizure frequency with >50% and then again, of 17 patients who had generalized seizure disorders, 8 patients (47.1%) did. Quality of life was also measured to be enhanced as follows, memory in 50.0% (8/16) patients, mood in 62.5% (10/16) patients, behavior in 68.8% (11/16) patients, alertness in 68.8% (11/16) patients, achievement in 37.5% (6/16) and verbal skill in 43.8% (7/16) patients. Adverse events included hoarseness in 7 patients, dyspnea at sleep in 2 patients, wound infection in 1 patient but were transient and successfully managed by careful follow‐up and an adjustment of parameters. Conclusions: Our data show that VNS roles as one of alternative therapy for refractory pediatric epilepsy without any significant adverse events. 1 Nicolas Specchio, 1 Lucia Fusco, 1 Giuseppe Gobbi, 1 Federico Vigevano, and 1 Tiziana Granata ( 1 Division of Neuroscience, Bambino Gesù Children's Hospital, Rome, Rome, Italy; Division of Neuroscience, Bambino Gesù Children's Hospital, Rome, Rome, Italy; Division of Neuropediatrics, Ospedale Maggiore Cà Pizzardi, Bologna, Bologna, Italy; Division of Neuroscience, Bambino Gesù Children's Hospital, Roma, Roma, Italy; and Division of Child Neurology, Istituto Nazionale Neurologico C. Besta, Milano, Milano, Italy ) Rationale: Rasmussen encephalitis (RE) is an acquired neurological disease characterized by intractable partial seizures and symptoms of progressive dysfunction of one cerebral hemisphere. Laboratory findings suggested the possibility of a immune‐mediate component in the pathogenesis of the disease implying. However, surgical treatment still remains the preferred treatment. We report our experience on the surgical treatment of 10 RE patients collected in the context of an Italian collaborative study. Methods: The series included 10 patients with RE followed in the centers of Bambino Gesù Children's Hopsital (Roma), Neurological Institute “C. Besta” (Milano) and Ospedale Maggiore (Bologna). Pre‐ and post‐operative evaluations included: EEG recordings with video monitoring, MRI, assessment of motor function and cognitive evaluation based on clinical observation and standardized tests. The surgical procedures was based on functional disconnection of the affected hemisphere (hemispherotomy), performed according to vertex or peri‐insular approach. Results: The series includes 10 patients (6 M,4 F), aged between 10 and 23 years (16.3 ± 3.4) observed between 1993 and 2000. The age at onset ranged between 3 and 11 years (6.5 ± 2.8); surgery was performed 8 months to 9 years from disease onset (4.1 ± 2.7). Post‐surgical follow‐up ranged between 5 and 10 years (6.9 ± 1.5). The disease onset was marked by partial seizures, epilepsia partialis continua and unilateral limb paresis followed by status epilepticus. Neurologic deficits appeared in all patients within 24 months. EEG recordings consistently showed slow and epileptic abnormalities over the affected hemisphere. At the time of surgery patients were treated with three or more AED. All, but one, patients became seizure‐free after surgery, AE treatment has been stopped in 7 patients, and reduced in 3. Patients still treated are given a mean of 1.5 ± 1 drugs. The need for neurological evaluations and EEG monitoring reduced from monthly to yearly. Conclusions: After surgery seizures ceased in all patients, thus demonstrating the successfulness of the operative procedures. It is noteworthy that postural control improved in all patients and regained unassisted gait. Mental abilities, as evaluated by standardized scales, improved in all cases. None of them performed anymore habitual monthly EEG and neurological evaluations. Surgical treatment of RE appeared in our series as a gold standard because of its effectiveness in seizure control, dramatic improvement in cognitive evaluations and reduction in the general healthcare assistance. 1 Marino M. Bianchin, 1 Vera C. Terra‐Bustamante, 1 Tonicarlo R. Velasco, 2 Neife H. Deghaide, 1 Hélio R. Machado, 2 Eduardo A. Donadi, and 1 Américo C. Sakamoto ( 1 Department of Neurology, Psychiatry, and Medical Psychology, Ribeirão Preto School of Medicine, University of São Paulo, Ribeirão Preto, São Paulo, Brazil ; and 2 Division of Clinical Immunology, Ribeirão Preto School of Medicine, University of São Paulo, Ribeirão Preto, São Paulo, Brazil ) Rationale: Rasmussen encephalitis (RE) is a rare disorder of unknown etiology. The disease has been attributed to possible viral infections or autoimmune mechanisms. HLA is the landmark of autoimmune disorders and specific alleles have been associated with different neurological diseases. Here we report the results of HLA Class‐Ia study in Brazilian RE patients. Methods: After approval by the Ethic Committee of our institution and after informed consent was obtained from the patients, DNA samples were obtained from 13 Brazilian patients with pathologically confirmed RE. For the study of allele prevalence, data were compared with those for ethnically matched control population. The Fischer test was used to determine possible differences. The level of significance was established at p < 0.05. Results: No specific HLA Class‐Ia was found to be strongly associated with RE in our patients. HLA class Ia genes in Rasmussen Encephalitis HLA A Rasmussen Control p Positive Negative Positive Negative A1 2 24 229 1961 0.89 A2 5 21 264 1926 0.42 A3 3 23 226 1964 0.84 A23 2 24 133 2057 0.73 A24 3 23 227 1963 0.96 A25 2 24 41 2149 0.15 A26 3 23 67 2123 0.16 A30 2 24 124 2066 0.98 A32 1 25 71 2119 0.86 A33 2 24 79 2111 0.56 A69 1 25 5 2185 0.06 HLA B Positive Negative Positive Negative p B7 2 24 153 2161 0.83 B8 2 24 128 2186 0.96 B13 1 25 37 2277 0.90 B14 3 23 117 2197 0.30 B15 1 25 179 2135 0.71 B18 1 25 120 2194 0.76 B35 6 20 274 2040 0.15 B38 2 24 42 2272 0.14 B39 2 24 66 2248 0.38 B44 1 25 269 2045 0.35 B45 1 25 45 2269 0.49 B49 1 25 71 2243 0.82 B51 1 21 200 2114 0.61 B52 2 24 36 2278 0.09 Conclusions: HLA‐Class‐Ia molecules are expressed in most somatic cells including certain neurons, making some neural population potentially susceptible to autoimmune attacks under triggering stimuli from cytotoxic T‐cells. Additionally, HLA Class‐Ia molecules might have other roles in the CNS in terms of neuroplasticity and neuroprotection, or might be involved in normal brain development. In fact, HLA Class‐Ia molecules have been associated with some neurological disorders like schizophrenia, autism, or dyslexia. RE is an autoimmune disorder observed mostly during the years of brain development and associated with neuroplastic alterations that leads to an intriguing form of epilepsia partialis continua. However, in spite of multiple potential roles for HLA molecules in RE, in this group of patients no specific HLA Class‐Ia was found to be strongly associated with RE. We conclude that none of the specific HLA Class‐Ia alleles studied here is particularly involved in RE. However, larger samples involving multicentric studies are necessary to confirm our findings. (Supported by FAPESP.) 1 Mary L. Zupanc, 2 Ellianne Dos Santos Rubio, 1 Rhonda Roell‐Werner, 1 Michael J. Schwabe, 3 Wade Mueller, 3 Sean Lew, 1 Kathy Eggener, and 1 Maria Chico ( 1 Department of Neurology, Medical College of Wisconsin, Milwaukee, WI ; 2 Department of Pediatric Neurology, Erasmus Medical College, Sophia Children's Hospital, Rotterdam, Netherlands ; and 3 Department of Neurosurgery, Medical College of Wisconsin, Milwaukee, WI ) Rationale: To study seizure outcome and determine the correlation between seizure outcome and quality of life measurements in children after epilepsy surgery. Methods: A consecutive, retrospective analysis of seizure outcome and quality of life assessment was conducted in 83 pediatric patients who underwent epilepsy surgery for medically refractory epilepsy at Children's Hospital of Wisconsin between 2002–2005. Data was collected via a chart review and examination of post‐operative patients. Data collection included seizure outcome measurements as described in the Engel classification system. Quality of life assessment was performed by sending quality of life questionnaires to caregivers of our patients (QOLCE and QOLIE‐AD‐48). Results: The seizure outcomes for all surgery types were Class I‐68.7%; Class II‐12%; Class III‐19.3%. If corpus callosotomies were excluded, Class I outcome was 72%. Seizure free outcome was highest in temporal lobectomies (84.2%) followed by hemispherectomies (76.2%), focal resections (61.2%), and finally corpus callosotomies (25%). Temporal lobectomies were more common among older patients. Seizure free outcome was 50% in infants; 67.9% in young children, and 78.4% in adolescents. Cortical dysplasia was predictive of a less favorable seizure free outcome. 62.7% of the QOLCE surveys and 45.8% of the QOLIE‐AD‐48 surveys were returned. The quality of life measures for the surveys paralleled seizure outcome. The measures of physical activity, cognition, social activity, general health, and overall quality of life were significantly better in children with a seizure free outcome than children who were not seizure free. Conclusions: Epilepsy surgery in medically refractory epilepsy in children significantly improves quality of life as well as seizure control. 1 Warren W. Boling ( 1 Neurosurgery, West Virginia University, Morgantown, WV ) Rationale: The author describes a novel keyhole approach to selective amygdalohippocampectomy (SAH). The surgical technique has proven to be successful in the treatment of medically intractable mesial temporal lobe epilepsy (MTLE). Image guidance is required in the keyhole approach to orient the cranial opening, verify cortical landmarks for positioning the corticectomy, and assess the extent of hippocampal disconnection. The restricted surgical exposure has not limited the ability to perform SAH with favorable results on reducing the seizure tendency. Methods: All patients with one year or more follow‐up after keyhole SAH at West Virginia University (WVU) were included in this report. Forty‐eight surgeries for temporal lobe epilepsy have been performed since 2002 by the author with over one year follow‐up. Ten had keyhole SAH performed and are the subjects reported here in. Results of surgery in keyhole SAH subjects were contrasted with individuals undergoing SAH at WVU using standard craniotomy approach. All individuals since 2002 who underwent standard SAH surgery by the author for epilepsy and with over 1 year of postoperative follow‐up were included in the evaluation. Results: Ten subjects had keyhole approach for SAH performed. Seven (70%) had an excellent outcome from surgery becoming seizure free (Class I). One subject had over 90% seizure frequency reduction (Class III), and 2 had less than 90% reduction in seizure frequency (Class IV). Three subjects had no stereoelectroencephalography (SEEG) performed prior to keyhole SAH (all seizure free), and the remainder had SEEG performed for further confirmation of MTLE. There were no surgical complications encountered. Twenty‐One surgeries were performed for SAH using a standard craniotomy approach. 81% were seizure free after surgery. Excellent outcomes compared favorably between keyhole and non‐keyhole approaches (p = 0.54). Hospital stay after surgery was determined for the subjects who underwent keyhole SAH at WVU and an age matched control of patients who had SAH performed via a non‐keyhole (pterional craniotomy) approach at WVU. In keyhole SAH subjects the mean length of hospital stay was 3 days (range 2 – 4 days). Non‐keyhole SAH patients stayed in the hospital a mean of 4 days (range 2 – 6 days) after surgery representing a trend to shorter hospital stays after keyhole SAH (p = 0.11). Conclusions: The keyhole approach to SAH is a strictly image guided procedure. The corticectomy is positioned at the upper bank of T2 anterior to the precentral suclus in the dominant hemisphere and in front of the central sulcus in the non‐dominant hemisphere. These cortical landmarks can only be confirmed with image guidance. The results of surgery on the seizure tendency depend on an accurate diagnosis of MTLE, and compare favorably with non‐keyhole SAH and corticoamygdalohippocampectomy for temporal lobe epilepsy. The keyhole SAH is an effective approach in patients undergoing surgery for MTLE. 1 Richard C. Burgess ( 1 Neurology, Cleveland Clinic, Cleveland, OH ) Rationale: The Cleveland Clinic has a 25 year history as a tertiary referral center for patients with epilepsy. Rigorously kept records of seizure/epilepsy classification and neurophysiologic results are saved electronically. With the advent of a hospital‐wide electronic medical record, picture archiving and communication system, digtal video recordings, and other scattered electronic repositories, data about our patients is entered electronically in many areas, but is not available in an integrated fashion when needed to generate reports to referring doctors or to carry out sophisticated research queries. We sought to put all of this information at the neurologist's fingertips for clinical and research purposes, by developing a system that provides access via standard web browser. Methods: The database contains both text and graphics. Data is imported from other servers, as well as maintained by pointers. For access we employed the Faces framework deployed on a J2EE application server. Because of its deployment in a very busy clinical center, careful attention was paid to real‐time performance issues, e.g. by employing asynchronous Javascript and XML to handle page display simultaneous with data retrieval. In addition to the carefully classified neurophysiologic and epilepsy information, a range of graphical and tabular information has been organized for incorporation into the database. This includes pictorial information, formerly available ad hoc from a variety of sources, such as: grid placement maps, depth placement maps, stimulation maps, ictal and nterictal discharge maps, MRI & PET segments, ictal SPECT, results of volume reconstructions, coregistrations showing grids, intra‐operative photos, video clips, evoked potentials, pathology slides, etc. Results: The application is a complex mix of Java files, JSP files, JSP fragments, Javascript files, configuration files, etc. — over 2000 files and approximately14 megabytes. In addition to ongoing data input the database has been back‐filled with data on more than 65,000 patients from our previous codified textual database. Data output is provided in standardized formats that are familiar to users, easy to work with, and aesthetically pleasing, like HTML and PDF. It is possible to create and view the inpatient evaluation reports in their entirety. New classes of reports (such as for patient management conferences) can be derived from the fundamental information by abstracting and using a built‐in templating engine. Conclusions: Learning from expeience in previous large database development projects, we focused our efforts during this software design on a) obtaining rapid results, quickly visible to and useable by the users, and b) employing modern web‐based tools to speed development and performance. This philosophy has yielded a clinical workflow aid for communication with many disparate sources and for manipulation of data in order to flexibly present results. It has also been designed to include powerful data aggregating tools for research purposes. 1 Hélène Catenoix, 2 Marc Guénot, 1 François Mauguière, 2 Marc Sindou, and 1 Jean Isnard ( 1 Functional Neurology and Epileptology Department, Neurologic Hospital, Lyon, Rhône, France ; and 2 Functional Neurosurgery Department, Neurologic Hospital, Lyon, Rhône, France ) Rationale: Intracranial EEG recordings in presurgical epilepsy assessment are done in our unit using stereotactically implanted depth electrodes. Five years ago, we explored use of these electrodes to produce RFTC lesions inside the epileptogenic area. We report the results in 42 patients. Methods: 42 consecutive patients explored by video‐SEEG recordings were enrolled in this study between June 2001 and February 2006. RFTCs were produced by inducing a Joule effect between electrode contacts where discharge onsets had been recorded. Four to 31 lesions (mean, 12.5) were performed per patient. The epileptic focus (EF) was located in the frontal (8 patients), occipital (7), insular (4), parietal (3) and temporal (20) lobe. Several etiologies were found: cortical dysplasia (CD) (21), hippocampal sclerosis (4), periventricular heteropia (PH) (3), posttraumatic focal atrophy (1) and cryptogenic epilepsy (13). The results were evaluated on the decrease in the number of seizures: good outcome (GO, >70%) and using a self‐administered questionnaire, each patient estimated the presence (a) or absence (b) of improvement in their quality of life (QL). Results: The GO rate 1 month after the intervention (57%) decreased to 35% at the 6‐month evaluation and then stabilized. The QL evaluation showed comparable mean results (a = 38%) but rose to 66% on the subgroup of patients who were inoperable after SEEG exploration. The GO rate was positively correlated with: 1) anterior location of the EF: 75, 37 and 33% of GO for insular, frontal and temporal lobe epilepsy, respectively, with no change for parietal and occipital epilepsy. 2) the etiology, with the best results for symptomatic epilepsy (37 and 33% GO in CD and PH, respectively) vs 23% GO in cryptogenic forms. No general or neurological complication occurred during the procedures other than two transient deficits (paresthetic sensations in the mouth and mild apraxia of the hand). 20 unsatisfied patients had a secondary classical intervention (19 now in Engel class 1 and 1 in class 2) Conclusions: This study reports the results at 5 years of RFTC treatment given to 42 epileptic patients during SEEG exploration. This procedure is: 1) sure: complications are rare, minor and always reversible despite treatment often targeting high‐value functional cortical and/or inaccessible areas 2) effective: benefit in terms of seizure control concerns 1/3+ patients with GO maintained long term from the 6th month after treatment 3) offers improvement in quality of life, especially for the 2/3 of patients considered inoperable after invasive exploration 4) no surgical burden on prognosis: 19/20 patients operated after RFTC failure are in Engel class 1 These results changed our SEEG exploration strategy by combining diagnosis and therapeutic potential until now unknown with this technique. (Supported by HCL.) 1 Stephan Chabardes, 2 Lorella Minotti, 1 Dominique Hoffmann, 1 Eric Seigneuret, 2 Philippe Kahane, and 1 Alim L. Benabid ( 1 Neurosurgery Department, CHU Michallon, Grenoble, France ; and 2 Neurology Department, CHU Michallon, Grenoble, France ) Rationale: Temporal lobe epilepsy (TLE) is the most common form of medically intractable partial epilepsy in adults, and surgery (anterior temporal lobectomy or a selective amygdalo hippocampectomy) is effective in the majority of patients. We have developed since 1998 a new surgical technic to treat non lesional TLE which consists in disconnecting the whole temporal lobe from the insula and surrounding structures. The aim of this study was to assess the safety and results of temporal surgical disconnection of patients suffering from non lesional TLE. Methods: Since 1998, 62 patients suffering from TLE, or extra TLE have been treated by surgical disconnection of the epileptogenic zone. Among this group, 45 patients (mean age: 35 y ± 10) suffered from TLE and underwent temporal disconnection (left and right disconnection in 20 and 25 patients respectively). There was 16 males and 29 females, 12 were left handed and 33 were right handed. All patients underwent a comprehensive pre‐surgical evaluation with MRI, chronic video‐EEG recording and neuropsychological assessment. 16 patients (35%) underwent pre‐surgical evaluation with depth electrodes (SEEG). All patients underwent a post operative MRI, neuropsychological testing, repeated scalp EEG and visual field examination. Temporal disconnection was performed under neuronavigation and consisted in disconnecting the lateral temporal neocortex, the whole hippocampus, para‐hippocampus and the amygdala. Veins and arteries were spared in order to avoid post‐operative temporal lobe ischemia and subsequent temporal lobe swelling. Results: At two years follow‐up, 84.4% of patients were Engel's I (58% were Ia). Four patients were class II and 3 class IV. Post‐operative clinical morbidity was as follow: persistent mild hemiparesia (n = 1) or mild facial paresia (n = 1), definitive quadranopsia (n = 23), hemianopsia (n = 1). In dominant temporal lobe, verbal memory was worsened in 69%. Radiological post‐operative morbidity was as follow: 2 thalamic or pallidal limited ischemia without clinical consequences, 2 temporal horn cystic dilatation (one requested reintervention without clinical consequences). At 6 months and one year follow up, scalp EEG showed temporal spikes respectively in 27% and 19% of non cured patients compared to 8.5% and 6%of patients seizure free. Conclusions: Temporal disconnection was effective in 85% of patients (Engel's Class I). Post –operative mild contro‐lateral motor deficit was found in about 4%, quadranopsia in 51% and verbal memory deficit in dominant hemisphere in about 69%. Presence of spikes in post‐operative scalp EEG was less frequent in patients seizure free. Comparative studies are required to evaluate temporal disconnection as an alternative to lobectomy in non lesional TLE. 1 Anita Datta, 1 Matt B. Wheatley, 1 Tom Snyder, 1 S.N. Ahmed, 1 Donald Gross, and 1 David B. Sinclair ( 1 Comprehensive Epilepsy Program, University of Alberta, Edmonton, AB, Canada ) Rationale: The objective of the study was to review our experience and compare the results of the selective amygdalohippocampectomy in children versus adults with refractory temporal lobe epilepsy. Methods: A retrospective case series was used in the setting of a tertiary care hospital which provides care to both children and adults. All patients underwent a selective amygdalohippocampectomy performed by a single neurosurgeon and had at least one year of follow up. Patient charts were reviewed for pertinent clinical information and the data was compared between children and adults. Outcome measures included seizure frequency, neuropsychological assessment, EEG, neuro‐imaging and anti‐epileptic drugs pre and post‐operatively. Results: Twenty three patients, 9 children and 14 adults were studied. Age at surgery varied from 6 to 58 years. Follow‐up ranged from 1 to 4 years. Surgical outcome was variable between the two groups. Amongst the younger cohort (6–18 years of age), 5 patients had good outcome (Engel Class I or II), two patients had a poor outcome (Engel class III) and two patients had refractory seizures (Engel Class IV) that required returning to the operating room for a standard anterior temporal lobectomy. This differed from the adults, who all had a good outcome. Ten patients were seizure‐free (Engel Class I) and the remainder had rare seizures (Engel Class II). Non‐MTS pathology was more common in children. Thirteen of 14 (93%) had MTS in the adult population, while only 4 of 9 children (44%) had MTS. The pathology at surgery in this group included mesial temporal sclerosis (MTS), brain tumors, focal encephalitis and normal pathology. Conclusions: Selective amygdalohippocampectomy can lead to favorable seizure and neuropsychological outcome in adults with refractory temporal lobe epilepsy. However, preliminary results show less favorable results in children. The difference is probably related to the different pathology between the two groups. Adults most commonly have MTS, whereas children more commonly have non‐MTS temporal lobe pathology. Clinicians should be aware of this difference when considering surgical management in children with refractory temporal lobe epilepsy. 1 Gena R. Ghearing, 1 Drahomira Sencakova, 2 Barbara Schauble, 1 Gregory D. Cascino, 1 Elson L. So, 3 Clifford R. Jack Jr., and 1 Gregory A. Worrell ( 1 Neurology, Mayo Clinic, Rochester, MN ; 2 Section of EEG and Epileptology, Department of Neurology, University Hospital of Berne, Inselspital, Berne, Switzerland ; and 3 Radiology, Mayo Clinic, Rochester, MN ) Rationale: Unilateral hippocampal atrophy is associated with an excellent surgical outcome after temporal lobectomy. The management of patients with bitemporal disease is more controversial and many undergo intracranial EEG (iEEG) with bitemporal depth electrodes. The purpose of this study is to evaluate the usefulness of intracranial monitoring for predicting surgical outcomes in patients with both asymmetric and symmetric hippocampal volumes. Methods: Volumetric hippocampal analysis was performed on the MRIs of all consecutive intractable temporal lobe epilepsy patients between the ages of 16 and 45 years who underwent bitemporal depth electrode recording at our institution over a seven year period. Symmetric hippocampal volumes were defined by a difference between the right and left hippocampus between the range – 0.1 and + 0.3 cm3. The hippocampal volumes were compared to age matched controls, and significant atrophy defined as greater than two standard deviation difference from control. The localization of ictal and interictal discharges on scalp and iEEG were reviewed, and surgical outcome determined for each patient. Results: Of the 27 patients, 14 had asymmetric and 13 had symmetric hippocampal volumes. The surgical outcome was available for 22 patients (four did not have surgery and one was lost to follow up) with an average follow up of 51 months. Ten of 12 surgical patients with asymmetric atrophy (83%) had an Engel class 1 outcome. Nine of these patients with a favorable outcome had significant atrophy on the resected side. The predominant side of seizure onset during iEEG was concordant with the smaller hippocampus in all patients who had significant atrophy. Four out of ten surgical patients with symmetric atrophy (40%) had a favorable outcome. The degree of atrophy did not predict surgical outcome. Two of the patients with a favorable outcome (50%) and three of the patients with an unfavorable outcome (50%) had significant bilateral atrophy. Of the four patients who had a favorable surgical outcome, three had a predominance of seizures arising from the resected side during iEEG, and three of the patients who had unfavorable outcomes also had a predominance of seizures from the resected side. Conclusions: In this limited retrospective study, patients with asymmetric hippocampal atrophy had a high probability (83%) of seizure free outcome. The smaller hippocampus was concordant with the predominant side of iEEG seizure onset in all patients with significant asymmetric atrophy. Conversely, only 40% of patients with symmetric hippocampal atrophy had favorable surgical outcomes, and iEEG did not prove helpful for predicting outcome. 1 Jorge A. Gonzalez, 2 Dileep Nair, and 1 William Bingaman ( 1 Neurological Surgery, Cleveland Clinic, Cleveland, OH ; 2 Neurology, Cleveland Clinic, Cleveland, OH ; and 3 University Heights, OH ) Rationale: Treatment of patients who fail epilepsy surgery is problematic. Selected patients may be candidates for further surgery, potentially leading to a significant decrease in seizures frequency and severity. The purpose of this manuscript is to present our long‐term outcome series of highly investigated patients who failed resective epilepsy surgery and subsequently underwent re‐operative resective procedures. Additionally, our intention is to identify possible factors associated with unsatisfactory seizure outcome, improving selection criteria for re‐operation in failed epilepsy surgery. Methods: We performed a retrospective consecutive analysis of patients who underwent re‐operative procedures due to medically intractable epilepsy at our institution from 1990 to 2001. Seventy patients underwent re‐operative epilepsy surgery with 57 patients having a minimum follow‐up of 2 years. We assessed the relationship between seizure outcome and categorical variables using Chi‐square and Fisher's exact tests, and the relationship between outcome and continuous variables using Wilcoxon rank‐sum test. For all analyses, statistical significance was set at a p value of 0.05. Results: Of the 57 patients, there were 29 males and 28 females. Age of seizure onset ranged from 3 months to 39 years (mean 10.7 ± 10.3 years, median 7 years). The mean age at re‐operation was 24.7 ± 12 years (range 4–50 years). Follow‐up ranged from 24 to 228 months (mean: 128 months, mode: 132 months). Seizure outcome were classified according to Engel's classification and divided in two groups: Favorable outcome (Engel's classes I and II) and non‐favorable outcome (Engel's classes III and IV). Fifty two percent of patients had favorable outcome. Patients with tumors as initial pathology had better outcome in comparison with malformations of cortical development (MCD) and mesial temporal sclerosis (MTS) (p < 0.05). Conclusions: Re‐operation in selected patients failing epilepsy resective surgery should be considered as approximately 50% of patients may have benefit. Patients with MCD and MTS are less likely to improve after re‐operation. 1 Robert E. Gross, 1 Edward K. Sung, 1 David LaBorde, 1 Michele Johnson, and 1 Klaus Mewes ( 1 Department of Neurosurgery, Emory University School of Medicine, Atlanta, GA ) Rationale: Frameless image‐guidance techniques for implantation of depth electrodes for intracranial monitoring are increasingly used, but few reports have assessed the accuracy and precision of these techniques. Methods: Frameless image‐guidance using the Stealth® (Medtronic) neuronavigational hardware and Cranial® software was used exclusively for the implantation of depth electrodes since 1998. Electrodes were implanted through burr holes using an orthogonal (trans‐temporal) approach (n = 32 patients), typically into the amygdala and 2 – 3 electrodes in the hippocampus or parahippocampal gyrus on each side, or occipital approach (n = 3 patients). We analyzed the implantation of all electrodes for which the intra‐operative planning trajectories (saved on the pre‐operative MRI scan by the Stealth) and post‐operative MR imaging were available (n = 155 electrodes: 150 orthogonal, 5 occipital). The pre‐ and post‐operative image sets were registered using the navigational software, and a vector was constructed from the “intended” target to the “actual” location of the electrode tip (taking into account imaging artifact around the electrode). Results: The discrepancy between the “intended” and “actual” location of the orthogonal electrode tips was 5.6 ± 2.9 mm. There was no significant difference between accuracy of electrodes implanted on the first vs. second side as might be expected from brain shift, nor on the left (n = 80) vs. right (n = 70) side. The direction of error could be characterized as “too deep” for 93 electrodes as opposed to “too shallow” for 50 electrodes probably due to difficulty in securing the electrode precisely during placement. Five electrodes implanted via an occipital approach had an error of 7.7 ± 8.0 mm, reflecting lower accuracy and greater variability likely due to the longer passage. No patients implanted via an orthogonal approach required reoperation in contrast to 1 of 3 patients implanted occipitally. No complications definitively related to post‐operative MR imaging were detected over the 150 electrodes and 32 patients, although 2 patients recently implanted experienced transient neurologic deficits which in one patient may have been infectious in etiology but in the other was uncertain. Conclusions: Frameless image‐guidance implantation of depth electrodes via an orthogonal approach is an effective technique with sufficient accuracy for intracranial monitoring. 1 Marc Guenot, 2 Jean Isnard, 2 Philippe Ryvlin, 2 Helene Catenoix, 2 Francois Mauguiere, and 1 Marc Sindou ( 1 Department of Functional Neurosurgery, Hop. P. Wertheimer, Hospices Civils de Lyon, Lyon, France ; and 2 Department of Functional Neurology, Hop. P. Wertheimer, Hospices Civils de Lyon, Lyon, France ) Rationale: Except in case of Mesial Temporal Lobe Epilepsy Syndrom (MTLE), and despite their more recent refinement, the noninvasive techniques still often fail to clearly localize the epileptogenic zone in a large number of patients presenting with drug‐resistant temporal lobe epilepsy. Different techniques of intracranial recordings are currently available. StereoElectroEncephaloGraphy (SEEG), which consists in the stereotactic orthogonal implantation of depth electrodes (5 to 15, 11 on average) into the brain, is the only one allowing to record the mesial as well as the lateral aspects of the temporal lobe, including the depth of the sulci. The object of this paper is to display the curent usefulness of SEEG in terms of surgical strategy related to temporal resection, in 100 consecutive procedures. Methods: 100 patients, suffering from drug‐resistant epilepsy arising from a temporal onset, underwent a SEEG procedure between 2000 and 2005. For each of these patients, the non invasive investigation, which included clinical and neuropsychological examination, long‐term video‐EEG recording, functional imaging, and MRI scan, were not in favor of a MTLE syndrom, and were not congruent enough to clearly identify the epileptogenic focus. Consequently, the sites of implantation of the electrodes were chosen in order to precise: either the side of the onset of seizures, or their uni or multilobar feature, or a possible strictly neocortical or insular onset of the seizures, and also, using direct electrode stimulation, the proximity of speech area. Results: Complications occurred in 3 patients (1 superficial infection, and 2 breakages of electrodes). ‐ SEEG was helpful in most (91%) of the 100 patients to confirm, or cancel, surgical indication, and to adjust the extent of the resection: ‐ In 33% of cases, SEEG‐recordings led to an indication of a strictly extra‐limbic lobectomy, i.e. a tailored (temporo‐basal or lateral) neocortectomy sparing the temporo‐mesial structures. ‐ In 20% of cases, SEEG led to a limited limbic resection, which included the amygdala, as well as the pole, while sparing most of the (non‐atrophic) hippocampus. ‐ In 25% of cases, SEEG allowed to plan an enlarged temporal resection, despite the fact that it showed a wide, or even multiple, epileptogenic zone. ‐ In 13% af cases, SEEG allowed to definitly exclude the patients from surgery. ‐ In 9% of cases, SEEG confirmed the indication of a conventional limbic resection (anterior temporal lobectomy). Conclusions: SEEG proved to be a safe and a very useful method for complex cases candidate for temporal lobe epilepsy surgery. It should be considered as soon as the non invasive investigation do not display a clear MTLE syndrome. 1 Hirochika Imoto, 1 Masami Fujii, 2 Jouji Uchiyama, 1 Nobuhiro Tanaka, 3 Kimihiko Nakano, 1 Sadahiro Nomura, 1 Hirosuke Fujisawa, 2 Takashi Saito, and 1 Michiyasu Suzuki ( 1 Neurosurgery, Yamaguchi University School of Medicine, Minamikogushi, Ube, Yamaguchi, Japan ; 2 Applied Medical Engineering Science, Graduate School of Medicine Yamaguchi University, Ube, Yamaguchi, Japan ; and 3 Mechanical Engineering, Faculty of Engineering Yamaguchi University, Ube, Yamaguchi, Japan ) Rationale: Local brain cooling using an implantable cooling system is an attractive prophylactic technique for seizures since it has a potential to be an alternative to a surgical resection of the epileptogenic foci. In this report, we applied the newly devised local brain cooling system to intractable epilepsy in humans. Methods: We utilized a Peltier chip as the cooling device and therefore applied this cooling device to two patients with medically intractable epilepsy. During surgery, cooling was performed for two minutes in the human cortex where the epileptiform discharges (EDs) were recorded and therefore it had to be resected. Electrocorticograms and the temperature just beneath the cooling site were recorded before and during cooling. Changes of EDs between before and during cooling were quantitatively analyzed. Results: The first case was a 12‐year‐old girl with temporal lobe epilepsy, while the second case was a 2‐year‐old boy with parietal lobe epilepsy caused by tuberous sclerosis. In both cases, EDs diminished during the cooling process when the temperature of the brain surface reached less than 25°C. Conclusions: This is the first report demonstrating the effectiveness of the cooling device using a Peltier chip for the treatment of human epilepsy. Owing to recent advances in the precision machinery industry, an implantable local cooling system in humans is therefore expected to become a reality in the near future. 1 Floor E. Jansen, 1 Alexander C. Van Huffelen, 1 Ale Algra, and 1 Onno Van Nieuwenhuizen ( 1 Neurology, Pediatric Neurology, Clinical Neurophysiology and Clinical Epidemiology Julius Center for Health Sciences and Primary Care, University Medical Center, Utrecht, Netherlands ) Rationale: Cortical tubers are associated with epilepsy and intellectual disability in the majority of patients with tuberous sclerosis. Seizures are often drug resistant. Uncertainties exist concerning the selection of patients suitable for epilepsy surgery. We performed a systematic review of the available literature to assess the overall outcome of epilepsy surgery and to identify factors related to good seizure outcome. Methods: A comprehensive literature search of Medline, Embase, and bibliographies of reviews and book chapters published from 1960 until May 2006 was performed. Studies were included if they at least reported quantitative seizure outcome, and a description of the type of surgery. Seizure outcome was analysed both as seizure freedom and good seizure outcome (seizure reduction of >90%) versus poor seizure outcome. Results: Twentyfive studies, including 177 patients, were found to be eligible. Seizure freedom was achieved in 101 patients (57%). Seizure frequency was reduced by > 90% in another 32 patients (18%). Moderate or severe intellectual disability (IQ < 70) (RR 1.8; 95% CI 1.2–2.8), the presence of tonic seizures (RR 1.7; 95% CI 1.2–2.4) multifocal abnormality on SPECT examination (RR 7.0; 95% CI 1.1–43) and corpus callosotomy (corpus callosotomy versus resective surgery RR 2.5; 95% CI 2.1–3.0) predicted an unfavourable outcome. Conclusions: A good outcome of surgery (seizures reduction of > 90% or complete cure) was achieved in 75% of patients. Resective epilepsy surgery should be considered in all drug resistant TSC patient, as a relation between multiple seizure types with early onset, multiple cortical tubers and multifocal or generalized epileptogenicity and a poor outcome is not supported. (Supported by the Epilepsy Fund of the Netherlands (grant number 02–13, F.E.J.).) 1 Andres M. Kanner, 2 Richard Byrne, 1 Ada V. Chicharro, 1 Marlis Frey, and 1 Michael C. Smith ( 1 Neurological Sciences, Rush University Medical Center, Chicago, IL ; and 2 Neurosurgery, Rush University Medical Center, Chicago, IL ) Rationale: Following an antero‐temporal lobectomy, one third to one half of patients who become free of any disabling seizures continue experiencing auras. The cause of temporal lobe epilepsy (TLE), duration of the seizure disorder and the presence of generalized tonic‐clonic (GTC) seizures have been identified as predictors of post‐surgical outcome folloiwng a temporal lobectomy. A comorbid psychiatric history has also been associated with a worse post‐surgical seizure outcome in these patients. Predictors of persistence of post‐surgical auras in patients free of disabling seizures are yet to be identified. Methods: 97 consecutive patients, 58 men and 39 women with a mean age of 30.6 ± 10.7 years who underwent an antero‐temporal lobectomy at the Rush Epilepsy Center were included in the study. The cause of TLE was: mesial temporal sclerosis in 60 (62%), lesional TLE in 18 (18.5%) and idiopathic TLE in 19 (19.5%). The mean duration of the seizure disorder was 19.7 ± 8.8 years. All patients had undergone a psychiatric evaluation as part of their presurgical evaluation (not out of suspicion of psychiatric disorder). All patients had to have a minimal post‐surgical follow‐up period of 2 years before being included in the study (mean post‐surgical follow‐up: 6.9 ± 3.0 years). The folloiwng variables were entered into a logistic regression model to identify the predictors of achieving a seizure‐free state without auras: cause of TLE, duration of seizure disorder, occurrence of GTC seizures and life‐time history of depression. Results: Among the 97 patients, 37 (38%) were free of any disabling seizures and auras since the time of surgery (Class IA of Engel). An additional 43 patients (44.3%) were free of disabling seizures but had auras (Class I‐B). A life time prevalence of depression of 48.5% (n = 47) was identified among the 97 patients. The absence of a life‐time history of depression was the only variable that predicted the achievemnet of a seizure‐free outcome without auras (p < 0.0001). This model correctly classified 76% the data. A lifetime history of depression continued to be a predictor, but to a lesser degree when the post‐surgical outcomes consisted of freedom of disabling seizures with or without auras (p = 0.001) and freedom of disabling seizures in the last two years (p = 0.06). On the other hand the cause of TLE became a predictor in the setting of these two outcomes (p = 0.04 and p = 0.009, respectively). Conclusions: A presurgical history of depression appears to predict the peristence of auras after an antero‐temporal lobectomy in patients that become free of diabling seizures. The pathogenic mechanisms that mediate thsi phenomenon are yet to be identified. 1 Keith W. MacDougall, 1 Jorge G. Burneo, 1 Richard S. McLachlan, 1 Andrew G. Parrent, and 1 David A. Steven ( 1 Department of Clinical Neurological Sciences, University of Western Ontario, London, ON, Canada ) Rationale: In many patients being considered for epilepsy surgery, a standard pre‐surgical evaluation consisting of a scalp electroencephalogram (EEG), magnetic resonance imaging (MRI), functional neuro‐imaging and a neuropsychological assessment is sufficient to identify the epileptogenic zone. When this evaluation cannot identify the epileptogenic focus, invasive intracranial electrodes are often required. Although there is considerable data about the outcome of patients undergoing epilepsy surgery, most of these investigations examine all patients (with and without invasive recordings) together. This purpose of this study was to specifically examine patients who underwent invasive EEG and to determine how many of these patients eventually underwent resective surgery. In those who had surgery, we wished to examine the eventual outcome. Methods: We retrospectively reviewed charts for all patients admitted to the London health Sciences Centre Epilepsy Monitoring Unit between January 2000 and July 2004 for intracranial electrode monitoring. Surgical outcomes were reported using the Engel classification. We used a multivariate analysis to determine which factors if any were associated with successful surgery. Results: There were 103 subdural electrode implantations performed by our two epilepsy surgeons in 100 patients. Of these, 78 patients went on to have resective surgery. Temporal lobectomy was performed in 39 (50%) patients. One‐year follow‐up was available on 66 of the 78 patients (85%). Thirty patients (45%) realized an Engel I outcome at 1 year. An additional 20 (30%) had their seizure frequency significantly reduced. Predictors of an Engel I outcome included younger age (p < 0.0001) and having a temporal lobectomy (adjusted OR = 4.17; CI = 1.12 – 12.8). Factors that did not significantly affect outcome included the duration, area, and location of electrode coverage, the presence of a pre‐operative lesion or the final pathology. Conclusions: Excellent results from eventual resective surgery can be achieved in the subset of patients needing invasive recordings. Younger patients with temporal lobe epilepsy seem to have the highest likelihood of seizure freedom. The need for bilateral or multilobar electrodes did not negatively affect prognosis. 1 Taketoshi Maehara, 1 Tadash Nariai, 1 Toshiya Momose, 1 Chihiro Hosoda, 2 Kenji Ishii, 2 Kiich Ishiwata, and 1 Kikuo Ohno ( 1 Neurosurgery, Tokyo Medical and Dental Universuty, Tokyo, Japan ; and 2 Positron Medical Cente, Tokyo Metropolitan Institute of Gerontology, Tokyo, Japan ) Rationale: With the advent of intraoperative navigation systems, structural lesions disclosed by MRI can now be removed more easily and accurately. But in cases receiving epilepsy surgery, the seizures can only be stopped by removing the surrounding epileptogenic areas together with the lesions themselves. To accomplish this, we need to functionally detect peri‐lesional epileptogenic areas during surgery. Our group recently introduced a method of F18‐fluorodeoxyglucose (FDG) and C11‐flumazenil (FMZ) PET co‐registration as a reference for the detection of epileptogenic areas during image‐guided epilepsy surgery for patients with organic lesions. In this study we investigated the clinical role of multimodal navigation epilepsy surgery using FDG‐PET and FMZ‐PET. Methods: We studied a series of 10 patients who suffered from intractable epilepsy and were found to have organic lesions on MRI. All of the subjects underwent FDG‐PET and FMZ‐PET studies before their operations. Surgical specimens demonstrated cavernous angiomas in 2 patients and benign tumors in 8 (3 low‐grade astrocytomas, 2 DNTs, 2 gangliogliomas, and 1 monomorphous angiocentric glioma). The lesions were located in the mesial temporal lobe (mT) in 6 patients and in the extra‐mesial‐temporal lobe (extra‐mT) in 4. Our first steps were to examine the relationships between the organic lesions and hypometabolic areas on FDG‐PET, and then to compare the FDG‐hypometabolic areas and low‐uptake areas on FMZ‐PET. Next, we performed intraoperative electrocorticography (ECoG) in and around the hypometabolic areas to examine the epileptogenicity of the hypometabolic areas on FDG‐PET and the low‐uptake areas on FMZ‐PET. Results: 1) The hypometabolic areas on FDG‐PET were much wider than the low‐uptake areas on FMZ‐PET in 8 cases, and slightly wider in 3 cases. 2) Spikes on intraoperative ECoGs were recorded in the hypometabolic areas in all 6 of the mT cases. 3) The hypometabolic areas were much wider than the ECoG‐spiking areas in 2 of the 4 extra‐mT patients, and slightly wider in the other 2. ECoG‐spiking areas were wider than the low‐uptake areas on FMZ. 4) Four of the 6 mT patients underwent anterior temporal lobectomy and the other 2 underwent lesionectomy along with anterior temporal resection. The extra‐mT patients underwent lesionectomy along with resection of any ECoG‐spiking areas which exhibited hypometabolism on FDG‐PET and low‐uptake on FMZ‐PET. 5) All 10 patients became seizure‐free during follow‐up periods ranging from 6 to 24 months. Conclusions: The use of multi‐modal navigation surgery by FDG‐PET and FMZ‐PET in combination with intraoperative ECoG may become an important method for focus resection in epileptic patients with organic lesions, especially in patients with extra‐mesial temporal lobe epilepsy. 1 Michiharu Morino, 1 Yumiko Urano, 1 Masaki Yoshimura, and 1 Kenji Ohata ( 1 Neurosurgery, Osaka City University Graduate School of Medicine, Osaka, Japan ) Rationale: Although selective amygdalohippocampectomy was developed in epilepsy surgery to spare unaffected brain tissue minimizing the cognitive consequences of temporal lobe surgery, several previous studies have shown that it was equivocal whether selective amygdalohippocampectomy for intractable epilepsy could lead to a better memory outcome or not. The aim of the present study was to investigate the effects of selective surgery to short‐term memory outcome in patients of intractable mesial temporal lobe epilepsy who underwent transsylvian selective amygdalohippocampampectomy (TSA). Methods: The 51 patients received comprehensive neuropsychological testing of verbal and nonverbal memory, attension, and delayed recall before and 1 months after unilateral TSA. Results: Significant improvements in verbal memory and delayed recall were found particularly for the right resected group. For right TSA, the outcome of verbal memory was not related to hippocampal pathology, however, the risk for a nonverbal memory decline was increased when patients had non hippocampal sclerosis. After left TSA, there were no significant decline and it indicated that all aspects of memory were preserved after surgery although the risk for a verbal memory decline was increased of patients without hippocampal sclerosis. Memory before and 1 month after TSA Right TSA p values Left TSA p values Preop. 1MP Preop. 1MP Memory Verbal 74.9 (14.5) 87.1 (18.5) p < 0.001 77.8 (18.5) 73.8 (16.1) NS Non verbal 81.0 (17.5) 87.2 (16.1) NS 90.8 (17.6) 92.1 (21.1) NS General 73.3 (14.8) 84.6 (18.1) p < 0.001 78.9 (18.6) 76.2 (17.3) NS Delayed Recall 72.0 (15.9) 78.8 (20.0) p = 0.0015 75.2 (19.8) 72.4 (17.6) NS Attension 83.7 (16.6) 86.8 (16.7) NS 91.2 (16.4) 90.8 (20.8) NS Cells provide the mean standardized scores with the standard deviations in parentheses. NS: not significant (p>0.05). MP: month postoperative. Conclusions: The results indicate that left TSA can lead to preserve the memory function, particularly right TSA can lead to a significant improvements in memory functions. 1 William Murphy, 1 Terry Myles, 1 Walter Hader, 1 Neelan Pillay, 1 Lorie Hamiwka, 1 Elaine Wirrell, 1 Paolo Federico, 1 Nathalie Jette, and 1 Samuel Wiebe ( 1 Clinical Neurosciences, University of Calgary, Calgary, AB, Canada ) Rationale: Reports on the diagnostic accuracy of structural MRI in detecting putative lesions in patients undergoing epilepsy surgery vary substantially, and are probably influenced by MRI protocols, types of lesions, and expertise. Using histopathology as the gold standard, we assess the overall sensitivity and specificity of structural MRI in an unselected cohort of patients undergoing epilepsy surgery, and provide post‐tests probabilities for negative and positive studies. Methods: In a retrospective cohort of consecutive cases undergoing resective epilepsy surgery, we identified those who had histopathological tissue examination,ie., the gold standard. MRIs protocols contained a combination of general and epilepsy‐specific sequences, and experts and non‐experts performed MRI interpretation. The presence or absence of a putative lesion on structural MRI, using epilepsy specific protocols, was compared with the presence or absence of a histopathological lesion. We obtained sensitivity (true positive rate), specificity (true negative rate), positive and negative likelihood ratios, and post‐test probabilities using a range of plausible pre‐test probabilities. We also obtained 95% confidence intervals around each estimate. Factors influencing the diagnostic accuracy were assessed when the sample size allowed. Results: Of 135 consecutive patients evaluated for epilepsy surgery, 117 had MRI and histopathological analyses. Histopathological abnormalities occurred in 104 patients (prevalence of 89%), of whom the MRI found a lesion in 93, yielding a false negative rate of 11%, and a sensitivity of 89% (95% CI = 83%, 95%). Of 13 patients with normal histopathology, the MRI showed a lesion in eight, with a false positive rate of 62%, and a specificity of only 38% (95% CI = 12%, 64%). The likelihood ratios for a positive and a negative MRI were 1.44 (low) and 0.3 (medium‐high), respectively. Indicating that in this analysis, the MRI seems to be more useful when negative, because its false positive rate was too high. When the chance of finding a lesion is 50%, a positive MRI increases the probability to only 59%, and a negative MRI decreases is to 22%. Conclusions: In unselected patients undergoing epilepsy surgery Structural MRI has a sensitivity of almost 90% and a specificity of only 38%. The causes for the high false positive rate of MRI in our cohort are explored. Reference Von OJ, Urbach H, Jungbluth S, et al. Standard magnetic resonance imaging is inadequate for patients with refractory focal epilepsy. J Neurol Neurosurg Psychiatry 2002;73:643–647. 1 Yu‐tze Ng, 2 Norman C. Wang, 2 Steve C. Chung, 3 Erin C. Prenger, 1 Harold L. Rekate, 2 Iman Feiz‐Irfan, and 1 John F. Kerrigan ( 1 Pediatric Neuroscience, Barrow Neurological Institute/St. Joseph's Hospital, Phoenix, AZ ; 2 Neurology, Barrow Neurological Institute, Phoenix, AZ ; and 3 Neuroradiology, Barrow Neurological Institute, Phoenix, AZ ) Rationale: Hypothalamic hamartomas (HHs) are rare developmental abnormalities of the inferior hypothalamus that often cause refractory, symptomatic, mixed epilepsy. Multiple surgical approaches have been described. We present a series of 39 patients where a transcortical, transventricular, endoscopic resection was performed. Methods: Forty‐two patients with refractory epilepsy underwent surgical resection of HH between October 2003 and April 2005 at our institution, i.e. all patients with more than 12 month follow‐up. All patients underwent endoscopic resection and were prospectively studied. Neurosurgery was performed with the endoscope held by an articulated pneumatic arm and tracked with a frameless stereotactic neuronavigation system. Data collection and follow‐up was performed by personal interview. Three patients were excluded because they underwent combined surgical approaches (subfrontal and transcallosal). Results: The average age of the patients was 15.8 years (range 8 months to 55 years). There were 24 males. All 39 patients had very frequent seizures, usually multiple types. The average number of each patient's total daily seizures was 18.2 (range 0.1–192) and an average of 2.2 seizure types each. 33 (85%) patients had trademark gelastic seizures at the time of surgery. The patients were on an average of 2.0 antiepileptic drugs (AEDs). 25 patients had either cognitive and/or behavioral impairment. Postoperative MRI demonstrated 100% resection and disconnection of the HH from the hypothalamus in 12 patients. There were no mortalities. Significant morbidity was limited to one patient with residual hemiparesis. Two patients were lost to follow‐up. At latest follow‐up (average 20.1 months (range 13 to 28 months)), 18/37 (48.6%) patients were seizure‐free, 26/37 (70.3%) patients had >90% reduction in their seizures and another 8/37 (21.6%) of patients had 50–90% seizure reduction. Fourteen patients were off all AEDs. Twenty‐five (67.6%) patients reported either improved cognition and/or behavior. The percentage of disconnection trended toward seizure freedom (p = 0.069). The percentage of resection did not significantly correlate with seizure freedom (p = 0.11). In an endoscopic resection group of 33 patients, the average length of post‐surgical hospital stay was 5.8 days (SD = 3.1) as compared to 18 patients who underwent transcallosal resection and stayed for 8.7 days (SD = 4.7). This overall difference i.e. shorter hospital stay amongst the endoscopically resected patients was statistically significant (p < 0.03). Conclusions: Endoscopic resection of HH is safe and effective for the treatment of seizures. Its efficacy appears to be comparable to transcallosal resection of HH, however with significantly shorter patient recovery time post‐operatively. 1 Tomonori Ono, 1 Keisuke Toda, 1 Shiro Baba, 1 Hiroshi Baba, and 2 Kenji Ono ( 1 Neurosurgery, National Nagasaki Medical Center, Omura, Nagasaki, Japan ; and 2 Yokoo Hospital, Isahaya, Nagasaki, Japan ) Rationale: Surgery for bilateral temporal lobe epilepsy (Bi‐TLE), herein defined as seizures arising independently from both temporal lobes, has been demonstrated to have poor efficacy. However, lateralized MRI abnormality and a preponderance of interictal spikes on EEG have been identified as factors for good surgical outcome. Ictal EEG findings are also a decisive factor for determining the resection side. Predominant ictal onset side (PIOS) is considered to be a direct measure of laterality of ictogenisity, however, this has yet to be quantitative analyzed. In this study, we reviewed our surgical series of Bi‐TLE patients and determined the correlation between quantitatively assessed PIOS and postoperative outcome. Methods: We analyzed data from 23 patients with Bi‐TLE who had undergone anterior temporal lobectomy. Patients' ages at the time of the operation ranged from 14 to 55 years (mean 31.2 years). Diagnoses of Bi‐TLE were confirmed using subdural electrode recordings, and the side of resection was determined by PIOS. Postoperative seizure outcome was assessed using Engel's outcome classification between 9 and 161 months (mean 80 months) after surgery. Percentages of ictal onset side calculated on the basis of all recorded seizures and other clinical characteristics were correlated with postoperative seizure outcome. Results: Regarding postoperative seizure outcome, 9 patients (39.1%) were class I, 8 patients (34.8%) were class II, and 6 patients (26.1%) were class III. The following items were found to be statistically significant factors for good surgical outcome: 1) no history of severe encephalitis (p = 0.03), 2) higher percentage of PIOS (class I vs. class II‐III, 87.6 ± 3.1% vs. 73.7 ± 3.3%, p = 0.009), and 3) lateralized lesion (p = 0.01), or absence of bilateral abnormalities on MRI (p = 0.02). Furthermore, multivariate analyses showed that factors 1) and 2) were independent predictors of good outcome. Conclusions: Surgery for Bi‐TLE results in a greater reduction of seizures if preoperative evaluations show good predictive factors. In particular, when the percentage of PIOS is greater than 85%, a more favorable outcome can be obtained. 1 Andre Palmini, 1 Eliseu Paglioli, 1 Jaderson Costa, 1 Caroline Torres, 1 Jose Victor Martinez, 1 Eduardo Portela, 1 Helio F. Silva‐Filho, 1 Eduardo Paglioli, 1 Mirna Portuguez, 1 Jose‐Rubiao Hoeffel‐Filho, 1 Ligia Coutinho, 1 François Dubeau, and 1 Frederick Andermann ( 1 Department of Neurology and Neurosurgery, Pontificia Universidade Católica do RS, Porto Alegre, RS, Brazil; and Department of Neurology and Neurosurgery, McGill University, Montreal, QC, Canada ) Rationale: Incomplete resections are not uncommon in surgery for epilepsy due to FCD, however, the dynamics of long‐term surgical outcome following these resections have not been investigated in detail. We report on the variable patterns of seizure recurrence following incomplete resections for FCD and relate these to the histopathologic type. Methods: Forty‐two children and adult patients with FCD identified through MRI and/or confirmed on histopathology were followed for a mean of 54 months. In 25, a complete resection of the dysplastic tissue was not possible in a first operation. They all underwent preoperative evaluation and surgery was guided by acute ECoG and cortical electrical stimulation. Residual lesion was identified either at the end of operation or at post‐op MRI. The degree of seizure control was evaluated periodically, and classified according to Engel's outcome scale. FCD lesions were classified histopathologically in types I, IIA, and IIB. Results: Three recurrence patterns were identified: (i) in 6 patients, seizures recurred within 7 days of operation, with higher frequency and intensity than in the preoperative period. Three needed ICU admission for status epilepticus; (ii) 9 presented recurrent seizures during the first post‐op month and continued seizing with similar frequency as before surgery over the years or until reoperation. Finally (iii) 10 patients had a post‐op seizure free interval between 3 months and 2 years, following which seizures recurred. Histopathologic types IIA and IIB were similarly represented in all 3 recurrence patterns. Conclusions: Lasting seizure freedom was incompatible with incomplete lesion resections in FCD types II A and B. Interference with inhibitory networks associated with dysplastic lesions may explain the patterns of running up of seizure frequency and intensity, particularly early post‐resection severe deterioration leading to status epilepticus. (Supported by FAPERGS.) 1 Tarek A. Radwan, 1 David W. Roberts, 1 Vijay M. Thadani, 1 Peter D. Williamson, 1 Richard P. Morse, 1 Ann‐Christine Duhaime, and 1 Barbara C. Jobst ( 1 Neurosurgery & Neurology, Dartmouth‐Hitchcock Medical Center, Lebanon, NH ) Rationale: Interhemispheric electrodes (IHE's) are increasingly used in epilepsy surgery, but the use of IHE's has been associated with heightened concern about risk of bridging vein disruption & haemorrhage. We hypothesize that the use of IHE is safe and indispensable in localizing epilepsy originating from the mesial surface of the brain. Methods: We performed a retrospective review of complication rate, type of resection, EEG findings and surgical outcome of all patients who had undergone interhemispheric intracranial electrode monitoring. The interhemispheric location was anterior in 22 (69%), and posterior in 10 (31%). Results: 21 patients had IHE implanted. 17 had interhemispheric grids (curvilinear‐doublesided Adtech 2 × 8 or 3 × 8 grids), four had 1 × 8 or 1 × 4 strip electrodes placed interhemispherically. One patient had two interhemispheric 3 × 8 grids implanted. The interhemispheric location used was anterior in 22 (69%), and posterior in 10 (31%). Complications as a result of IHE's implantation consisted of temporary leg weakness in two patients which resolved over several days. No hemorrhagic complications or other complication related to electrode implantation were observed. Of all the patients implanted with IHE's, 20 (95%) underwent resective surgery, with 12 (57%) undergoing mesial resection. Out of those 12 patients, 7 (58%) had normal MRI, while 5 (41%) had a lesion identified on MRI. In the seven patients with normal MRI, localization and resection was only possible due corticography obtained from IHE. In the remaining patients with abnormal MRI IHE determined extent of resection. Four patients had not clearly defined developmental abnormalities on MRI and in one patient with a tumor IHE clearly localized seizure onset zone. Of the 9 patients that did not localize mesially 8 (89%) underwent resection after localizing to other cortical areas. In those patients IHE were helpful in excluding mesial seizure onset. One patient (11%) did not localize and did not undergo any resection. Outcome by Engel Class for all patients that had IHE's implanted and underwent resection was: Class I‐ 13 (Ia‐12/Ib‐1) (62%); Class II‐ 3 (14%); Class III‐ 2 (10%); Class IV‐ 3 (14%) Conclusions: Interhemispheric intracranial monitoring electrodes are a safe and effective tool in epilepsy surgery. 1 Fortino Salazar, 2 Imad Najm, 3 Prakash Kotagal, 1 Ann Warbel, and 1 William Bingaman ( 1 Neurosurgery/Epilepsy Surgery, Cleveland Clinic Foundation, Cleveland, OH ; 2 Neurology/Adult Epilepsy, Cleveland Clinic Foundation, Cleveland, OH ; and 3 Pediatric Neurology/Epilepsy, Cleveland Clinic Foundation, Cleveland, OH ) Rationale: Subdural grids (SDG) have been increasingly used in the evaluation of some patients with pharmacoresistant epilepsy. In a small number of patients the placement of SDG is not followed by a surgical resection. There is a paucity of data about these patients and the causes behind the lack of success in their invasive evaluation. The purpose of this study is to retrospectively review the data on patients who were evaluated at our center with SDG. Methods: Adult and pediatric patients who underwent the placement of SDG, followed by prolonged video‐EEG monitoring but did not have resective surgery were identified from the Cleveland Clinic database and their medical records were reviewed. A total of 233 patients underwent SDG placement between January 1997 and June 2005. The placement of SDG and prolonged invasive recordings did not lead to resective surgery in 29 (12%) patients who were included in this review. Mean age was 25 years (7 were < 18 years old). There were 15 males and 14 females. Average duration of seizures was 15 years. The MRI was interpreted as normal in 12 patients, other MRI findings included malformation of cortical development, schinzencephaly and focal brain atrophy. All patients had PET studies and when indicated a Wada test. The number of electrodes used varied between 56 and 160. Average monitoring duration was 8 days (between 4 and 19). Results: Reasons for NOT pursuing surgical treatment were: a) Identification of multiregional or non‐localizable epileptogenic foci (12 pts), b) presence of epileptogenic foci in eloquent cortex (8 pts), c) lack of adequate coverage of the suspected ictal onset zone (5 pts), or d) the lack of seizure recording after SDG (2 pts). In two patients the evaluation was terminated because of surgical complications. Following the SDG evaluation, 6 patients had VNS implanted and 5 others are considering this option. Three patients are considering another invasive evaluation, one had a resective surgery at a later time, and 7 patients are only considering medical treatment. Conclusions: Approximately 12% of patients were not suitable candidates to undergo surgical resection. The two most common reasons were involvement of eloquent cortex and evidence of diffuse epileptogenic zone. It is important to refine or develop better ways to identify this group of patients prior to implantation, such as the use of fMRI, MEG and sophisticated EEG to avoid costly and risky procedures. 1 Kostas N. Fountas, and 1 Joseph R. Smith ( 1 Department of Neurosurgery, Medical College of Georgia, Macon, GA ) Rationale: Implantation of subdural strip and grid electrodes represents a common methodology in the invasive evaluation of patients with medically refractory epilepsy. Although in the vast majority of cases their implantation is safe, the occurrence of implantation‐associated complications can be occasionally troublesome. The purpose of our current communication is to identify the most commonly occurred complications and calculate their frequency in our institution. Methods: In a retrospective study, 185 patients (104 males and 81 females) undergoing subdural strip/grid electrode implantation during a 20‐year period, were examined. Their age ranged between 16 and 48 years, while their mean age was 23.6. AdTech (AdTech, Racine,WI,USA) strip and grid electrodes were used in all our cases. The anatomical location of implantation were:26 implants in orbito‐frontal area, 41 in mesial frontal,14 in mesial occipital, 13 in antero‐basal area, 44 in interhemispheric areas while 180 in convexity areas (4 temporo‐polar, 83 parietal, 17 occipital and 76 temporal). The monitoring period ranged from 2–25 days (mean 10.8 and median 7 days). Results: The most commonly occurred complication in our series was the development of post‐operative epidural hematoma in 3/185 (1.6%) patients, requiring emergent surgical evacuation. In 2/185 (1.1%) cases transient aphasia occurred, which was spontaneously resolved. A surgical subdural hematoma developed in 2/185 (1.1%) while in another 2/185 (1.1%) an infection occurred. Interestingly, in 2 patients (1.1%) significant brain edema was post‐operatively evident; one of these patients was conservatively managed with no further consequences while the other one succumbed despite aggressive management of his intracranial hypertension. Finally, one patient developed post‐operatively stiff lung syndrome secondary to aspiration and eventually expired. Another interesting observation in our series was the recording of non‐habitual seizures in 5/185(2.7%) patients. Conclusions: Subdural electrode complications although rare can be troublesome and occasionally disastrous. Knowledge, early identification and prompt management of potential complications can minimize the risks associated with the surgical implantation of subdural electrodes. 1,2 William O. Tatum IV, 2 Kumar Anthony, 2 Leanne Heriaud, 1,2 Selim R. Benbadis, 2 Adele Haber, and 1,3 Fernando L. Vale ( 1 Neurology, University of South Florida, Tampa, FL ; 2 Neurology, Tampa General Hospital, Tampa, FL ; and 3 Neurosurgery, University of South Florida, Tampa, FL ) Rationale: Epilepsy Surgery is a standard of care in the treatment of intractable epilepsy. Intracranial electrodes can help identify surgical candidates when discordant information is present from the extracranial evaluation. We analyzed our experience with intracranial electrodes over 6 years. Methods: Two‐hundred and fifty‐eight resective epilepsy surgeries were performed from 2000–2006, with a mean follow‐up of 2.6 years (range 2–65 months). Twenty‐nine females and 21 males with a mean age of 29.4 years were implanted using an electrode array based upon the results of a standardized, non‐invasive, pre‐surgical evaluation. All patients were implanted at a single tertiary care epilepsy center between 2000–2006 by one neurosurgeon. Discordant information following a non‐invasive evaluation including scalp‐based video‐EEG monitoring was obtained in each case. Resective surgical procedures, implanted array, and outcome were assessed. Outcome was reported as seizure‐free, seizure improved (>50% of baseline), and no improvement, and was based upon information obtained at the time of last contact. Results: Fifty epilepsy patients (19.4%) were implanted with intracranial electrodes. Twenty‐four patients (48%) had bitemporal placement, 15/50 (30%) were unilateral, and 11/50 were bilateral‐extratemporal (22%). Thirty‐seven (74%) underwent resective epilepsy surgery, and 13/50 (26%) were rejected as surgical candidates. Seizure freedom or improvement was noted in 9/14 (64.3%) implanted with unilateral electrodes, 14/24 (58.3%) with bitemporal electrodes, and 4/10 (40.0%) with bilateral‐extratemporal electrodes. Resections were temporal in 27/37 (73.0%), and extratemporal in 10/37 (27%) including frontal in 6/37 (12.5%), multi‐lobar in 2/37 (5.4%), parietal in one (2.7%), and occipital in one. At last follow‐up, 21/27 (81.4%) of the temporal resections were either seizure free (33.3%) or seizure improved (48.1%). Only 5/27 (13.5%) failed temporal surgery, compared with 3/8 (37.5%) extratemporal resections. MRI demonstrated an abnormality in 19/50 (38%) with neuronal migrational disorders accounting for the most common cause. The presence of a lesion did not have a more favorable outcome (8/19; 42.9%) than those with cryptogenic causes (18/29; 62.1%) (p = 0.24, Fisher's exact test). Conclusions: Most epilepsy patients requiring intracranial electrodes will be candidates for resective surgery. Bitemporal electrode placement for lateralization, and temporal resections were the most frequent surgeries performed. Patients implanted only with bitemporal electrodes, were not more likely than those with a lateralized array to benefit from resective surgery. Those with lesional epilepsy had no more favorable outcome than those with a cryptogenic etiology in our series. 1 Samuel Wiebe, 1 Walter Hader, 1 Neelan Pillay, 1 Terry Myles, 1 Lorie Hamiwka, 1 Elaine Wirrell, 1 Paolo Federico, 1 Nathalie Jette, and 1 William Murphy ( 1 Clinical Neurosciences, University of Calgary, Calgary, AB, Canada ) Rationale: There is a notion that patients whose epilepsy is associated with a putative lesion have better surgical outcomes than those without such lesions. However, cases are often selected and other prognostic often remain unexplored. We explore seizure outcomes in unselected cases of lesional versus non‐lesional epilepsy surgery, as well as the variables that impact outcomes in the Calgary Comprehensive Epilepsy Program. Methods: A retrospective cohort study of consecutive, unselected cases assessed post‐surgical seizure freedom at the time of last follow‐up in patients whose MRI showed and in those in whom it did not show putative lesions. We used survival analyses for the main comparison and single and multivariate analyses to explore factors that may influence outcome, such as age, duration of epilepsy, etiology, type of surgery and location of surgery. Results: Of 113 consecutive patients, 91(81%) had putative MRI lesions. An equal proportion (70%) of lesional and non‐lesional cases had temporal lobe surgery. Patients without lesions were significantly more likely to have a family history of epilepsy (p = 0.03) and they had a trend towards a younger age at surgery. Two years after surgery 57% of lesional and 36% of non‐lesional cases were seizure free (p = 0.07). As compared with temporal lesional cases, extratemporal lesional cases were younger at presentation (20 vs 38 years) and at surgery (17 vs 35 years) (p < 0.01), and their seizure free rate was lower (not significant). Preoperative deficits were more frequent in non‐lesional extratemporal than in temporal cases. Of patients with frontal resections, most were non‐lesional (p = 0.02). Conclusions: In unselected patients undergoing epilepsy surgery those without MRI lesions have overall poorer seizure outcomes, are younger (especially extratemporal cases), more often have a family history of epilepsy, and more often have frontal lobe epilepsy. 1 Charles M. Zaroff, 1 Deepak Madhavan, 2 Alexis Arzimanoglou, 2 Florence Renaldo, 1 Josiane LaJoie, 1 Howard L. Weiner, 3 Eva Andermann, 4 Frederick Andermann, 4 Francois Dubeau, 4 Andre Olivier, 4 Alexei Yankovsky, 5 David N. Franz, 5 Jennifer Leonard, 6 Mary Connolly, 7 Greg D. Cascino, and 1 Orrin Devinsky ( 1 Department of Neurology, NYU Medical Center, New York, NY ; 2 Child Neurology & Metabolic Diseases Department, Hôpital Robert Debré, Paris, France ; 3 Department of Neurology and Neurosurgery and Human Genetics, Montreal Neurological Institute, Montreal, QC, Canada ; 4 Department of Neurology and Neurosurgery, Montreal Neurological Institute, Montreal, QC, Canada ; 5 Department of Neurology, Cincinnati Children's Hospital Medical Center, Cincinnati, OH ; 6 Division of Neurology, Vancouver BC Children's Hospital, Vancouver, BC, Canada ; and 7 Department of Neurology, Mayo Clinic, Rochester, MN ) Rationale: Tuberous Sclerosis Complex (TSC) is associated with medically refractory epilepsy. Epilepsy surgery, with the goal of resecting epileptogenic brain tissue for seizure relief, has become a viable therapeutic option in TSC. The purpose of the current study is to determine factors associated with seizure relief postoperatively in TSC patients utilizing multicenter data. Methods: A retrospective analysis of patients with TSC following surgery for relief of epilepsy was performed from multiple centers. Seventy subjects were analyzed. We studied patient characteristics (presence/absence of mental retardation, age at TSC diagnosis), electrophysiological characteristics (seizure onset age, ictal/interictal focality), seizure type (i.e., presence/absence of infantile spasms), tuber number, electrophysiological/neuroimaging concordance, surgical variables (age at surgery, surgical method, and number of stages), and postoperative residual dominant tuber as determined by neuroimaging. Data were analyzed for all 70 subjects as a whole and for subjects assessed two years or less postoperatively and more than 2 years postoperatively. Results: 27/70 subjects were seizure free (Engel Class I). Subjects with ictal multifocality, mental retardation, and discordant EEG/MRI data were less likely to achieve seizure freedom. However, only associations between seizure freedom and age at seizure onset, history of infantile spasms, and interictal focality were significant at all postoperative durations. The mean age at seizure onset was 2 years in seizure free subjects compared to 7 months in those with persistent seizures. Only 5/32 subjects with a history of infantile spasms were seizure free postoperatively. Focal interictal discharges were seen in 85% of seizure free subjects versus 32% of subjects with persistent seizures. The number of tubers was significantly associated with poor outcome when measured less than 2 years postoperatively. Conclusions: In patients experiencing seizure freedom following surgery for epilepsy in TSC, ictal focality, normal development, and concordant electrophysiological and neuroimaging data are common findings. However, similar features can also be observed in subjects with postoperative seizures. The current study suggests that later age at seizure onset, no history of infantile spasms, interictal focality, and fewer cortical tubers may be determining factors in seizure freedom. 1 Carolina R. Almeida, 2,3 Kette D. Valente, 2 Silvia Vincentiis, 4 Juan Castro, 4 Hung T. Wen, 1 Renato L. Marchetti, 1 Wagner F. Gattaz, and 1 Elida B. Ojopi ( 1 Laboratory of Neuroscience, Psychiatry Institute, USP, Sao Paulo, SP, Brazil ; 2 Laboratory of Neurophysiology, Psychiatry Institute, USP, Sao Paulo, SP, Brazil ; 3 LIM 21, Psychiatry Institute, USP, Sao Paulo, SP, Brazil ; and 4 Neurosurgery, USP, Sao Paulo, SP, Brazil ) Rationale: Mesial temporal lobe epilepsy with hippocampal sclerosis is one of the most common medically intractable epilepsy syndromes and its pathogenesis remains obscure. The polymorphisms of single nucleotide (SNPs), alterations in a single nucleotide of DNA sequence, when non‐synonymous, are capable of cause changes in protein's structure and have been studied in association studies. Recent studies demonstrated controversial results about the relationship between gene polymorphism and epilepsy. Several association studies involving more than 30 genes involved in Mesial Temporal Sclerosis (MTS) are described. However, it remains the uncertainty in associations and the imperfection in replications. The objective of this study is the molecular characterization of MTS. Methods: Thirty patients with MTS, well characterized by MRI, were pared by sex and age with 30 controls without family history of neurological or psychiatric disorders to be genotyped. The authors investigated, through search of new DNA SNPs, susceptibility genes to developmental of MTS. Initially, more than 100 genes coding to ionic channels and other genes related to neuronal migration and neurogenesis were evaluated. From these, 82 genes were selected. These genes were previously described in association studies of epilepsy or studies of families. To the moment, we found 20 non‐synonymous polymorphisms in 11 different genes. Results: In the initial experimental analysis, 14 SNPs find in 5 different genes were evaluated. The data show that for the first 30 patients, 8 SNPs were confirmed in 5 different genes (EFHC1, ME2, BRD2, BDNF and TMEM1). Conclusions: The number of patients and controls limited the statistical power of the ongoing study. Knowledge of these SNPs and their effect on genes involved with MTS may be important to help a better understanding of complexes pathways that characterize the neuronal hyperexcitability and the development of new strategies for treatment and diagnosis. (Supported by FAPESP 05/50135–3.) 1 Danielle M. Andrade, 1 Stephen W. Scherer, and 1,2 Berge A. Minassian ( 1 Department of Medical and Molecular Genetics, The Hospital for Sick Children, Toronto, ON, Canada ; and 2 Department of Paediatrics, The Hospital for Sick Children, Toronto, ON, Canada ) Rationale: Protein transduction domains (PTDs) may be capable of transducing full‐length fusion proteins across plasma cell membranes and the blood‐brain‐barrier (BBB). They are therefore considered a promising tool for the development of protein therapy aimed at the central nervous system. In this study we analysed the characteristics of TAT‐PTD genetically fused to cystatin B (CSTB), the protein responsible for the Progressive Myoclonus Epilepsy of Unverricht‐Lundborg disease (ULD). Methods: A PCR product containing the EPM1 coding region was cloned into the pRSET vector containing the TAT‐PTD sequence. A control protein was assembled by insertion of EPM1 cDNA into this vector, followed by removal of TAT‐PTD. TAT‐PTD‐CSTB and CSTB were expressed in bacterial cells, purified and transduced into COS‐7 cells and human lymphoblasts. Whole cell lysates and cellular fractions were obtained for study by polyacrylamide gel electrophoresis and immunoblotting. Transduced (and control) cells were also used for analysis through immunofluorescence microscopy. Results: Initial results pointed to an apparent time and concentration‐dependent transduction of TAT‐PTD‐CSTB. However, careful investigation demonstrated that the nuclear localization of TAT‐PTD‐CSTB was an artifact of fixation and that the fusion protein was adhered to the plasma membrane. Conclusions: TAT‐PTD‐CSTB does not penetrate the cells. For this reason, even if TAT‐PTD is likely able to cross the BBB, the fusion of this protein to CSTB cannot be used as a form of replacement of the intracytoplasmic protein missing in ULD. Importantly, we discuss precautions to avoid false‐positive results when working with TAT‐PTD for protein therapy of neurological diseases. 1,2 Anna‐Kaisa Anttonen, 2 Erika Ahonen, 3 Eija Gaily, 3 Elina Liukkonen, 4 Marjo‐Riitta Aine, 5 Eila Herrgård, 3 Liisa Metsähonkala, 6 Heikki Rantala, 2 Auli Sirén, and 2 Anna‐Elina Lehesjoki ( 1 Department of Medical Genetics, University of Helsinki, Helsinki, Finland ; 2 Folkhälsan Institute of Genetics and Neuroscience Center, University of Helsinki, Helsinki, Finland ; 3 Department of Pediatric Neurology, Hospital for Children and Adolescents, Helsinki University Central Hospital, Helsinki, Finland ; 4 Department of Child Neurology, Lapland Central Hospital, Rovaniemi, Finland ; 5 Department of Pediatrics, Division of Child Neurology, Kuopio University Hospital, Kuopio, Finland ; and 6 Department of Pediatrics, University of Oulu, Oulu, Finland ) Rationale: Mutations in SCN1A are associated with many epilepsy syndromes such as severe myoclonic epilepsy of infancy (SMEI) and generalized epilepsy with febrile seizures plus (GEFS+). Mutations in SCN2A2 have been described in benign familial neonatal‐infantile seizures and GEFS+ but to date only one SCN2A2 mutation has been associated with intractable epilepsy. We evaluated the contribution of SCN1A and SCN2A2 mutations to infantile onset intractable epilepsy in Finnish patients. Methods: SCN1A and SCN2A2 were screened for mutations in 33 patients, 21 of which had a clinical suspicion of SMEI. We amplified all exons and exon‐intron boundaries of SCN1A and SCN2A2 from genomic DNA and sequenced the PCR products with an ABI 3730 DNA Analyzer. Sequences were analyzed using the Sequencher 4.5 program. Results: We found six different de novo mutations in SCN1A: One nonsense, two splice site, one frameshift‐causing and two missense mutations that change highly conserved amino acids. Two of the mutations were previously described. The clinical presentation in three of the mutation positive patients is SMEI. Two patients have borderline SMEI and one has a milder phenotype with normal cognitive development. We also found one previously undescribed de novo missense mutation of a conserved amino acid in SCN2A2 in a patient with infantile onset epilepsy and episodic ataxia. None of the mutations were observed in 192 control chromosomes. Conclusions: Our findings are compatible with previous findings showing that de novo SCN1A mutations are a major cause of SMEI. SCN2A2 mutations are found only in a small subset of infantile onset intractable epilepsy. 1 Christie Barker‐Cummings, 1 W. Allen Hauser, and 1 Ruth Ottman ( 1 Sergievsky Center and Epidemiology Department, Columbia University, New York, NY ) Rationale: Epilepsy and febrile seizures are complex disorders with both distinct and overlapping genetic influences. Febrile seizures are associated with an increased risk of epilepsy (individual comorbidity). In addition, each disorder aggregates in families, and an increased risk of each disorder has been reported in the relatives of probands with the other. No single study has comprehensively examined the contributions of shared and distinct genetic influences to these observations. We aimed to evaluate the overlapping genetic contributions to epilepsy and febrile seizures by studying patterns of co‐occurrence, within individuals and within families, in a large genetic epidemiologic study. Methods: The study population comprised 2279 full siblings of 879 probands with epilepsy from the Epilepsy Family Study of Columbia University. Data on febrile seizures were obtained by interview of at least one of the proband's parents. We evaluated the risks of epilepsy in siblings without febrile seizures, and febrile seizures in siblings without epilepsy, in relation to the proband's febrile seizure history and family history of febrile seizures and epilepsy in other first‐degree relatives. We considered four types of family history: epilepsy alone, febrile seizures alone, both epilepsy and febrile seizures, and neither. We used Cox proportional hazards models to compute rate ratios (RRs) and 95% confidence intervals (CI) for epilepsy, and logistic regression to compute odds ratios (ORs) for febrile seizures. Results: Epilepsy risk in siblings was not associated with a history of febrile seizures in the proband (RR = 0.5, CI 0.17–1.35). As expected, epilepsy risk was associated with a family history of epilepsy in other relatives (RR = 3.6, CI 1.78–7.47). Epilepsy risk was not associated with a family history of febrile seizures alone (RR = 1.4, CI 0.47–3.85), but was strongly associated with a family history of both febrile seizures and epilepsy (RR = 8.9, CI = 3.92–20.05). The risk of febrile seizures in siblings was associated with a history of recurrent febrile seizures in the proband (OR = 2.3, CI 1.17–4.60), but not with a history of a single febrile seizure in the proband. As expected, febrile seizure risk was associated with a family history of febrile seizures (OR = 4.3, CI 2.54–7.39). Febrile seizure risk was also associated with a family history of epilepsy alone (OR = 2.3, CI 0.99–5.16). As in the analysis of epilepsy, febrile seizure risk was strongly associated with a family history of both epilepsy and febrile seizures (OR = 6.1, CI 2.46–15.18). Conclusions: These results provide evidence for familial clustering of epilepsy and febrile seizures that is not attributed to individual comorbidity of the two disorders. Familial clustering was strongest in families containing multiple individuals with epilepsy. The findings are reminiscent of the syndrome GEFS+, which has previously been described only in highly selected pedigrees. (Supported by NIH R01 NS43472 and an Epilepsy Foundation predoctoral fellowship.) 1 Eylert Brodtkorb, and 2 Fabienne Picard ( 1 Department of Neurology and Clinical Neurophysiology, St. Olav's Hospital, Trondheim, Norway ; and 2 Department of Neurology, University Hospital and Medical School of Geneva, Geneva, Switzerland ) Rationale: Autosomal dominant nocturnal frontal lobe epilepsy (ADNFLE) has been linked to mutations in nicotinic acetylcholine receptors (nAChR) in many families. A beneficial effect of transdermal nicotine has been reported in one single case. We aimed to investigate the influence of tobacco habits on seizure activity in two families with ADNFLE. Methods: Interviews focusing on the pattern of nicotine intake and seizure activity were performed in 22 adult mutation carriers belonging to two ADNFLE pedigrees with different mutations in the gene coding for the nAChR α4 subunit. Results: Ten of 14 tobacco consumers were seizure free, whereas all seven non‐smokers with manifest ADNFLE had persistent seizures. In several patients, remissions and fluctuations of epileptic symptoms clearly appeared to correspond to changes in tobacco habits. Furthermore, one pharmacoresistant patient who had recently begun treatment with transdermal nicotine had experienced improvement. Conclusions: Tobacco appears to be an environmental factor which influences the course of ADNFLE. An inactivation of the mutant nAChRs by desensitization during chronic nicotine exposure may explain the beneficial effect. In pharmacoresistant patients with ADNFLE, treatment with transdermal nicotine should be considered, weighing efficacy against potential long‐term harmful effects. Prospective studies are needed. (Supported by Swiss National Foundation n° 3100A0–104190/1 (Fabienne Picard).) 1 Russell J. Buono, 2 Michael R. Sperling, 3 Dennis J. Dlugos, 1 Michael D. Privitera, 4 Jacqueline A. French, 5 Warren Lo, 6 Steven C. Schachter, 7 Patrick Cossette, 8 Hongyu Zhao, 8 Zeny Feng, 1 Nancy J. Collins, 4 Theresa Scattergood, 9 Wade H. Berrettini, and 9 Thomas N. Ferraro ( 1 Neurology, University of Cincinnati, Cincinnati, OH ; 2 Neurology, Thomas Jefferson University, Philadelphia, PA ; 3 Neurology, Children's Hospital of Philadelphia, Philadelphia, PA ; 4 Neurology, University of Pennsylvania, Philadelphia, PA ; 5 Neurology, Columbus Children's Research Institute, Columbus, OH ; 6 Neurology, Beth Israel Deaconess Medical Center, Boston, MA ; 7 Neurology, CHU Hospital Notre Dame, Montreal, QC, Canada ; 8 School of Public Health, Yale University, New Haven, CT ; and 9 Psychiatry, University of Pennsylvania, Philadelphia, PA ) Rationale: The ABCB1 gene encodes P‐Glycoprotein, a membrane transporter of xenobiotics and natural molecules in many tissues. Of seven studies reporting association tests between variation in ABCB1 and response to anti epileptic drugs (AEDs), three showed positive association, while four did not. Two papers showed association in a case control design, suggesting ABCB1 variation is related to seizure susceptibility. We report the first association study on a cohort of patients and controls from the US and Canada. Methods: Genotyping single nucleotide polymorphisms (SNPs) c1236t and c3435t was performed using TaqMan assays. Samples were collected at 7 sites and all subjects (ages 3–70) were recruited according to local institutional review board policies. Blood samples were obtained and most are available as cell lines from the Coriell Institute. Both markers were typed in 297 generalized patients (70 AED resistant), 253 focal patients (88 AED resistant) and 184 controls, all of European ancestry. Hardy‐Weinberg equilibrium (HWE) was checked and alleles tested for association using chi square tests. Linkage Disequilibrium (LD) and haplotype effects were calculated using likelihood ratio and CHAPLIN respectively. AED response was defined as one year of seizure freedom since last checkup. Results: All markers were in HWE and the SNPs were in strong LD, p = 1 × 10−3. Single marker analyses showed a lack of association for drug response or seizure susceptibility. However, an association was observed between AED response and the CC haplotype in generalized patients (p = 0.007). Haplotypes CC and TT were strongly associated with seizure susceptibility comparing all cases to controls p = 1 × 10−7. Conclusions: We conclude that variation in ABCB1 is weakly associated with AED response in our generalized patients and is strongly associated with seizure susceptibility in all patients. The same haplotpye that confers drug resistance also increases risk for seizures suggesting a relationship between susceptibility and AED response. (Supported by NIH R01NS40396 to RJB, The University of Cincinnati Department of Neurology and The Neuroscience Institute, and The Center for Neurobiology and Behavior at the University of Pennsylvania.) 1 Gianpiero L. Cavalleri, 1 Mike E. Weale, 1 Kevin Shianna, 2 John Lynch, 2 Rinki Singh, 3 Kevin Murphy, 1 Roderick Radtke, 1 James McNamara, 4 Samuel Berkovic, 4 Ingrid Scheffer, 4 John Mulley, 2 Nicholas Wood, 5 Chantal Depondt, 5 Kristl Claeys, 3 Colin Doherty, 3 Norman Delanty, 2 Sanjay Sisodiya, and 1 David Goldstein ( 1 Institute of Genome Sciences and Policy, Duke University, Durham, NC ; 2 The Department of Clinical and Experimental Epilepsy, Institute of Neurology, University College London, London, United Kingdom ; 3 The Division of Neurology, Beaumont Hospital, Dublin, Ireland ; 4 Department of Medicine, The University of Melbourne, Melbourne, Australia ; and 5 Department of Neurology, Universite Libre du Bruxelles, Bruxelles, Belgium ) Rationale: Although great success has been experienced in mapping genes that harbour causative variants for familial forms of epilepsy, little progress has been made in identifying variants that contribute to the development or treatment of more complex forms of the condition. Several key shortcomings in the study design of previous studies have been identified such as insufficient sample sizes, insufficient coverage of genes and lack of replication of detected associations. Here we attempt to integrate these factors in to the largest genetic mapping study to date in epilepsy. Methods: Through the establishment of a multinational collaborative effort, a collection of 3107 epilepsy patients phenotyped for the principal seizure and syndrome types and 1797 healthy controls was assembled. Two hundred and seventy nine candidate genes were selected. For each of these genes bioinformatically‐inferred functional variants were identified. Tagging SNPs were also selected for 198 of the 297 genes. In total 4608 SNPs were genotyped using the Illumina Golden Gate platform across all 4904 participants. Results: Preliminary analyses show an enrichment for significant p values (p < 0.05) but no one p value is strong enough to withstand correction for multiple testing. Conclusions: This result suggests that the data contain a number of risk factors but the effect of any one of these risk factors is marginal. Data analyses will be enhanced by additional genotyping currently ongoing. (Supported by: This study was funded by the Medical Research Council grant G0400126 and the Programme for Human Genomics funded by the Higher Education Authority of Ireland.) 1 Fabio F. Conte, 1 Patricia A.O. Ribeiro, 1 Lourenco Sbragia, 1 Rovilson Gilioli, 1 Fernando Cendes, and 1 Iscia L. Cendes ( 1 Department of Medical Genetics, Universidade Estadual de Campinas, Campinas, São Paulo, Brazil ) Rationale: Missense mutations in the mRib72–1/Efhc1 gene have been identified to co‐segregate with juvenile myoclonic epilepsy (JME) patients. In addition, functional studies have demonstrated that expression of mRib72–1/Efhc1 gene induces apoptosis in neurons in culture. However, the exact relationship between mRib72–1/Efhc1 function and epileptogenesis is still unclear. The objective of this study was to determine the expression pattern of mRib72–1/Efhc1 gene in mouse brain in order to investigate its possible role in programmed cell death during development. Methods: Balb/c mice were submitted to programmed mating to obtain animals in different developmental stages. The brains of embryos aged 15, 17 and 18 days old and of neonates aged 1, 7, 14 and 28 days old were subsequently removed. We used brain of three animals for each age studied. After RNA extraction, the relative gene expression of mRib72–1/Efhc1 was determined by the real‐time PCR technique using TaqMan™ assay. GAPD gene expression was used as an endogenous control. Results: Preliminary results demonstrate that the expression of mRib72–1/Efhc1 is higher in embryos than in neonates (p = 0.001), with a progressive decrease from embryos aged 18 days to neonates. Conclusions: The highest expression of the mRib72–1/Efhc1 gene in embryo brains in comparison to neonates suggests that the putative pro‐apoptotic role of this gene may be necessary for the correct organization of the central nervous system during its initial phases of development. (Supported by CAPES – Coordenacao de Aperfeicoamento de Pessoal de Ensino Superior.) 1 Yaman Z. Eksioglu and 1 Masanori Takeoka ( 1 Neurology/Epilepsy, Children's Hospital Boston, Harvard Medical School, Boston, MA ) Rationale: The Aristaless‐related homeobox gene, ARX, is a transcription factor with a crucial role in the development of forebrain, pancreas and testes. Located in the X chromosome (Xp22.13), ARX mutations cause a pleiomorphic array of phenotypes such as mental retardation, lissencephaly, infantile spasms, and abnormal genitalia in males. Based on human genetics and loss‐of‐function studies in mice, ARX appears to be crucial in the regulation of GABAergic interneuron development and maintenance of circuitry. Ohtahara syndrome consists of epileptic encephalopathy, suppression‐burst EEG‐pattern and refractory epilepsy with tonic spasms in young infants. Despite reports of neonatal epilepsy, Ohtahara syndrome has not been previously described in infants with ARX mutations. Methods: We describe a boy with Ohtahara syndrome who presents with refractory seizures since day‐1‐of‐life caused by a novel mutation in the ARX‐gene. His initial presentation with startle and jerking movements, evolve into brief tonic seizures, refractory to multiple antiepileptic drugs including phenobarbital, zonisamide and levetiracetam. Despite a normal MRI, EEG reveals suppression‐burst‐pattern. He is normocephalic without dysmorphism, except for micropenis and bilateral cryptorchidism, in addition to diffuse hypotonia without weakness. Later, he develops infantile spasms with partial response to vigabatrin. Results: ARX nucleotide sequence analysis, by Athena diagnostics (Worcester, MA), including entire coding regions and intronic sequences of exon‐intron splice junctions, revealed that the proband was hemizygous for a novel mutation comprising 1 base pair insertion of C after nucleotide position 1471, resulting in a frameshift at codon 491. Analysis of mother's DNA revealed that she was heterozygous for the same mutation in ARX‐gene, hence, demonstrating her as a carrier(City of Hope Clinical Molecular Diagnostic Laboratory, Duarte, CA). Conclusions: This is the first reported case of Ohtahara syndrome as consequence of a novel mutation in exon 5 of the ARX‐gene. Insertion of a C after position 1471 in the ARX gene leads to a frameshift that completely changes the structure of the ARX protein from amino acid 491, affecting the composition of the C terminal end. This leads to the disruption of the aristaless‐domain, which is known to play crucial role in central nervous system regionalization with respect to its potential role in transactivation activity. The molecular pathology observed in this novel mutation reflects the complexity associated with ARX protein structure and function. The aristaless domain, localized to exon 5 in the C terminal end of the ARX protein, stands out as an important region in ARX protein signaling. Hence, further genetic/molecular studies are necessary to better understand the function of the aristaless‐domain and its association with the Ohtahara syndrome in our patient with this novel mutation. 1,2,3 Marco Fedi, 1,4 Leon Bach, 1,3 Samuel F. Berkovic, 1,3 Ingrid Scheffer, and 1,5 David C. Reutens ( 1 Department of Medicine, The University of Melbourne, Austin Health, Heidelberg, Victoria, Australia ; 2 Clinical School, The University of Melbourne, Austin Health, Heidelberg, Victoria, Australia ; 3 Department of Neurology, Austin Health, Heidelberg, Victoria, Australia ; 4 Department of Endocrinology, Alfred Hospital, Monash University, Melbourne, Victoria, Australia ; and 5 Southern Clinical School, Monash University, Clayton, Victoria, Australia ) Rationale: The hypothalamo‐pituitary system is innervated by cholinergic fibers and activation of the neuronal nicotinic acetylcholine receptors (nAChRs) modulates the release of anterior pituitary hormones. Mutations of nAChRs subunits are linked to autosomal dominant nocturnal frontal lobe epilepsy (ADNFLE), a syndrome characterized by clusters of nocturnal frontal seizures with no overt endocrinological manifestations. Here we assessed the effect of central cholinergic activation on the release of anterior pituitary hormones in individuals with ADNFLE. Methods: Thirteen subjects with the α4‐Ser248Phe mutation (4 male, mean age 43.2 ± 16.8) and 41 healthy volunteers (24 male, mean age 36.2 ± 12.2) were studied. Serum levels of GH, LH, FSH, PRL, TSH, thyroxine and cortisol were measured at baseline, 30 and 60 minutes after infusion of physostigmine. Results: In controls, cholinergic activation increased the serum levels of GH (mean change + 993%, p < 0.001), PRL (169%, p < 0.001) and reduced the levels of FSH (−7.9%, p < 0.002). In contrast, in subjects with the a4‐Ser248Phe mutation no changes in the release of these hormones were observed (GH 104% p > 0.2; PRL 5.2%; p > 0.8, FSH – 0.8% p > 0.2). Physostigmine increased cortisol levels both in controls (p < 0.01) and in subjects with the mutation (p < 0.03). Between subjects repeated measure ANOVA showed a selective difference in the release of GH and PRL (time*group interaction GH p < 0.04; PRL p < 0.01) following administration of physostigmine. Conclusions: The α4‐Ser248Phe mutation selectively affects cholinergic‐mediated GH and PRL secretion, a finding that may be of pathophysiological significance given the role of somatostatin in experimental models of focal epilepsy. 1 Tracy A. Glauser, 1 Katarzyna Bryc, 1 Todd G. Nick, 1 Donald L. Gilbert, 1 Ericka Holt, 1 Steven Fordyce, and 2 Frank R. Sharp ( 1 Pediatric Neurology, Cincinnati Children's Hospital, Cincinnati, OH ; and 2 Neurology, University of California at Davis, Sacramento, CA ) Rationale: Specific genomic profiles have been noted in the blood of patients with various neurological disorders including Down syndrome, Neurofibromatosis, Tuberous Sclerosis, and Tourette syndrome. This study aimed to determine if specific genomic profiles existed in the blood of children with newly diagnosed and untreated idiopathic epilepsy compared to normal children. Methods: Patients with untreated idiopathic generalized epilepsy (IGE), untreated idiopathic partial epilepsy (IPE) or healthy children with no history of epilepsy (NML) were enrolled in this IRB approved study. RNA was extracted from whole blood and gene expression was determined using Affymetrix U133 2.0 PLUS microarray chips. Using microarray expression data, an analysis of variance (ANOVA) model with epilepsy type (IGE v. IPE v. NML), gender and age (where age was dichotomized over the median of 10 years) was used to identify genes (with a 90% confidence of having false discovery rate, FDR < 0.1) that differentiated between the three groups (IGE, IPE and NML). Pairwise comparisons were performed to identify significantly differentiated genes between groups. Gene ontology (GO) analysis was performed on the IGE and IPE groups from the genes that differentiated between the three groups. GO classes and parent classes with at least 5 observations in the selected subset and with an Observed vs. Expected ratio of at least 2 were reported. Results: A total of 79 children were enrolled; 21 with IPE, 30 with IGE and 28 NML patients. After adjustment for age and gender, 460 genes (with a 90% confidence of having FDR < 0.1) were differentially expressed among the epilepsy groups. Pairwise comparisons showed: IGE v. NML had 301 significantly differentiated genes, (overall p = 0.002), IPE v. NML had 308 significantly differentiated genes (overall p = 0.003) and IPE v. IGE had 26 significantly differentiated genes (overall p < 0.001). From the 460 genes which differentiated the IGE v. IPE v. NML with FDR < 0.1, two gene ontologies were found in the comparison of IGE v. IPE, where the 61 most significant genes contained no more than 5 false discoveries with 90% confidence. The two gene ontologies were intracellular non‐membrane‐bound organelle (GO id 0043232) and non‐membrane‐bound organelle (GO id 0043228) each with Observed vs. Expected ratios of 2.67. Conclusions: Specific genomic profiles exist in children with idiopathic generalized epilepsy and idiopathic partial epilepsy that differentiate from normal controls and from each other. These profiles may provide insight into the underlying process and may represent an additional tool for proper diagnosis and classification of pediatric epilepsy. (Supported by NIH‐R21‐NS044956.) 1 Alica M. Goldman, 2 Daniel L. Burgess, 3 Richard A. Gibbs, 1,4 Kevin E. Chapman, 2,4 Angus A. Wilfong, 1,5 Richard A. Hrachovy, 1,5 Perry J. Foreman, 1 Ian L. Goldsmith, 1 Amit Verma, 2 Marcia J. Hill, and 2 Jeffrey L. Noebels ( 1 Peter Kellaway Section of Neurophysiology, Department of Neurology, Baylor College of Medicine, Houston, TX ; 2 Department of Neurology, Baylor College of Medicne, Houston, TX ; 3 Human Genome Sequencing Center, Baylor College of Medicne, Houston, TX ; 4 Department of Pediatrics, Baylor College of Medicine, Houston, TX ; and 5 Michael E. Debakey VA Medical Center, Houston, TX ) Rationale: Mutations of ion channel genes are recognized as an important cause of defined inherited syndromes in epilepsy and cardiac arrhythmias. We are reporting preliminary data on the following; (1) the prevalence of cardiac co‐morbidities and (2) the frequency and spectrum of genetic variation in the five major genes associated with long QT syndrome in patients with idiopathic epilepsies (IE). Methods: 153 patients with IE were systematically recruited and detailed phenotypic data regarding epilepsy and cardiac arrhythmias were collected. DNA samples were analyzed by sequencing across five major LQT genes (SCN5A, KCNQ1, KCNH2, KCNE1, and KCNE2). Results: Routine ECG was performed in 41/153 individuals. The study was abnormal in 13/41 (32%) subjects. 9/41 (22%) patients had both an abnormal ECG and a medical history positive for syncope, cardiac palpitations or skipped beats. 2/153 (4%) patients were lost due to sudden unexplained death in epilepsy (SUDEP) within one year of follow‐up. Interim DNA sequence analysis of the five major LQT genes in the IE population showed that 27% of all the LQT single nucleotide polymorphisms (SNPs) were exon bound variants of which 45% was represented by protein altering/non‐synonymous SNPs (NS‐SNPs). NS‐SNPs were most commonly found in the SCN5A (33%) and KCNH2 (43%) genes. SCN5A and KCNH2 NS‐SNPs were also the most frequent genetic variants found in patients with dual phenotype of cardiac arrhythmias and epilepsy. Conclusions: The prevalence of inter‐ictal cardiac arrhythmias in our cohort of idiopathic epilepsy patients was similar to the prevalence of ictal arrhythmias cited in the literature ((39%) by Nei et al.Epilepsia.2004;45(4):338–345, and (35%) by Opherk et al. Epilepsy Res.2002;52(2):117–27). The presence of relatively frequent non‐synonymous coding variants in the SCN5A and KCNH2 genes in IE patients with a cardiac phenotype suggests their involvement in the comorbid dual clinical phenotype of cardiac arrhythmias and seizures. (Supported by NINDS 5K08NS47304–3 and NS049130.) 1 An Jansen, 2 Marijke Bauters, 3 Marjan De Rademaeker, 1 Linda De Meirleir, 2 Guy Froyen, 4 Jean‐Pierre Fryns, and 4 Hilde Van Esch ( 1 Department of Child Neurology, AZ Vrije Universiteit Brussel, Brussels, Belgium ; 2 Human Genome Laboratory, Centre for Human Genetics, VIB, Leuven, Belgium ; 3 Medical Genetics, AZ Vrije Universiteit Brussel, Brussels, Belgium ; and 4 Centre for Human Genetics, University Hospital Gasthuisberg, Leuven, Belgium ) Rationale: Mutations in the cyclin‐dependent kinase‐like 5 gene (CDKL5) have been detected in girls with the atypical Rett syndrome variant (MIM 312750) that is characterized by early‐onset infantile spasms and severe neurodevelopmental retardation. We present a young boy with early‐onset infantile spasms, severe encephalopathy, tetralogy of Fallot, and bilateral cataract, in whom a deletion at Xp22 comprising the Nance Horan Syndrome (NHS) gene and CDKL5 gene was detected by array‐CGH. Methods: Brain MRI, video EEG, metabolic and infectious work‐up, karyotype, subtelomeric screening, and array‐CGH were carried out in the proband. Results: The boy was diagnosed shortly after birth with a Tetralogy of Fallot, mild facial dysmorphism, extreme hypotonia, and bilateral microphthalmia with microcornea and congenital nucleocortical cataract. At age 7 weeks he developed myoclonic seizures, occurring most frequently post‐prandial or during drowsiness. EEGs during wakefulness and sleep were normal. At age 4 months, the seizure pattern was consistent with infantile spasms, but no hypsarrhythmia was seen on EEG. At age 5 months, he developed tonic spasms, sometimes preceded by crying or accompanied by laughter or oral automatisms. A video EEG at age 6 months showed multiple myoclonic jerks without EEG correlate, as well as several tonic spasms characterized by sudden onset low voltage rhythmic diffuse 15 Hz activity followed by generalized slowing. No typical epileptiform activity was seen. Seizures were resistant to treatment with valproic acid, vigabatrine, topiramate, phenobarbital and levetiracetam. At age 10 months, he is severely retarded with no reaction to light stimuli, severe axial hypotonia, and absent midline organization. Additional investigations were normal, except for a de novo deletion of approximately 2.8 Mb at Xp22 including 16 genes of which the only known disease‐associated genes are the NHS gene that causes the ophthalmologic abnormalities and the CDKL5 gene that is responsible for the severe epileptic encephalopathy. Conclusions: This is the first description of a male patient with a deletion of the NHS and CDKL5 genes, showing the involvement of CDKL5 in severe epileptic encephalopathy in males. The epilepsy phenotype was characterized by drug‐resistant early‐onset infantile spasms, tonic spasms and myoclonic jerks, without clear epileptiform EEG correlate. (Supported by: H.V.E. is a post‐doctoral researcher of the FWO‐Vlaanderen. M.B. is a doctoral student of the IWT‐Vlaanderen. This work was supported by a research grant G‐0229–01 of the FWO‐Vlaanderen, Belgium.) 1 Ruth Ottman, 1 Kay Kamberakis, 2 Edyth Wiggs, 2 Pat Reeves‐Tyer, 2 Eva H. Baker, 2 John A. Butman, 2 Sadat Shamin, and 2 William H. Theodore ( 1 G.H. Sergievsky Center, Columbia University, New York, NY ; and 2 Clinical Epilepsy Section and Diagnostic Radiology Department, National Institutes of Health, Bethesda, MD ) Rationale: Autosomal dominant partial epilepsy with auditory features (ADPEAF) is a form of idiopathic lateral temporal lobe epilepsy frequently caused by mutations in the leucine‐rich, glioma inactivated 1 (LGI1) gene. Subjects with epilepsy in families with LGI1 mutations have not generally been found to have mesial temporal sclerosis or other structural brain abnormalities detected on MRI. However, in one previously reported family with an LGI1 mutation, a lateral temporal malformation was identified in 10 individuals. This result was important because it strongly supported the concept that LGI1 influences risk through a neurodevelopmental mechanism. We attempted to replicate these findings in other families. Methods: We studied 15 subjects (10 affected, 2 clinically unaffected mutation carriers, 3 unaffected non‐carriers) from six ADPEAF families, each of which had a different LGI1 mutation. Subjects underwent high‐resolution structural MRI on a Philips 3‐tesla scanner, including MPRAGe, FLAIR, t1 and t2 sequences, and neuropsychological testing with a special emphasis on language measures. MRIs were analyzed by radiologists blinded to clinical and genetic data. Results: Subjects ranged in age from 20–81 years; nine were male and six female; and education ranged from 13–21 years. These factors were similar in affected and unaffected subjects. We found no evidence of mesial temporal sclerosis, malformations of cortical development, or abnormal gyral patterns in any of the 15 subjects. One affected individual (age 63) had bilateral T2 increases, and two unaffected subjects, neither of whom carried a mutation, had incidental findings thought to be unrelated to epilepsy (a bifrontal white matter abnormality, and marginally lower right temporal neocortical volume). Verbal IQ scores were significantly lower in subjects with epilepsy than in those without epilepsy (ave. 108 vs. 124, p = 0.012). Among subjects without epilepsy, however, verbal IQ was similar in subjects who carried a mutation (ave. = 124) and those who did not (ave. = 125), suggesting that the difference in verbal IQ was more likely to be due to epilepsy or its treatment than to the mutation per se. In contrast, average scores for the oral word association test were lower both in subjects with epilepsy (94) and clinically unaffected mutation carriers (95) than in noncarriers (120) (p = 0.048). Conclusions: The results of this study do not support the previously reported association of LGI1 mutations with a structural brain abnormality detectable on high‐resolution MRI. Although the number of subjects tested is small, findings from the oral word association test may reflect an impact of mutations on language function. (Supported by NIH R01 NS36319 and the NINDS Division of Intramural Research.) 1 Marcelo A. Kauffman, 2 Dolores Gonzalez Moron, 1 Damian Consalvo, and 1 Silvia Kochen ( 1 Centro de Epilepsia, Hospital Ramos Mejia, Buenos Aires, Argentina ; and 2 Neurology Residency, Hospital Ramos Mejia, Buenos Aires, Argentina ) Rationale: Previous studies have examined the role of a single nucleotide polymorphism (SNP) at the promoter region of Interleukin 1b gene (IL1b −511T) on Temporal Lobe Epilepsy (TLE) and Febrile Seizures (FS) susceptibility, but they have been inconclusive. The aims of the present study were to define the role of this polymorphism on TLE and FS susceptibility and to obtain more precise estimates of the risk associated with this genetic variant. Methods: Published studies up to Febuary 2006 of TLE, FS and IL1b −511T SNP were identified by searches of PubMed, Biosis, Web of Science and Embase databases. Meta‐analysis of TLE and FS case‐control data were performed to assess the association of IL1b −511T SNP with TLE, TLE with Hippocampal Sclerosis (TLEHS) and FS. Pooled Odds Ratios were estimated by means of a genetic‐model free approach. Quality of Studies included was assessed by a score. Publication Bias was investigated by Egger and Begg tests. Results: A total of 12 studies were included where 7 examined the association with TLE (with a total of 1798 subjects) and 5 examined the association with FS (with a total of 1299 subjects). The data analyzed by the genetic model‐free approach indicated that the best model of inheritance is recessive. This model for allele T produced significant results in the TLE sample (fixed effects odds ratio 1.30 (1.01–1.69)) and in the TLEHS sample (fixed effects odds ratio 1.69 (1.26–2.27)) (Figure below), whereas for FS the results were non‐significant (fixed effects odds ratio 1.33 (0.95–1.84)). There was not evidence of Publication bias (Egger and Begg tests non‐significant for TLE, TLEHS and FS samples). Quality was heterogeneous between studies. Conclusions: Overall, these findings support a role for IL1b −511T gene variant in TLE under a recessive mode of inheritance. Further explorations of the involvement of the IL1b gene in the susceptibility to FS, with a greater number of studies with larger sample sizes, are needed to fully establish the role of the IL1b gene in this condition. (figure 1) (Supported by: Dr. Kauffman has a PhD scholarship from CONICET.) 1 Ok Joon Kim and 2 Nam Keun Kim ( 1 Neurology, Pochon CHA University, Sungnam, Kyunggi‐do, Republic of Korea ; and 2 Clinical Research Institute, Pochon CHA University, Sungnam, Kyunggi‐do, Republic of Korea ) Rationale: Elevated plasma homocysteine level (hyperhomocysteinemia) is associated with various vascular diseases, abortion, and cancer. In epilepsy, prolonged antiepileptic drugs induce hyperhomocysteinemia, especially in the methylenetetrahydrofolate reductase (MTHFR) TT type, which induce cerebrovascular disease and congenital fetal abnormalities such as neural tube defects. Because folic acid decreases homocysteine level, folic acid is useful treatment in patients with hyperhomocysteinemia.We studied to assess factors that have affects on the decrease of homocysteine levels in epilepsy patients, especially according to single nucleotide polymorphisms. Methods: We prospectively included 75 epilepsy patients admitted at Pundang CHA Hospital between 2002 and 2005. All patients treated with antiepileptic drugs during one year and examined homocysteine levels at two times: initial and last time (one year later). When hyperhomocysteinema (homocysitne > 12μmol/L) was found at initial time, patients were supplemented with daily 400μg folic acid for one year. All patients were divided into two groups; increase group (IG) and decrease group (DG) according to the change of homocysteine levels between initial and last time. Also, all patients examined various polymorphisms, such as MTHFR C677T, MTHFR A1298C, methionine synthase A2756G, and thymidylate synthase gene polymorphisms. Results: Among 75 patients, IG was 31 patients and DG was 44 patients. Decrease of homocysteine levels during one year was observed in young age, initial high level of homocysteine, use of folic acid, presence of family history, and semiology of complex partial seizure. There were no significant differences of homocysteine levels according to polymorphisms. Conclusions: Our results suggest that folic acid can cheaply decrease homocysteine levels in epilepsy patients with hyperhomocysteinemia. We recommend administration of folic acid in old epilepsy patients with generalized and symptomatic seizures since homocyteine level did not increase in the young age patients with complex partial seizure and family history. 1 Patrick Kwan, 2 Virginia Wong, 3 Ping W. Ng, 4 Colin H.T. Lui, 5 Ngai C. Sin, 1 Andrew C.F. Hui, 1 Lawrence K.S. Wong, 1 Yu Evelyn, and 1 Larry Baum ( 1 Division of Neurology, Chinese University of Hong Kong, Hong Kong, Hong Kong ; 2 Department of Paediatrics, University of Hong Kong, Hong Kong, Hong Kong ; 3 Department of Medicine, United Christian Hospital, Hong Kong, Hong Kong ; 4 Department of Medicine, Queen Elizabeth Hospital, Hong Kong, Hong Kong ; and 5 Department of Paediatrics, Prince of Wales Hospital, Hong Kong, Hong Kong ) Rationale: There is conflicting data whether CC genotype of the C3435T single nucleotide polymorphism (SNP) of the ABCB1 gene is associated with multi‐drug resistance in Caucasian epilepsy patients. There is wide ethnic variation in the genotype frequencies of this SNP. We investigated whether the association might be observed in ethnic Chinese. Methods: ABCB1 3435 was genotyped in 518 Chinese patients with epilepsy, classified as drug resistant in 221 and drug responsive in 279, and 179 control subjects without epilepsy. Phenotypic definitions were similar to previous studies in Caucasian subjects. Drug resistance was defined as at least 1 seizure per month despite treatment with 2 or more antiepileptic drugs, and drug responsive as seizure free for at least 1 year on treatment. Genotype was carried out by polymerase chain reaction‐restriction fragment length polymorphism assays. Results: As compared with patients with drug responsive epilepsy, patients with drug resistant epilepsy were more likely to have the TT genotype at ABCB1 3435 than CT or CC genotypes (odds ratio 2.2, 95% confidence interval 1.2 – 4.1, p = 0.004). The frequency of CC genotype was similar between patients with drug resistant and drug responsive epilepsy. There was no difference in genotype frequencies between the control subjects and the two patient groups. Conclusions: In Chinese patients with epilepsy, ABCB1 3435TT was found to be associated with drug resistance. The results contradict those reported in Caucasian subjects, and highlight the complexity of the potential role of ABCB1 3435 SNP in the pathogenesis of drug resistant epilepsy. (Supported by: Partially supported by the HKU Small Project Funding.) 1 Walter J. Lukiw, 1 Victor L. Marcheselli, and 1 Nicolas G. Bazan ( 1 LSU Neuroscience Center, Louisiana State University Health Sciences Center, New Orleans, LA ) Rationale: Experimentally induced epileptogenesis elicits profound changes in early response gene (ERG) expression in the mammalian brain. In these studies we analyzed early (0–6 hr), medium (1–7 da) and long term (2–8 wk) changes in the DNA‐binding of transcription factors (TFs) AP1, AP2, Egr1, HIF‐1, STAT1, NF‐kB, NFIL‐6, SP1 and TFIID to target DNA sequences in adult rat hippocampus after a single kainic acid (KA) injection. These TFs are enriched in the immediate promoters of the cFOS and cyclooxygenase‐2 (COX‐2) ERGs. Methods: To trigger epileptogenesis, male albino Wistar rats were injected i.p. with 10 mg/kg KA using saline as a vehicle. At indicated time points (0, 1, 3, 6 hr; 1, 3, 7 da and 2, 4, 8 wk) rats were sacrificed and hippocampal and cortical cellular and nuclear proteins and total RNA were co‐isolated from the same samples. TF‐DNA binding was studied using gel shift and super‐shift assay; RNA message levels were determined using DNA array analysis, RT‐PCR and Northern blots. cFOS and COX‐2 protein levels were quantified using Western immunochemistry. Results: Of the TFs examined, AP1, HIF‐1, NF‐kB and STAT1 binding were found to display oscillatory DNA binding profiles over the short, medium and long term. AP1‐DNA binding kinetics strongly correlated with COX‐2 gene activation over the short term while NF‐kB‐DNA binding paralleled COX‐2 gene expression at longer time points. Sustained increases in TFIID‐DNA binding suggested prolonged de novo induction of transcription from TATA‐containing rat hippocampal genes. Conclusions: Taken together, these results suggest that induction of the pro‐inflammatory and pro‐apoptotic TFs AP1, HIF‐1, NF‐kB and STAT1, and their binding to specific target DNAs in ERG promoters, drives inflammatory and apoptotic gene expression in a phasic fashion. These data also support the notion of a global re‐programming of hippocampal gene expression patterns long after a single KA‐induced triggering of epileptogenic events. (Supported by NIH AG18031 and NIH NS23002.) 1 Carla Marini, 2 Paolo Aridon, 3 Chiara Di Resta, 4 Elisa Brilli, 2 Maurizio De Fusco, 3 Fausta Politi, 1 Elena Parrini, 2 Irene Manfredi, 5 Tiziana Pisano, 5 Dario Pruna, 3 Giulia Curia, 5 Carlo Cianchetti, 4 Massimo Pasqualetti, 3 Andrea Becchetti, 1,6 Renzo Guerrini, and 2 Giorgio Casari ( 1 Division of Child Neurology and Psychiatry, IRCCS Stella Maris Foundation, Pisa, Italy ; 2 Human Molecular Genetics Unit, Dibit San Raffaele Scientific Institute, Milan, Italy ; 3 Dipartimento di Biotecnologie E Bioscienze, Università di Milano‐Bicocca, Milan, Italy ; 4 Unità di Biologia Cellulare E Dello Sviluppo, Dipartimento di Biologia, Università di Pisa, Pisa, Italy ; 5 Child and Adolescent Neuropsychiatric Clinic, University of Cagliari, Cagliari, Italy ; and 6 University of Pisa, Pisa, Italy ) Rationale: Autosomal dominant frontal lobe epilepsy is characterized by clusters of sleep‐related hypermotor seizures, and has been associated with mutations of the alpha4 and beta2 subunits of the neuronal nicotinic acetylcholine receptor (nAChRs) genes (CHRNA4 and CHRNB2). An Italian family from Sardinia with sleep‐related seizures with prominent fear and wandering is presented. Methods: Clinical information, EEG studies and blood samples were obtained. A genome‐wide linkage analysis was performed, followed by electrophysiological studies of transfected HEK293 cells expressing either the newly identified nAChR alpha2 mutant subunit or the wild type receptor. Expression pattern of the genes coding for the alpha2 and 4, and beta2 nAChRs in areas of autoptic human brain was also performed. Results: Ten family members had clusters of seizures during sleep. Mean age at seizure onset was 10 years. The proband had since 6 years of age, clusters of sleep‐related brief episodes with awakening, vocalization, fearful expression and movements of the tongue. The remaining individuals had nocturnal clusters of similar episodes including complex motor behaviors such as walking outside the room. The genome‐wide linkage analysis showed a locus on chromosome 8p12.3–8q12.3, and a heterozygous missense mutation in the nAChR alpha2 subunit gene (CHRNA2) was detected in all 10 affected individuals. Whole‐cell recordings of transfected HEK293 cells showed that the CHRNA2 mutation markedly increased the receptor sensitivity to acetylcholine. The expression pattern of alpha2 mRNA in autoptic human brain exhibited the highest level in the thalamus, but a diffuse expression was detected in the investigated samples. beta2 and alpha4 mRNA were uniformly distributed throughout the samples. Conclusions: CHRNA2 is the third nAChRs gene to be associated with familial sleep‐related epilepsies. CHRNA2 mutation causes a complex ictal behavior with prominent fear and wandering. The alpha2 subunit showed the highest expression in the thalamus, hence confirming that mutations in nAChRs genes are involved in the epileptogenesis related to sleep‐arousal mechanisms. (Supported by the Italian Minister of Health; Telethon Foundation; Italian Ministry of Education, University and Research (MIUR).) 1 Claudia V. Maurer‐Morelli, 1 Rodrigo Secolin, 1 Romenia R. Domingues, 1 Rafael B. Marchesini, 1 Neide F. Santos, 2 Eliane Kobayashi, 2 Fernando Cendes, and 1 Iscia Lopes‐Cendes ( 1 Medical Genetics, UNICAMP, Campinas, São Paulo, Brazil ; and 2 Neurology, UNICAMP, Campinas, São Paulo, Brazil ) Rationale: The association between temporal lobe epilepsy and mesial temporal sclerosis (MTS) has been well established; as well as the use of hippocampal atrophy (HA) on magnetic resonance imaging (MRI) as an in vivo surrogate marker of MTS. One of the risk factors associated to MTS is childhood prolonged febrile seizures. We have described a type of mesial temporal lobe epilepsy with evident familial recurrence associate with HA but low frequency of febrile seizures. Previous pedigree and complex segregation analyses provided evidence for the presence of a major gene predisposing to HA in familial mesial temporal lobe epilepsy (FMTLE). The objective of this study was to identify the region harboring the main gene associated with HA in FMTLE by a genome‐wide linkage study. Methods: Genome‐wide scan was performed using a total of 332 microsatellite markers at ∼12cM intervals. An additional 14 markers were genotyped in the candidate region. Two‐point and multipoint LOD scores were calculated with the LINKAGE computer package. We assumed an autosomal dominant inheritance with incomplete penetrance. Results: We identified linkage to chromosome 18p11.3–11.2, with a maximum LOD score of 3.60 at θ= 0.0 for the D18S976 marker in a single family (F‐10) with 11 affected individuals with HA. Multipoint and haplotype analyses localized the locus within a 6 cM interval flanked by markers D18S976 and D18S452. No other significantly positive LOD score was detected for this family in the entire genome. Conclusions: This is the first conclusive evidence that HA may be caused by genetic factors, which can have major implications in the study of the pathophysiological mechanisms underlying MTS and its relationship with temporal lobe epilepsy. (Supported by FAPESP.) 1 Marcelo J. Murai, 1 Ricardo Horiuchi, 2 Daniel Martins, 1 Cláudia Maurer‐Morelli, 2 José Camillo Novello, 3 Fernando Cendes, and 1 Iscia Lopes‐Cendes ( 1 Medical Genetics Department, Unicamp, Campinas, São Paulo, Brazil ; 2 Biochemical Department, Unicamp, Campinas, São Paulo, Brazil ; and 3 Neurology Department, Unicamp, Campinas, São Paulo, Brazil ) Rationale: Mesial temporal lobe epilepsy (MTLE) is the most common and severe type of partial epilepsy, representing ∼50% of all adult epilepsy patients and frequently associated with pharmaco‐resistance. The relationship between MTLE and hippocampal sclerosis (HS) is well established. However, the precise pathogenesis of HS and its relationship with MTLE is not completely clarified. Two‐dimensional electrophoresis (2DE) is a powerful fractionation method for complex protein mixtures. In difference gel electrophoresis (DIGE) based proteomics, the experimental and control samples are labeled with different fluorophores and are run in the same gel, thereby minimizing gel preparation variation. DIGE is one of the few techniques that is capable to perform quantitative proteomics, generating statistical data to differences in protein abundances. Methods: We analyzed the proteome of three hippocampus removed from patients with refratory MTLE who underwent epilepsy surgery. Results: 2DE‐DIGE identified 2 up‐regulated and 10 down‐regulated proteins as determined by Student's T‐test (p ≤ 0.01). The observed molecular weight of the spots ranged from 28 to 93 kDa. Conclusions: The identity of these spots will be elucidated by mass spectrometry in order to gain additional information in a global scale about the mechanism of epileptogenesis in MTLE. (Supported by Fapesp, CAPES and CNPq.) 1,2 Rima Nabbout, 2 Stéphanie Baulac, 1 Nadia Bahi‐Buisson, 1 Catherine Chiron, 1 Olivier Dulac, and 2 Eric Leguern ( 1 Neuropediatrics Department, INSERM U663, Hopital Necker‐Enfants Malades, Paris, France ; and 2 INSERM U679, Hôpital de La Pitié Salpêtrière, Paris, France ) Rationale: We report the clinical and genetic study of a large French family with febrile seizures (FS) and childhood absence epilepsy (CAE). Methods: This family was identified through a national French campaign for familial epilepsy. It spans 4 generations and consists of 51 members with 13 affected. The medical history of all members was obtained by personal information and by consulting the medical files of each affected member. All family members gave a written consent to participate and 26 DNA were available for genetic study. Results: Clinical study: All affected members presented FS with CAE in 5 members and TLE in one. All FS stopped before the age of 6 and they recurred less than 4 times. Patients presenting CAE had recorded absences and characteristic EEG with 3 Hz spike waves. FS were simple except in one patient who had a long lasting complex FS (45 minutes) that occurred at 8 months of age. He presented later pharmaco‐resistant TLE. He had a left hippocampal sclerosis on brain MRI. All family members have a normal psychomotor development.Genetic study: The FS trait segregates as autosomal dominant trait. The genetic study allowed the exclusion of any linkage with reported loci for FS and FS plus (FS+), especially with the locus of GABRG2 gene on chromosome 5q reported in a family presenting FS+ and absences. A genome wide search with 380 markers allowed us to localize a new gene responsible for FS on 3p. We could not exclude another genomic segment with positive Lodscore (Z = 1.94 at θ= 0.00). All patients presenting further epilepsy (CAE and TLE) shared a common haplotype on this locus in addition to the FS haplotype on 3p. Conclusions: These findings emphasize the genetic heterogeneity of FS. Epilepsy in association with FS could result in this family from an interaction between at least 2 genes: the gene on 3p with a modifier gene. (Supported by ARGE, FFRE.) 1 Ruth Ottman, 1 Christie Barker‐Cummings, 2 Vincent M. Vasoli, 2 James P. Burke, 3 Janet L. Sobell, 1 W. Allen Hauser, and 2 Jeffrey R. Buchhalter ( 1 G.H. Sergievsky Center, Columbia University, New York, NY ; 2 Mayo Clinic, Rochester, MN ; and 3 Zilkha Neurogenetic Institute, University of Southern California, Los Angeles, CA ) Rationale: Valid screening methods to identify people with a history of seizures or epilepsy are crucial for epidemiologic studies. They are also essential to identify affected family members in genetic studies. Screening relies on careful history‐taking, but it is not clear what questions should be asked, or how accurately they identify affected individuals. Methods: We designed an 8‐question screening interview to identify people with a history of seizures. The first question asks, “Have you ever had, or has anyone ever told you you had, epilepsy or a seizure disorder?” If this question is answered “no,” it is followed by a second question that asks, “Have you ever had a seizure, convulsion, fit or spell under any circumstances?” This is followed by six additional questions that screen for symptoms possibly related to seizures (e.g., uncontrolled body movements, unexplained change in mental state, etc.). Subjects who screen positive are given a detailed diagnostic interview to confirm seizure occurrence and obtain additional clinical details. The interview was administered by telephone to 148 subjects with medical record documented incidence of unprovoked seizures from 1935–1994 while residing in Rochester, Minnesota, and 168 age‐ and sex‐matched controls. Interviewers were blinded to case‐control status. Results: Sensitivity was defined as the proportion who screened positive (“yes” or “possible” to any screening question) among those with either epilepsy (n = 124) or an isolated unprovoked seizure (n = 24); specificity was defined as the proportion who screened negative among the controls. Sensitivity was 86% for epilepsy and 75% for isolated unprovoked seizures. Specificity was 92%. Sensitivity was significantly higher for subjects with incidence in 1955 or later than for those with earlier incidence (epilepsy: 89% vs. 71%, isolated unprovoked seizure: 82% vs 0%(0/2)). Among subjects with epilepsy who screened positive, 87% answered “yes” or “possible” to the first question (asking about epilepsy or a seizure disorder), 9% to the second question (asking about seizures under any circumstances), and 4% to another question. Among subjects with isolated unprovoked seizures who screened positive, 56% answered “yes” or “possible” to the first question and 44% to the second question. Conclusions: This screening instrument is highly effective (89% sensitivity) for identifying people with onset of epilepsy up to 50 years before interview. Subjects are more likely to be missed if they had onset of epilepsy even earlier, or had an isolated unprovoked seizure rather than epilepsy. Further analyses should address the relation of sensitivity to clinical epilepsy features, so that sensitivity can be improved by targeting new questions to particular subgroups. (Supported by NIH R01 NS043472.) 1 Sumit Parikh, 2 Deepak K. Lachhwani, 2 Ajay Gupta, 2 Elaine Wyllie, and 2 Prakash Kotagal ( 1 Neurology, Section of Pediatric Neurology, Cleveland Clinic Foundation, Cleveland, OH ; and 2 Neurology, Section of Pediatric Epilepsy, Cleveland Clinic Foundation, Cleveland, OH ) Rationale: Up to 30% of children with mental retardation (MR) have epilepsy and up to 45% of children with epilepsy have MR (Epilepsia. 2000 Sep;41(9):1214–20). The report of the Amalfi group raised concern about missing treatable causes of epilepsy; as a result diagnostic metabolic testing in children with developmental delay and epilepsy is commonly performed . (J Child Neurol. 2002 Dec;17 Suppl 3:3S98–102) A standardized approach to such testing does not yet exist, and the prevalence of metabolic disease in this population is not known. Methods: Records of the 429 children admitted to the pediatric epilepsy monitoring unit during 2005 were reviewed. We noted the presence or absence of developmental delay, type of EEG discharges (focal‐mixed or generalized), and presence or absence of metabolic testing. Metabolic studies included: a) Plasma amino acids, biotinidase, lactate/pyruvate, acylcarnitines, ammonia levels, transferring isoelectric focusing, peroxisomal testing, and uric acid; b) CSF amino acids, glucose and lactate/pyruvate levels; and c) Urine organic acids, uric acid and creatine/guanidinoacetate levels. Results: 20% of our patients received some metabolic testing (85/429). Metabolic abnormalities were found in 48% of these patients (40/83), with 31% of these individuals having findings of mitochondrial dysfunction (27/85) with elevations in a) plasma alanine/proline/sarcosine (amino acids); b) plasma lactic acid; and c) urine lactic acid, 3‐methyl‐glutaconic acid and Kreb cycle intermediates (organic acids). Metabolic abnormalities were most common in those with focal discharges (70%; 21/30), when compared to multi‐focal (53%; 8/15) and generalized (40%; 13/32) discharges. Secondary carnitine deficiency (26%, 22/85) and valproate related isolated elevations in plasma glycine (63%, 5/8) were also commonly seen. A single patient with a disorder of creatine metabolism was identified. Only 11% of the patients had CSF analysis done (10/85) with a 30% yield of abnormalities (3/10), including elevations in alanine, glutamate or glutamine. Of 9 patients with seizures but normal development, 6 had metabolic abnormalities identified (66%). Only 17 individuals received their first MRI at our institiution, with 86% (12/14) having structural abnormalities found, and one of these individuals, with a cortical heterotopia, having mitochondrial disease diagnosed. Conclusions: Metabolic abnormalities may represent a significant finding in individuals with epilepsy, especially in those with developmental delays, even despite findings of a focal structural brain abnormality. A systematic approach to the diagnoses of these individuals is needed. A consult with a metabolic specialist is encouraged in these patients if a diagnosis is not known. Disorders of mitochondrial metabolism should be placed high on the differential list. 1 Patricia A.O. Ribeiro, 2 Lourenço Sbragia, 3 Rovilson Gilioli, 4 Fernando Cendes, and 1 Iscia Lopes‐Cendes ( 1 Medical Genetics, UNICAMP, Campinas, São Paulo, Brazil ; 2 Pediatric Surgery, UNICAMP, Campinas, São Paulo, Brazil ; 3 CEMIB, UNICAMP, Campinas, São Paulo, Brazil ; and 4 Neurology, UNICAMP, Campinas, São Paulo, Brazil ) Rationale: Mutations in LGI1 gene were described in patients with autosomal dominant lateral temporal lobe epilepsy and preliminary functional studies point to a possible involvement of LGI1 with migration and/or neuronal proliferation. However, the precise function LGI1 remains unknown. The objective of the present study was to determine the expression pattern of the Lgi1 gene in mice brain during development and in adult animals. Methods: Programmed mating was carried with Balb/c mice in order to obtain embryos of different ages. The brains of three animals at the following ages were removed: E15, E17, E18 days (E: embryo), P1, P7, P14, P28, P42 and P56 days (P: post‐natal). Gene expression assays were carried out using real time PCR with the TaqMan™ system. In addition we used an endogenous control (Gapdh gene) and all experiments were performed in duplicates. Results: Lgi1 gene expression was significantly low during the intra‐uterine ages increasing gradually until P56 (adult animal). Samples from P28, P42 and P56 presented a seven fold increase in expression as compared to E15 samples. Conclusions: The pattern of Lgi1 gene expression that we observed suggests a predominantly inhibitory function during development of the central nervous system (low expression during embryonic stages). In addition, we may speculate that in more advanced ages, when neurons are already differentiated, its inhibitory function could be also essential, which can be suggested by the high expression in adult mouse brain. (Supported by CNPQ and FAPESP.) 1 Graeme J. Sills, 1 Mohd Makmor‐Bakry, 1 Nikolas Hitiris, 1 Elaine Butler, and 1 Martin J. Brodie ( 1 Epilepsy Unit, Division of Cardiovascular & Medical Sciences, University of Glasgow, Glasgow, Scotland, United Kingdom ) Rationale: The dose of carbamazepine (CBZ) required to achieve optimal seizure control varies widely from patient to patient. Genetically‐determined variability in the activity and/or capacity of various drug metabolising enzymes (DMEs) is believed to influence the pharmacokinetics, and thereby dosage requirements, of a number of therapeutic agents. We have investigated polymorphic variants in the genes encoding DMEs involved in the metabolism of CBZ in an effort to identify predictors of CBZ maintenance dose. Methods: A total of 70 epilepsy patients (49% male; median age 34 years, range 14 – 72 years) who had benefited (≥50% reduction in seizure frequency) from treatment with CBZ monotherapy were included in the analysis. Individuals who did not tolerate the drug and/or those who did not experience significant efficacy were specifically excluded. Common genetic variants in CBZ‐related DMEs, including specific cytochrome P450s (CYP3A4, CYP3A5 and CYP1A2) and microsomal epoxide hydrolase (EPHX1), were screened by conventional polymerase chain reaction – restriction fragment length polymorphism (PCRRFLP) or direct sequencing. Associations were identified and characterised by multiple logistic and linear regression analyses, respectively. Results: All genotype frequencies were consistent with Hardy‐Weinberg equilibrium (p>0.05). No single clinical characteristic or genetic variant was sufficient to predict CBZ maintenance dose. However, a multiple logistic regression model incorporating patient age at time of optimal seizure control (OR = 1.03, 95%CI 1.00–1.07, p = 0.024), genotype of EPHX1 c.337T>C (OR = 0.44, 95%CI 0.22–0.87, p = 0.018) and genotype of EPHX1 c.416A>G (OR = 0.46, 95%CI 0.22–0.98, p = 0.044) demonstrated a significant association with maintenance dose (r2= 0.16; p = 0.009). Conclusions: This analysis suggests that a combination of age and genetic variants of the EPHX1 gene can be used to predict optimal maintenance doses of CBZ. Further investigations are required in an effort to strengthen the predictive value of this observation before its clinical utility can be realistically assessed. (Supported by: MMB is supported by a studentship from the Government of Malaysia and Universiti Kebangsaan Malaysia.) 1 Lisa J. Strug, 1 Martina Durner, 3 Effie Cayanis, 1 Fengli Zhang, 1 Dana Politis, 1 Irene Klotz, 1 Elisa Dicker, and 1,2,3 David A. Greenberg ( 1 Biostatistics, Division of Statistical Genetics, Columbia University, New York, NY ; 2 Psychiatry, Columbia University, New York, NY ; and 3 Genome Center, Columbia University, New York, NY ) Rationale: In linkage genome scans of Idiopathic Generalized Epilepsy (IGE) we found evidence for linkage on chromosome 5p with the juvenile absence phenotype. There are at least two candidate genes in this region: Succinate dehydrogenase complex subunit A (SDHA) and Programmed cell death 6 (PDCD6). Methods: We examined this region for association with Juvenile Absence Epilepsy (JAE) and Childhood Absence Epilepsy (CAE), genotyping single nucleotide polymorphisms in and around SDHA and PDCD6. We used case‐control and family‐based association methods to determine the degree of association. Results: We found that CAE is strongly associated with all SNPs tested in the PDCD6 gene (maximum chi‐squared = 21.22, p < 0.0001 at rs4957014). JAE, in contrast, was not significantly associated with PDCD6 or any SNPs in the region studied. These results were confirmed in the family‐based association analysis. The odds of having CAE if one carries at least one copy of the associated allele at the rs4957014 locus is 3.2 times the odds for an individual without a copy of the allele at this locus (95%CI: 1.91 – 5.37). Conclusions: Our results suggest that not only does PDCD6 play a role in CAE susceptibility, but that different genes contribute to susceptibility in CAE and JAE; no association was seen in the JAE group. This implies that CAE and JAE cannot necessarily be treated as having the same genetic basis. Although there may be susceptibility genes common to the two (or more) syndromes, treating them as one genetic entity could mask the existence of genes for only one of the syndromes. It is especially interesting that this is at least the second apoptosis gene that appears to be related to a form of IGE. We wish to acknowledge our many collaborators who were critical to this study. Space limitations make it impossible to name them here. (Supported by NIH RO1 NS37466 (to MD) and NIH RO1 NS027941 (to DAG).) 1,6 Miyabi Tanaka, 2 Marco T. Medina, 2 Reyna M. Duron, 3 Ramon Castro, 4 Iris J. Martinez, 1 Machado J. Salas, 4 Maria Elisa Alonso, 1,5 Julia Bailey, 6 Richard W. Olsen, and 1 Antonio V. Delgado‐Escueta ( 1 Comprehensive Epilepsy Center, UCLA & VA GLAHS, Los Angeles, CA ; 2 Neurology, National Autonomous University, Tegucigalpa, Honduras ; 3 University of Sonora, Hermosillo, Mexico ; 4 Neurology, National Instiutute of Neurology & Neurosurgery, Mexico City, Mexico ; 5 Neuropsychiatrics Institute, UCLA, Los Angeles, CA ; and 6 Pharmacology, UCLA, Los Angeles, CA ) Rationale: Because GABRB3 deficient mice shows absence like features including EEG characteristics and pharmacological response because atypical absences in Angelman syndrome correlates with deletions which includes GABRB3, and because Transmission Disequilibrium Test showed possible genetic association between GABRB3 and CAE, we screened for mutations in GABRB3 in families with CAE. We detected two heterozygous missense mutations (P11S, S15F) in the alternative signal peptide, exon1a of GABRB3 segregating with CAE affected members of three Hispanic families. The same missense mutations were not present in 34 other probands with CAE and 440 Hispanic healthy controls Here, we report on the functional consequences of these missense mutations. Methods: Expression constructs: We used mammalian expression vectors, pCMV‐SPORT 6 into which was cloned exon1–9 of GABRB3 and pFP‐N1 (GFP fusion vector). We amplified the mutation containing Exon1a with designed primers with restriction enzyme sites, and replaced Exon1 in pCMV‐SPORT6 with Exon1a with/without mutations. The construct which has the region from exon1a to exon9 without stop codon was amplified from the replaced vector. After digestion, each construct was cloned into pFP‐N1. Immunoblot analysis. HeLa cells were transfected with expression constructs and examined 48 hours post‐transfection and homogenized. Protein samples were run on a 10% Tris‐HCl gel and transferred onto a nitrocellulose filter. After blocking, the filter was processed through sequential incubations with the primary antibody for 1 h, and then with horseradish secondary antibody for 1h. Immunoreactive proteins on the filter were visualized using Typhoon software 9410. Results: Immunoblot studies revealed that both P11S and S15F mutations in exon1a decreased the expression of the protein level when compared with controls. The average densitometric image quantitation was 33%(±0.014) of control for P11S and 69%(±0.14) of control for S15F, normalized to actin density(n = 3). Conclusions: Our results indicate that both P11S and S15F mutations in exon1a decrease expression of protein levels and suggest less trafficking of the nascent peptide from the endoplasmic reticulum to cell membrane. GABRB3 is a main component of GABAR in reticular nucleus of thalamus and expressed transiently in other thalamic nuclei in the perinatal period, decreasing in adult. We speculate that the decreased expression of GABRB3 in developmental brain leads to absence seizures. 1 Simone S. Tsuneda, 1 Fabio R. Torres, 2 Maria A. Montenegro, 2 Marilisa Guerreiro, 2 Fernando Cendes, and 1 Iscia T. Lopes‐Cendes ( 1 Medical Genetics, UNICAMP, Campinas, SP, Brazil ; and 2 Neurology, UNICAMP, Campinas, SP, Brazil ) Rationale: Recent studies have demonstrated that mutations in FLN1 gene are responsible for bilateral periventricular nodular heterotopia (BPNH). We identified a novel mutation (1159G>C), in a family segregating BPNH. However, the exact molecular mechanism by which this mutation lead to abnormal FLN1 protein was unclear. The purpose of this study was to investigate the molecular mechanism of this new mutation. Methods: Total RNA was obtained from peripheral blood samples, from 2 BPNH patients (mother and daughter) and control individuals. RT‐PCR was performed by standard techniques and the cDNA was amplified using specific primers spanning the region containing exon 6 and exon7. Amplicons were cloned into pGEM‐T vector, gel‐purified and subsequently sequenced, using SP6 and T7 primers. Results: Analysis of the cDNA amplicon demonstrated a different pattern of electrophoretic migration between patients and controls. The sequencing of these fragments showed that amplicons from individuals with BPNH kept the intronic sequence between exons 6 and 7. Conclusions: Our data clearly showed that the molecular mechanism of the mutation 1159G>C is the abolishment of the exon 6 donor splicing, resulting in an alternative stop codon and, possibly, in a truncated protein. Thus an aminoacid substitution, as suggested previously, is not the mechanism involved in the etiology of BPNH sydrome in the patients analysed. (Supported by FAPESP.) 1 V. Udani, 2 R. Shukla, 1 P. Munot, 1 L. Iyer, 2 S. Prakash, 1 S. Pujar, and 2 G. Chandak ( 1 Child Neurology & Epilepsy, PD Hinduja National Hospital, Mumbai, India ; and 2 Genome Research Group, Centre for Cellular and Molecular Biology (CCMB), Hyderabad, India ) Rationale: Cohen syndrome (MIM 216550) is an autosomal recessive disorder with features of psychomotor retardation, microcephaly, characteristic facial features, hypotonia, joint laxity, progressive retinochoroidal dystrophy, myopia and neutropenia. Characteristic facial features include high‐arched or wave‐shaped eyelids, a short philtrum, thick hair, and prominent incisors. Clinically homogenous in the Finnish population, non‐Finnish patients show considerable variability. Seizures have never been described in Cohen's syndrome. The genetic locus for Cohen syndrome was recently mapped to chromosome 8q and truncating mutations in the COH1 gene were identified. In Finnish patients a founder mutation, C1117fsI1124X has been consistently found while other mutations have been associated with the atypical non‐Finnish Cohen's syndrome. One of us (VPU) observed several clinical features of Cohen's syndrome in Indian patients with cryptogenic epilepsy. A pilot study confirmed eight Cohen‐like features in a group of children with cryptogenic epilepsy when compared with controls. This prompted us to study the COH1 gene in such patients. Methods: We systematically studied patients with cryptogenic epilepsy and a Cohen‐like phenotype for all features of Cohen's syndrome and included only those with at least 5/8 clinical features that we had found to be significant in our previous study. We initially screened for all the reported common mutations and later sequenced all 62 exons of the COH1 gene in this cohort of 24 patients. We also collected DNA from age, sex and ethnically matched controls. Results: We found several novel mutations such as Y413X, A829T and G3407R in the COH1 gene in 8 of 24 patients (33%). Interestingly, we did not find any of the earlier reported mutations including the most common mutation, C1117fsI1142X. The commonest mutation Y413X was identified in 4 patients, of which one was a compound heterozygote with the G3407R mutation. These mutations were not observed in any of the 100 controls. The mutations were mostly observed in the heterozygous condition except in one patient who was homozygous for the G3407R mutation. We could not find any clinical differences in the patients with and without mutations. However the patients with mutated COH1 had a significantly higher number of Cohen's syndome features than those without the mutations. Conclusions: We conclude that novel COH1 mutations are strongly associated with Indian patients of cryptogenic epilepsy with a Cohen – like phenotype. There is however a possibility of additional mutations in the non‐coding regions of the COH1 gene or in other genes.Functional analysis will throw more light on the mechanisms by which these mutations may be causally related to the epilepsy which is a characteristic feature in these patients. (Supported by Research Society, PD Hinduja National Hospital.) 1,3 Kette D. Valente, 2 Monica Varela, 2 Cintia Fridman, 2 Celia Koiffmann, and 4 Maria J. Marques‐Dias ( 1 Neurophysiology Laboratory‐Psychiatry, USP, Sao Paulo, SP, Brazil ; 2 Biology, USP, Sao Paulo, SP, Brazil ; 3 LIM 21 – Psychiatry, USP, Sao Paulo, SP, Brazil ; and 4 Pediatrics, USP, Sao Paulo, SP, Brazil ) Rationale: Angelman syndrome (AS) results from deletion (DEL), uniparental disomy,imprinting anomalies or UBE3A mutations. Patients determined by DEL have a severer phenotype. In 95% of AS patients caused by DEL,two main classes are found: Class I with breakpoints at BP1 (proximal) and BP3 (distal), and Class II with breakpoints at BP2 (proximal) and BP3 (distal).The remaining 5% have the distal breakpoint at BP4 and BP5.We formerly described that BP1‐BP3 patients are more severely affected,on clinical grounds,than BP1‐BP2 patients. The study of electroclinical data in these distinct groups of DEL patients has not been done. Therefore, we aimed to evaluate the importance of breakpoint on the electroclinical phenotype in patients with DEL. Methods: The authors evaluated 18 DEL patients (6 were Class I; 10 were Class II and 2 were Class III and IV). Diagnosis was made by methylation pattern analysis of exon 1 of the SNRPN‐SNURF gene and by microsatellite profiling of loci within and outside the 15q11‐q13 region. We characterized epilepsy by history obtained with a questionnaire including:occurrence of epilepsy;age of onset;seizure type;epilepsy aggravated by fever;severity of epilepsy determined by:daily seizures; disabling/injurious seizures;more than 3 seizure types;status epilepticus (SE) and;history of refractory epilepsy. These data were corroborated by medical records, personal contact with previous physicians, and video‐EEG monitoring. Patients underwent a mean of 2.6 EEGs. Suggestive EEG patterns for AS were classified according to Boyd et al.and were studied as to morphology, duration, occurrence, frequency, amplitude and distribution. Results: Class I had more daily (p 0.0357) and disabling seizures (p 0.0350). They also presented a higher frequency of SE(p 0.0357). Both groups had similar rates of epilepsy aggravated by fever (p 0.5879); however, in Class I these events evolved to SE(p 0.0357). Polytherapy was more frequent in Class I (p 0.0350), associated with the refractoriness observed in this group.No statistical difference was observed as to age of onset, age of seizure control, seizure type and diversity of seizures presented by these patients. Background abnormalities, interictal and ictal epileptiform discharges were similar in both groups. Suggestive EEG patterns of AS were high in both groups. Occurrence of prolonged and frequent runs of these patterns did not show statistical differences. Conclusions: Epilepsy was more severe and refractory to treatment in patients with large deletions. Therefore, deletion is not a homogeneous group and breakpoint is predictive for epilepsy severity in AS, representing an important factor in parents' counseling. On the other hand, EEG patterns of AS seem to be reliable for diagnosis, despite the breakpoint. (Supported by FAPESP.) 1 Patrick Van Bogaert, 2 Regis Azizieh, 1 Alec Aeby, 3 Linda De Meirleir, 4 Florence Christiaens, 2 Julie Desir, and 2 Marc J. Abramowicz ( 1 Pediatric Neurology, ULB‐Erasme Hospital, Brussels, Belgium ; 2 Medical Genetics, ULB‐Erasme Hospital, Brussels, Belgium ; 3 Pediatric Neurology, AZ‐VUB, Brussels, Belgium ; and 4 Pediatric Neurology, UCL‐Saint Luc, Brussels, Belgium ) Rationale: Rare, mendelian forms of epilepsy amenable to positional cloning help understand the molecular mechanisms of epilepsy. Methods: Three female patients from two consanguinity loops in a large inbred Morroccan family were investigated for progressive myoclonic epilepsy (PME). After exclusion of known causes, linkage study and analysis of candidate genes were performed. Results: Epilepsy started between 18 and 24 months of age after normal psychomotor development. Seizure types were multifocal myoclonus affecting limbs and face and aggravated by movements, and febrile generalized tonic‐clonic in one patient. EEG showed slow dysrythmia, multifocal epileptiform discharges without reproducible temporal relationship with myoclonus and occasional generalized epileptiform discharges. Photosensibility was present in one patient. Giant somatosensory evoked potentials were not found. Cerebral MR imaging and fundoscopy were normal. In 2 patients, seizures were refractory to anti‐epileptic drugs. Evolution was unfavourable in both of them: one patient died from septic shock in the course of a corticosteroids trial, and the other one, now aged 9 years, showed progressive dysarthria, dementia and ataxia with lost independent walking at age 7. Epileptic seizures were controlled under lamotrigine and clonazepam in the third patient, now aged 16 years, who presented profound mental retardation and dysarthria without ataxia. GeneChip analysis of 11K SNPs showed a 15 Mb homozygous haplotype common to the 3 patients. Further analysis using microsatellite markers confirmed linkage of the disease to a 5 cM pericentromeric region of chromosome 7, with a maximum multipoint LOD of 4.2 at D7S663. In silico inspection showed the gene for a Potassium Channel Tetramerization Domain homolog, KCTD7, close to D7S663 within the segment. KCTD7 contains a BTB/POZ‐type domain and has strong homology with the T1 domain of the voltage‐gated potassium channels. Direct sequencing of KCTD7 revealed a C to T transition creating a stop codon in exon 2, homozygous in patients and heterozygous in parents. Conclusions: PME in our patients was associated with a mutation of the KCTD7 gene encoding a putative potassium channel tetramerization domain, consistent with a defect of membrane repolarization. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Epilepsia Wiley

Monday, December 4, 2006 Poster Session IV 7:30 a.m. – 4:30 p.m.

Epilepsia , Volume 47 – Oct 1, 2006

Loading next page...
 
/lp/wiley/monday-december-4-2006-poster-session-iv-7-30-a-m-4-30-p-m-KVK6SocWVT

References (0)

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Wiley
Copyright
Copyright © 2006 Wiley Subscription Services, Inc., A Wiley Company
ISSN
0013-9580
eISSN
1528-1167
DOI
10.1111/j.1528-1167.2006.00001_8.x
Publisher site
See Article on Publisher Site

Abstract

1 Véronique M. André, 1 Carlos Cepeda, 1,3 Vinters V. Harry, 1,2 Mathern W. Gary, and 1 Levine S. Michael ( 1 Mental Retardation Research Center, David Geffen School of Medicine at UCLA, Los Angeles, CA ; 2 Division of Neurosurgery ; and 3 Department of Neuropathology, UCLA, Los Angeles, CA ) Rationale: Based on histopathology, cortical dysplasias (CD) are classified into mild CD if the cortex shows architectural abnormalities only and severe CD if the cortex also presents cytomegalic neurons and balloon cells in addition to architectural changes. Both types induce intractable seizures in children. We showed that GABAergic cell numbers and GABA terminals were altered in severe CD but not in mild CD. GABA peak currents, densities and desensitization time constants were also differentially altered in mild and severe CD, suggesting that mechanisms of seizure induction might be different in those two types of pediatric CD (Andre et al., Epilepsia 46 Suppl. 8: 5, 2005). The present study characterized further postsynaptic GABAA receptor function in non‐CD, mild and severe CD human brain. Methods: Cortical samples resected for the treatment of pharmaco‐resistant epilepsy were collected from non‐CD (n = 10), mild (n = 6) and severe CD (n = 10) patients. Brain tissue slices were acutely dissociated and electrophysiological recordings were performed on isolated pyramidal neurons. Patch electrodes were filled with N‐methyl‐D‐glucamine for whole‐cell voltage clamp recordings. Different concentrations of GABA and GABAA receptor modulators were applied to the cells. Results: EC50 values were similar in non‐CD and severe CD but higher in mild CD cells suggesting less sensitive GABAA receptors. Zinc reversibly decreased GABA peak currents in all cells. However, severe CD cells had a smaller sensitivity to zinc compared to non‐CD cells. Zolpidem, a benzodiazepine (BZ) of type I that binds specifically to α1 subunits, reversibly enhanced GABA peak currents in all cells. Zolpidem sensitivity was significantly smaller in severe CD cells compared to mild and non‐CD cells. Conclusions: GABA sensitivity determined by EC50 values was similar in non‐CD and severe CD cells. However, GABAA receptors responded differently to modulators in severe CD vs non‐CD. Sensitivity to zinc was smaller in severe CD, suggesting higher expression of γ2 subunits. Zolpidem effect was also decreased in severe CD vs non‐CD indicating a decreased expression of α1. As α1 and γ2 subunits are part of the BZ binding site, alteration in their expression is likely to change the efficacy of BZ used to treat seizures in severe CD. In mild CD, effects of modulators binding to α1‐ and γ2‐containing receptors were not different from non‐CD indicating those subunits are not altered. However, GABA sensitivity was lower in mild CD, which could implicate a loss of inhibition and lead to seizures. Together, these results indicate that GABAA receptor subunit composition is differentially altered in mild and severe CD and suggest that different drugs targeting GABA function might be needed to treat the two types of CD. (Supported by NIH NS38992.) 1 Karin Boer, 1 Sandra Redeker, 2 Wim G.M. Spliet, 3 Peter C. Van Rijen, 1 Dirk Troost, and 1 Eleonora Aronica ( 1 Department of (Neuro) Pathology, Academic Medical Center, University of Amsterdam, Amsterdam, Amsterdam, Netherlands ; 2 Department of Pathology, University Medical Center Utrecht, Utrecht, Utrecht, Netherlands ; and 3 Department of Neurosurgery/Rudolf Magnus Institute for Neuroscience, University Medical Center Utrecht, Utrecht, Utrecht, Netherlands ) Rationale: Developmental glioneuronal lesions, such as gangliogliomas (GG) are increasingly recognized causes of chronic pharmaco‐resistant epilepsy. The cellular mechanism(s) underlying the epileptogenicity of GG remain largely unknown. It has been postulated that chronic epilepsy in patients with malformations of cortical development is associated with dysfunction of the inhibitory GABAergic system. We aimed to identify the subtypes of interneurons present within GG specimens and the expression and cellular distribution patterns of GABA receptors (GABAR) and GABA transporter 1 (GAT1). The expression of the various components of the GABAergic system were also analysed in the perilesional cortex. Methods: We investigated immunocytochemically the expression of parvalbumin, calbindin, calretinin, neuropeptide Y (NPY), GABAAR (a1 subunit), GABAB (R1 and R2) and GAT‐1 in 30 specimens of GG obtained during epilepsy surgery, including 10 cases with sufficient amount of both perilesional and adjacent normal control cortex. Results: Immunocytochemistry for parvalbumin, calbindin, calretinin, neuropeptide Y demonstrate the presence of inhibitory neurons of different subtypes within the GG specimens. Most interneurons were NPY positive, whereas fewer reacted to calretinin, calbindin and parvalbumin. Both GABAAR and GABABR (R1 and R2) subtypes were detected within the neuronal component of GG specimens. In addition, GABABR2 immunoreactivity (IR) was observed in glial cells. GG specimens displayed also expression of GAT‐1 IR. Compared to normal cortex, reduced perilesional IR for parvalbumin was observed in 70%, for calbindin in 60% and for GABAR and GAT‐1 in 30% of the specimens. Increased staining was detected for NPY in 80% of the cases. Conclusions: The cellular distribution of components of the GABAergic system in GG, together with the perilesional changes suggest that alterations of the GABAergic system may contribute to the complex abnormal functional network of these highly epileptogenic developmental lesions. (Supported by National Epilepsy Fund –“Power of the Small”/Hersenstichting Nederland (NEF 02–10; NEF 05–11, E. Aronica and K. Boer) and by the Epilepsy Institute of the Netherlands (Heemstede, The Netherlands; E. Aronica).) 1 Michael L. Bell, 2 Jeffery R. Buchhalter, 3 Eugene B. Morris, and 4 Joseph E. Parisi ( 1 Neurology, Mayo Clinic, Rochester, MN ; 2 Child and Adolescent Neurology, Mayo Clinic, Rochester ; 3 Division of Neuro‐Oncology, St. Jude Children's Hospital, Memphis, TN ; and 4 Pathology, Mayo Clinic, Rochester ) Rationale: Epilepsy surgery outcomes may differ among the focal malformations of cortical development (MCDs). Specifically, most pathologic studies have suggested that balloon cells portend a better prognosis compared to other “focal cortical dysplasias.” This could impact surgical decision making, since there are several MRI features highly suggestive of balloon cells. For example, focal cortical thickening with T2 signal abnormality extending towards the ventricle has been reported to be 94% specific for balloon cell pathology. We have previously proposed a categorization scheme for focal malformations of cortical development, discarding the term “focal cortical dysplasia” since these lesions are not precancerous. This study correlates the histopathologic type of focal malformation with epilepsy surgery outcomes. Outcome of each focal MCD category Pathologic subtype Engel class I outcome (free of disabling seizures) Architectural disorganization 1 of 6 (17%) Cytoarchitectural dysmorphism without balloon cells 9 of 12 (75%) Cytoarchitectural dysmorphism with balloon cells 12 of 20 (60%) Polymicrogyria 4 of 8 (50%) Indeterminate 0 of 1 (0%) p = 0.12 Methods: Pathologic specimens were retrospectively reviewed from 49 consecutive focal MCDs resections at Mayo Clinic Rochester. These cases were all removed for intractable epilepsy. Specimens were systematically classified using microscopy and standardized immunohistochemical techniques, according to our recently reported classification scheme. Clinical outcomes were compared between different histopathologic subtypes. Results: Of the 49 cases, three were excluded since there was no post‐operative follow‐up. Of the remaining 46, 26 patients (57%) remain free of disabling seizures (Engel class I). Average follow‐up was 34 months. Patients with cytoarchitectural dysmorphism (with or without balloon cells) trended towards better outcomes compared to architectural disorganization and polymicrogyria. However, balloon cells did not portend the best prognosis. Only 60% of patients with balloon cells (“cytoarchitectural dysmorphism with balloon cells”) were free of disabling seizures compared to 75% of those with cytologically abnormal neurons and architectural disorganization without balloon cells (“cytoarchitectural dysmorphism without balloon cells”). Conclusions: Epilepsy surgical outcomes remain suboptimal in focal MCDs. Other focal lesions causing epilepsy, such as cavernomas or mesial temporal sclerosis, are associated with excellent outcomes in up to 90% of patients. In contrast, we found that only 57% of patients with focal MCDs remained seizure‐free. In our cohort, balloon cells did not portend a better prognosis than other focal MCDs. 1 Kristina A. Fenoglio, 1 Timothy A. Simeone, 2 Frank Schottler, 1 Do Young Kim, 1 Harold Rekate, 1 Jack Kerrigan, and 1 Jong M. Rho ( 1 Barrow Neurological Institute, St Joseph's Hospital & Medical Center, Phoenix, AZ ; and 2 Alpha Med Sciences Co., LTD, Osaka, Japan ) Rationale: The hypothalamic hamartoma (HH) represents a rare but important model of subcortical epileptogenesis. Clinical studies, based primarily on intracranial electrode recordings have established that the HH itself is epileptogenic, but the mechanisms are unknown. In the present study, we examined the network properties of surgically‐resected human HH tissue using a planar multielectrode array recording system. Methods: HH tissue was obtained with patient consent and upon resection was immediately submerged in ACSF bubbled with 95% O2/5% CO2. Tissue slices (400 μm) were placed in a microelectrode dish (Alpha Med Systems, Osaka, Japan) and perfused with warmed (35°C) oxygenated ACSF. The electrodes were arranged in an 8 × 8 grid with 150 μm separation. Slices were arbitrarily placed over the entire electrode grid. Results: Paired‐pulse stimulation (50 ms interval) to random points within the tissue elicited small negative field potentials and stimulation‐evoked single unit firing within specific regions of the tissue. A 5–20% paired‐pulse depression of the field potential was evident in most areas, and was converted to a similar magnitude of paired‐pulse facilitation by picrotoxin (100 μM) suggesting the presence of both excitatory and inhibitory synaptic components. Furthermore, the stimulation‐evoked single unit firing increased (∼4‐fold) with the second pulse consistent with paired‐pulse facilitation of an excitatory component. Picrotoxin substantially increased evoked unit firing suggesting that endogenous GABAergic neuron activity is present normally, and inhibition or shunting of an excitatory component is prominent. In addition, we recorded spontaneous single unit firing of individual cells from multiple electrodes with frequencies ranging from 1–14 Hz and spontaneous negative slow wave activity with amplitudes ranging from 5–300 mV. Interestingly, the frequency of spontaneous slow wave activity decreased with application of picrotoxin and increased with muscimol (30 μM). Conclusions: Our observations indicate that HH tissue is composed of neurons that are strongly modulated by GABAA receptor‐mediated mechanisms. This is consistent with earlier observations demonstrating positive immunoreactivity of HH tissue to glutamic acid decarboxylase (GAD65/67). A novel finding in this study is that populations of neurons also give rise to spontaneous slow wave discharges. It is of interest to determine if the spontaneous activity will become synchronous across the tissue and develop into epileptiform activity in response to provocation. Future studies will characterize the spontaneous activity and the functional architecture of HH tissue. (Supported by NIH and the Barrow Neurological Foundation.) 1 Jan A. Gorter, 2 Karin Boer, 1 Erwin Van Vliet, 2 Sandra Redeker, 3 Fernando H. Lopes da Silva, 3 Wytse J. Wadman, and 2 Eleonora Aronica ( 1 Swammerdam Institute for Life Sciences/Stichting Epilepsie Instellingen Nederland, University of Amsterdam, Amsterdam, Amsterdam, Netherlands ; 2 Department of (Neuro)Pathology, Academic Medical Center, University of Amsterdam, Amsterdam, Amsterdam, Netherlands ; and 3 Swammerdam Institute for Life Sciences, University of Amsterdam, Amsterdam, Amsterdam, Netherlands ) Rationale: The complement cascade has been suggested to be involved in the inflammatory responses observed in a variety of disorders of the central nervous system. Recent evidence suggests that complement activation may contribute to seizure activity and promote neurodegeneration. We investigated the possible involvement of the complement cascade during epileptogenesis in a rat model of temporal lobe epilepsy (TLE) and in the chronic phase of spontaneous seizures in both experimental as well as human TLE. Methods: RNA was obtained from three different brain regions (CA3, entorhinal cortex and cerebellum) at three different time points (1 day = acute phase; and 1 week = latent phase; 3–4 months = chronic epileptic phase) after electrically induced status epilepticus (SE) in rat. Gene expression analysis was performed using the Affymetrix Gene Chip System (230A). The immunohistochemical pattern of C1q, C3c and C3d was assessed in human hippocampal specimens from controls and medically intractable TLE with or without hippocampal sclerosis (HS). Results: Genes involved in the complement pathway (e.g. C1qa, C1qc, C3, C4a, Cfh) were upregulated after induction of SE in rat and remained elevated in the chronic phase. C1qa and C1qc were already induced in the acute phase. C3, C4a and C8b were induced in the latent phase and showed the highest expression at the end of the latent period. In human TLE with HS, glial expression of C1q, C3c and C3d was observed particularly within regions where neuronal cell loss occurs (e.g. CA1 and hilus). Conclusions: Our data indicate that within the immune response the complement activation represents a prominent process during epileptogenesis. The persistence of complement activation is supported by the data on human HS specimen showing increased expression of complement factors in activated glial cells. Although the complement system may be useful in eliminating aggregated and harmful proteins associated with epilepsy, abnormal activation of the complement system can also have damaging effects through activation of inflammatory processes. (Supported by National Epilepsy Fund –“Power of the Small”/Hersenstichting Nederland (NEF 02–10; NEF 05–11, E. Aronica and K. Boer; 03–03 J.A. Gorter).) 1 Stephen T. Hantus, 1 Dileep Nair, 1 Richard Burgess, and 1 Imad Najm ( 1 Neurology, Cleveland Clinic, Cleveland, OH ) Rationale: One third of epilepsy patients are refractory to pharmacologic treatment. Malformations of cortical development (MCD) are a well‐recognized cause of refractory epilepsy that can often be cured by surgical resection. The presence of periventricular heterotopia (PVH) is associated with the development of epilepsy, although the role of these lesions in the generation of epileptogenicity and their cellular/molecular characteristics remain unclear. Methods: In this report, we present the histological and immunocytochemical (ICC) characteristics of a periventricular nodule that was resected from a 17 years old female with a history of pharmacoresistant epilepsy, a lateral temporal MCD and multiple PVH. Invasive video electrographic encephalography (VEEG) with subdural electrodes and a depth electrode targeted at the PVH did not confirm the PVH as the seizure onset zone. The patient underwent a temporal lobectomy that included the removal of a subcortical nodule. She remained seizure free on no antiepileptics for four years, then presented with a single generalized seizure. Results: Histological and ICC examinations of the resected nodule showed dysmorphic neurons lacking any laminar or columnar organizations. These neurons were intermixed with GFAP positive glial cells, immature neurons (TUJ1 positive) and immature astrocytes (vimentin positive). Some subcortical neurons expressed doublecortin, a marker of migrating cells. The expression of GABA (Calbindin, parvalbumin) and NMDA (NR1, NR2A, NR2B) receptors were also examined. Conclusions: Our results suggest that PVH is constituted of a mixture of neuronal and glial cells at various levels of differentiation that failed to mature and migrate to their final position in the neocortex. It is unclear if these nodules are epileptogenic. Further studies of the role of PVH in epilepsy are needed in order to further characterize the electrophysiology and molecular biology of the constituent cells of these lesions and their overlying neocortical areas. 1 David C. Henshall, 2 Clara K. Schindler, 2 Akitaka Yamamoto, 1 Carmen Bellver‐Estelles, 1 Brona M. Murphy, 1 Seiji Hatazaki, 2 Norman K. So, 2 Roger P. Simon, and 2 Robert Meller ( 1 Physiology & Medical Physics, Royal College of Surgeons In Ireland, Dublin, Ireland ; 2 RS Dow Neurobiology Laboratories, Legacy Research, Portland, OR ; and 3 Oregon Comprehensive Epilepsy Program, Neurological Sciences Center, Portland, OR ) Rationale: The role of tumor necrosis factor receptor 1 (TNFR1) signaling in temporal lobe epilepsy (TLE) is incompletely understood. Experimental modeling suggests seizures activate TNFR1 and downstream pro‐ and anti‐apoptotic signaling cascades, although little is known about whether these pathways function in human TLE brain. Methods: We examined TNFR1 signaling in hippocampus surgically obtained from refractory TLE patients (n = 10), and compared results to matched autopsy controls (n = 6). TNFR1 signaling was also examined in hippocampus obtained from mice following focally‐evoked limbic seizures. Samples were processed for detection of protein in whole cell lysates or subcellular fractions, immunoprecipitation studies and immunohistochemistry. Results: Western blotting established total protein levels of the TNFR1 proximal signaling adaptor TNFR‐associated protein with death domain (TRADD) and cleaved caspase‐8 were higher in TLE samples than matched autopsy controls. Intracellular distribution analyses determined cytoplasmic levels of TNFR1, TRADD and the caspase‐8 recruitment adaptor Fas‐associated protein with death domain (FADD), and microsomal levels of TRADD, FADD and cleaved caspase‐8, were raised in TLE samples. Immunoprecipitation studies, supported by fluorescence microscopy revealed increased TNFR1‐TRADD and TRADD‐FADD association in TLE samples. Expression and localization of these proteins was similar in mouse hippocampus. Seizures evoked by intraamygdala kainic acid increased microsomal levels of TNFR1 signaling components and triggered TNFR1‐TRADD and TRADD‐FADD complex formation and caspase‐8 processing. Conclusions: These data show seizures in mice activate hippocampal TNFR1 signaling and reveal this pathway to be engaged in patients with temporal lobe epilepsy. Accordingly, the TNFR1 pathway may be a potential target for adjunctive neuroprotective therapy following status epilepticus or in the setting of recurrent epileptic seizures. (Supported by NIH/NINDS grants NS39016 and NS41935, Science Foundation Ireland, Health Research Board, Marie Curie Foundation and the Program for Human Genomics, PRTLI & Higher Education Authority, Ireland.) 1 Ludmyla Kandratavicius, 1 Jaime Eduardo Hallak, and 1 Joao Pereira Leite ( 1 Neurology, FMRP‐USP, Ribeirao Preto, Sao Paulo, Brazil ) Rationale: Heat shock protein 70 (HSP70) is a stress inducible protein. In animal models of epilepsy, increased HSP70 expression is proportional to seizure severity. HSP70 neuroprotective mechanisms are not yet elucidated, but HSP70 can attenuate intracellular calcium influx and diminish seizure‐induced apoptosis. HSP90 has received little attention in epilepsy research. HSP90 interacts with calcium and calmodulin, and regulates nitric oxide synthase, proteasome, Tau, actine and tubulin, among others. There are few studies about the role of HSP70 in epilepsy and so far nothing regarding to HSP90. Due to the probable HSP70 protective role and the HSP90 relation with key‐factor proteins in epileptogenesis, we looked at HSP70 and HSP90 immunohistochemical expression in temporal lobe epilepsy (TLE). Methods: 47 hippocampi were obtained from medically intractable TLE patients: 16 TLE, 18 TLE+depression and 13 TLE+psychosis. 6 control hippocampi were from necropsy cases without past history of neurological disorders, with pos‐mortem delay no longer than 4 hours. Specimens were equally treated and submitted to imunohistochemistry to HSP70 and HSP90. Hippocampal regions examined were fascia dentata, hilum, CA4, CA3, CA2, CA1, prosubiculum, subiculum, parahippocampal gyrus and entorhinal cortex. Positive immunoreactivity (puncta) was estimated using the software ImageJ. Results: Analyzes of variancy (ANOVA) of our results showed significant differences. Dunn's post‐hoc tests (p < 0.05) of our ANOVA showed significant lower expression of HSP70 and HSP90 in epileptic patients when compared to controls in all hippocampal regions, except fascia dentata and subiculum. Regarding HSP70, the same occurred in all subicular subfields. As to HSP90, there are also differences between TLE+depression and TLE+psychosis at hilum, prosubiculum and hippocampal gyrus. TLE+psychosis always showed values closed to controls. Conclusions: Unlike the reports in animal models the present results indicate that chronic seizures in TLE patients are not sufficient to induce HSP70 and HSP90 activation. Neuronal loss, gliosis and synaptic reorganization, which are typical attributes of TLE, may be determinants of low HSP expression. Inherent to our findings and conclusions, it is important to notice methodological differences between experimental and control groups. Although an effort was made to include only necropsy patients with short post mortem delay, the cause of patients death could be, at least partially, responsible to some degree of HSP activation. Nevertheless, the neurological complications on this group led us to suggest that low expression of HSPs in epileptic groups is related to seizure maintenance. We found no explanation for the differential HSP90 expression in TLE patients with psychosis. (Supported by Fapesp (CInAPCe project # 05/56447–7), CNPq and FAEPA.) 1 Orfa Y. Galvis‐Alonso, 1,2 Jose E. Peixoto‐Santos, 3 David Jr. Araujo, 3 Renata Caldo‐Standiuzzi, and 3 Joao P. Leite ( 1 Molecular Biology, Medicine School of Sao Jose do Rio Preto (FAMERP), Sao Jose do Rio Preto, Sao Paulo, Brazil ; 2 Biology, University of Sao Paulo State (UNESP), Sao Jose do Rio Preto, Sao Paulo, Brazil ; and 3 Neurology, University of Sao Paulo at Ribeirao Preto (USP), Ribeirao Preto, Sao Paulo, Brazil ) Rationale: Patients with temporal lobe epilepsy (TLE) usually present hippocampal sclerosis that can be detected by magnetic resonance imaging (MRI). Most of the resected hippocampi of patients with TLE resistant to pharmacological treatment are characterized by neuronal loss and abnormal mossy fiber sprouting in the inner molecular layer of the fascia dentata. The main goal of this work was to study neuron density and mossy fiber distribution in the hippocampi of TLE patients with different MRI volumetric hippocampal profiles. Methods: Using MRI hippocampal volumetry, patients were classified as: 1) normal volume (NV; n = 8; hippocampal volume with at least 2.5 cm3); 2) mild atrophy (MA; n = 9; hippocampal volume between 2 and 2.5 cm3); and 3) severe atrophy (SA; n = 10; hippocampal volume with less than 2 cm3). Corresponding hippocampal histological sections were stained with cresylecht violet or hematoxylin/eosin for hippocampal cell counts and neo‐Timm histochemistry for mossy fiber detection. Histological controls were autopsy patients (n = 13; without antemortem clinical seizures or evidence of brain pathology on postmortem examination). Results: 1) Compared to autopsies, all three TLE groups showed loss of granular cells (NV: 44%; MA: 43%; SA: 48%) and CA1 (NV: 83%; MA: 69%; SA: 72%), CA2 (NV: 39%; MA: 39%; SA: 47%); CA3a (NV: 64%; MA: 70%; SA: 62%) and CA4 (NV: 55%; MA: 63%; SA: 64%) pyramidal cells, p < 0.001; 2) all TLE groups presented higher neo‐Timm staining in the inner molecular layer of the fascia dentata than their autopsy controls (gray level: Control: 7,6 ± 4; NV: 72,6 ± 8; MA: 90,8 ± 10,8; SA: 95,5 ± 8; p < 0.001). Conclusions: Atrophy profile observed on MRI does not necessarily reflect either cell loss intensity or axonal reorganization in patients with TLE. (Supported by BAP‐FAMERP; FAPESP – CInAPCe Project # 05/56447–7; CNPq.) 1 Mireille Lerner‐Natoli, 1,2 Valerie Rigau, 1 Melanie Morin, 1,3 Arielle Crespel, 1 Marie‐Claude Rousset, 3 Philippe Coubes, 1,3 Michel Baldy‐Moulinier, and 1 Joel Bockaert ( 1 Neurobiology, Institute of Functional Genomics, CNRS, INSERM, University of Montpellier, Montpellier, France ; 2 Unit of Pathological Anatomy and Cytology, CHU Gui de Chauliac, Montpellier, France ; and 3 Epilepsy Unit, CHU Gui de Chauliac, Montpellier, France ) Rationale: In adult patients with intractable temporal lobe epilepsy (TLE), we previously described inflammatory processes and neurogenesis; unexpectedly, we observed the presence of immature micro‐vessels. The present study was designed to demonstrate: i) a pathological vascularization associated with TLE, ii) angiogenic processes active in chronic focus, iii) an impact of vascular remodeling on epileptogenicity. Methods: Hippocampi were collected from 40 patients after surgery for intractable TLE with various etiologies and from 5 non epileptic (NE) patients (autopsies or hippocampectomy for parahippocampal tumor). By immunohistochemistry or immunoblot, we evaluated: microvessel density, expression of markers of immature endothelial cells, of angiogenic factors and their receptors. We also checked impairments of blood brain barrier (BBB). In parallel, we studied the kinetics of these processes in the rat model of lithium‐pilocarpine‐induced limbic epilepsy. Results: We observed division markers in vascular cells in neurogenic areas of epileptic hippocampi. The microvessel density was significantly higher in TLE patients than in NE patients. More, there was no quantitative differences between TLE with hippocampal sclerosis and TLE with other etiologies. Most of these microvessels were positive for markers of immature endothelial cells. The angiogenic factor VEGF was expressed mainly by neurons and often by astrocytes, and its receptor Flk‐1 was localized on thin vascular sprouts. In all cases of TLE, we observed impairments of BBB, allowing IgG leakage and blood cell extravasation. In rats, the degree of vascularization increased progressively from 1 week after status epilepticus. VEGF was rapidly and strongly overexpressed after seizures, whereas the other markers appeared during the silent period. The BBB was early damaged. In the chronic stage, these phenomena were still obvious, as for human TLE. Conclusions: For the first time, we demonstrate a neo‐vascularization of the TLE focus, independently of etiology. We suggest that the pathological angiogenic processes are related to the persistent inflammation and are maintained by recurrent seizures. The consequences of this vascular remodeling may concern hemodynamics, inflammation, neurotoxicity, BBB permeability and drug refractoriness. (figure 1) (Supported by French Fondation for Research on Epilepsy.) 1 Michael Majores, 1 Susanne Schoch, 2 Matthias Simon, 1 Gudrun Engels, and 1 Albert J. Becker ( 1 Department of Neuropathology, University of Bonn Medical Center, Bonn, Germany ; and 2 Department of Neurosurgery, University of Bonn Medical Center, Bonn ) Rationale: Focal cortical dysplasias (FCD) are characterized by a localized malformation of the neocortex and the underlying white matter. Balloon cells, similar to those observed in tuberous sclerosis (TSC) are present in a significant number of cases (FCDIIb). Histopathological similarities indicate that FCDIIb may be pathogenetically related to the tuberous sclerosis complex (TSC), caused by mutations in either TSC1 or TSC2, that normally constitute a tumor suppressor mechanism. Accumulation of coding allelic variants affecting exons 5 and 17 of TSC1 as well as loss of heterozygosity of the second TSC1 allele have been observed in many FCDIIb patients (Becker et al., Ann Neurol 2002). Here, we have studied the potential functional relevance of coding allelic variants of TSC1 found in FCDIIb with respect to the interaction of the TSC1/TSC2 complex. Methods: Sequence variants of the TSC1 allele found in FCDIIb patients were introduced in the full‐length TSC1 cDNA (kindly provided by Dr. M. Nellist, Amsterdam) by site‐directed mutagenesis. We have further generated TSC1 cDNA variants with mutations found in TSC patients, i.e. by introducing premature stop codons. Protein interaction assays were performed to analyze a potentially compromised interaction between the TSC1/2 proteins hamartin and tuberin. Results: TSC1 sequence variants were successfully introduced by site‐directed mutagenesis. Compromised interaction of the hamartin‐tuberin complex was most pronounced in TSC1 mutants bearing the premature stop codon at position R692X (p < 0.01). Sequence alterations of TSC1 present in FCDIIb (H732Y) also resulted in significantly compromised TSC1/2‐interaction compared to wild‐type TSC1 (p < 0.05). The latter was comparable to findings obtained by the expression of the TSC‐determining allele variant R786X. Conclusions: Our data suggest, that allelic variants with amino acid exchange of the TSC1 gene in FCDIIb may be associated with compromised TSC1/TSC2 interaction. The functional relevance is currently studied in cell culture and animal systems. (Supported by DFG (SFB TR3, Emmy‐Noether‐Program), BONFOR, the BMBF German Israeli program and the Deutsche Krebshilfe.) 1 Shailaja Srinath, 2 Richard Prayson, 3 William Bingaman, and 1 Prakash Kotagal ( 1 Department of Neurology, Cleveland Clinic Foundation, Cleveland, OH ; 2 Department of Neuropathology, Cleveland Clinic Founation, Cleveland, OH ; and 3 Department of Neurosurgery, Cleveland Clinic Foundation, Cleveland, OH ) Rationale: The cause of seizures in patients with arachnoid cyst (AC) is thought to be secondary to the compression of adjacent cortex, associated cortical dysplasia (CD) or subpial gliosis according to anecdotal reports in the literature. The aim of our study is to systematically review the pathology around the cyst in patients with epilepsy treated with AC and cortical resection. Methods: We retrospectively reviewed patient data from epilepsy surgeries at the Cleveland Clinic Foundation between 1978 and 2006. Only those patients treated with resection of cyst and adjacent cortex were included in the study. Patients' records were reviewed for demographic data, follow‐up duration, EEG, MRI or CT imaging, and histopathology reports. Results: Of 1880 patients, only 10 patients had AC and six were treated with cyst resection. There were five males and one female patient. The locations of the AC were in the left temporal fossa in four, left peri‐rolandic region and right sylvian fissure in one each. Ictal onset zone was localized to the AC region in all patients. The mean duration of epilepsy was 12.2 + 8.8 years and the mean cyst size was 5.7 × 4.2 × 4 cms. Review of pathology showed mild subpial gliosis around the AC in all patients. Four of these showed normal adjacent cortex. However, one patient showed mild cortical dysplasia and another showed hippocampal sclerosis. At follow‐up ranging from 9 ± 0.24 months, all 6 patients were seizure free (one patient had 3 seizures at 2 months and none thereafter). Patients with abnormal cortical pathology were found to have earlier onset, longer duration, and lower frequency of seizures. Conclusions: A cortical abnormality either independently or in association with subpial gliosis appears to be the cause of seizures in patients with AC. Identifying the cortical pathology preoperatively with the help of high‐resolution MRI and a dedicated epilepsy imaging protocol could influence the type and extent of resection required. It is important to have proper orientation of the cyst and adjacent cortex in order to verify the pathological substrate for the seizures. 1 Mary T. Silvia, 3 Alexander R. Judkins, 2 Phillip B. Storm, 1 Robert G. Kalb, and 1 Brenda E. Porter ( 1 Department of Pediatrics and Neurology, Children's Hospital of Philadelphia, Philadelphia, PA ; 2 Department of Neurosurgery ; and 3 Department of Neuropathology ) Rationale: Cortical dysplasia (CD) is the cause of intractable pediatric epilepsy in up to 60% of surgical case series.1 Previously identified histologic features of CD include large dysmorphic neurons and prominent disorganized neurites. We hypothesize that there are differences in dendrite morphology (total dendrite length and branching patterns) between CD tissue and controls. To better understand the cellular morphology of CD, we quantitatively characterized the dendritic tree of dysmorphic neurons from Palmini grade II A and II B dysplasia from children with refractory epilepsy.2 Methods: Tissue from pediatric CD patients was collected at the time of epilepsy surgery and stained using the Golgi‐cox method. We have analyzed dendrite length and branching order from 3 refractory patients. Dendrites attached to the cell body are first order dendrites and those arising from the first order are called second order dendrites, and so on. Branching order is a measure of the neuron's dendrite tree complexity. Dendrite length refers to the summed total of the basilar dendritic tree. Slides were analyzed for total dendrite length, and dendrite order for at least 10 neurons per patient and compared against literature autopsy controls.2,3 Results: Patients 1, 2, and 3 were 11% (461μm ± 74.7 SEM), 30.3% (1271 μm ± 182.3 SEM) and 17.0% (714.8 μm ± 121.5 SEM) of the total dendrite tree length compared to literature controls (4193.4 μm ± 250 SEM). The most complex of our patients neurons branched to the 5th order, compared to the 8th order in controls. The largest difference in branching order occurred in 3rd and 4th branches, with Patients 1, 2, and 3 at 8%, 22%, and 2.7% of the control values, respectively, for number of dendrites per cell that branched to a 4th order. Conclusions: Pediatric CD tissue varied significantly from literature autopsy controls', with shorter total dendrite length per cell and decreased branching, indicating that CD dendrite trees are smaller and simpler in structure. Our findings suggest that the large and disorganized appearance of neurites in type II CD is associated with lack of dendrite organization rather than dysmorphic neurons with enlarged dendrite trees. Our data also suggest the dysmorphic neurons are similar to immature neurons with less complex dendrite trees. We are analyzing autopsy specimens from multiple ages to address this issue. References 1. Neurology. 2003 Aug 12; 61(3)365–8. 2. Neurology. 2004 Mar 23; 62(6 Suppl 3):S2 3. Cereb Cortex. 2001 Jun; 11 (6):558–71. (Supported by: This work was funded by grants from the Epilepsy Foundation of America for Silvia MT and by the CURE Foundation for Porter BE.) 1 Hidenori Sugano, 1 Madoka Nakajima, 1 Akihide Kondo, 1 Ikuko Ogino, 1 Hajime Arai, and 2 Tatsunori Seki ( 1 Neurosurgery, Juntendo University, Hongo Bunkyo‐ku, Tokyo, Japan ; and 2 Anatomy, Juntendo University, Hongo Bunkyo‐ku, Tokyo, Japan ) Rationale: It is known that neurogenesis still continues in the dentate gyrus even after reaching adulthood. Several reports have described the neurogenesis from surgical specimens of temporal lobe epilepsy, not only in an animal epilepsy model. However, the function of neurogenesis remains to elucidate. We herein examined the neurogenesis in the human hippocampus from temporal lobe epilepsy patients by immunohistochemical staining, and compared the neurogenesis between the specimens from hippocampal sclerosis and non‐hippocampal sclerosis. Methods: Six surgical specimens were examined by immunohistochemical staining using NeuroD, PSA‐NCAM, GFAP, Ki‐67. Three cases thus demonstrated hippocampal sclerosis in both magnet resonance imaging and pathological evaluations, and another 3 showed non‐hippocampal sclerosis. The cell counting of positive cells of each type of staining was applied in the ammon horn and the dentate gyrus. We also observed the nature of the cell and axon with positive staining. Results: PSA‐NCAM positive cells were mostly detected in the hilus, and also in the granule cell and CA1 pyramidal cell layers in both the hippocampal sclerosis and non‐hippocampal sclerosis groups. The PSA‐NCAM positive immature neurons were co‐stained with NeuN, but not with either NeuroD or Ki‐67. The PSA‐NCAM positive immature neurons were frequently detected in the non‐hippocampal sclerosis group. Conclusions: The PSA‐NCAM and NeuN co‐stained immature neurons were located in the hilus of human temporal lobe epilepsy. However, the ability of neurogenesis was found to be higher in the non‐hippocampal sclerosis group than in the hippocampal sclerosis group. Figure 1 (figure 1) (Supported by Juntendo University.) 1 Kiyotaka Suwa, 2 Thomas L. Beaumont, 2 Bin Yao, 3 Eishi Asano, 1 Aashit Shah, 4 William J. Kupsky, and 1,2 Jeffrey A. Loeb ( 1 Department of Nuerology, Wayne State University School of Medicine, Detroit, MI ; 2 Center for Molecular Medicine and Genetics ; 3 Department of Pediatrics ; and 4 Department of Pathology ) Rationale: Cortical malformations are one of the most common causes of medically intractable epilepsy. One of the more common malformations is polymicrogyria (PMG), where up to 80% of patients have epilepsy. The underlying mechanisms responsible for this high incidence of epilepsy are not known, nor is it known if epileptic regions of neocortex differ between patients with PMG and patients with non‐lesional neocortical epilepsy. Recently, we used a functional genomic strategy to identify a small group of activity‐depdnent genes are induced in neocortical epileptic foci, regardless of the underlying cause. Here, we hypothesize that both common and as well as unique patterns of gene expression can be identified in children with PMG compared to patients with non‐lesional neocortical epilepsy. Localization of these induced genes can identify neuronal populations involved human epilepsy associated with malformations. Methods: Electrically mapped neocortical tissue was obtained from 8 children with intractable epilepsy (4 with PMG, and 4 with gliosis) who underwent a two‐stage surgical resection. The seizure onset region and a nearby area without spontaneous epileptiform activity as a “control” were compared by histology, real‐time qPCR, and Agilent genome wide microarrays. Genes identified as differentially expressed were characterized further by in situ hybridization to identify populations of activated neurons. Results: Despite a marked difference in tissue histology between PMG patients and those with gliosis, there was a remarkably similar pattern of gene induction at regions of seizure onset and maximal interictal spiking. Patients without malformations expressed these activity‐dependent genes predominantly in normal‐appearing layers II‐IV pyramidal neurons. In patients with PMG, these same genes were also restricted to layer II‐IV neurons, in highly epileptic regions directly adjacent to thickened layer I regions in the polymicrogyri. In addition to common gene expression changes, we also identified genes uniquely altered in PMG neocortex that could underlie some of the morphollgical features of this malformation. Conclusions: Results genomic studies of PMG suggest that similar molecular pathways are activated in PMG as in other non‐lesional forms of neocortical epilepsy, however, the expression of some of these activity‐dependent genes in PMG localize to neurons directly adjacent to malformations. In addition, there are unique patterns of gene expression in PMG that are likely to relate directly to the malformation. (Supported by: 1. NIH NINDS R01 NS45207.; 2. Postdoctoral fellowships from the Center for Molecular Medicine and Genetics, Wayne State University School of Medicine, Detroit, MI, USA; 3.Japan research foundation for clinical pharmacology, Tokyo, Japan.) 1 Delia M. Talos, 2 Joseph R. Madsen, 2 Peter M. Black, and 1 Frances E. Jensen ( 1 Neurology, Children's Hospital/Harvard Medical School, Boston, MA ; and 2 Neurosurgery, Children's Hospital/Harvard Medical School, Boston, MA ) Rationale: The most common neurological symptom of human neocortical dysplasias is early‐life epilepsy, including infantile spasms. Frequently, these patients develop pharmaco‐resistant seizure disorders, which require surgical excision. The exact mechanisms of epileptogenesis and high incidence of pharmaco‐resistance to conventional antiepileptic drugs (AED) in cortical dysplasias are poorly understood. GABA receptors (GABARs), which are normally hyperpolarizing, can become depolarizing under certain circumstances and actually increase neuronal excitation. This depends on the intracellular chloride concentration, regulated by the chloride‐importer NKCC1 and the chloride‐exporter KCC2. Increased NKCC1 and decreased KCC2 expression in hippocampal subiculum from temporal lobe epilepsy (TLE) patients renders GABARs depolarizing (PNAS 103:8465;2006). We hypothesized that similar to TLE, neurons from neocortical dysplastic tissue would demonstrate higher NKCC1 expression and lower KCC2 levels relative to control neurons, a profile consistent with the presence of depolarizing GABARs. Methods: Cortical tissue from 6 patients ages 1–6 years were obtained during brain surgery for drug‐resistant epilepsy. In all cases, malformations of cortical development were confirmed by histopathological examination (tuberous sclerosis, n = 3; focal cortical dysplasia, n = 2; hemimegalencephaly, n = 1). Age‐matched autopsy cortical samples from cases with normal neurologic history were used as controls (n = 4). Tissue was collected and handled in accordance with the Clinical Research Committee at Children's Hospital, Boston. 4% paraformaldehyde‐fixed tissue blocks were cut at 50 microns and immunohistochemically double labeled with cell markers (NeuN, non‐phosphorylated neurofilament, vimentin, GFAP) in combination with transporter antibodies NKCC1 (1:100, Chemicon) and KCC2 (1:500, Upstate). Results: Staining of dysplastic tissue with neuronal and glial markers demonstrated loss of cortical lamination, the presence of disoriented, misshaped neurons, an increase in number of white matter neurons and a variable degree of gliosis. In addition, in cortical tubers and hemimegalencephaly tissue, cytomegalic dysplastic neurons could be easily identified. In all cases, increased NKCC1 expression associated with a moderate decrease in KCC2 levels was observed in most normal‐size and cytomegalic dysplastic neurons, compared to controls. Notably, strong NKCC1 immunoreactivity was observed in glial cell population as well. Conclusions: Our results demonstrate an imbalance in neuronal NKCC1 and KCC2 expression (increased NKCC1: KCC2 ratio) in human neocortical dysplasias that may render GABARs depolarizing. This mechanism may play a critical role in increased seizure susceptibility and relative refractoriness to GABAR agonists in these patients. (Supported by TS Alliance, Boston Neurosurgical Fdn, NIH NS31718.) 1 Cristiane Q. Tilelli, 2 Jean A. Tkach, 2 Yuguo Li, 1 Roger Oghlakian, 2 Mark Pagel, and 1 Imad M. Najm ( 1 Department of Neurology, Section of Adult Epilepsy, Cleveland Clinic, Cleveland, OH ; and 2 Department of Biomedical Engineering, Case Western Reserve University, Cleveland, OH ) Rationale: Magnetic Resonance Imaging (MRI) is a non‐invasive method often used for the assessment of patients with epilepsy. More than a quarter of patients who present with drug‐resistant epilepsy due to cortical dysplasia (CD) show no MRI abnormalities. The role of differential expression of glutamate receptors as a substrate for hyperexcitability in epilepsy has been shown. We intended to develop a non‐invasive and sensitive imaging technique to be used with specific biomarkers for epileptic tissue. Methods: Anti‐NMDA receptor subunit 1 (NR1) antibody (Chemicon AB1516) was labeled with Gd‐DOTA (NR1/Gd‐DOTA) using an activated carboxyl to create an amide coupling. In a first study, ex vivo human brain slice tissue was maintained alive immediately following surgical resection, and optimized T1‐weighted MRI was acquired using a 9.4T MRI scanner. Following baseline MRI, each tissue was treated with NR1/Gd‐DOTA or NR1 without Gd‐DOTA for two hours and washed for one hour to remove non‐specific binding. In a second study, MRI (7.0T) was acquired before and after ex vivo live rat brain slices were treated with NR1/Gd‐DOTA or control treatments (vehicle only, Gd‐DOTA only or anti‐NR1 only). Additional 3‐hour washing followed by MRI was conducted in order to verify signal variation over time. Results: The first study (Figure) demonstrated that optimized high‐field MRI and antibody‐based MRI probes can be used to non‐invasively detect cell receptors within freshly resected brain tissue. The second study validated that MRI contrast was due to specific binding of the labeled antibody to cell receptors. Fading in the signal detected after additional washing revealed a temporal dependence on the amount of antibody probe bound to tissue slices. Conclusions: Our preliminary data show that NR1 protein can be detected in freshly harvested human and rat brain tissue samples using high field MRI techniques. The development and application of new antibody‐based MRI probes that can be selectively detected and quantified will provide a new platform technology for visualizing molecular substrates of epileptogenic brain areas and other pathologies. (figure 1) (Supported by Northeastern Ohio Animal Imaging Resource Center, an NIH funded program #R24CA110943, part of the Case Center for Imaging Research; and R21 NS42354 from NINDS to IN.) 1 W. Saskia Van Der Hel, 2 Peter C. Van Rijen, 2 Cees W.M. Van Veelen, and 1 Pierre N.E. De Graan ( 1 Pharmacology and Anatomy, University Medical Center Utrecht, Utrecht, Netherlands ; and 2 Neurosurgery, University Medical Center Utrecht, Utrecht, Netherlands ) Rationale: Elevated levels of extracellular glutamate have been linked to the initiation and maintenance of seizures in patients with pharmaco‐resistant temporal lobe epilepsy (TLE). Impairment of the glutamate‐glutamine cycle, responsible for the inactivation and recycling of extracellular glutamate, has been implicated.One of the key neuronal components of this cycle is the vesicular glutamate transporter (VGLUT), which mediates the uptake of glutamate into synaptic vesicles. Three types of VGLUTs have been identified so far. VGLUT1 is the major type in the hippocampus. Little is known about VGLUT expression in the human hippocampus. The aim of this study is to investigate changes in VGLUT expression in the hippocampus of patients with TLE. Methods: We examined the hippocampal distribution of the VGLUT subtypes by immunohistochemistry, immunofluorescence and in situ hybridisation in 7 μm paraffin sections. Protein levels were assessed by semi‐quantitative western blotting. Measurements were performed in hippocampal biopsies obtained during surgical resection as treatment for TLE. Comparisons were made between hippocampi of TLE patients with (HS group) and without hippocampal sclerosis (nonHS group) and autopsy controls without neurological symptoms. Results: VGLUT1 immunoreactivity (IR) showed a diffuse neuropil staining of the subiculum, the pyramidal cell layer of all CA fields, the hilus, and the inner molecular layer of the dentate gyrus (DG). Large punctate structures around neuronal cell bodies and axons were observed in CA4 and CA3. No IR was detected in cell bodies. VGLUT1 mRNA was abundant in granule cells and pyramidal neurons. In nonHS and HS hippocampi VGLUT1 mRNA levels were decreased compared to controls, especially in areas with neuron loss. However, VGLUT1 protein levels were significantly increased in the nonHS hippocampus compared to both the HS and control hippocampus. This upregulation indicates translational regulation within hippocampal neurons. In addition, VGLUT1 mRNA and protein were upregulated in the DG of the HS hippocampus, consistant with the increased number of mossy fiber synapses onto granule cells due to mossy fiber sprouting. VGLUT2 IR was very weak in controls, but it was upregulated in the molecular layer of the DG in TLE patients. In the nonHS group VGLUT2 IR was only found in the outer molecular layer, whereas in the HS group VGLUT2 was also present in the inner molecular layer. These data indicate an upregulation of VGLUT2 associated with mossy fiber sprouting. Conclusions: Hippocampal mRNA expression of VGLUT1 is decreased and correlates with loss of glutamatergic synapses due to neuron loss. The increase in VGLUT1 and VGLUT2 protein found in TLE patients is most likely due to new glutamatergic synapse formation. This increase in glutamatergic signalling may contribute to the progression and pathology of TLE. 1 Jie Wu, 1 Jamie DeChon, 2 Nasim Bahadorani, 3 Harold Rekate, 1 John F. Kerrigan, and 2 Yongchang Chang ( 1 Neurology, Barrow Neurological Institute, St. Joseph's Hospital and Medical Center, Phoenix, AZ ; 2 Neurobiology, Barrow Neurological Institute, St. Joseph's Hospital and Medical Center, Phoenix, AZ ; and 3 Neurosurgery, Barrow Neurological Institute, St. Joseph's Hospital and Medical Center, Phoenix, AZ ) Rationale: Human hypothalamic hamartomas (HH) are associated with precocious puberty, gelastic seizures and severe refractory epilepsy. Emerging evidence suggests that the HH itself plays a pivotal role in this syndrome. γ‐aminobutyric acid is the major inhibitory neurotransmitter in the brain and plays an important role in modulation of fast inhibition through GABAA receptors. Abnormal GABAA receptor function is an important mechanism of epileptogenesis in various types of human epilepsy and animal epileptic models. However, it is unknown whether functional GABAA receptors are natively expressed in HH neurons – and if so, what are the functional, pharmacological and molecular properties of these receptors. Methods: Acutely enzymatic/mechanical dissociation of HH neurons was performed. Fresh resected HH tissues were immediately placed in ice‐cold dissection solution containing (in mM): 136.7 NaCl, 5 KCl, 0.1 NaH2PO4, 9.84 HEPES, 16.6 glucose and 21.9 sucrose. After incubation at 22 ± 1 °C for at least 1 h, tissue sections were treated with papain (4–6 mg/ml at 31 °C for 50–60 min), and then one tissue fragment was mechanically dissociated. Isolated single HH cells usually maintained function for 2–6 h. Perforated patch‐clamp whole‐cell recording combined with U‐tube drug application was applied to characterize function and pharmacology of GABAA receptors. Results: Under perforated patch‐clamp recording in voltage‐clamp mode, GABA induced an inward current (IGABA) at a holding potential of –50 mV. The IGABA was mimicked by a GABAA receptor agonist (muscimol) and blocked by GABAA receptor antagonist (bicuculline), suggesting that the IGABA was mainly mediated through the activation of the GABAA receptor. The concentration‐response relationship curve of IGABA showed that the EC50 and Hill coefficient were 1.3 mM and 0.9, respectively. The current‐voltage relationship current was linear at a reversal potential close to zero with a symmetric external and internal chloride concentration. HH neurons exhibited heterogeneity for allosteric modulations. Although most HH neurons presented an enhancement of IGABA by pentobarbital and pregnenalone, only ∼60% of neurons showed potentiation of IGABA by diazepam. RT‐PCR showed normal GABAA receptor subunit expression in HH tissues. Conclusions: We are the first to describe the functional, pharmacological and molecular properties of the GABAA receptor expressed in human HH. An understanding of functional GABAA receptors in HH neurons provides new insights into the epileptogenesis of gelastic seizures. (Supported by Women's Board Fund of the Barrow Neurological Institute and NS056104.) 1 Irene Yamazaki, 1 Véronique M. André, 1 Nanping Wu, 2 Harry V. Vinters, 3 Gary W. Mathern, 1 Michael S. Levine, and 1 Carlos Cepeda ( 1 Mental Retardation Research Center, David Geffen School of Medicine, UCLA ; 2 Department of Neuropathology ; and 3 Division of Neurosurgery, David Geffen School of Medicine, UCLA, Los Angeles, CA ) Rationale: Tuberous Sclerosis Complex (TSC) is an autosomal dominant disease caused by mutations in TSC genes that encode hamartin (TSC1) and tuberin (TSC2). These proteins affect cellular migration, proliferation, and differentiation in multiple organs, including the brain. In particular, it has been suggested that these proteins may represent growth suppressors. Epilepsy is one of the major complications of TSC and affects about 85% of patients. TSC shares histopathologic similarities with severe forms of focal cortical dysplasia (CD), specifically the presence of cytomegalic neurons and balloon cells. Methods: In this study we examined the morphological and electrophysiological properties of cells in cortical tissue samples from pediatric cases with TSC (n = 11, ages 1.8–10.1 yr) or severe CD (n = 15, ages 0.2–14 yr), that underwent surgery for the treatment of pharmacoresistant epilepsy. Normal‐ and abnormal‐appearing cells in slices were identified morphologically by infrared videomicroscopy and biocytin labeling (n = 72 in TSC and n = 151 in severe CD). Electrophysiological membrane properties were assessed with whole‐cell patch clamp recordings in current and voltage clamp modes. Results: In both TSC and severe CD cases, abnormal‐appearing cells were observed in the cortical mantle and classified into four principal groups. The first consisted of very large, pyramidal‐shaped cells corresponding to cytomegalic neurons. The second group included large, non‐pyramidal cells with atypical somatodendritic morphology that correspond to balloon cells. The third group included misoriented and dysmorphic pyramidal neurons, and the fourth consisted of immature‐looking pyramidal cells. Morphologically, abnormal cells from TSC and severe CD cases were very similar. Electrophysiologically, the most abnormal properties were found in cytomegalic neurons and balloon cells. Cytomegalic neurons were hyperexcitable, whereas balloon cells had no active inward currents and lacked synaptic inputs. Balloon cells were found more frequently in TSC cases, whereas cytomegalic pyramidal neurons occurred more frequently in severe CD cases. Conclusions: Besides the frequency of occurrence, there were no clear morphological or electrophysiological differences in cytomegalic neurons and balloon cells from TSC and CD patients, suggesting a link between the lesions produced by TSC gene mutations and the cellular alterations observed in severe forms of CD. (Supported by NIH NS38992.) 1 Andreas V. Alexopoulos, 1 Cristiane Q. Tilelli, 1 Roger Oghlakian, and 1 Imad M. Najm ( 1 Neurology, Cleveland Clinic Epilepsy Center, Cleveland, OH ) Rationale: Synaptic vesicle protein 2A (SV2A) was recently identified as the specific binding site for levetiracetam in the brain and as a promising target for drug development. We have previously reported the differential expression of SV2A in the neocortex of patients with cortical dysplasia (CD)‐associated pharmacoresistant epilepsy (Alexopoulos et al., Epilepsia 46 Suppl. 8:273, 2005). The objective of this study is to investigate the expression of the same protein in a well‐characterized animal model of CD. Methods: Sprague‐Dawley time‐pregnant rats received 145cGy whole body irradiation at embryonic day 17, resulting in pups with multifocal hippocampal (HIP) and cortical abnormalities. These animals exhibit recurrent, spontaneous seizures following “a second hit” in the form of a single dose of pentylenetetrazol (PTZ 30 mg/kg i.p.; postnatal day 60). Occurrence of seizures was confirmed by video‐EEG and behavioral monitoring. Age‐matched, nonirradiated controls received the same subconvulsive dose of PTZ. Video‐EEG recordings were obtained for 1 month before and after PTZ injection in all animals. Brains were examined using Cresyl Violet and immunostaining for SV2A and neuron‐specific nuclear protein. Results: Irradiated rats (n = 8), but not control rats (n = 3) developed spontaneous behavioral and electrographic seizures following PTZ (“second hit”), characterized by extended periods of arrest, facial or forelimb clonus and occasionally rearing and falling. Immunocytochemical (ICC) analysis in the cortex and HIP of non‐irradiated controls showed SV2A to be localized in regions surrounding cell bodies. Punctate staining was observed in HIP areas occupied by mossy fiber terminals. Histological examination of stained slices of irradiated animals confirmed the presence of multifocal CD and neuronal dispersion in HIP. ICC analysis of irradiated epileptic animals displayed a redistribution of the SV2A protein from the cell bodies to the neuropil (figure 1). SV2A in HIP of irradiated rats occupied the area of the hilus and apical dendrites in the dentate gyrus and cornus ammonis. Conclusions: The differential changes in SV2A immunoreactivity following recurrent seizures in dysplastic but not control animals are reminiscent of those observed in human epileptic and dysplastic tissue. Further studies will examine the influence of seizures in the SV2A redistribution, and its potential role in the expression of epileptogenesis and pharmacoresistance. (figure 1) (Supported by Investigator‐initiated study grant to A.V.A. sponsored by UCB, Inc.) 1 Atif Ali, 1 Yashomati Dua, and 1 F. Edward Dudek ( 1 Physiology, School of Medicine, University of Utah, Salt Lake City, UT ) Rationale: Animal models of chronic epilepsy with spontaneous seizures may be useful for antiepileptic (AED) drug development. Convenient and effective methods for testing potential AEDs over days and weeks should facilitate many experiments. This study investigated the effect of carbamazepine (CBZ) administered in food pellets on rats with kainate‐induced epilepsy. Methods: Adult male rats (Sprague‐Dawley; 150–250 g) were given repeated low‐dose kainate treatment (5 mg/kg ip hourly until convulsive status epilepticus persisted >3 h). Food pellets with CBZ were provided (30, 100 or 300 mg/kg/day) in three 2‐week trials (n = 7–9 rats) involving 5 days of CBZ or control treatment, separated by 2 recovery days. The total amount of food provided corresponded to normal caloric diet of 60 gm/kg/day. Results: All animals consumed the entire daily drug‐in‐food dose intermittently but continuously throughout the 24‐h day. CBZ significantly reduced mean seizure frequency in a dose‐dependent manner. At 30 and 100 mg/kg/day, CBZ treatment consistently failed to block all motor seizures, but no motor seizures were observed in any of the rats at 300 mg/kg/day. Conclusions: This AED‐in‐food protocol appears to be an effective means of chronic drug administration. When administered over 5‐day periods, CBZ reduced the frequency of spontaneous motor seizures in a dose‐dependent manner, but very high doses were required to block all motor seizures. The effects on electrographic seizures remain to be determined. (Supported by NS 049620 (FED).) 1 Michael R. Kasten, 1 Sanghoon Lee, and 1 Matthew P. Anderson ( 1 Neurology, Harvard Medical School, Beth Israel Deaconess Medical Center, Boston, MA ) Rationale: Recent studies have found epilepsy‐associated mutations in non‐ion channel genes. LGI1, a secreted protein of unknown function, was found to be mutated in patients with autosomal dominant lateral temporal lobe epilepsy (ADLTE). To investigate the functional effects of this protein on native neural circuits, we developed methods to study wild‐type and mutant LGI1 protein in living brain tissue. Methods: Transgenic mice were created using a full‐length LGI1 gene engineered to contain the epilepsy‐associated LGI1 gene mutations. We also created transgenic mice using the wild‐type, full‐length LGI1 gene. In order to measure the effect of this mutant protein on neuron excitability and synaptic communication, we prepared acute brain slices from adult LGI1 transgenic and control mice and compared the electrophysiological properties of individual neurons. Because ADLTE is characterized by auditory hallucinations during seizures, we studied neurons in a second order thalamic nucleus that relays auditory sensation to the cortex, the dorsal medial geniculate nucleus (DMGn). Results: Mutant LGI1 transgene increased the intrinsic excitability of DMGn thalamic neurons. Mutant LGI1 transgene also increased phosphorylation of a number of protein kinases. Conclusions: These findings support our current hypothesis that mutant LGI1 causes epilepsy through a unique mechanism involving altered signal transduction. (Supported by: Funded by National Institute of Neurologic Disease (MPA), Burrough Wellcome Fund Career Award (MPA), and Beth Israel Deaconess Medical Center.) 1 Ramin Atefy, 1 Sanghoon Lee, 1 Takatoshi Mochizuki, 1 Thomas E. Scammell, 1 Donald L. Schomer, and 1 Matthew P. Anderson ( 1 Neurology, Harvard Medical School, Beth Israel Deaconess Medical Center, Boston, MA ) Rationale: Patients with autosomal dominant lateral temporal epilepsy (ADLTE) with auditory features were recently found to harbor mutations in the non‐ion channel protein LGI1. To determine if mutations of this protein are sufficient to produce epilepsy, we created a transgenic mouse from the full‐length LGI1 gene engineered to contain an epilepsy‐associated gene mutation. As a control, other transgenic mice were also created that carry extra copies of the wild‐type, full‐length LGI1 gene. We predicted that ADLTE mutant LGI1 will produce epilepsy in mice. Methods: Mice of each genotype were implanted with electroencephalogram (EEG) and electromyogram (EMG) electrodes for polysomnogram recording. After > 9 days of recovery from surgery, they were transferred to recording cages in a sound‐attenuated chamber with a 12 hr light/dark (LD) cycle (30 lux; lights on at 7:00 A.M. and off at 7:00 P.M.) and a constant temperature (22–24°C) for 5 days of continuous recording. Video recordings were also performed during the last day in a subset of mice. The signals were digitally filtered (EEG, 0.3–30 Hz; EMG, 2–100 Hz) and semi‐automatically scored in 10 sec epochs as wake, NREM sleep, REM sleep, or seizures. Sleep analyses were focused on the last recording day. The preliminary scoring was visually inspected and corrected when appropriate. Results: Mutant LGI1 transgenic (n = 5), full‐length LGI1 transgenic (n = 4), and wild‐type littermate (n = 5) mice showed no significant differences in the hourly amounts of NREM, REM, or wakefulness. Seizures induced by pentylene‐tetrazol (60 mg/kg, i.p.) were also examined. Conclusions: ADLTE mutant LGI1 transgene does not significantly influence the pattern of sleep. We are currently testing the hypothesis that mutant LGI1 is sufficient to cause epilepsy. (Supported by: Funded by National Institute of Neurologic Disease (MPA), Burrough Wellcome Fund Career Award (MPA), and Beth Israel Deaconess Medical Center.) 1 Paven K. Aujla, 2 Michael Fetell, and 1 Frances Jensen ( 1 Neurology, Children's Hospital and Harvard Medical School, Boston, MA ; and 2 IVAX Pharmaceuticals, Key Biscayne, FL ) Rationale: Hypoxic encephalopathy is a major cause of neonatal seizures, and can lead to long term neurologic deficit and epilepsy. Neonatal seizures can be refractory to conventional AED medication. AMPARs are overexpressed on cortical neurons in the neonatal rat and term human, when the susceptibility to hypoxic seizures is highest (J. Comp. Neur. 2006; 497:42). We have shown that the AMPA subtype of neuronal glutamate receptors in cortex and hippocampus play a critical epileptogenic role in a rodent model of neonatal seizures (J. Neurosci. 2005; 25:3442), and AMPAR antagonists such as NBQX and topiramate effectively suppress the acute and long term epileptogenic effects of hypoxia in the perinatal rodent (Ann. Neurol. 2001; 50:366; Epilepsia 2004; 45:569). Here we evaluated the efficacy of talampanel (GYKI‐53773), a novel orally‐available non‐competitive AMPAR antagonist in our rodent model. Methods: P10 rat pups were exposed to global hypoxia per our previous protocols (J. Neurosci. 2001; 21:8154). Rats were pretreated i.p. with talampanel or vehicle 30 minutes prior to hypoxia and seizure time and severity were recorded by videotape and scored blinded. A second outcome measure was evaluated by assessing susceptibility of rats to KA‐induced seizures 20 days (P30) after they were exposed to hypoxia. KA was administered i.p. (10 mg/kg) and seizure activity was recorded over a 3‐hour period. Rats given KA on P30 were killed on P33 by transcardiac perfusion with 4% paraformaldehyde. Each brain was removed, postfixed, cyroprotected in 20% sucrose overnight, and mounted onto a freezing microtome. To quantify DNA fragmentation, counts of positively stained cells per 12,500μm2 in CA1, CA3 and amygdala were made as previously described (Ann. Neurol. 2001; 50:366). Results: Acute hypoxia‐induced seizures at P10 were suppressed by talampanel in a dose‐related manner. Anticonvulsant activity was maximal at 7.5 and 10 mg/kg, where seizures were blocked 74.6% at 10 mg/kg (25.4 ± 7.3, n = 17; p < 0.001) and 86.7% at 7.5 mg/kg (13.3 ± 3.2, n = 17; p < 0.001). Rats pretreated with vehicle prior to hypoxia showed significantly more neuronal injury in amygdala following KA‐induced seizure at P30 than did nonhypoxic controls or rats pretreated with talampanel (7.5 and 10 mg/kg) prior to hypoxia at P10 (p < 0.001, one‐way ANOVA). Conclusions: Talampanel exhibited a dose‐dependent effect on hypoxia‐induced hypoxic seizures in the immature rat. Futhermore, acute talampanel treatment during hypoxia protected the rats from later life increases in seizure susceptibility that are observed in vehicle treated hypoxic controls. These data suggest talampanel may have clinical potential in the treatment of neonatal seizures. (Supported by Epilepsy Project, IVAX, and NINDS RO1 NS31718.) 1,2 Natacha Porta, 1,2 Stephane Auvin, 1,2 Cecile Lecointe, 2 Regis Bordet, and 1,2 Louis Vallee ( 1 Pediatric Neurology, Lille University Hospital, Lille, France ; and 2 EA 1046, School of Medicine, Lille, France ) Rationale: Peroxisome proliferator‐activated receptor α (PPARα) is one of the three subtypes of the nuclear receptor PPAR family. PPARα activation by fenofibrate can protect cerebral injury by anti‐oxidant and anti‐inflammatory mechanisms. In this study, we tested PPARα activation on the cerebral excitability before study its potent neuroprotective effects in status epilepticus. Methods: Wistar rats (280g‐320g) were separated in two groups (n = 8). Sham group (Sh) received a complete powder diet ad libidum, while experimental group (F) received powder diet containing 0.2% fenofibrate. After 14 days of treatment, pentylenetetrazole threshold (PTZth) was used to assess anti‐convulsive effect. PTZ was dissolved in a saline solution (10 mg/ml) and was infused i.v. at 5 ml/h. Two threshold doses (mg/kg) were studied: the first bilateral forelimb myoclonus and the start of generalized seizure. Liver weights were recorded in each animal to assess that fenofibrate was pharmacologically active. All values are given as mean ± sem. Results: Rats receiving diet with 0.2% Fenofibrate had an increase of liver weight (liver weight/body weight): 0.063 ± 0.004 versus 0.036 ± 0.002 (p = 0.001). Myoclonic and seizure PTZ thresholds were increased in F group in comparison to Sh group: myoclonic PTZth were 40.85 ± 3.06 mg/kg in F group vs. 34.28 ± 1.26 mg/kg (Sh) (p = 0.028); seizure PTZth were 46.85 ± 3.58 mg/kg in F group vs. 37.29 ± 1.8 mg/kg (Sh) (p = 0.021). Conclusions: A treatment with 0.2% fenofibrate significantly increases myoclonic and seizure PTZ thresholds suggesting anticonvulsive proprieties. The role of PPAR activation in brain excitability should be investigated. (Supported by AEAC association and NP was supported by the grant “Conseil Régional Nord‐Pas‐de‐Calais et CHRU de Lille”.) 1,2 Thomas L. Babb, 1 Weili Zhou, and 1 Thomas P. Miller ( 1 Pediatrics, Wayne State University, Detroit, MI ; and 2 Neurology, Wayne State University, Detroit, MI ) Rationale: We used the rat in utero radiation model of cortical dysplasia (CD), which is characterized postnatally (P) by neuronal dyslaminations, polarity disorientations, and daily increases in cortical thickness. At P30 these rats may exhibit cortical seizures from the region of the permanent CD. Because the CD is present at birth and increases daily in size with the growth of neurons and dendrites, we hypothesized that a permanent increase in NMDA receptor densities may develop early in the first week to enhance CD neurons' excitability. Methods: Dams were radiated with 145 Rads (cGy) at embryonic day 17 (E17), their offspring were studied at 24 hour intervals, postnatal ages P0–10 and P21–36. The development of CDs measured neuron sizes and neuron densities in cortical layers II‐V. Densities were compared to age matched controls using stereology. In siblings, expressions of NMDA receptor subunits NR1 and NR2B were measured by immunocytochemistry and Western blotting. Statistics between control and radiated age matched pups were made between treatments and across ages. Results: Daily increases in neuron sizes were found from P0 through P10 in both groups, and after that there were no greater cytoplasmic increases in either group. At P0 and P1 nearly all neurons were spherical or oval in shape, lacking dendritic processes. At P2 neurons developed larger cytoplasms and larger proximal dendrites, which gave the neurons a pyramidal shape in all cortical layers. Growth of dendritic and axonal processes began at P2 and continued after P10. The expression levels for NR1 and NR2B at P0 and P1 were low, but increased significantly by P2; however, this increase plateaued after P2 and through P10. It is important to note that these developmental patterns were the same for both radiated and nonradiated pups. Conclusions: We used the rat in utero radiation model to identify postnatal malformations of cortical neurons and their NMDA receptor subunit expressions. During the earliest postnatal period, P0‐P10 the cortex is expanding with increasing neuron sizes and neuritic outgrowths. Only P0 and P1 had significantly immature neurogenesis, with small, round neurons lacking dendrites and having the lowest NMDA receptor expressions. At P2‐P10 neuron sizes and dendritic lengths increased. Also, at P2 NMDA expressions were maximal, and neither NR1 nor NR2B expressions increased on subsequent days. These developmental patterns in neurons and receptors were the same for control and radiated rats that always had permanent CD starting at birth. Hence, the later (P30) process of epilepsy likely occurs when all the axo‐dendritic synapses, including aberrant excitatory connections within the CD, are fully mature. (Supported by: 1) NIH Grant NS41375; 2) The Frankel Family Endowed Chair in Pediatric Neuroscience.) 1 Mihyun Bae, 1 Melissa Brunckhorst, 4Wendy M. Mars, 4George K. Michalopoulos, 4Cristian Achim, and 1,2,3 Elizabeth M. Powell ( 1 Department of Anatomy and Neurobiology, University of Maryland, School of Medicine, Baltimore, MD ; and 3 Department of Pathology, University of Pittsburgh, School of Medicine, Pittsburgh, PA ) Rationale: Inhibitory neurotransmission functions in neural circuits by repressing of excitatory output. Local GABAergic interneurons are the main source of inhibition in the cerebral cortex. Loss of these neurons leads to epilepsy along with other neurological disorders. Interneuron development is tightly regulated by molecular cues including hepatocyte growth factor/scatter factor (HGF/SF). HGF/SF signals via its receptor MET to promote cellular survival, differentiation, mitogenesis and migration dependent upon cellular context. HGF/SF, which is secreted as a single chain precursor form, must be cleaved for biological function by serine proteases such as urokinase plasminogen activator (uPA). uPAR, the receptor of uPA, efficiently increases the protease cleavage by uPA. Adult uPAR−/− mice have decreased numbers of neocortical GABAergic interneurons and severe behavioral dysfunction such as spontaneously generated seizure and increased susceptibility to PTZ(pentylenetretrazol)‐induced seizures. These defects of GABAergic interneurons have been attributed to a reduction of HGF/SF. Methods: The uPAR−/− mouse was bred with transgenic mouse overexpressing HGF/SF under the astrocytic glial fibrillary acidic protein promoter (GFAP). Immunohistochemistry and seizure behaviors to PTZ (pentylenetetrazol) were used to evaluate anatomical and functional recovery of GABAergic interneuron in wildtype, GFAP‐Hgf, uPAR−/−, and uPAR−/−:GFAP‐Hgf mice. Results: The loss of interneurons in the uPAR−/− mouse is restored in uPAR−/−xGFAP‐Hgf mouse, in which the uPAR−/− mouse was bred with an HGF‐overexpressing mouse line (GFAP‐HGF). Reduction of PV(parvalbumin), a marker of GABAergic neuron, in the uPAR−/− mouse is rescued in uPAR−/−xGFAP‐Hgf mouse (Figure 1). The seizure susceptibility to PTZ of uPAR−/− mice is decreased in the uPAR−/−xGFAP‐Hgf mouse, to frequencies and severity observed in wildtype animals. The latency to seizure activity after PTZ treatment is also increased in the uPAR−/−xGFAP‐Hgf mouse as compared to uPAR−/− mouse. Conclusions: These data suggest that restoring HGF/SF levels can recover the anatomical interneuron defects and neocortical circuitry, leading to prevention of the seizure activity of uPAR−/− mice. This study will help uncover molecular mechanisms critical for seizure onset and their possible therapeutic targets for epilepsy. (figure 1) (Supported by Epilepsy Foundation (EMP and MB) and by the Women's Health Research Group (EMP).) 2 Justin M. Keener, 1 H. Steve White, and 2 Joanna C. Beachy ( 1 Pharmacology and Toxicology, University of Utah, Salt Lake City, UT and 2 Pediatrics, University of Utah, Salt Lake City, UT ) Rationale: Moderate to severe hypoxic‐ischemic encephalopathy (HIE) occurs in 1‐ 2/1000 live births. Approximately 40% of the infants with HIE and neonatal seizures develop epilepsy by 4 years of age. The Levine rat pup model of hypoxia/ischemia (H/I) insult is a well characterized model of HIE. Recently, it has been reported that 40% of rat pups exposed to H/I insult at 7 days of age exhibit spontaneous seizures by 6 months of age (Epilepsia 2004; 45(10):1210–1218). We hypothesize that a H/I insult will alter seizure threshold and, thus, predispose the rat to the development of spontaneous seizure activity. Methods: Seven day old Sprague‐Dawley rat pups were anesthetized with isoflurane and subjected to right carotid artery ligation. After recovery, the pups were returned to their dams for 2 hours. Pups were then exposed to 8% oxygen/92% nitrogen for 2 hours. Temperature was maintained at 34.5–36°C. Convulsive current curves for clonic (forebrain) and tonic (hindbrain) seizures were generated at 1 and 2 months after H/I insult and compared to those obtained from age‐ and sex‐matched naive control rats (> 25 rats per treatment group). Data were compared using Probit analysis (Minitab statistical software program). Results: Data for each seizure type is presented as CC50, the current at which 50% of the rats exhibited seizures ± standard error. A decrease in CC50 suggests increased seizure susceptibility. In naive male rats, CC50 significantly increased over time. Female naive rats had significantly deceased CC50 at 2 months of age (clonic 18.2 ± 0.4 mA, tonic 29.2 ± 0.9 mA) when compared to male naive rats (clonic 27.3 ± 0.9 mA, tonic 52.6 ± 1.4 mA). When compared to naive female rats, the CC50 for clonic seizures in female rats was significantly decreased (p = 0.005) at 1 month post H/I insult but increased at 2 months (p = 0.001). CC50 for tonic seizures in female naive and H/I rats did not differ. Male rats exhibited increased CC50 for tonic seizures (p = 0.007) at 1 month (35.5 ± 1.2 mA) and 2 months (61.6 ± 1.7 mA) when compared to naive male rats (28.3 ± 1.7 mA and 52.6 ± 1.4 mA, respectively). Conclusions: Alteration in seizure susceptibility is dependent on sex and time after H/I insult. Female rat pups exhibited enhanced clonic (forebrain) seizure susceptibility at 1 month after H/I insult but deceased susceptibility 2 months after insult. Interestingly, H/I insult in male rats decreased tonic seizure susceptibilty at 1 and 2 months after insult. Thus, H/I insult does not appear to consistently increase seizure susceptibility as evaluted by the MES method. Based on these results, it would appear that electrical population seizure threshold testing is not predictive of long‐term outcome (e.g. spontaneous seizure activity). (Supported by Primary Children's Medical Center Foundation Innovative Research Grant.) 1 Walter G. Besio, 1 Kanthaiah Koka, 2 Andrew J. Cole, and 1 Fei Zhu ( 1 Biomedical Engineering, Louisiana Tech University, Ruston, LA ; and 2 Epilepsy Service, Massachusetts General Hospital, Harvard Medical School, Boston, MA ) Rationale: We sought to evaluate the effect of transcutaneous electrical stimulation (TcES) via concentric ring electrodes on ictal electrographic and behavioral activity and mortality in rats with pilocarpine‐induced status epilepticus (SE). Methods: Male Sprague‐Dawley rats (290–330 g) were briefly anesthetized and 3 concentric ring electrodes were afixed to their scalps one day before the experiment. Scopolamine methylnitrate (2 mg/kg i.p.) was given 30 minutes prior to pilocarpine. Pilocarpine HCl (310 mg/kg i.p) was given to cause long lasting SE. Laplacian EEG was recorded from tri‐polar concentric electrodes on the scalp. TcES was applied 5 minutes after the onset of SE. Time‐frequency analysis was performed on the Laplacian EEG signals to compare the electrographic activity before and after the application of TcES. Behavior was monitored by inspection. Survival was assayed at 24 h after administration of pilocarpine. Results: Control rats (n = 8) followed the classic electrographic stages of pilocarpine‐induced status epilepticus described by Treimen (1987) and expired on average 15 hours after the pilocarpine injection. TcES‐treated rats (n = 8) lived significantly longer (p = 0.013, Two‐Sample t‐Test), which were euthanized on average at 48 hours. Twenty‐four hours after the pilocarpine injection, six (75%) TcES‐treated rats versus one (12.5%) control rat were alive (p = 0.041, Mann‐Whitney U test). Both electrographic and behavioral manifestations of seizure activity were reduced or abolished after the application of TcES. All surviving TcES‐treated rats recovered to baseline activity, including eating and drinking. By contrast, none of the control rats ate or drank after they entered SE. The time‐frequency analysis showed evident differences before and after TcES. Conclusions: The application of TcES increased survival of rats with pilocarpine‐induced SE. Positive TcES effects on electrographic and behavioral manifestations of seizure activity were significant and persistent. TcES may represent a novel and effective early treatment for SE. (Supported by Louisiana Tech University Center for Entrepreneurship and Information Technology and the Louisiana Board of Regents.) 1 Devin J. Cross, 1 Svitlana Kravchenko, and 1 Jose E. Cavazos ( 1 Neurology and Pharmacology, University of Texas Health Science Center at San Antonio, San Antonio, TX ) Rationale: Early life status epilepticus is detrimental to the behavioral and cognitive development in childhood and increases the risk to develop epilepsy in later life. Recurrent seizures during the first two weeks of life in rats induce limited neuronal injury and synaptic reorganization of the mossy fiber pathway in the CA3 region of the hippocampus (Holmes et. al., 1999; Cilio et. al., 2003). These neurobiological alterations enhance the degree of neuronal injury and the duration of status epilepticus experienced after a second excitotoxic injury later in life (Koh et al., 1999; Schmid et al., 1999). The present study examines the frequency of late spontaneous seizures one year after early life kainic acid induced status epilepticus. Methods: Kainic acid was administered in age specific doses for the following experiments. On postnatal day (P) 15, kainic acid was administered to induce convulsive status epilepticus and control rats received equal volumes of saline. On P45, half of rats from each group received kainic acid induced status epilepticus. Status Epilepticus was monitored for 6 hours and behavioral seizures were scored using a previously described scale. These four groups were allowed to mature for 1 year and then underwent surgical implantation of bipolar hippocampal electrodes to monitor the occurrence of late onset spontaneous seizures. Rats were then monitored for 8 hours to determine the frequency of late onset spontaneous seizures. After rats were monitored they were perfused with sodium sulphide and paraformaldehyde/gluteraldehyde for histological examination with Timm's reagent and Cresyl Violet. Results: After one year of maturation, rats that received kainic acid or saline on P15, did not demonstrate late onset spontaneous seizures over several 8 hour monitoring sessions, however mossy fiber sprouting was detected in the oriens of the CA3b region. Rats that receive saline on P15 and kainic acid on P45 developed late onset spontaneous seizures with an hourly seizure frequency of 0.22 ± 0.13 (mean ± SEM). This excitotoxic insult induced neuronal injury and synaptic reorganization of the mossy fiber pathway into the dentate gyrus and CA3b regions. Rats that received a “double hit” kainic acid on P15 and P45 developed severe late onset spontaneous seizures, with an hourly seizure frequency of 0.62 ± 0.31. These rats also developed widespread neuronal injury and synaptic reorganization of the mossy fiber pathway into the dentate gyrus and the CA3b regions. Conclusions: Early life status epilepticus induced by kainic acid induces limited neuronal injury and synaptic reorganization in the CA3b region, which is not sufficient to initiate the process of epileptogenesis. However, alterations after early life seizures shift the neurobiology in a pro‐epileptic direction that enhances the susceptibility to seizures and accelerates epileptogenesis after subsequent insults during adulthood. 1 Marion Czapp, 1 Jens P. Bankstahl, 2 Mark Helm, 3 Ulrich Massing, 3 Vittorio Ziroli, 4 Uta Huelsermann, 4 Gert Fricker, and 1 Heidrun Potschka ( 1 Department of Pharmacology, Toxicology, and Pharmacy, University of Veterinary Medicine Hannover, Hannover, Germany ; 2 Department of Chemistry, University of Heidelberg, Institute of Pharmacy and Molecular Biotechnology, Heidelberg, Germany ; 3 Clinical Research, Clinic for Tumorbiology, Freiburg, Germany ; and 4 Department of Pharmaceutical Technology and Pharmacology, University of Heidelberg, Institute of Pharmacy and Molecular Biotechnology, Heidelberg, Germany ) Rationale: Despite the launch of several new antiepileptic drugs (AED) during the last two decades pharmacoresistance of epilepsy remains an unsolved problem. It is hypothesized that local overexpression of multidrug transporters (MDT) in the epileptic tissue of pharmacoresistant individuals results in limited access of AED to the epileptic focus region. Downregulation of MDT by RNA interference (RNAi) would be a possible approach to overcome pharmacoresistance. Aim of the present study was to test if specific delivery of siRNA targeting the mRNA sequence of the major MDT P‐glycoprotein (Pgp) to brain‐capillary endothelial cells is possible. Methods: Liposomes were coupled with ApoE4 peptides which are discussed to mediate an uptake of liposomes into brain‐capillary endothelial cells via LDL receptors. ApoE4‐liposomes loaded with fluorescence‐labelled siRNA targeting the rat Pgp mRNA sequence were administered i.v. to female Wistar rats. Distribution of siRNA was investigated at different time points (4, 10, 24 h) post administration by confocal microscopy in rat brain sections. Based on a fluorescence resonance energy transfer (FRET) effect the integrity of siRNA was controlled (siRNA was labelled with two fluorescent dyes). The influence of siRNA molecules on Pgp expression was evaluated 24 h post administration by immunhistochemistry and compared with control groups. Results: Fluorescence microscopy of rat brain sections after siRNA administration demonstrated a successful targeting to brain‐capillary endothelial cells. An accumulation of labelled siRNA in the cells was obvious at all time points. The intensity of fluorescence increased from 4 to 24 h post administration. FRET analysis indicated that the siRNA was still intact 24 h following administration. Preliminary results of the immunhistochemical investigations gave evidence for a reduced Pgp expression in rats treated with siRNA in comparison with control groups. Conclusions: It was demonstrated that specific delivery of siRNA encapsulated in immunoliposomes to brain‐capillary endothelial cells can be achieved. Thus, the study substantiates the suitability of this novel approach for siRNA delivery. Reduction of Pgp expression in brain‐capillary endothelial cells will be investigated in further experiments to ensure that an RNAi‐based downregulation of Pgp takes place and enhances AED efficacy in the amygdala kindling model of temporal lobe epilepsy. (Supported by CURE (Citizens United for Research in Epilepsy).) 1 Veerle De Herdt, 2 Jan De Waele, 1 Jeroen Van Aken, 1 Kristl Vonck, 3 Jean Delbeke, 1 Robrecht Raedt, 1 Wytse Wadman, and 1 Paul Boon ( 1 Laboratory for Clinical and Experimental Neurophysiology, Ghent University Hospital, Ghent, Flanders, Belgium ; 2 Department of Intensive Care, Ghent University Hospital, Ghent, Flanders, Belgium ; and 3 Neural Rehabilitation Engineering Laboratory, Université Catholique de Louvain, Brussels, Brussels, Belgium ) Rationale: The precise mechanism of action of vagus nerve stimulation (VNS) in suppressing seizures remains to be elucidated. This study investigated whether VNS affects the threshold for provoking focal motor seizures using cortical electrical stimulation. Methods: Male Wistar rats (n = 8) were implanted with a custommade cuff‐electrode around the left vagus nerve. Stainless steel electrodes were bilaterally implanted on the rat motor cortex for stimulation and recording of EEG. Motor cortex stimulation (50 Hz, biphasic current pulses) with increasing amplitude (0 to max 2000 μA) was performed. At the time of the first clinical seizure symptoms, the stimulation train was acutely aborted and the threshold intensity was determined by means of continuous video‐EEG monitoring. The vagus nerve was stimulated for one hour with an intensity of 0,75 mA and a maximum duty cycle. For each rat, seizure thresholds (ST) were assessed before and immediately after VNS during 2 or 3 sessions on different days. Results: In 7/8 rats, ST increased after one hour of VNS. In 1/8 rat ST remained unchanged. In total, 21 sessions of VNS were performed in the 8 rats; for each session, ST were assessed at least 3 times before and 3 times after VNS. There was a significant increase (99% CI) of the ST after one hour of VNS (mean: 1424 μA; SD: 315; range: 869–1999 μA) compared to the baseline value (mean: 1077 μA; SD: 244; range: 673–1754 μA). Conclusions: In this study VNS significantly increased ST in a cortical stimulation model used for evaluating efficacy of antiepileptic treatment for motor seizures. These data indicate that VNS is capable of modulating cortical excitability. (Supported by V. De Herdt is supported by a grant (‘aspirant’) from the FWO‐Flanders.) 1 Stefanie M. Dedeurwaerdere, 2 Marie‐Claude Gregoire, 3 Peter Roselt, 3 David Binns, 2 Andrew Katsifis, 3 Rod Hicks, 1 Terrence J. O’Brien, and 1 Damian E. Myers ( 1 The Department of Medicine/Royal Melbourne Hospital, The University of Melbourne, Melbourne, Vic, Australia ; 2 Radiopharmaceuticals Research Institute, Australian Nuclear Science and Technology Organisation, Sydney, NSW, Australia ; and 3 Centre for Molecular Imaging, Peter MacCallum Cancer Centre, Melbourne, Vic, Australia ) Rationale: Changes in GABAergic inhibition are known to be associated with epilepsy. Human positron emission tomography (PET) studies have shown altered GABAA receptor expression using flumazenil (FMZ), a ligand with a high affinity for the central benzodiazepine receptor (cBZR). Small animal PET, using serial scans in the same epileptic animal, is potentially a powerful technique to investigate the evolution and pathophysiolgical basis of these changes during epileptogenesis. The aim of this study was to explore the potential of 2’‐(18F)fluoroflumazenil ((18F)FFMZ) to study the GABAA/cBZR complex in rats utilizing small animal PET. Methods: In five male non‐epileptic control rats (350–375 g), two serial dynamic PET scans were acquired under ketamine/xylazine anaesthesia after bolus injection of (18F)FFMZ (66.6 ± 3.7 MBq). Six scans were acquired after injecting a tracer dose. In addition, a pre‐saturation scan, consisting of a cold FMZ injection followed 10 min later by (18F)FFMZ, and two displacement studies (cold FMZ injection during the dynamic scan) were performed. PET/MRI co‐registration was performed to delineate different regions of interest (whole brain, hippocampus, pons). Results: In the hippocampus, a structure with high GABAA/cBZR density, the signal intensity was higher (10%) than in the whole brain (p < 0.05) with no difference between left and right hippocampus. In the pons, which has low GABAA/cBZR density, the signal intensity was approximately 20% lower than in the whole brain (p < 0.01). Pre‐saturation and displacement studies showed a high non‐specific component in the measured signal, which may be due to recirculation of (18F)‐fluoroethanol, a metabolite of (18F)FFMZ. (figure 1) Conclusions: This study has demonstrated differential uptake of (18F)FFMZ into key structures in the brain, with the highest activity in the hippocampus. However, significant non‐specific activity in the brain will limit the ability to detect changes in GABAA/cBZR density during epileptogenesis and therefore alternative ligands need to be investigated. (Supported by the CRC for Biomedical Imaging Development.) 1,2 Karine Bressand, 1 Corine Roucard, and 2 Antoine Depaulis ( 1 Co, SynapCell, Grenoble, France ; and 2 INSERM, U704, Grenoble, France ) Rationale: The syndrome of mesial temporal lobe epilepsy (MTLE) represents a major challenge in the clinical management of focal seizures mainly due to drug refractoriness. Both morphological and electroclinical features of MTLE can be mimicked in the adult mice by a unilateral injection of kainic acid (KA) in the dorsal hippocampus. Neuronal loss in CA1, CA3 and hilus as well as gliosis and mossy fiber sprouting are observed in the injected side, as well as dispersion of the dentate gyrus, as observed in MTLE patients. In addition, recurrent focal hippocampal discharges occur two weeks after KA which share many similarities with human focal discharges. Whether these discharges are suppressed by antiepileptic drugs (AEDs) was addressed in this study where the effects of different classical AEDS were compared to a new AED: pregabalin. Methods: C57BL/6 adult mice were stereotaxically injected with KA (1 nmol in 50 nl) in the dorsal hippocampus and equipped with cortical and hippocampal electrodes. After 3 weeks, mice were injected with different doses of AEDs or vehicle in a random order and EEG recordings were performed for 20 min pre‐injection and up to 120 min post‐injection. Diazepam (0.5, 1, 2, 3 mg/kg) (n = 6), valproate (100, 200, 400 mg/kg) (n = 6), lamotrigine (30, 60, 90 mg/kg) (n = 7), carbamazepine (25, 50 mg/kg) (n = 12) and pregabalin (n = 8) (10, 50, 100 mg/kg) were injected i.p. Only animals with cell loss and dentate gyrus dispersion were used in the analysis. Cumulated durations of hippocampal discharges during 20 min periods were compared between the different drug conditions and versus vehicle using ANOVA for repeated measures. Results: Hippocampal discharges were suppressed in a dose‐dependent way following diazepam (1, 2 and 3 mg/kg). Suppression of seizures was also observed with high doses of valproate (400 mg/kg) and lamotrigine (90 mg/kg) without dose‐dependency and with occasional behavioral side‐effects. No suppression was obtained following carbamazepine (25 and 50 mg/kg) and aggravation was observed after a low dose of lamotrigine (30 mg/kg). Pregabalin significantly suppressed hippocampal discharges in a dose‐dependent way (50 and 100 mg/kg). Conclusions: Our study shows that besides benzodiazepines, only pregabalin suppresses focal discharges in a dose‐dependent way in the kainate‐mouse model of MTLE. On the contrary, classical AEDs were effective only at high doses (valproate, lamotrigine) or were without effects (carbamazepine). In addition to the histological, electrophysiological and behavioral features of this model, the resistance to classical AEDs suggest a form of drug‐refractoriness often observed in human MTLE. Therefore this model fulfils most criteria for an efficient preclinical development of antiepileptic therapeutical strategies for MTLE. The suppressive effects of pregabalin suggest that this compound could be effective on human MTLE. (Supported by Pfizer France, Inserm and Ligue Française Contre l’Epilepsie.) 1 Taku Doi, 1 Yuto Ueda, 1 Keiko Nagatomo, and 2 James L. Willmore ( 1 Department of Psychiatry, Miyazaki Medical College, Kiyotake, Miyazaki, Japan ; and 2 School of Medicine, Saint Louis University, St. Louis, MO ) Rationale: Kindling is a form of epileptogenesis that can be induced with pentylenetetrazol (PTZ). We undertook this study to evaluate the contribution of glutamate and GABA transporters in the process of PTZ kindling. Methods: Rats were kindled with PTZ (40 mg/kg) injected i.p. three times per week until two consecutive stage five seizures were observed. Experiment 1 Animals underwent pentobarbital anesthesia and decapitation at 24 hours (Group A) or at 30 days (Group B) after their last PTZ induced seizure. Rats in Group A and B were compared with each control rats (Group C and D). Experiment 2 We collected early‐established animals (Group E) and kindling resistant animals (Group F). Early‐established rats had consecutive stage five seizures with less than 5 injections of PTZ. Kindling resistant animals did not have fully kindled seizures by 12 injections of PTZ. Under anesthesia with sodium pentobarbital brain tissue was collected at 30 days after the last PTZ injection. We used western blotting to measure the changes in hippocampal glutamate and GABA transporters in all animals. Results: GLAST, GLT‐1 and EAAC‐1 of Group A were elevated significantly compared with Group C. No change was found in any transporters of Group B compared with Group D. However, at 30 days (Group B) after their last PTZ induced seizure, animals had stage 5 seizure with PTZ. In early‐established rats (Group E), levels of EAAC1 and GAT1 were decreased by 30% when compared to kindling resistant animals (Group F). In these animals, the level of GLT‐1, GLAST and GAT3 equaled control. Conclusions: Activation of NMDA receptors is critical to the kindling process. Since all glutamate transporters were increased at 24 hours after establishing kindling with PTZ we considered this effect a response to seizure induced glutamate turnover. While animals remain kindled at the chronic phase, the lack of sustained changes in glutamate transporters suggest these proteins are not involved in the maintenance of epileptogenesis. We wonder if the glutamate and GABA transporters might be operant in the convulsion threshold set factor or as a pace factor for kindling. (Supported by a Grant‐in‐Aid for Encouragement of Young Scientists (17591223) from the Ministry of Education, Science, Sport and Culture, Japan (to T.D.).) 1,2 Rastislav Druga, 1 Hana Kubova, and 1 Pavel Mares ( 1 Developmental Epileptology, Institute of Physiology, Academy of Sciences, Prague, Czech Republic ; and 2 Anatomy, Charles University, 2nd Medical Faculty, Prague, Czech Republic ) Rationale: Pilocarpine induced status epilepticus (SE) in adult and immature animals results in neurodegenerative changes in many thalamic nuclei In the group of midline thalamic nuclei massive neuronal degeneration was found in immature animals in the reuniens nucleus. The nucleus reuniens (RE) is the largest of the midline nuclei and major source of projections to the hippocampus. Afferent projections originate mainly from limbic structures. Immunocytochemical studies have revealed calretinin (CR) and calbindin (CB) immunoreactive neurons in the RE and 70% of CR‐immunoreactive cells also showed CB‐immunoreactivity. The present study was focused on distribution and dynamics of neuronal degeneration in the RE and on changes in CR‐immunoreactivity within this nucleus in immature rats surviving up to one week after lithium‐pilocarpine SE. Methods: Experiments were performed in Wistar rats 12, 15, 18, 21 and 25 days old. Lithium‐pilocarpine model of SE was used. Only animals exhibiting convulsive SE were included in this study. Rats of all age groups were sacrificed 4, 8, 12, 24, 48 hours and 1 week after SE. Animals were deeply anesthetized with urethane (2.5 g/kg i.p) and perfused with 0.01 PBS (pH 7.4) followed by 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4). The brains were postfixed and cryoprotected. Blocks of the brain were sectioned in a coronal plane (50 μm), mounted on gelatin‐coated slides and processed for FluoroJade B (FJB) histochemistry. Another group of 25‐day‐old animals surviving 1 week after SE was processed for localization of CR. To determine the CR‐expressing cells the standard immunocytochemical techniques were used. Sections were observed under an epifluorescence microscope Olympus Provis AX 70 using FITC filter set. Results: Since P15 massive neuronal degeneration was evident in the RE. Damage was found in all age groups (P15, 18, 21, 25) and number of degenerated (FJB positive) neurons¨increased with age of SE elicitation and with survival interval. Neuronal degeneration was evident in all parts of nucleus, but the majority of FJB‐positive cells was in the rostral half of the RE. A significant decrease in number of CR‐positive neurons was apparent in animals surviving 1 week. CR‐reactive neurons were absent in the lateral portions of the RE, while the medial part of the nucleus exhibited a few of CR‐positive cells. Conclusions: The reuniens nucleus belong among thalamic nuclei where massive neuronal degeneration was evident since P15. The number of degenerated neurons increased with age and survival time after SE. Significant decrease in number of CR‐immunoreactive neurons indicate that calretinin has no neuroprotective effect within the reuniens nucleus. (Supported by a grant No.304/04/0464 of the Grant Agency of the Czech Republic.) 1 Mark J. Dunleavy, 2 Satchiko Shinoda, 2 Clara Schindler, and 1 David C. Henshall ( 1 Department of Physiology and Medical Physics, Royal College of Surgeons In Ireland, Dublin, Ireland ; and 2 Legacy Research, Portland, OR ) Rationale: Prolonged seizures in adults can damage the hippocampus and cause epilepsy. However, the effects of prolonged seizures on the immature brain remain poorly understood. A small number of studies have suggested seizures in rats < P14 cause damage, however the date at which vulnerability emerges is contraversial. If these seizures were able to cause hippocampal sclerosis, a characteristic of temporal lobe epilepsy (TLE), this could have significant implications for seizure treatment and approaches to anti‐epileptogenesis. Methods: We have developed a novel animal model in P10 Sprague‐Dawley rat pups that provides the unique ability to investigate the molecular pathogenesis of hippocampal sclerosis following seizures in the developing brain. The model features early‐life prolonged seizures, induced by intraamygdalar injection of kainic acid (KA), that trigger acute neuronal death in the hippocampus, leading to hippocampal sclerosis at adulthood. EEG was recorded both acutely and over a period of weeks at adulthood to confirm the presence of seizures. Results: Seizures were induced and hippocampal injury was then examined at various time points thereafter. Seizures induced largely unilateral hippocampal damage, affecting CA1 and CA3 regions. Western blot revealed that cell death was associated with activation of apoptotic pathways. By adulthood animals exhibited profound unilateral hippocampal sclerosis. Long‐term EEG confirmed the presence of spontaneous seizures in these animals. Administration of a neuroprotectant FK506 during the initial seizures provided protection from hippocampal sclerosis at adulthood. Conclusions: We have successfully developed a novel model of unilateral hippocampal sclerosis and TLE in P10 rat pups. Using this model valuable insights may be gained into epileptogenesis in the developing brain and provide targets for anti‐epileptogenic treatments. (Supported by Wellcome Trust and NIH.) 1 Tobias Engel, 1 Seiji Hatazaki, 1 Ina Koegel, 1 Carmen Bellver‐Estelles, 1 Jochen H.M. Prehn, and 1 David C. Henshall ( 1 Physiology, Royal College of Surgeons In Ireland, Dublin, Ireland ) Rationale: Epilepsy is one of the most common neurologic disorders, affecting approximately 50 million people worldwide and its characteristics are a predisposition to recurrent unprovoked seizures caused by abnormal hypersynchronous discharges of neurons. Through experimental modelling and clinical neuroimaging it has been shown that seizures are capable of causing neuronal death and this may contribute to epileptogenesis, impairments in cognitive function or the epilepsy phenotype. Neuronal death in this setting is, at least in part, due to the induction of the molecular machinery of apoptosis. Studies using clinical material from patients who suffer recurrent seizures and animal epilepsy models have confirmed changes in expression and activation of pro‐ and anti‐apoptotic proteins of the caspase and Bcl‐2 protein family. Another consistently up‐regulated pro‐apoptotic protein in seizure models is the tumour suppressor protein p53. Recently a new p53 up‐regulated target gene has been identified named PUMA. PUMA, a BH3‐only protein, is one of the most potent pro‐apoptotic proteins of the Bcl‐2 family which is also expressed in the central nervous system. In this study we examine the role of PUMA in seizure‐induced neuronal death using PUMA knock‐out mice. Methods: Seizures were focally evoked in adult C57Bl/6 mice by unilateral stereotaxic microinjection of kainic acid into the basolateral amygdala nucleus. The EEG was continuously monitored until lorazepam was administered to terminate seizures after 40 minutes. For Western blotting, samples were homogenized in lysis buffer and loaded on a SDS‐PAGE gel. For immunohistochemical analysis, coronal brain sections were stained with NeuN and analysis of cells exhibiting DNA fragmentation was performed using fluorescein‐linked TUNEL. Data are presented as mean ± SEM. Data were analyzed using one‐way ANOVA with post hoc Fisher's paired least significant difference test as appropriate. Significance was accepted at p < 0.05. Results: In C57Bl/6 wild‐type mice, p53 and PUMA were up‐regulated in the hippocampus shortly after seizure induction. Assessment of the phenotype of PUMA knock‐out mice revealed normal numbers of hippocampal and amygdala neurons, and normal expression of the kainate receptor in the amygdala and hippocampal site of seizure elicitation. Seizures elicited by intra‐amygdala kainate were of similar frequency and severity in the PUMA knock out mice. However, neuropathological analysis revealed neuron survival was significantly greater in PUMA knock‐out mice than in wild‐type mice and TUNEL counts were significantly lower in PUMA knock‐out mice than in wild‐type mice. Conclusions: The present data suggest that PUMA contributes to the neurodegeneration taking place during seizure‐induced neuronal death and may therefore represent a therapeutic target for adjunctive therapy to prevent hippocampal damage following status epilepticus. (Supported by Health Research Board, Ireland; Science Foundation Ireland.) 1 Dario J. Englot, 1 Ulrich Schridde, and 1,2 Hal Blumenfeld ( 1 Department Neurol., Yale University School Med, New Haven, CT ; and 2 Department Neurobiol., Yale University School Med, New Haven, CT ) Rationale: Partial seizures in patients with temporal lobe epilepsy (TLE) are often classified as either complex, characterized by deficits in consciousness, or simple, involving no loss of consciousness. Prior intracranial EEG studies of complex‐partial seizures in TLE patients have shown large amplitude slow waves in the frontal and parietal cortices ictally, most prominent in the orbitofrontal cortex. This neocortical slowing is associated with decreased cerebral blood flow in these areas measured using single photon emission computed tomography (SPECT). It has been hypothesized that ictal neocortical slowing may reflect abnormal cortical and subcortical network interactions during temporal lobe seizures, responsible for the deficits in consciousness seen in TLE patients. However, an animal model is needed to further study the mechanisms and pathophysiology of neocortical slowing. Methods: Sprague‐Dawley rats were implanted with a bipolar stimulating/recording electrode in the dorsal hippocampus, and a frontal recording electrode in the association (n = 5), cingulate (n = 6), or lateral orbital (n = 5) cortex. After one week of recovery, animals were kindled with a daily 1s train of 1ms, 60 Hz biphasic pulses at the threshold current necessary to produce an afterdischarges of >3s duration. EEG recordings from the hippocampal and frontal electrodes were compared. Behavior was rated using the standard Racine scale. Results: Hippocampal EEG showed seizure activity, consisting of polyspike discharges, during all events. While frontal recordings in the association and cingulate cortices showed rapidly propagating fast (polyspike or sharp) activity within 1–2s of hippocampal stimulation, large amplitude slow waves of approximately 3–4 Hz were seen ictally in the lateral orbital cortex. Behaviorally, these seizures did not involve motor convulsive activity. As daily stimulations proceeded, the slow rhythm in the lateral orbital cortex progressively converted to fast spike activity, suggesting seizure propagation. Conclusions: Our results show neocortical slowing in the lateral orbital cortex during hippocampal seizures in rats. These data suggest that hippocampal kindling may provide a good animal model of complex‐partial seizures in patients with TLE, demonstrating neocortical slow activity outside the region of seizure initiation. Further studies into the fundamental mechanisms of neocortical slowing in this model may provide insight into pathophysiology of altered consciousness in complex‐partial seizures, and may lead to improved treatments of TLE. (Supported by NIH R01 NS049307, the Betsy and Jonathan Blattmachr Family, and NIH Medical Scientist Training Program T32 GM07205.) 1 Ebru Erbayat‐Altay, 1 Kelvin A. Yamada, 1 Michael Wong, and 1 Liu Lin Thio ( 1 Neurology, Washington University, St. Louis, MO ) Rationale: Leptin is a neurohormone that helps regulate body weight and also has complex effects on neuronal excitability. Because its predominant effect on neuronal excitability in vivo is unclear, we examined the susceptibility of leptin deficient ob/ob mice to pentylenetetrazol (PTZ) induced seizures. Methods: We studied PTZ induced seizures in 31 wild type (C57BL/6J) and 30 ob/ob (B6.V‐Lepob/J) male mice at 44–58 days of age using continuous video EEG recordings. After 15 minutes of baseline recording, a 25, 50, 75 or 100 mg/kg dose of PTZ was injected intraperitoneally. The recording continued for 30 minutes after the injection or until death, whichever came first. PTZ induced generalized absence, myoclonic, clonic and clonic‐tonic seizures. Maximal clinical seizure severity, latency to clinical seizure onset, latency to the first epileptiform abnormality and the normalized cumulative epileptiform duration were determined. Clinical seizure severity was rated as: 0 = no seizures, 1 = absence seizures with normal clinical behavior and EEG interictally, 2 = myoclonic seizures with normal clinical behavior and EEG interictally, 3 = generalized clonic or clonic‐tonic seizures with normal clinical behavior and EEG interictally, 4 = repetitive absence or myoclonic seizures without normal interictal behavior or EEG throughout the 30 minute recording, 5 = death. The normalized cumulative epileptiform duration was the sum of the duration of all epileptiform abnormalities divided by the total time of the recording after PTZ injection. Comparisons were made using a t‐test with statistical significance set at p < 0.05. Results: The background EEG for both wild type and ob/ob mice showed 6–7 Hz theta activity without any epileptiform abnormalities. In both groups, PTZ induced the same types of clinical and electrographic seizures in a dose‐dependent manner. The latency to the first clinical seizure and to the first epileptiform abnormality was the same in wild type and ob/ob mice at all PTZ doses. The maximal clinical severity at 25 mg/kg and 100 mg/kg PTZ were the same in wild type and ob/ob mice. However, at 75 mg/kg PTZ the clinical severity was 5.0 ± 0.0 for the ob/ob mice because all mice died whereas it was 3.5 ± 0.8 (p < 0.05) for wild type mice. At 75 mg/kg PTZ, the normalized cumulative epileptiform duration was 9 ± 3% for ob/ob mice and 5 ± 1% for wild type mice (p < 0.05). ob/ob mice also showed a higher maximal clinical severity and normalized cumulative epileptiform duration at 50 mg/kg PTZ. A greater number of generalized clonic and clonic‐tonic seizures accounts for the greater severity of seizures in the ob/ob mice. Conclusions: These results indicate that PTZ induces more severe seizures in ob/ob mice. These results are consistent with leptin acting as an endogenous anticonvulsant and support a role for leptin in the treatment of epilepsy. (Supported by Washington University McDonnell Center for Cellular and Molecular Neurobiology, NS 42744 and JDRF 1–2004–594.) 1,3 Chris J. Feeney, 1,3 Miron Derchansky, 1 Shanthini Mylvaganam, 1 Yana Adamchik, and 1,2,3,4 Peter L. Carlen ( 1 Division of Cellular and Molecular Biology, Toronto Western Research Institute, Toronto, ON, Canada ; 2 Department of Medicine (Neurology), University of Toronto, Toronto, ON, Canada ; 3 Department of Physiology, University of Toronto, Toronto, ON, Canada ; and 4 The Epilepsy Research Program, University of Toronto, Toronto, ON, Canada ) Rationale: Fragile X syndrome is the most common inherited form of mental retardation, affecting 1 in 1200 males and 1 in 2500 females. The syndrome is characterized primarily by intellectual impairment, ranging from mild learning disability to profound mental retardation. We are interested in studying the cellular electrophysiological factors underlying the pathophysiology of seizure activity in the brain tissue of Fragile X mental retardation protein (FMRP) gene knockout mice (KO). Methods: We have characterized the induction of seizure‐like events (SLE) in young adult mouse hippocampal slices exposed to low Mg‐ACSF and tetanic stimulation bouts, as with our previous KO mice. All recording were performed on transverse hippocampal slices (400 μm) from mice aged 6–9 weeks, in oxygenated ACSF, at 33°C. SLEs and extracellular field responses were recorded from the CA1 pyramidal layer of the hippocampal slices, in response (were appropriate) to Schaffer collateral stimulation. Results: The KO mice hippocampal slices were more easily induced into SLEs with either low Mg‐ACSF, or by repetitive tetanizations. KO slices showed a shorter lag time to SLE of 15.4 min. vs 26.5 min. in wt slices, required fewer tetanic stimuli to produce SLEs. A brief application of the mGluR5 antagonist MPEP (at 1 or 10 μM delayed the onset of low Mg‐induced, but not tetanic stimulation‐induced SLEs in KO hippocampal slices by more than 3‐fold. However, we found that longer application of this same drug leads to diminished (with low Mg‐ACSF exposure) or increased susceptibility (with tetanic stimuli) to SLEs in the KO hippocampus. Also, we have found that inhibitory network that regulates paired‐pulse inhibition in CA1 of the hippocampus, is dramatically reduced in the KO hippocampus, leading to a complete lack of PPI that can be elicited in the wt slices. Conversely, we have found that paired‐pulse facilitation is significantly enhanced in the KO hippocampus as compared to the wt, and that this may point to alterations in presynaptic calcium handling in these FMRP knockout mice. Conclusions: We have found that hippocampi from FMRP knockout mice are more easily induced into seizure‐like events in vitro. We are continuing work to further elucidate the nature of mGluR regulation of SLEs in this model, as well as the influence of inhibitory inputs in the KO hippocampus on seizure susceptibility in these mice. (Supported by the Fragile X Research Foundation of Canada (CJF), The Savoy Foundation (MD), and CIHR (PLC).) 1 Neil M. Fournier, 1 Devon R. Andersen, 1 Sumeer A. Mann, 2 Hector J. Caruncho, 3 G. Campbell Teskey, 4 Dan C. McIntyre, and 1 Lisa E. Kalynchuk ( 1 Psychology, University of Saskatchewan, Saskatoon, SK, Canada ; 2 Cell Biology, University of Santiago de Compostela, Santiago, Spain ; 3 Psychology, University of Calgary, Calgary, AB, Canada ; and 4 Psychology, Carleton University, Ottawa, ON, Canada ) Rationale: Reelin is a large secreted glycoprotein that has important roles in the developing nervous system as well as in the adult nervous system. It is believed that extracellular reelin may act as either a stop or as a chemoattracting signal for radially migrating neurons. Deficits in reelin signaling have been found in a variety of neurodevelopment and neurodegenerative disorders, including epilepsy. However, in the adult nervous system, the specific role of reelin is not fully understood. Reelin is synthesized and released by GABAergic interneurons and recent studies have suggested that reelin might modulate and even enhance long‐term potentiation in the adult brain. Importantly, enhancement in synaptic transmission is thought to play a crucial role in the intensification of afterdischarges and progressive development of convulsions in the amygdala kindling model of epilepsy. Methods: To determine if amygdala kindling was associated with changes in reelin expression, rats were surgically implanted with an electrode into the basolateral amygdala and subjected to 99 kindled or sham stimulations. An additional group of non‐kindled genetically prone FAST kindled rats was also processed in order to determine if basal reelin expression was altered in rats that were genetically more susceptible to enhanced amygdaloid excitability and the effects of kindling. Brains were processed for immunohistochemical determination of reelin, NeuN, and doublecortin. Results: The total number of reelin positive cells in the hilus was not substantially different between all groups; however, there was a slight decrease in the number of these cells within the subgranular zone for the kindled and non‐kindled FAST rats. A decrease in reelin expression within the stratum lacunosum‐moleculare was also found for kindled and non‐kindled FAST rats. Doublecortin immunohistochemistry revealed that some newly born neurons migrated ectopically into the hilus for kindled rats but not for sham‐stimulated controls or for non‐kindled FAST rats. Interestingly, we often observed several abnormally located reelin positive cells scattered throughout the dentate granule cell layer for non‐kindled FAST rats suggesting a possible pre‐existing GABAergic circuit malformation in this group of rats. Conclusions: These findings suggest that changes in reelin expression within the hippocampus during development might influence network formation that may influence susceptibility to epilepsy. (Supported by NSERC.) 1 Patrick D. Fox, 1 Hyokwon Chung, and 1 Sookyong Koh ( 1 Neurology, Children's Memorial Hospital, Northwestern Medical School, Chicago, IL ) Rationale: Chronic epilepsy often starts with an isolated early life seizure, a period of remission, and then the re‐emergence of seizures later in life. To mimic this event in rats, we developed a “two‐hit”‐kainic acid (KA) model and demonstrated that an early life seizure increases susceptibility to seizures and to seizure‐induced neuronal injury in adulthood. We have found that activation of resident microglia and subsequent increases in mRNA of inflammatory cytokines and complements appear to be the key initiating events for seizure‐induced inflammatory responses. We hypothesize that the inflammatory reaction provoked by early‐life seizures primes the developing brain and microglia, leading to rapid reactivation by a second seizure in adulthood. Blocking of this microglia activation through pharmaceuticals such as minocycline may prevent some neuronal injury typically seen following these later life seizures. Methods: Postnatal day 21–25 LE rats were injected with KA (10 mg/kg, I.p.) or saline. The first dose of minocycline (10 mg/kg, i.p.) or vehicle was administered immediately after KA injection, and 6 more doses were given 24 hours apart over the next 7 days. Two weeks after KA/PBS injections, all remaining rats were given KA. Rats were perfused two days later and all brain sections were processed for immunocytochemistry (ICC) using IbA1/Aif antibody to visualize activated microglia. CA1 subregions of the hippocampal sections were analyzed using MetaMorph (Universal Imaging Corp.). An average percent of area above the set threshold (threshold area containing immunoreactive cells captured at 20x) and the number of stained cells were calculated per animal. Results: Cellular hypertrophy of microglia suggestive of activated state was observed within 24 h after KA‐induced seizures. The group that experienced early‐life seizures (K/K) had a marked increase in the number of activated microglia compared to control littermates who received single KA in adulthood (S/K). These microglia were larger and formed a complex, web‐like network localized within the hippocampus especially over the CA1 region. Over two fold increase in immunoreactive areas was noted in K/K compared to S/K (KA: 9.0 ± 1.3 vs. Saline: 4.0 ± .1.0, p < 0.04, n = 8). The microglia count was also significantly greater (KA: 233.1 ± 12.1 vs. Saline 153.5 ± 9.2, p < .039, n = 8). This hypertrophic, hyperplastic, reactivating effect of an early‐life seizure on microglia was abolished in minocycline treated rats. Conclusions: Our results indicate that the experience of early life seizures primes the microglia and elicits a much more robust activation in response to later life seizures. Whether microglial reactivation is causally related to the epileptogenic effect of early‐life seizures to increase seizure susceptibility and cell death will be determined. (Supported by Child Neurology Foundation and K02NS048237.) 1 Brita Fritsch, 1 Maciej Gasior, 1 Rafal M. Kaminski, and 1 Michael A. Rogawski ( 1 Epilepsy Research Section, NINDS/NIH, Bethesda, MD ) Rationale: Kainate (KA) receptors are formed from homomeric or heteromeric combinations of subunits, including GluR5 and GluR6. Studies with transgenic mice indicate that KA receptors containing GluR6 subunits play a role in KA‐induced seizures (1). Here we used kainate receptor knockout (KO) mice to determine the clinical seizure and EEG correlates of GluR5 kainate receptor activation, and the roles of kainate receptor subunits in other seizure models. Methods: Video‐EEG monitoring was carried out during slow tail vein infusion of the selective GluR5 agonist ATPA in wild type (WT), GluR5‐/‐, and GluR6‐/‐ mice. We also assessed seizure thresholds using the i.v. PTZ and 6 Hz models. Results: Slight myoclonic jerks/twitches associated with single sharp waves occurred at lower doses of ATPA in WT than in GluR5‐/‐, but generalized seizures occurred at similar doses. Myoclonic jerks occurred at comparable doses in WT and GluR6‐/‐ mice; paradoxically, EEG seizure patterns began with lower doses in GluR6‐/‐ mice. There were no significant differences between any of the genotypes in their thresholds for clonic seizures induced by i.v. PTZ or in the 6 Hz seizure model. Conclusions: GluR5 kainate receptors contribute to ATPA‐induced myoclonic jerks, but not to generalized seizures. The increased sensitivity of GluR6 KO mice to ATPA was unexpected. In the KO animals, the existence of GluR5 homomers instead of GluR5/GluR6 heteromers may account for the phenomenon since GluR5 homomers are about 2.4‐fold more sensitive to ATPA than are GluR5/GluR6 heteromers (2). GluR5 and GluR6 kainate receptors do not appear to play a role in seizure activity in the PTZ and 6 Hz models. 1 Mulle C, Sailer A, Perez‐Otano I, et al. Nature 1998;392:601–605. 2 A. Alt, B. Weiss, A.M. Ogden, et al. Neuropharmacology 2004; 46:793–806. 1 Michael A. Galic, 2 James G. Heida, 3 Campbell G. Teskey, and 2 Quentin J. Pittman ( 1 Neuroscience, University of Calgary, Calgary, AB, Canada ; 2 Neurology, Albert Einstein College of Medicine, Bronx, NY ; and 3 Psychology, University of Calgary, Calgary, AB, Canada ) Rationale: There is increasing evidence that immune and inflammatory processes in early life are capable of producing lasting effects on physiology and CNS function and raise the possibility that a single inflammatory episode during development could evoke changes in seizure susceptibility later in adulthood. Methods: To explore this hypothesis, postnatal day 14 male rats received either the bacterial endotoxin lipopolysaccharide (LPS) (25–250 μg/kg), or the viral mimetic Polyinosinic:Polycitydylic acid (POLY I:C) (1 mg/kg) to induce a mild inflammatory process. Two months later, rats received lithium and pilocarpine (LI‐PILO), kainic acid (KA) or were subjected to amygdala kindling to discern any changes to seizure susceptibility. Following KA seizures, subsets of animals were killed and brains evaluated for neurodegeneration using Fluoro jade histochemistry. Results: Neonatally LPS‐treated, but not POLY I:C‐treated rats, showed significantly faster seizure onset times (SOTs), by about 25% compared to controls. Adult rats given LPS and then LI‐PILO two months later, did not show the same reduction in SOT as found with the neonatal treatment. Separate experiments determined a comparable susceptibility of neonatally LPS‐treated rats to adult convulsions evoked with KA, but not to amygdala kindling. Behavioural sensitivities to KA were also reflected in the amount of neuronal degeneration in the CA3 and CA1 hippocampus 24 h after SOT. Conclusions: Taken together, we have observed that long‐lasting changes in seizure susceptibility can be evoked following a mild inflammatory episode during development, possibly through a mechanism involving, or specific to, the Toll‐like receptor (TLR)‐4 (LPS‐activated), but not TLR‐3 (POLY I:C‐activated) pathways. (Supported by Canadian Institutes of Health Research (CIHR), the Natural Sciences and Engineering Council of Canada (NSERC), and the Alberta Heritage Foundation for Medical Research (AHFMR).) 1,2 Levi B. Good, 3 Shivkumar Sabesan, 2 Trevor D. Boone, 2 Leon D. Iasemidis, and 1,2 David M. Treiman ( 1 Neurology Research, Barrow Neurological Institute, Phoenix, AZ ; 2 Harrington Department of Bioengineering, Arizona State University ; and 3 Department of Electrical Engineering, Arizona State University, Tempe, AZ ) Rationale: In the last decade, substantial progress has been made in the study of the human epileptic brain by utilizing concepts and measures from nonlinear dynamics. The hallmark of this research is the ability to predict seizures prior to their clinical or electrographic onset (IEEE TBME 2003; 50:616–627). As the ability to predict leads to the possibility of control, research in controlling seizures with closed‐loop systems is expected to flourish in the near future and will most likely include studies performed in animal models of epilepsy. We have thus applied the concepts from nonlinear dynamics, namely the Largest Short‐Term Lyapunov exponent (STLmax), to evaluate the effectiveness for real‐time seizure prediction in the lithium pilocarpine rat model of chronic epilepsy as a precursor for a closed‐loop seizure control system that utilizes deep brain stimulation. Methods: Three male Sprague‐Dawley rats (300–350g) were stereotaxcially implanted with a customized array of Tungsten depth wire electrodes which included four cortical, two hippocampal, and two thalamic contacts. Chronic epilepsy was established several weeks after an episode of prolonged status epilepticus using the lithium pilocarpine model (3 mmol/kg LiCl followed by 30 mg/kg pilocarpine 24 hrs later). Continuous EEG/video recordings were made during the entire experiment and all data analyzed in real‐time. STLmax values were calculated for each electrode and entrainment of STLmax was evaluated utilizing a pair t‐statistic (T‐index). True and false predictions were noted with a seizure prediction horizon set at 180 minutes. Results: Continuous EEG data from Rat1 (55 seizures in 166 hours), Rat2 (25 seizures in 384 hours), Rat3 (16 seizures in 739 hours) resulted in a total of 96 seizures in 1289 hours of recording. The seizure prediction results from each rat were: Rat1, prediction rate 35/55 = 63.4%, 11 false positives (0.0663/hr), mean prediction time = 111 minutes; Rat2, prediction rate 22/25 = 88.0%, 50 false positives (0.1302/hr), mean prediction time = 149 minutes; Rat3, prediction rate 10/16 = 62.5%, 114 false positives (0.1543/hr), mean prediction time = 114 minutes. Overall prediction rate‐ 67/96 = 69.8%, 175 false positives (0.1358/hr), mean prediction time = 125 minutes. Conclusions: While the overall prediction rate of ∼70% is less than the one reported in human data (∼80%), similar false positive rates (∼1 every 8 hours of recording), and an extended prediction time were noted. Given these results, prediction of seizures in the lithium pilocarpine rat model of chronic epilepsy appears feasible as a model of chronic epilepsy for testing real‐time feedback control systems utilizing thalamic deep brain stimulation. (Supported by Epilepsy Research Foundation of America and Ali Paris Fund for LKS Research, and Barrow Neurological Foundation.) 1 Heidi L. Grabenstatter, 2 Suzanne Clark, and 3 F.E. Dudek ( 1 Biomedical Sciences, Colorado State University, Fort Collins, CO ; 2 School of Pharmacy, University of Wyoming, Laramie, WY ; and 3 Physiology, University of Utah, Salt Lake City, UT ) Rationale: Animal models of chronic epilepsy, such as the kainate‐treated rat, may be useful in the discovery of new antiepileptic drugs (AEDs). The present study evaluated the effectiveness of carbamazepine on the frequency of spontaneous motor seizures in rats with kainate‐induced epilepsy. Another aim of the study was to of develop a paradigm for continuous, long‐term oral administration of potential AEDs. Methods: Single intraperitoneal (IP) injections of carbamazepine (10–100 mg/kg) were compared to vehicle (i.e., 20% (2‐Hydroxypropyl)‐β‐cyclodextrin) injections via six AED‐versus‐vehicle tests using a repeated‐measures cross‐over protocol. To establish an effective dose and the time‐course of the anticonvulsant effect, two trials (8–9 rats per trial) evaluated single per os(i.e., oral‐ by mouth (PO)) doses of 30 mg/kg and 100 mg/kg carbamazepine. Each PO trial included 3 AED‐versus‐vehicle tests comprised of carbamazepine‐containing or control pellet feedings on alternate days with a recovery day between each treatment day. Based on the single‐feeding studies, carbamazepine at 100 mg/kg was administered three times per day (TID) for 5 days. Results: Carbamazepine significantly reduced seizure frequency at 10, 30, and 100 mg/kg after single IP injections, and caused complete seizure cessation during the 6‐h post‐drug epoch in 25% and 70% of animals at 30 and 100 mg/kg. Single IP injections and single oral feedings of 30 mg/kg and 100 mg/kg carbamazepine significantly (p < 0.0001) reduced seizure frequency relative to control treatments during 6‐h and 22‐h epochs following drug administration. A single feeding of carbamazepine (100 mg/kg PO) was significantly effective for 20 h and full recovery was complete within 23 h. Continuously administered carbamazepine (100 mg/kg PO, TID) significantly reduced seizure frequency by >50% for 24 h and completely suppressed motor seizures in 50% of the animals tested. Conclusions: High doses of carbamazepine (30 and 100 mg/kg) were highly effective at suppressing spontaneous seizures, and were associated with a high percentage of motor seizure freedom. Administration of higher doses of carbamazepine did not further suppress spontaneous motor seizures. Oral administration of 30 mg/kg and 100 mg/kg carbamazepine was as effective as IP injections. Significant anticonvulsant effects of 100‐ mg/kg PO carbamazepine were sustained for the duration of the 5‐day trial. (Supported by NIH (HLG) and (FED).) 1 Jan Grashof, 1 Julika Pitsch, 1 Sandra Bauerkämper, 1 Albert Becker, and 1 Susanne Schoch ( 1 Neuropathology, University of Bonn Medical Center, Bonn, NRW, Germany ) Rationale: The pathogenesis of temporal lobe epilepsy (TLE) is characterized by activity‐dependent changes in synaptic plasticity. Certain forms of synaptic plasticity have been shown to depend on modifications of the presynaptic release machinery. However, our knowledge about the underlying molecular mechanisms and the role of the presynaptic release machinery in functional changes that underlie epileptic activity at the level of the synapse is very limited. Therefore, we have analyzed whether the expression patterns of components of the presynaptic release machinery are altered in different stages of an experimental model of temporal lobe epilepsy. We have furthermore characterized knock‐out mice deficient for three proteins that are integral components of the release machinery, RIM1α, RIM2α or Synaptotagmin 10 (Syt 10), in an experimental model of TLE. Methods: Status epilepticus (SE) was induced by systemic application of pilocarpine to rats as well as to three mutant mouse lines. Changes in mRNA levels were analyzed by Real‐time PCR and in situ hybridization (ISH). Seizure susceptibility was determined after induction of SE. The frequency and severity of chronic recurrent seizures was analyzed with a telemetric EEG system (DSI) and parallel video analysis. To study hippocampal damage, the amount of segmental hippocampal loss of neurons and gliosis were analyzed in wild‐type and knock‐out mice after SE. Results: To investigate changes in the mRNA expression level of various components of the presynaptic release machinery hippocampal subregions (dentate gyrus (DG), CA3 and CA1) were isolated from rats 4, 6 and 8 hours and 5 and 28 days after SE induced by pilocarpine. Analysis of these samples by Real‐time PCR, ISH and Affymetrix‐Micro‐Arrays showed distinct changes in the expression patterns of several constituents of the release machinery; e.g. Syt 10 could not be detected in the CA1 region of control rats/mice, neither by single cell PCR nor by ISH, however, its mRNA was strongly expressed in CA1 cells 6 h after SE. To study if ablation of individual components of the presynaptic release machinery affects epileptogenesis we are analyzing three mutant mouse lines in the pilocarpine model of TLE: knock‐out mice for the active zone proteins RIM1α and RIM2α, which exhibit defects in short‐ and long‐term synaptic plasticity, and Syt 10, whose mRNA expression is induced by seizures. So far the results obtained by Video‐ and EEG‐Monitoring of these animals suggest that RIM1α knock‐out mice have a higher susceptibility to seizures than wild‐type littermates (n = 5). Furthermore, the latency to the first seizue and to SE seem to be reduced. Conclusions: Our results suggest that components of the presynaptic release machinery are modified during epileptogenesis, may play a pathogenetic role and constitute targets for new therapeutic approaches in TLE. (Supported by DFG (SFB TR3, Emmy‐Noether‐Program), BONFOR.) 1 Alberto Lazarowski, 2 Veronica Trida, 3 Christian Höcht, 2 Nelida Gonzalez, 3 Javier Opezzo, 3 Guillermo Bramuglia, 3 Carlos Taira, and 2 Elena Girardi ( 1 Clinical Biochemistry, Faculty of Pharmacy and Biochemistry, UBA., Buenos Aires, Argentina ; 2 Celular Biology Institute, University of Buenos Aires, Buenos Aires, Argentina ; and 3 Pharmacology, Faculty of Pharmacy and Biochemistry, UBA., Buenos Aires, Argentina ) Rationale: The present work addressed possible participation of the efflux transporter P‐glycoprotein (P‐gp) in the alterations of the central pharmacokinetics of phenytoin in an experimental model of epilepsy induced by the administration of mercaptopropionic acid. Methods: Male Wistar rats were administered for 7 days with physiological solution (SF) or mercaptopropionic (MP) acid. The day of the experiment, a shunt micrordialysis probe or a concentric probe was inserted into the carotid artery or the hippocampus respectively, in order to monitor extracellular phenytoin (PHT) levels. In SF and MP treated rats, PHT (30 mg/kg, i.v.) was administered 30 after an intraperitoneal administration of SF or nimodipine (NIMO, 2 mg/kg). Pharmacokinetic analysis of the data was made using TOPFIT program. Results: No differences were found in PHT plasmatic levels comparing all experimental groups. In rats pretreated with SF, hipocampal PHT concentrations were signicantly lower in MP treated rats (maximal concentration (Cmax): 2.7 ± 0.3 μg/ml, p < 0.05 vs SF rats) compared to SF animals (Cmax: 5.3 ± 0.9 μg/ml). Whilst pretreatment with NIMO did not modify central pharmacokinetics of PHT in SF treated rats (4.5 ± 0.8 μg/ml), PHT levels were significantly higher in MP rats pretreated with NIMO (Cmax: 6.8 ± 1.0 μg/ml, p < 0.05 vs MP rats pretreated with SF) compared to MP animals with previous administration of SF. Conclusions: Thus, our results suggest that central pharmacokinetics of PHT was altered in epileptic rats by MP administration. The effect of NIMO on hipocampal concentrations of PHT suggested that P‐gp is implicated in the reduced central bioavailability of PHT in MP epileptic rats. (Supported by University of Buenos Aires.) 1 Susan L. Campbell, and 1 John J. Hablitz ( 1 Department of Neurobiology, University of Alabama at Birmingham, Birmingham, AL ) Rationale: Elevated glutamate levels have been reported in several epilepsy models. The electrophysiological changes that occur when transporter function is disrupted in experimental models of epilepsy have not been examined. The freeze‐induced model of cortical dysplasia results in abnormal lamination of cortical structures associated with hyperexcitability in the region adjacent to the microsulcus. Here we used this model to test the role of glutamate transporters in synaptic transmission in the hyperexcitable region. Methods: Focal freeze lesions were induced in postnatal day (PN) 0 Sprague‐Dawley rats using methods described previously. In freeze‐lesioned cortex, recordings of layer II/III pyramidal cells were obtained 0.3–1.5 mm lateral to the microsulcus. Synaptic responses were evoked with a bipolar stimulating electrode (twisted pair of 25 mm Formvar insulated nichrome wires) positioned 150–200 μm below the recording pipette. Stimuli were current pulses 10–100 μA in amplitude and 50–100 ms in duration. A stimulation frequency of 0.05 Hz was used. The glutamate uptake antagonists DL‐threo‐β‐benzylozyaspartic acid (TBOA) and dihydokainate (DHK) were bath applied. Results: TBOA prolonged postsynaptic currents (PSCs) and decreased the threshold for evoking spontaneous epileptiform activity in lesioned cortex. Uptake inhibition also induced an increase in holding current in the dysplastic cortex. In contrast, responses in sham operated and control animals were not affected. The effect of blocking uptake was mediated primarily by the glia glutamate transporter (GLT‐1) since the selective GLT‐1 blocker DHK mimicked the effects of TBOA. The increases in the holding current suggested that glutamate was accumulating extracellularly and activating ionotropic glutamate receptors. The increase in excitability following uptake inhibition appeared to be mediated by N‐methyl‐D‐aspartate (NMDA) receptors since prior application of D‐(‐)‐2‐amino‐5‐phosphonovaleric acid (APV) prevented TBOA‐induced effects and TBOA‐induced changes were reversed by APV. In the freeze‐lesioned cortex spontaneous epileptiform activity does not occur at a high rate. Under control conditions, the probability for the occurrence of spontaneous discharge was 0.09 ± 0.06 (n = 22) which increased significantly to 0.59 ± 0.1 (n = 22) when TBOA was applied. Conclusions: These results indicate that glutamate transporter function is altered in rat focal cortical dysplasia and these changes contribute to the observed hyperexcitability. Alteration in transporter function could result in increasing extracellular glutamate concentrations and over activation of NMDA receptors during neuronal activity. (Supported by NS22373.) 1 Kiyotaka Hashizume, 1 Seiji Takebayashi, and 1 Tatsuya Tanaka ( 1 Neurosurgery, Asahikawa Medical College, Asahikawa, Hokkaido, Japan ) Rationale: The aim of this study was a comparison of the antiepileptic effect between deep brain stimulation (DBS) of the subthalamic nucleus (STN) and of the anterior thalamic nucleus (ATN), using the same epilepsy model. Methods: A guide cannula was stereotactically implanted into the left sensorimotor cortex of male Wistar rats. Focal cortical seizures were induced by a microinjection of kainic acid (KA) into the left sensorimotor cortex. Using this focal seizure model, High‐frequency electrical stimulation of STN or ATN was performed intermittently when the animal developed into the seizure status. The seizure frequency of DBS‐on and DBS‐off was measured from a video‐EEG system. The data was analyzed statistically. Results: An intracortical injection of KA was induced focal and secondarily generalized seizures in all animals. The ipsilateral (the side of KA injection) STN‐DBS suppressed the seizure frequency to 50–60% of the DBS‐off phase. The effect of bilateral STN‐DBS showed no significance than the ipsilateral DBS. When the electrode was implanted into bilateral ATN, no seizure was induced by KA injection. The implantation of the ipsilateral ATN lead to seizure suppression, and the ipsilateral ATN‐DBS suppressed seizures more strongly. Conclusions: Both DBS of STN and ATN suppressed KA‐induced focal seizures. The antiepileptic effect of ATN‐DBS was stronger than STN‐DBS in this animal model. The seizure suppression of the electrode implantation alone may be caused by a focal destruction of ATN. 1 Henry J. Hasson, and 1,2 Solomon L. Moshe ( 1 Neurology, Albert Einstein College of Medicine, Bronx, NY ; and 2 Neuroscience and Pediatrics, Albert Einstein College of Medicine, Bronx, NY ) Rationale: Status epilepticus (SE) is a neurological emergency and a common problem in neonates. To date, effective treatments to stop ongoing status epilepticus in this age group have not been identified. To assess efficacy and morbidity from the treatment of prolonged seizures in the immature brain, it is necessary to develop screening assays. Methods: SE was induced in postnatal day 15(P15) and 21(P21) rats by systemic administration of kainic acid (KA) 3.5 mg/kg (P15) and 10 mg/kg (P21) or lithium‐pilocarpine (PI) 60 mg/kg (P15) and 30 mg/kg (P30). Rats were monitored by EEG with an electrode in the CA1 region of the hippocampus. In both models behavioral seizures continue for several hours and electrographic seizures continue for more than 24 hours if left untreated. After one hour of electroclinical SE, rats were treated with pentobarbital (PB), diazepam (DZ), or saline. Rats were then intermittently monitored for cessation of behavioral and electrographic seizures, and for mortality for 3 days following treatment to confirm that the seizures stopped and did not recur. Results: Both drugs can stop the ongoing seizures both behaviorally and electrographically. PB (50 mg/kg) was XX% effective in stopping KA and PI induced SE in both P15 and P21 rats. DZ (20 mg/kg) was effective in stopping SE in 50–64% of rats in 3 of the 4 groups. In the KA P21 group, it was not effective. Higher doses of diazepam did not improve the outcome in P15 pups in KA‐SE. When effective, DZ stopped SE much quicker than PB (p < .0001). In most groups, mortality from SE did not change irrespective of whether SE was stopped or not. In the P15 PI group, treatment with either drug had increased mortality compared to untreated SE. Conclusions: This study shows that it is possible to abort seizures in the developing rat even after 1 hour of SE using drugs at appropriate doses. The drugs, when effective, stop both the behavioral and the EEG seizures. The action of DZ is rapid, suggesting that subsequent treatments can be quickly initiated if SE does not stop within 1–2 minutes. On the other hand, PB takes significantly longer to work, and more time is needed to determine if the therapy will stop the seizures. Age and cause of SE may contribute to drug‐induced morbidity. Our protocol can be used to assess the efficacy and morbidity of new agents. (Supported by NIH NINDS grants K12‐NS048856 (NSADA) (HH) and R01‐NS20253 (SLM).) 1 Seiji Hatazaki, 1 Carmen Beller‐Estelles, and 1 David C. Henshall ( 1 Physiology, Royal College of Surgeon In Ireland, Dublin, Dublin, Ireland ) Rationale: Sublethal brain insults can confer protection against cell death induced by subsequent (prolonged) damaging insults. While best described for ischemia, there is some evidence brief seizures may precondition the brain against damage following prolonged seizures (e.g. status epilepticus). To study this so‐called “epileptic tolerance”, we examined neuronal damage following status epilepticus in mice that received various doses of sublethal preconditioning seizures and at various intervals between the conditioning and injurious stimuli. Methods: Seizure preconditioning activity was induced by intra‐peritoneal (i.p.) injection of kainic acid (KA) at various doses (5.0, 7.5 and 15.0 mg/kg; n = 5, plus vehicle controls, n = 9) and subsequent status epilepticus was induced by intra‐amygdala (i.a.) injection of KA (1.0 ug) at two time‐points (24 or 72 hours) after the first insult. Terminal deoxynucleotidyl transferase‐mediated dUTP nick end labeling (TUNEL) and NeuN immunostaining was used to detect neuronal damage within the hippocampus. Results: Preconditioning seizures did not induce significant hippocampal damage at any dose tested. Mice that received preconditioning seizures induced by KA i.p. at 15.0 mg/kg given 24 hours before i.a. KA showed significant reductions in TUNEL counts and significantly more surviving neurons as assessed by NeuN in the ipsilateral hippocampus, particularly the CA3 subfield (p < 0.01). There were no significant differences in the duration of high amplitude polyspike seizures evoked by i.a. KA in mice that received preconditioning. Surprisingly, increasing the time interval between preconditioning and prolonged seizures to 3 days resulted in raised mortality rates (e.g. 3/5) when mice underwent the second prolonged (i.a. KA) seizures: evidence of preconditioning‐induced sensitization to seizures. Conclusions: This study provides a new model with which to investigate the molecular mechanisms underlying endogenous neuroprotection programmes in brain. (Supported by Science Foundation Ireland grant B466.) 1 James G. Heida, 1 Jana Veliskova, 1 Libor Velisek, 1 Ondrej Cudomel, 1 Solomon L. Moshe, and 1 Aristea S. Galanopoulou ( 1 Neurology and Neuroscience, Albert Einstein College of Medicine, Bronx, NY ) Rationale: It is widely accepted that status epilepticus (SE) in early life does not result in major neuronal death as observed in adults. There is increasing evidence however that more subtle aspects of brain physiology are permanently affected. Of particular interest is the potential effect of repetitive early life seizures on the physiology of endogenous seizure controlling circuits, such as those involving the substantia nigra reticulata (SNR). The aim of the current study was to determine the effect of 3 episodes of SE, during early postnatal development, on the seizure controlling function of the SNR in rats. Methods: Three episodes of SE were induced in Sprague Dawely rat pups of both sexes on postnatal days 4 (P4), P5, and P6 by systemic injection of kainic acid ((KA) 1.25, 1.5, and 2 mg/kg i.p. respectively). Animals were left until P28 when they were implanted with bilateral cannulae aimed at the anterior SNR. At P30, rats received bilateral infusions of the GABAA receptor agonist muscimol (100ng/0.25 μl per side), which has been shown previously to have an anticonvulsant effect when infused into the anterior SNR at P30 (Veliskova & Moshé 2001). Control rats were infused with 0.25μl/SNR of saline. Flurothyl seizure threshold for the first clonic seizure was determined 30 minutes later as the amount of flurothyl necessary to induce the seizure. Subsequently, cannulae placements were confirmed histologically. Only results from animals with cannulae placement in the anterior SNR are reported. Results: Systemic injection of KA resulted in behavioral SE in all animals. Within 5 min of the injections, pups manifested scratching‐like behavior and hindlimb clonic movements, hyperactivity, tonic or tonic clonic seizures followed by swimming‐like movements intermixed with tonic or tonic clonic seizures. These seizures lasted for the entire observation period of 6 hrs. At P30 flurothyl thresholds were 179 ± 15.7 μl (saline n = 4) and 169 ± 10.9μl (muscimol n = 7) in males, and 174 ± 25.9μl (saline n = 5) and 181 ± 15.7μl (muscimol n = 8) in females. There were no statistically significant differences between any of the groups (two way ANOVA, p > 0.05, NS). Conclusions: Repetitive episodes of SE during early postnatal development alter the GABAA receptor mediated seizure controlling function of the SNR. The results suggest that early life SE produces long‐term alterations in the SNR GABAA receptor system, which leads to loss of the well‐known anticonvulsant effects of GABAA receptor activation in the SNR of rats without prior seizures. Such alterations in seizure controlling circuits may predispose the brain to future seizures and epileptogenesis. (Supported by NIH NINDS grants NS 20253, NS 045243 and a grant from the Rett Syndrome Research Foundation.) 1 William Gomes, 2 Fred Lado, 3 Nihal De Lanerolle, and 1 Hoby Hetherington ( 1 Radiology, Albert Einstein College of Medicine, Bronx, NY ; 2 Neurology, Albert Einstein College of Medicine, Bronx, NY ; and 3 Neurosurgery, Yale University, New Haven, CT ) Rationale: Reductions in N‐acetyl aspartate (NAA) have proven useful for the lateralization and localization of seizure foci in patients with intractable epilepsy. Since most patients to date have been studied only after years of intractable seizures, it is unclear, if the deficits in NAA precede the onset of overt seizures or, alternatively, are solely a consequence of chronic, intractable seizures. To determine when the reductions in NAA occur, we used a pilocarpine rat model and magnetic resonance spectroscopic imaging (MRSI). Methods: Status epilepticus (SE) was induced in male rats with pilocarpine and terminated after 1 hour with diazepam. Control rats received sham‐injections of saline (n = 10 control rats). The severity of the SE was graded according to Racine, and rats with severe generalized seizures (stages 3, 4 and 5) were included for analysis (n = 11). Two and seven days post‐SE, MRSI data were acquired with 2ul resolution. Three voxels within each hippocampus (6 voxels per rat) were selected for analysis. Following the day 7 MRSI study, the rats were maintained under anesthesia and transcardially perfused with 4% paraformaldehyde. Neuronal loss in the regions spanned by the spectroscopic images was assessed using stereologic methods as described by West. Results: Displayed in Fig. 1 are hippocampal spectra from a sham injected animal (control) and 2 and 7 days post‐SE in a pilocarpine treated rat. NAA is significantly reduced at 2 days post‐SE (27.5 ± 6.9% decrease, p < 0.001) and 7 days post‐SE (17.3 ± 6.9% decrease, p < 0.001) in comparison to sham‐injected controls (Fig. 1). Despite the substantial reductions in NAA 7 days post SE, neuronal counts averaged over the entire hippocampus showed a non‐significant reduction of 2% in comparison to sham treated animals. Conclusions: Hippocampal NAA content in this rat model of epilepsy is significantly reduced during the latent period prior to the development of overt spontaneous seizures. The reduction in NAA, 17%, is not due to neuronal loss, since neuronal numbers are not significantly decreased over the volume sampled. The presence of decreased NAA during the latent period suggests that spectroscopic studies of NAA provide an early marker of the processes underlying the development of epilepsy, prior to the manifestation of overt seizures. If validated in patients, this would provide a non‐invasive method for evaluating and identifying those patients at risk to develop epilepsy following a first provoked seizure. (figure 1) (Supported by National Institutes of Health R21 EB‐001748.) 1 Lan B. Hoang‐Minh, 1 Hector Sepulveda, 2 Mansi Parekh, 3 Angela Hadlock, 3 Wendy Norman, 3 Justin C. Sanchez, 1 William L. Ditto, 3 Michael A. King, 3 Paul R. Carney, 5 Zhao Liu, and 4 Thomas H. Mareci ( 1 Biomedical Engineering, University of Florida, Gainesville, FL ; 2 Neuroscience, University of Florida, Gainesville, FL ; 3 Pediatrics, University of Florida, Gainesville, FL ; 4 Biochemistry and Molecular Biology, University of Florida, Gainesville, FL ; and 5 Neurology, University of Florida, Gainesville, FL ) Rationale: Using MRI at 17.6 Tesla, we have studied structural changes that occur during the latent period of epileptogenesis in an animal model of mesial temporal lobe epilepsy (MTLE) and correlated these changes with histological analysis. Methods: Five 50‐day‐old adult Sprague‐Dawley male rats were anesthesized and implanted with stimulating electrodes to induce self‐sustained status epilepticus in 4 out of 5 rats. Subsequently, rats were video monitored to capture spontaneous seizures. After eight weeks, the animals were sacrificed, transcardially perfused, and the intact fixed brains imaged in vitro at 17.6 Tesla. After imaging, brains were processed for histology and stained with Fluoro‐Jade C, Timm, GFAP, Black Gold and Perl iron stains to correlate structural changes with those observed with MRI. Results: Three of 5 rats presented with observable spontaneous seizures during the epileptogenic period. These rats showed significantly decreased T2 in the pyramidal cell layer CA1 of the hippocampus and the dorsal thalamic nuclei bilaterally (Fig. 1). One presented with a cavity and increased T2 in the parahippocampal gyrus on the contralateral side of stimulation. No T2 changes were observed in the frontal cortices of these seizing rats. The two rats with no observable seizures either hadn't been stimulated or had stimulating electrode placement in the dorsal thalamus instead of the hippocampus. These rats showed no significant T2 abnormalities in any brain region. Histological analysis has been completed for one of the seizing rats and revealed mossy fiber sprouting around the dentate gyrus bilaterally, iron deposition in the dorsal thalamus and demyelination and astrocytic proliferation in the CA1 and CA3 hippocampal regions. Conclusions: These results suggest that there may be a correlation between T2 changes in the parahippocampal gyrus and the hippocampus (related to mossy fiber sprouting and gliosis on histology) and the seizure activity of animal models of MTLE. Future work will correlate these changes with in vivo MRI at different time points after electrical stimulation and structural changes using specific histological methods. (figure 1) (Supported by NIH grant R01 EB004752, the Wilder Epilepsy Research Center and University of Florida Alumni Foundation.) 1 Riikka J. Immonen, 1 Asla Pitkänen, and 1 Olli H.J. Gröhn ( 1 Department of Neurobiology, A.I.Virtanen Institute for Molecular Sciences, University of Kuopio, Kuopio, Finland ) Rationale: Mn2+‐enhanced magnetic resonance imaging (MEMRI) can reveal functional, structural, and connectional alterations in the brain with a high spatial resolution in vivo. We showed that Mn2+ injected in the entorhinal cortex is transported via perforant pathway to the dentate granule cells resulting in labeling that correlates with the density of mossy fiber sprouting (Nairismägi et al. Neuroimage 2006). Now we hypothesized that systemic Mn injection and MEMRI can be used for detection of mossy fiber sprouting. Methods: MnCl2 (45 mg/kg) was injected intraperitoneally to 10 adult Wistar rats. After 12 hours, status epilepticus (SE) was induced with kainic acid (KA) in 6 rats (MnSE group). T1 weighted (gradient echo, TE = 2.7ms, TR = 120ms) and quantitative T1 mapping (IR‐ fast spin echo, TI = 10, 400, 1000, 1600 ms) MRI was performed under 1% halothane anaesthesia at 4.7T, 3 and 25 hours after KA injection. Two months later, follow‐up MRI was performed before and 24 hours after injection of Mn. In the end, rats were perfused for histology, and brains were stained to detect neurodegeneration, gliosis, and mossy fiber sprouting. Results: At 15 h after KA injection, T1 relaxation times did not differ between the groups. This indicates that Mn accumulation during generalized seizures is not activity dependent. Two months after KA administration, Mn‐enhanced MRI indicated thickening of the dentate gyrus (DG) and proximal CA3 in T1‐weighted images in the MnSE group (260 ± 40% increase compared to controls, p < 0.05). The CA1/adjacent corpus callosum signal intensity was 1.37 ± 0.03 in the MnSE and 1.22 ± 0.03 in the control group (Fig. 1, p < 0.05), confirming the increased accumulation of Mn in the CA1 region of the KA‐injected animals. The initial histological analysis shows both mossy fiber sprouting (DG) and neuronal loss related gliosis (CA1) in the signal enhancement areas. (figure 1)Fig.1 Manganase enhanced T1‐wt images 25h (A,C) and 2 months after KA injection from two representative animals. Conclusions: MEMRI contrast after systemic Mn injection is not associated with brain activation during and after SE. Alterations in the thickness and shape of the DG and CA3 or CA1 regions of hippocampus proper can be attributed either to mossy fiber sprouting or gliosis, respectively. According to our data it is evident that MEMRI after systemic administration is able to reveal cellular level changes in the hippocampus. It remains to be studied if the accumulation of Mn into mossy fibers and/or glia can be differentiated. (Supported by Finnish Cultural Foundation of Northern Savo.) 1 John R. Ives, 2 Krista L. Gilby, and 2 Dan C. McIntyre ( 1 Department of Neuroscience, University of Western Ontario, London, ON, Canada ; and 2 Department of Psychology, Carlton University, Ottawa, ON, Canada ) Rationale: We tested the effects of TMS using a model MagPro X100 (Medtronic Canada Inc., Toronto, ON) on well‐established kindled seizures in rats as there seems to be no studies in this area. First we determined whether 60 Hz TMS could evoked seizures in kindled rats. Second, we determined whether low frequency (1 Hz) TMS (LFTMS) had any effect on the threshold or performance of kindled seizures. Methods: Since kindling was originally established in 21 rats using 2 sec of low intensity 60 Hz sinusoidal electrical stimulation, we set the TMS parameters to 2 sec of 60 Hz, biphasic, magnetic stimulation at 100%. The 21 rats, 7 amygdala and 14 ventral hippocampal kindled, previously received more than 10 generalized seizures and their thresholds were stabile. These 21 rats were also from 3 selectively bred groups, 6 “Fast kindler”, 5 “Slow kinders “ and 10 “F1 hybrids”. Two types of TMS coils were used to stimulate the rats, the round and the figure 8. The round coil had a central‐hole that was positioned over the rats' implanted headplug connector, when the kindling leads to the rat were not attached. The figure 8 coil was used when the leads were attached. In 12 ventral hippocampal kindled rats, we tested the efficacy of LFTMS against the kindled seizures in two experiments, A) 1 Hz, 70%, for 30sec, wait 1min and determine the seizure threshold and behavioral profile, and B) 1 Hz, 70%, for 5min, wait 24hrs before testing. Results: No TMS protocol alone triggered a convulsive seizure or focal after‐discharge in any rat. Further the focal after‐discharge threshold to the kindling stimulus was not affected by LFTMS. However, the latency to forelimb clonus of the kindled seizure was significantly lengthened by both LFTMS protocols, indicating a delay in the recruitment of the convulsive kindled seizure. Surprisingly, TMS did not block or even marginally obscure the online EEG from the implanted bipolar Nichrome electrodes. Conclusions: TMS alone does not trigger seizures in seizure‐prone or ‐resistant rats with a well‐established kindled focus in either the amygdala or hippocampus. 1 Hz TMS does not effect the AD threshold, but significantly affects the recruitment of the convulsive seizure, which is an anti‐epileptic effect. (Supported by Medtronic of Canada for loan of TMS equipment.) 1 Kenneth A. Jenrow, 2 Tavarekere N. Nagaraja, 3 Robert A. Knight, 1 Nancy W. Lemke, 2 Joseph D. Fenstermacher, and 1 Kost V. Elisevich ( 1 Neurosurgery, Henry Ford Health System, Detroit, MI ; 2 Anesthesiology, Henry Ford Health System, Detroit, MI ; and 3 Neurology, Henry Ford Health System, Detroit, MI ) Rationale: Epileptogenesis precipitated by status epilepticus (SE) is believed to evolve during a latent period which intervenes between SE and the emergence of spontaneous seizures. Vascular pathologies have been observed in several limbic structures in chronically epileptic animals following SE, including upregulated vascular endothelial growth factor (VEGF) among reactive astrocytes, angiogenesis, and increases in neovascular permeability. Here we have investigated the temporal evolution of these vascular pathologies following SE in the rat. Methods: SE was induced in male wistar rats (250 g – 350 g) by systemic kainic acid injection (10 mg/kg, i.v.), and was terminated four hours after onset by injection of pentobarbital (20 mg/kg, i.p.). The control group received vehicle injections and did not develop SE. Bromodeoxyuridine (BrdU) injections (50 mg/kg, i.p.) were administered daily for seven consecutive days to separate cohorts beginning 1, 7, 14, 21, 28, 35, 42, or 49 days post‐SE. Animals in each cohort were sacrificed 14 days after completing the BrdU injection series, and angiogenesis was subsequently assayed by double‐label immunohistochemistry for BrdU and von Willebrand factor. Prior to sacrifice, vascular permeability was assayed serially at 7, 14, 21, 24 and 56 days post‐SE, using gadolinium‐diethylenetriaminepentaacetic acid (Gd‐DTPA)‐enhanced magnetic resonance imaging (MRI). Gd‐DTPA was administered using a stepped down infusion protocol that maintained a relatively constant blood Gd‐DTPA level during 20 min of imaging. Pre‐contrast T1‐weighted images (T1WI) were subtracted from post‐contrast T1WI to identify enhancing areas. Results: Endothelial proliferation was significantly increased (p < 0.05) during the first two weeks (day 1 and day 7 cohorts) post‐SE, spanning the duration of the average latent period. At sacrifice, coincidently labeled cells were commonly found in abnormal neovessels within regions damaged by SE. MRI image subtraction indicated that vascular permeability increased toward the end of the latent period, coinciding with the formation of abnormal neovascular complexes within regions of epileptogenesis. Conclusions: The evolution of vascular pathologies following kainic acid‐induced SE occurred primarily during the latent period which preceded the onset of spontaneous seizures. These were observed exclusively in regions of the brain damaged by SE and consist of abnormally permeable neovascular complexes formed via angiogenesis. The timing of events in relation to SE suggests that the resultant increase in vascular permeability may bias the adjoining neuropil toward hyperexcitability and promote the emergence of spontaneous seizures. (Supported by Henry Ford Health Sciences Center: Mentored Scientist Grant #A10222) 1 Nigel C. Jones, 1 Gaurav Kumar, 4 Michael R. Salzberg, 3 Margaret J. Morris, 4 Sandra M. Rees, and 1 Terence J. O’Brien ( 1 Department of Medicine – Royal Melbourne Hospital, University of Melbourne, Melbourne, Vic, Australia ; 2 Department of Anatomy and Cell Biology, University of Melbourne, Melbourne, Vic, Australia ; 3 Department of Pharmacology and Physiology, University of New South Wales, Sydney, NSW, Australia ; and 4 Department of Psychiatry, University of Melbourne, Melbourne, Vic, Australia ) Rationale: Epilepsy patients commonly suffer from affective co‐morbidities, particularly anxiety and depression. This has widely been thought to be a consequence of the epileptic process, but recent evidence suggests there could be a common causal link. This study investigates whether the GAERS (Generalised Absence Epileptic Rats from Strasbourg) strain of rats, a phenotypic animal model of generalised absence epilepsy, also display behavioral abnormalities, and attempts to determine whether these behaviors are present prior to the development of absence seizures. Methods: Mixed gender GAERS rats were compared with Non‐Epileptic Controls in neurobehavioral tests for anxiety (elevated plus maze) and depression (sucrose‐preference test). Comparisons were made at 7 or 13 weeks, the equivalent neurodevelopmental age of adolescence in a rat, and also before and after the development of seizures. Results: GAERS rats showed increased anxiety in the elevated plus maze at both 7 and 13 weeks of age compared with NEC rats, as evidenced by a significantly lower number of open arm entries (at 7 weeks: 3.14 ± 0.53 entries (n = 14) vs 7.90 ± 1.01 (n = 20); at 13 weeks: 4.67 ± 1.03 entries (n = 15) vs 6.09 ± 0.73 (n = 11) – two‐way ANOVA, p = 0.0019). This strain of rats also displayed a significantly reduced preference for sucrose drinking water (at 7 weeks: 3.18 ± 0.66ml (n = 11) vs 6.67 ± 0.88ml (n = 3); at 13 weeks: 3.00 ± 0.84ml (n = 5) vs 4.83 ± 0.63ml (n = 12) – two‐way ANOVA, p = 0.0054), indicating an anhedonic state and suggestive of a depressive phenotype. Conclusions: This study provides strong evidence that GAERS rats, when compared with non‐epileptic controls, display anxious and depressive behaviors which are present prior to the onset of absence seizures. This suggests that affective disorders may not be a consequence of seizures, and also raises the possibility that there may be a common neurodevelopmental causality which predisposes humans to both affective disturbance and epilepsy. (Supported by an NHMRC project grant (400088), and a NARSAD Independent Investigator Award (TOB).) 1 Peter I. Jukkola, 2 Roger Strong, 2 Jaroslaw Aronowski, and 1,3 Kevin M. Kelly ( 1 Department of Neurology, Center for Neuroscience Research, Allegheny‐Singer Research Institute, Pittsburgh, PA ; 2 Department of Neurology, Stroke Program, The Medical Center at the University of Texas at Houston, Houston, TX ; and 3 Department of Neurology, Drexel University College of Medicine, Philadelphia, PA ) Rationale: Several rat strains have been studied extensively and are well‐known as genetic models of absence epilepsy. Less well‐known is the fact that many common laboratory rat strains frequently express spike‐wave discharges (SWDs) in a similar fashion. These SWDs are usually 6–9 Hz, and are associated with motor arrest of the animal. Generally, SWDs tend to be more frequent and longer in duration as the animals grow older. To our knowledge, however, no study in Long Evans rats has included a detailed observation of SWDs and how SWD expression changes throughout their lifetime. Previously, we reported that cortical infarction by middle cerebral artery/common carotid artery occlusion results in a significant reduction in the frequency and duration of SWDs compared to sham‐operated control animals (Kelly et al, 2006). In this study, we sought to extend these observations to gain a better understanding of the timecourse of SWDs and the effect of lesion‐associated changes on their expression in these animals. Methods: Five Long Evans rats lesioned by MCA/CCAO and four Long Evans sham‐operated controls were intermittently monitored by video‐EEG over a six‐month period as previously reported (Kelly et al, 2006). Fifty‐hour samples were then chosen from each animal's EEG record and visually reviewed for the presence of SWDs. The time and date of each SWD observed was documented, as well as the duration of the event. Preliminary results were obtained by plotting SWD data on a timeline for analysis by linear regression. Results: Preliminary results in control rats indicate that in 4 of 4 animals (100%) SWDs tend to increase in frequency of occurrence as the animal grows older. The average duration of SWDs tended to increase with age in 2 of 4 animals (50%), while decreasing in the other 2 animals (50%). Preliminary results in lesioned animals show a trend toward both greater frequency of occurrence and longer duration of SWDs in 2 of 5 (40%) of animals, and a decrease in both frequency of occurrence and duration in 2 of 5 animals (40%). One animal (20%) showed an increase in the frequency of occurrence, but a decrease in the average duration of SWDs. Conclusions: Since 7 of the 9 animals included in this study showed a trend toward an increase in the frequency of occurrence of SWDs over the six‐month monitoring period, these studies appear to be consistent with reports of an aging‐related increase in SWD expression in other rat strains. Changes in average duration of SWDs over time appear to be variable animal to animal. No reported changes in SWD expression are clearly related to the lesion. (Supported by RFA 01–07–26 to KMK.) 1 Tohru Kamida, 1 Hiroshi Ooba, 1 Mitsuhiro Anan, 1 Minoru Fujiki, and 1 Hidenori Kobayashi ( 1 Neurosurgery, Oita University Faculty of Medicine, Yufu, Oita, Japan ) Rationale: Nitric oxide (NO) is a free radical synthesized from L‐arginine by a Ca2+‐calmodulin‐dependent NO synthase (NOS). The functional meaning of NO and NOS in epilepsy remains controversial despite previous studies from animal models. In present study, we investigated NOS expression in the hippocampus of rat with pilocarpine‐induced seizures. Methods: Adult male Wistar rats (200–250g) were pretreated with atropine methylbromide (1 mg/kg s.c.) and 30 min later with pilocarpine hydrochloride (330 mg/kg i.p.). After 40 min, the Status epilepticus (SE) was terminated by injection of diazepam (4 mg/kg i.p.). Animals were sacrificed 1 day, 3days, 2 wk (silent phase) or 1 month (chronic phase) after status. Neuronal NOS (nNOS), inducible NOS (iNOS) and epithelial NOS (eNOS) expressions and the density of mossy fiber sprouting (MFS) in the hippocampus were evaluated using immunohistochemistry and neo‐Timm's histological procedures. Results: iNOS and eNOS expressions were more strongly upregulated than nNOS expression at both phases. nNOS expression gradually increased until 3 days and decreased after 2 wk. iNOS expression had no alteration at both phases. eNOS expression tended to decrease after 2 wk. The density of MFS gradually increased and was strongest at 1 month. Conclusions: Our data indicate that nNOS and eNOS with the strongest upregulation in the silent phase may have a correlation with epileptogenesis while iNOS with the durability of upregulation in both phases may have something of influence on the hippocampus also after the epileptogenesis was acquired. 1 Rafal M. Kaminski, 1 Hamid Sheikhi, and 1 Michael A. Rogawski ( 1 Epilepsy Research Section, NINDS/NIH, Bethesda, MD ) Rationale: 11‐Beta hydroxylase is involved in final steps of steroid hormone synthesis. Inhibition of this enzyme leads to accumulation of hormone precursors, which may eventually enter an alternative metabolic pathway and become neurosteroids. Neurosteroids are potent anticonvulsants acting via positive modulation of GABAA receptors. We have previously shown that they are particularly effective against 6 Hz seizures in mice. In the present study we investigated whether inhibition of 11‐beta hydroxylase protects against 6 Hz seizures and whether this effect is associated with increased neurosteroid synthesis. Methods: Male NIH Swiss mice were the experimental subjects. The animals were pretreated with either metyrapone or etomidate, which both are potent inhibitors of 11‐beta hydroxylase, and challenged at various time points with corneal 6 Hz electrical stimulation (32 mA, 3 sec.) to induce seizures. Separate groups of animals were additionally treated with finasteride to test whether seizure protection following administration of metyrapone or etomidate is attenuated by inhibition of neurosteroid synthesis. Results: Metyrapone afforded dose‐dependent protection against 6 Hz seizures 30 min. after injection (ED50 = 190.6 mg/kg). However, the potency of metyrapone was much higher (6‐fold increase) when the animals were tested 6 hrs. following injection (ED50 = 30 mg/kg). Etomidate displayed very similar pattern of protection against 6 Hz seizures with respective ED50 values of 4.5 and 1.7 mg/kg. Finasteride partially attenuated the anticonvulsant effects of both 11‐beta hydroxylase inhibitors. Conclusions: Inhibition of 11‐beta hydroxylase produces prolonged anticonvulsant actions in mice. This effect may be associated with an increased neurosteroid production, which could be responsible for protection from seizures. 1 Irina Kharatishvili, 1 Riikka Immonen, 1 Olli Grohn, and 1 Asla Pitkanen ( 1 A.I. Virtanen Institute, Kuopio University, Kuopio, Finland ) Rationale: The risk of posttraumatic epilepsy positively correlates with the severity of brain injury. Up to 50% of patients with penetrating head injury can develop epilepsy later in life, whereas for moderate injury the incidence of epilepsy is only 4.2%. Our previous study with lateral fluid percussion brain injury model (FPBI) demonstrated that severe impact strength resulting in 31–33% mortality, causes development of spontaneous recurrent seizures in 43–50% of experimental animals (Kharatishvili et al., Neuroscience 2006;140:685–97). In this study, we aimed to examine if in FPBI model, similarly to humans, probability of developing epilepsy is associated with injury severity. Methods: Head trauma was induced by lateral FPBI in adult Sprague‐Dawley rats (n = 15, 5 shams). The pressure of the applied impact, as recorded by a transducer, varied from 2.3 to 3.3 atm. The low mortality rate (23%) indicated moderate severity of the injury. Quantitative MRI was used to monitor the progression of the lesion starting at 3 hrs post‐injury and repeated at 3 d, 9 d, 2 month‐, 3 month‐ and 6 month post‐injury time‐points. Morris water maze test (MWM) was used to assess the long‐term behavioral outcome (spatial learning and memory) 7 months after FPBI. After that, 24h/day continuous video‐EEG monitoring was performed to assess the development of seizure activity; animals were implanted with depth electrode in the ventral hippocampus ipsilateral to the injury site and contralateral cortical electrode, and followed by continuous video‐EEG for 2 + 2 weeks with 3 week interval for up to 12 months. Results: The mean volume of the cortical damage at 6 months post‐injury was 40.8 ± 38.1mm3 (range 2.7–130.0 mm3). MWM test demonstrated no difference in spatial learning in injured rats compared to shams (p>0.05). EEG monitoring revealed 2–20 sec runs of 6–8 Hz monomorphic rhythmic activity with a spike‐like negative phase in 70% rats both in the FPBI and control groups. This rhythm was registered in drowsiness over the injury side and was responsive to movement. No epileptiform discharges/seizures were recorded during the entire monitoring period. Conclusions: Rats with moderate FPBI had substantial cortical lesion. However, the hippocampus‐dependent spatial learning and memory deficit was not as pronounced as previously found in animals which developed epilepsy after severe FPBI (p < 0.05). Also, none of the rats had electrographic seizures. In lateral FPBI, the moderate force of the impact results in progressive brain damage but does not increase the likelihood of posttraumatic epileptogenesis. (Supported by Academy of Finland, Sigrid Juselius Foundation, CIMO, EUCARE, Finnish Neurological Foundation.) 1 Sookyong Koh, 2 Carlo Condello, 2 Jaime Grutzendler, and 1 Hyokwon Chung ( 1 Neurology, Children's Memorial Hospital, Northwestern University School of Medicine, Chicago, IL ; and 2 Neurology and Physiology, Northwestern University School of Medicine, Chicago, IL ) Rationale: Microglia are the resident immune cells in the central nervous system (CNS) and constitute the first line of defense against pathological changes within the CNS microenvironment. Growing evidence suggests that pro‐ and anti‐inflammatory molecules are synthesized during seizures in the brain at the sites of seizure initiation and propagation, and that rapid induction of these inflammatory mediators may play a role in the pathophysiology of epilepsy. We hypothesize that activation of microglia and subsequent increases in cytokines, chemokines, and complements are the key initiating events for seizure‐induced inflammatory responses. Methods: In a total of 40 heterozygous Cx3cr1GFP/+ adult mice, Cx3cr1 encoding fractalkine chemokine receptor has been replaced by a green fluorescent protein (GFP) reporter gene by targeted deletion via homologous recombination in embryonic stem cells. Seizures were induced by intraperitoneal injection of kainic acid (KA, 20 mg/kg); control with phosphate buffered saline (PBS). The mice were sacrificed 30 min, 1day, 2 days, 7days, or 14 days after the injections. Three 50‐μm‐thick horizontal sections (6 hippocampal sections per brain) were selected for quantification of microglia. An average percent of area above the set threshold (containing green fluorescent cells/hippocampal hilar region capture at 20X magnification) and number of fluorescent element were calculated per animal, and comparison was made between KA and Control animals. Results: Enhanced fluorescent signaling indicative of microglial activation was already apparent in the hippocampus 30 minutes after KA‐induced seizures. Within 2 days, microglia in the hippocampus appeared enlarged and more numerous. A nearly 2 fold increase in the area of fluorescent cells was noted in KA treated mice compared to control littermates (PBS:1.67 ± 0.15 vs KA:3.08 ± 0.28, p < 0.005). Also, significant microglial proliferation occurred 2 days after KA‐induced seizures (113 ± 5 vs 189 ± 9, p = 0.0002). These increases in the area and number of fluorescent cells persisted to 7 days after KA (area%: 1.64 ± 0.10 vs 2.35 ± 0.21, p < 0.01; spot counts: 102 ± 3 vs 141 ± 5, p < 0.0001). Conclusions: Taking advantage of transgenic mice in which all microglia are fluorescently labeled, we have unequivocally demonstrated an acute and persistent microgliosis after KA‐induced seizures. Neuroinflammation provoked by seizures thus appears to be mediated via activation of resident CNS immune cells, microglia. Our data implicates microglia in seizure‐induced neuronal excitability and degeneration, and suggests an inhibition of microglia activation as a potential target for therapeutic intervention. (Supported by Child Neurology Foundation and K02NS048237.) 1 Seogkyoung Kong, 2 Ariana Lorenzana, 2 Jon E. Vanleeuwen, and 2 Paula E. Schauwecker ( 1 Biochemistry and Molecular Biology, University of Southern California, Los Angeles, CA ; and 2 Cell and Neurobiology, University of Southern California, Los Angeles, CA ) Rationale: Susceptibility to seizure‐induced excitotoxic cell death varies among inbred mouse strains due to variation in genetic background. Our QTL mapping studies have shown that distal chromosome 18 (Sicd1) contains a gene(s) conferring different susceptibility to seizure‐induced excitotoxic cell death between C57BL/6J (B6) and FVB/NJ (FVB) inbred mouse strains. We have verified that the causal gene(s) reside in the Sicd1 QTL utilizing congenic strains. We found the congenic strains having the Sicd1 QTL‐resistant locus from a cell‐death resistant strain (B6) and the remainder of their genome from a cell‐death susceptible strain (FVB) are less susceptible to seizure‐induced excitotoxic cell death as compared to the cell‐death susceptible strain (FVB). We hypothesized that a strain‐specific variation in one or more genes is responsible for the Sicd1 QTL effect. Methods: We have focused on two potential candidate genes, Galr1 and Mbp, based on their putative role in modulating seizure susceptibility and excitatory amino acid‐induced cell death. To test this hypothesis we sought to identify sequence variants in the Galr1 and Mbp genes between B6 and FVB strains using DNA sequencing. As a complimentary method to identify the Sicd1 QTL gene(s), we have examined changes in gene expression at the mRNA level of Galr1 and Mbp in hippocampal homogenates from both the parental inbred and the reciprocal congenic strains. Studies are underway to detect aberrant splicing based on cDNA sequencing and differential expression of Galr1 and Mbp at the protein level between our parental strains, B6 and FVB, as well as our reciprocal congenic strains. Results: Based on preliminary results, no coding sequence variants have been detected in the Galr1 and Mbp genes. We have found genotype‐dependent differences in Galr1 mRNA expression between B6 and FVB parental strains using qRT‐PCR (p = 0.023). Differential expression of hippocampal mRNA for Galr1 was also found by qRT‐PCR between the FVB.B6‐Chr18 congenic strain and the parental FVB strain. In contrast, no statistically significant differences in Mbp mRNA expression were observed either between the parental strains or recombinant congenic strains (p = 0.816). Conclusions: Based on the expression variation and its putative role in modulating seizure susceptibility and excitatory amino acid‐induced cell death we consider Galr1 as a strong candidate gene responsible for the Sicd1 QTL effect. Follow‐up studies whether there is functional expression differences in Galr1 gene between B6 and FVB would support Galr1 as a Sicd1 QTL gene. Results from the proposed studies will add to our understanding of the molecular determinants critical for the pathogenesis of seizure‐induced excitotoxic cell death. (Supported by NS038696 to P.E.S.) 1 Gaurav Kumar, 1 Nigel C. Jones, 2 Michael R. Salzberg, 4 Sandra M. Rees, 1 Abbie J. Coupar, 1 Terence J. O’Brien, and 3 Margaret J. Morris ( 1 Department of Medicine – Royal Melbourne Hospital, University of Melbourne, Melbourne, Vic, Australia ; 2 Department of Psychiatry, University of Melbourne, Melbourne, Vic, Australia ; 3 Department of Physiology and Pharmacology, University of New South Wales, Sydney, NSW, Australia ; and 4 Department of Anatomy and Cell Biology, University of Melbourne, Melbourne, Vic, Australia ) Rationale: We have previously demonstrated that low‐dose corticosterone (CS) supplementation, used as a model of the effect of chronic stress/depression, accelerates epileptogenesis in the amygdala kindling rat model of temporal lobe epilepsy (TLE). This study examined the contributions to this effect of the subtypes of glucocorticoid receptors: mineralocorticoid (MR, high‐affinity for CS) and glucocorticoid (GR, low‐affinity for CS). Methods: Female Non‐Epileptic Control rats 10–13 weeks of age were ovariectomized and implanted with a bipolar electrode into the left amygdala. Five experimental groups were studied: 1. Water‐control (n = 6), 2. CS treated (n = 6), 3. CS + Mifepristone (Mif – GR‐antagonist, 25 mg/kg sc, n = 7), 4. CS + Spironolactone (Spir – MR antagonist, 50 mg/kg, n = 5), and 5. CS + Mif (25 mg/kg) + Spir (50 mg/kg) (n = 6). Corticosterone was dissolved in drinking water (6 mg/100 ml) and administered ad libitum (groups 2–5) throughout the kindling period. Rats were injected subcutaneously with vehicle (Groups 1 and 2) or the relevant antagonist in vehicle (Groups 3–5) twice daily for the entire kindling period. Rapid Amygdala Kindling (RAK), which began two days after initial treatment, consisted of 10 sec bursts of 400μA every 15–20 mins, with up to 24 stimulations per day. Results: Groups differed significantly in the number of stimulations required to reach the ‘fully kindled state’ defined as five Class V seizures on the Racine scale (p = 0.04, Kruskal‐Wallis ANOVA). The CS‐treated group kindled faster than either the water‐treated control group (p < 0.05) or either of the three CS+antagonist groups (p < 0.10). No significant differences were found between the any of the CS+antagonist groups and the water‐treated controls. Conclusions: The data implicate both high affinity (MR) and low affinity (GR) glucocorticoid receptors in the mechanism by which low‐dose CS accelerates kindling epileptogenesis. This may have implications for understanding the potential effects of chronic stress and depression in initiating and/or exacerbating TLE. (Supported by an NHMRC project grant (400088), and a NARSAD Independent Investigator Award (TOB).) 1 Qizhi Zhang, 2,3 Zhao Liu, 2,3 Paul R. Carney, and 1 Huabei Jiang ( 1 The J. Crayton Pruitt Family Department of Biomedical Engineering, University of Florida, Gainesville, FL ; 2 Division of Pediatric Neurology, University of Florida, Gainesville, FL ; and 3 McKnight Brain Institute, University of Florida, Gainesville, FL ) Rationale: Photoacoustic tomography (PAT) is an emerging imaging modality that has the potential to image the dynamic function of the brain due to its unique ability of imaging biological tissues with high optical contrast and ultrasound resolution. Here we report the first application of our finite element based PAT method for imaging of seizure‐onset zone in an animal model of focal seizures. Methods: Male Harlen Sprague‐Dawley rats, weighing 50∼60g, were used in this study. The rats were anesthetized with urethane 1 mg/g in all the experiments. Focal seizures were induced by microinjection of a GABAA antagonist bicuculline methiodide (BMI) into the left parietal neocortex. A mechanical scanning photoacoustic system with single acoustic transducer was employed for collecting the light‐induced acoustic signals. A reconstruction algorithm was used to generate PAT images. Results: Following microinjection of BMI into the left parietal cortex, focal seizures were induced and confirmed with electroencephalogram recordings which showed high amplitude spike and wave discharges. These seizures were accurately localized tomographically by PAT. The dynamic changes in PAT imagings were comparable to EEG changes during seizures. Conclusions: Our result showed that PAT is an emerging non‐invasive imaging modelity to localize the seizure‐onset zone. Compared to other neuroimaging modalities, PAT has not only the advantage of high spatial and temporal resolution, but also the ease of use. It has the potential to bring brain imaging to the bedside, clinic and daily life, and coincide a scan with a seizure such that an epileptic focus and seizure occurrence can be visualized in real‐time. (Supported by: This work was supported in part by the National Institutes of Health (R01 AR048122) and Wilder Epilepsy Research Center of the University of Florida.) 1 Ariana Lorenzana, 1 Ruth I. Wood, and 1 Paula E. Schauwecker ( 1 Cell and Neurobiology, University of Southern California, Los Angeles, CA ) Rationale: Studies have shown a correlation between estrous cycle and seizure susceptibility in humans and most mammals. This effect is also seen during menopause, or reproductive senescence. It has been hypothesized that the hormonal fluctuations throughout the various stages of estrous exert some type of influence on seizure occurrence, prevalence, and severity. The goal of this study was to analyze the effect of ovarian hormone cyclicity at different stages of estrous, the loss of ovarian hormones and the relationship to kainate‐induced seizure susceptibility and seizure induced cell death. Methods: Control female mice of the FVB/NJ (FVB) strain were administered kainic acid (KA) at different stages of the estrous cycle. Estrous cycle was determined by vaginal cytology and only regularly cycling animals were chosen for the experiments. KA or saline was administered systemically to female FVB mice at various stages of the estrous cycle, to ovariectomized (OVX) females, OVX females treated with estradiol (OVX + E), and VCD‐treated female mice. Mice were monitored for 4 hours for seizure onset, severity, and activity and scored for latency to onset of first severe seizure and duration of severe seizures. After 7 days of recovery, neuronal damage was evaluated throughout the hippocampus proper. Results: Among ovary‐intact female FVB mice, estrous cyclicity was without effect on seizure susceptibility. In particular, no significant differences in latency to onset or duration of severe seizures were noted. In contrast, in the OVX + E female mice, seizure duration was significantly increased, and latency to onset of seizures was reduced, as compared to OVX and ovary‐intact female mice. The extent of hippocampal cell death among ovary‐intact mice was not affected based on estrous cycle staging. In particular, no significant differences in the extent of cell death were observed between ovary‐intact female FVB mice injected with KA during estrus, when gonadal hormone levels peak, or during diestrus, the nadir of gonadal hormone levels. Conclusions: Our results suggest that estradiol may lower the threshold for seizure activity. While we found no modulatory effects of estradiol on seizure susceptibility or cell death based on estrous cyclicity, it is important to note that two different models for the modulation of ovarian steroid levels were used. In one model, ovariectomy (OVX) causes a complete disruption of gonadal hormones. In contrast, ovarian steroid levels vary across the estrous cycle in ovary‐intact female mice. In addition, OVX + E mice have none of the protective effects of progesterone present. Studies are ongoing to compare two models of reproductive senescence: the OVX model and the chemically‐induced VCD model. Insight into the mechanisms of brain related‐changes as a result of the loss of ovarian steroids will help identify risk factors for seizure susceptibility and cell death in aging women. 1 Jesus Machado‐Salas, 3 Patricia Guevara, 1 Dongsheng Bai, 1 Reyna Duron, 2 K. Amano, 2 T. Suzuki, 2 Kazuhiro Yamakawa, 1 Eaian Cornford, and 1 Antonio Delgado‐Escueta ( 1 Epilepsy Center of Excellence, UCLA David Geffen School of Medicine, Los Angeles, CA ; 2 Laboratory for Neurogenetics, Riken Brain Science Institute, Saitama, Japan ; and 3 Neurociencias. Neurobiologia, Instituto Nacional de Neurologia y Neurocirugia, Mexico, DF, Mexico ) Rationale: We have identified two genes associated to Lafora disease.In our null mutant mice (EPM2A), we have studied the immunocytochemical architecture of the neurocytoskeleton and the Lafora bodies. Methods: Our Laforin‐deficient knockout mice were produced by deleting the Laforin domain coding region of the EPM2A gene (Hum Mol Genet 11:1252–1262.2002). Similar groups of control and homozygous null mutant mice were sacrificed at fixed intervals and their encephalons were fixed with glutaraldehyhde and processed with PAS and immunocytochemical techniques. Monoclonal antibodies were used against neurofilaments M and L (NFM/L) . Results: Our experimental animals showed neuronal death at 2 months of age, in the absence of apoptotic changes or accumulation of Lafora bodies (LfB). At 9 months of age the number of LfB was very prominent in large nerve cells. They were positive to PAS, Ubiquitin and AGEP. Large neurons from the brainstem, hippocampus and cerebellum showed abnormal immunoreaction for NFM/L, which, in some cases, very much resembled neurofibrillary degeneration. Clusters of fragmented and distorted neurites were seen, very much like “senile” (neurite) plaques. Analysis of LfB has revealed some unexpected structural components that appear to establish an anatomofunctional link between them and the neurocytoskeleton . Conclusions: Our observations apppear to indicate that our model of Lafora disease presents morphological evidence of a neurodegenerative process in absence of apoptotic cell death. Lafora bodies do not primarily participate in neuronal death. With age, they increase in size and number, accumulating poliglucosans and, probably, also misfolded proteins. The presence of a complex structure in the ring of Lafora bodies, allows us to postulate an active role for these still enigmatic inclusions. (Supported by NINDS 5RO1NS042376–03.) 1 Pavel Mares, 1 Denisa Lojkova, and 1 Anna Mikulecka ( 1 Department of Developmental Epileptology, Institute of Physiology, Academy of Sciences, Prague, Czech Republic ) Rationale: Many antagonists of NMDA receptors exhibit a powerful anticonvulsant action but also strong unwanted side effects. To utilize potential of these drugs attention is now focused on specific NMDA antagonists. We decided to study effects of representants of the two above mentioned classes low‐affinity noncompetitive antagonist memantine and NR2B‐specific antagonist ifenprodil in immature rats. Methods: Two models of epileptic seizures were used – pentetrazol(PTZ)‐induced motor seizures and cortical epileptic afterdischarges (ADs) induced by stimulation of the sensorimotor cortex. Male rats of the Wistar strain 12, 18, and 25 days old were tested. The same age groups were studied behaviorally to find possible side effects. Spontaneous behavior in the open field was registered and motor performance was checked by means of age‐specific tests. Memantine was administered intraperitoneally in doses of 10, 20 and 40 mg/kg, ifenprodil in doses of 10, 20 and 40 mg/kg. Each age and dose group in each experiment was formed by 8 to 10 animals. Results: Memantine suppressed generalized tonic‐clonic seizures (GTCS) in a dose‐dependent manner in all age groups. In contrast, it potentiated minimal clonic seizures. Ifenprodil suppressed only the incidence of the tonic phase of GTCS in 12‐ and 18‐day‐old rats. Memantine increased threshold intensities for elicitation of spike‐and‐wave type of AD in 12‐ and 18‐day‐old rats as well as threshold for the other (mixed) type of ADs (due to a transition of epileptic activity into the limbic system) in 25‐day‐old rats. Duration of ADs was shortened in all age groups. Ifenprodil pretreatment decreased threshold intensities for elicitation of mixed type of ADs in 18‐ and 25‐day‐old rats and increased duration of these ADs in 25‐day‐old rats. If administered when the animals allready passed repeated ADs, it significantly shortened ADs in 12‐ and 18‐ but not 25‐day‐old animals.All doses of memantine increased locomotor activity in 12‐ and 18‐day‐old rats, the change in the oldest group was not significant. The 20‐ and 40‐ mg/kg doses impaired motor performance in all age groups. Ifenprodil increased locomotor activity only in 12‐day‐old rats and motor performance was changed only with the highest dose in 12‐ and 25‐day‐old rats. Duration of the impairment of motor performance by either drug was shorter than duration of an anticonvulsant effect. Conclusions: Memantine exhibits anticonvulsant action as well as side effects described for common NMDA atagonists. Ifenprodil has an action mostly in the youngest group what is in agreement with developmenal decrease of NR2B subunits in the NMDA receptors. (Supported by research projects No. LC554 and AV0Z 50110509.) 1 Steven T. Marsh, 1,2 Levi B. Good, 1 Kevin J. Garvey, and 1,2 David M. Treiman ( 1 Neurology Research, Barrow Neurological Institute, Phoenix, AZ ; and 2 Harrington Department of Bioengineering, Arizona State University, Tempe, AZ ) Rationale: Status epilepticus (SE) is a dynamic state. There are progressive changes in clinical behavior, response to treatment, extent of neuronal damage, and EEG pattern during prolonged clinical and experimental status epilepticus. SE also causes deficits in learning and memory, but progressive impairment with increasing severity of SE has been little studied. We tested the hypothesis that the EEG stage during SE predicts the extent of subsequent cognitive impairment. Methods: Forty‐three 250–280 gm male Sprague‐Dawley rats were implanted with 4 stainless steel epidural electrodes for EEG recording. Status epilepticus was induced by 3 mmol/kg IP LiCl followed 24 hrs later by 30 mg/kg pilocarpine. EEG was continuously recorded before and after injection, and SE was stopped at EEG stage I (discrete seizures), III (continuous ictal activity), IV (continuous ictal activity punctuated by flat periods), or V (periodic epileptiform discharges) (Treiman et al., Epilepsy Res5:49–60, 1990) with IP diazepam (10 mg/kg) plus phenobarbital (25 mg/kg). Control rats were given PB and DZM prior to SE induction and thus did not develop SE. Spatial learning and memory were assessed one week prior to SE induction and one week after SE using a modification of the Morris Water Maze (Morris, J Neurosci Methods11:47–60, 1984). Learning impairment was tested in six trials on each of four sequential days. Effect of the SE stage on the mean escape times to a submerged platform were determined by a multiple analysis of variance. Pair‐wise multiple comparisons of groups were done by Fisher's least significant difference. The mean escape time for the trials on test day 1 post‐SE was compared with the mean of the trials on test day 4 pre‐SE to assess the effect of SE on spatial memory. Results: Mean time to stop SE from pilocarpine injection was: Stage I 25 min, Stage III 65 min, Stage IV 100 min, and Stage V 210 min. There were no significant differences between these groups in pre‐SE water maze performance. However, post‐SE day 1 mean escape times were significantly worse compared with pre‐SE day 4 (test of memory) as well as the mean escape times for all test days (test of learning) in Stage IV and Stage V rats when compared to controls, Stage I, and Stage III rats (p < 0.001). No significant differences were found between control, Stage I and Stage III rats, nor between Stage IV and Stage V rats. Conclusions: SE results in a marked impairment of both visual‐spatial memory and learning in the late stages of SE. Stage IV, and V rats demonstrated a greater impairment than controls, or Stage I‐III. Thus progressive cognitive impairment, in addition to other dynamic changes cited above, is also an effect of prolonged SE. These data provide further evidence that the EEG stage during prolonged SE is a marker of its severity. (Supported by the Barrow Neurological Foundation.) 1 Stella M. Valiensi, 1 Oscar A. Martinez, 1 Ricardo C. Reisin, 1 Rubens Granillo, 1 Silvia Christiansen, and 1 Federico J. Bottaro ( 1 Neurology, Hospital Italiano, Buenos Aires, Buenos Aires, Argentina; Neurology, Hospital Britanico, Buenos Aires, Buenos Aires, Argentina; Neurology, Hospital Britanico, Buenos Aires, Buenos Aires, Argentina; Neurology, Hospital Italiano, Buenos Aires, Buenos Aires, Argentina; Pathology, Hospital Italiano, Buenos Aires, Buenos Aires, Argentina; and Internal Medicine, Hospital Britanico, Buenos Aires, Buenos Aires, Argentina ) Rationale: S.E. can produce brain damage. Propofol is an anaesthetic with anticonvulsant effect, but its use is limited by complications such as Rhabdomyolisis® There are no experimental studies that can demonstrate neuroprotective activity of Propofol in S.E, neither R. attributed to this drug . Objectives: 1 – To evaluate if Propofol in an animal model of Pilocarpine – induced S.E can prevent neuronal damage in hippocampal structures. 2 – To evaluate if Propofol in S.E. in this experimental model can cause R. Methods: Adult male Wistar rats, (200 to 300 g) were used. Biosecurity rules defined by the Institute of Basic Sciences and Experimental Medicine of our Hospital were followed. 1.CONTROL GROUP : n = 9 rats: 4 cm3 saline solution intraperitoneal (i.p) 2.PILOCARPINE GROUP : n = 10: 1 mg/kg N‐Hyoscine butylbromide+ 350 mg/kg of i.p Pilocarpine. 3.PROPOFOL GROUP : n =10: 1 mg/kg N‐Hyoscine butylbromide+ 350 mg/kg of i.p Pilocarpine. After 30 minutes of S.E., 20 mg/kg Propofol. Behaviour and Scalp EEG monitoring were evaluated. Brains were removed after 2 hours,and stained with hematoxylin‐eosin (H.E). The hippocampus was assessed. 4 histologic groups with signs of neuronal damage(dark and shrunken) were identified . 1) No damage 2) Mild damage: (+) 2–3 neurons per field (400 X) 3) Moderate damage (++) 3–5 neurons per field 4) Severe damage (+++) 5–7 neurons per field The muscle quadriceps femoralis was removed and stained with H.E. The presence of macrophage in and around the fibres was evaluated . The pathologist performed analyses blinded to the treatment received by the rats. Results: Brain damage in every area was analyzed with a non‐parametric test (Wilcoxon sum rank).A p < 0.05 was considered a significant difference. Between Group1 and Group 2, greater difference was found in temporal neocortex and subiculum in Group 2. Significant damage decrease was observed in Group 3, with less damage in the CA3–4 regions (p = 0.0019). BEHAVIOUR All the rats had S.E. EEG: Group1: normal EEG. Group 2: continuous epileptiform activity . Group 3: continuous epileptiform activity during S.E. Propofol produced suppression of the epileptiform activity. R.: 2 rats in each group(Group 2 and Group 3) had signs of R.There were no statiscally significant differences between treatment groups. Conclusions: Neuronal damage was less severe in the rats treated with Propofol, with decrease neuronal damage regarding the CA region, specifically in the CA3–4 regions, which suggests its neuroprotective effect. R. was not considered a statiscally significant finding. 1 Brona M. Murphy, 2 Clara K. Schindler, 2 Sachiko Shinoda, 1 Carmen Bellver‐Estelles, 1 Mark Dunleavy, 2 Robert Meller, 1 Jochen H.M. Prehn, 2 Roger P. Simon, and 1 David C. Henshall ( 1 Physiology and Medical Physics, Royal College of Surgeons In Ireland, Dublin, Ireland ; and 2 Robert S. Dow Neurobiology Laboratories, Legacy Research, Portland, OR ) Rationale: Seizures whether brief or prolonged are capable of causing neuronal death. Apoptosis signaling pathways play a role is such neuronal loss. Previous studies have demonstrated that experimental seizures can activate the apoptotic intrinsic mitochondrial cell death pathway and that components of this pathway are altered in involved brain structures in patients with temporal lobe epilepsy. We have developed a mouse model of seizure‐induced neuronal death with features of programmed cell death. Presently, we examined the activation of the mitochondrial apoptotic pathway in this model and contrasted this to events within human epilepsy brain. Methods: The experimental methods used in this study included, a mouse seizure model: intraamygdala kainic acid injection in adult mice, terminated after 40 min by diazepam, human temporal lobe specimens: hippocampal resections obtained following surgery for intractable epilepsy compared to matched autopsy controls, Western blotting and Immunofuorescence. Results: Seizures evoked by intraamygdala kainic acid in C57BL/6 mice caused ipsilateral death of CA1 and CA3 neurons within the hippocampus. Western blotting revealed seizures induced overexpression of Bax, cytochrome c release and activation of caspase‐9 and −7. Analysis of hippocampi from patients with intractable epilepsy revealed cleaved caspases 9 and 7 were present. In searching for relevant anti‐apoptotic proteins we found no changes to expression of Bcl‐2 and Bcl‐xl but a rapid decline in Bcl‐w levels at both 0.5 (p < 0.01) and 4 h following seizures. Reasoning that loss of Bcl‐w may promote cell death after seizures we investigated hippocampal injury following seizures in mice lacking Bcl‐w. Bcl‐w knockout mice exhibited no obvious neuroanatomical differences to their wildtype littermates and expressed similar hippocampal levels of a range of apoptosis‐associated proteins including Bcl‐2 and similar amygdala levels of the kainic acid (KA) receptor GluR5–7. Seizures evoked by intraamygdala KA induced significantly more hippocampal neuronal loss and DNA fragmentation (terminal deoxynucleotidyl dUTP nick end labeling) in Bcl‐w knockout mice compared to wildtype mice. Extending these data, we found higher cytoplasmic levels of Bcl‐w in hippocampus from patients with TLE compared to autopsy controls. Conclusions: These data identify Bcl‐w as an endogenous neuroprotective gene that may be a therapeutic target for the treatment of seizure‐induced brain injury and temporal lobe epilepsy. (Supported by the Health Research Board, Ireland, Wellcome Trust, UK and NIH/NINDS, USA.) 1 Anna Tostevin, 1 David Liu, 1 Terence O’Brien, 2 Rod Hicks, 3 David Williams, and 1 Damian Myers ( 1 Department of Medicine, Royal Melbourne Hospital, University of Melbourne, Melbourne, Victoria, Australia ; 2 Centre for Molecular Imaging, Peter MacCallum Cancer Institute, Melbourne, Victoria, Australia ; and 3 Department of Physiology, The University of Melbourne, Melbourne, Victoria, Australia ) Rationale: Reduced GABAA/central benzodiazepine receptor (GABAA/cBZR) expression, mossy fibre sprouting (MFS) and hippocampal sclerosis are common pathological changes seen in both human and animal models of temporal lobe epilepsy (TLE). This study was designed to investigate immediate (24 hours) and long‐term changes (2, 4 and 6 weeks) in GABAA/cBZR expression, neuronal loss and MFS in the hippocampus in the kainic acid (KA)‐induced status epilepticus (SE) model of TLE. Methods: Age‐matched non‐epileptic control male rats were randomized into control and KA groups (at least n = 6/group). MFS and GABAA/cBZR expression, neuronal loss and MFS were measured in hippocampal subregions, dentate gyrus (DG), CA3 and CA1 in both the control and epilepsy groups. GABAA/cBZR expression (Bmax value) was measured by saturation‐binding analysis using (3H)‐Flumazenil, a cBZR antagonist. Stereological counting was performed for cell counting and volume analysis of the CA3 SPc subregion of the hippocampus. MFS into the dentate molecular layer was quantified from Timm‐stained brain slices. Results: At 24 hours post‐SE, Bmax was significantly increased compared to controls in the whole hippocampus (+29.4%± 5.7; p < 0.001) and parietal cortex (Par Ctx) (+30.6%± 12.8; p < 0.05) as well as in all subregions of the DG, CA1, and CA3 with the exception of CA3 SPc (Repeated measure ANOVA, p < 0.02). Calculated from absolute values, no changes in receptor expression were seen over 2–6 weeks. There was no difference in MFS 24 hours post‐SE compared with controls but MFS increased in the dentate molecular layer at 2 weeks post SE (20% above control; p < 0.05) and showed a progressive increase over time reaching 70% above controls (p < 0.001) at 6 weeks. At 24 hours, statistically significant neuronal loss occurred (−24 ± 9.3%) in the CA3 SPc region (control = 6, KA group = 6, p < 0.05) associated with a significant increase in GABAA/cBZR expression per surviving neuron. No recovery of neurons was apparent through to the end of 6 weeks post‐SE. Conclusions: Changes in GABAA/cBZR expression, cell loss and synaptic reorganization occur during epileptogenesis and may influence the severity of spontaneous recurrent seizures (SRS). Elevated receptor expression observed early after status epilepticus (within 24 hours) may be a protective response by the brain to dampen excitability and increase resistance to future seizures. MFS, a form of neuronal reorganization, is a prominent feature in the KA model of TLE progressively increasing during the latent epileptogenic period up to onset of SRS. GABAA/cBZR expression and MFS were not correlated until 6 weeks implying that MFS may affect seizure activity in late stages of epilepsy onset. 1 Sandeep P. Nair, 4 Wendy M. Norman, 1 Linda K. Dance, 1,3 Panos M. Pardalos, 1,2 Jose C. Principe, and 1,4 Paul R. Carney ( 1 Biomedical Engineering, University of Florida ; 2 Electrical and Computer Engineering ; 3 Industrial and Systems Engineering ; and 4 Pediatrics, University of Florida, Gainesville, FL ) Rationale: Gradual, progressive preictal dynamical convergence among brain regions has been reported in human epilepsy (IEEE TBME 2003; 50; 616–27) and in an animal model of chronic limbic epilepsy (CLE) (Epilepsia 45 (S7): 211–212). We have used a spontaneously seizing animal model of limbic epilepsy to investigate the occurrence of spatiotemporal transitions involving such dynamical convergence among different regions in an epileptogenic brain. We hypothesize that the frequency of these transitions increases progressively from the time of initial insult to the subsequent manifestation of epilepsy. Methods: Young adult male Sprague Dawley rats (n = 4) were used in the study. The model was developed using chronic hippocampal stimulation (CHS). The animals animals, after allowing sufficient time for recovery from the stimulation procedure,were connected to an automated system that monitored the T‐index calculated from short‐term maximum Lyapunov exponents estimated from multiple brain regions. A (dynamical convergence (transition) was defined as a drop in T‐index value from a preset upper threshold (UT= 10) to a lower threshold (LT= 2.662). The time leading to the first seizure was divided into blocks of 12 hours and the frequency of transitions in each block was calculated. The null hypothesis is rejected if the regression slope is found to be significantly positive (p < 0.05). Results: Two out of the four animals developed spontaneous seizures after CHS. The first showed a progressive increase in frequency of spatiotemporal transitions leading up to the first seizure (recording time before seizure∼232.9 hrs, slope = 0.09, p = 0.0005), the second animal showed an abrupt increase in incidence immediately preceding the seizure (recording time before seizure∼142.5 hrs, slope = 0.07, p = 0.0228). No dynamical transitions were observed in the two animals that did not develop spontaneous seizures. Conclusions: These results suggest that spatiotemporal transitions involving convergence of dynamical values among brain regions may be a marker for epileptogenesis. (Supported by NIH grant R01EB002089, Children s Miracle Network, University of Florida Division of Sponsored Research and Development and Department of Veterans Affairs.) 1 Jari Nissinen, and 1,2 Asla Pitkanen ( 1 Department of Neurobiology, A.I.Virtanen Institute/University of Kuopio, Kuopio, Finland ; and 2 Department of Neurology, Kuopio University Hospital, Kuopio, Finland ) Rationale: Prevention of epileptogenesis in patients with acute brain damaging insults like status epilepticus (SE) is a major challenge. We investigated whether carbamazepine (CBZ), levetiracetam (LEV), and valproic acid (VPA) as monotherapy are antiepileptogenic or disease modifying. To mimic clinical study design, treatment was started 24 h after the beginning of electrically induced SE. Methods: SE was induced by stimulating the lateral nucleus of the amygdala of adult Sprague‐Dawley rats (n = 98) electrically for 20–40 min. One group of vehicle treated non‐stimulated animals served as controls for histology and behavioral analysis (SHAM, n = 37). SE was stopped with diazepam (20 mg/kg, intraperitoneally, i.p.) at 4 h after SE induction (additional injection of DZP 10 mg/kg was given at 3h after the first injection). Administration of compounds (CBZ 120 mg/kg/d, n = 17; LEV 300 mg/kg/d, n = 18; VPA 600 mg/kg/d, n = 16; vehicle VEH, n = 47) was started 24 h after the beginning of SE. On the first day, AEDs were administered intraperitoneally, and thereafter i.g. All compounds were administered at 8 h intervals for 7 d. SE and the development of spontaneous seizures were monitored with video‐EEG. The occurrence of epilepsy and seizure frequency were assessed at 10 weeks after SE induction with a continuous video‐EEG monitoring (24h/day) for 2 weeks. To confirm antiepileptogenesis or disease modification, the second 2‐weeks continuous video‐EEG was started 14 weeks after SE. Thereafter, animals underwent behavioural testing (Morris water‐maze and fear conditioning) and were perfused for histology. Results: Altogether, 62% of VEH, 89% of LEV, 53% of CBZ, and 88% of VPA treated animals developed epilepsy. The mean seizure frequency in the VEH group was 6.3 ± 12.8 sz/d (range 0.04–50, median 0.39), in the LEV group 4.8 ± 7.6 sz/d (range 0.04–24.2, median 0.88), and in the VPA group 7.6 ± 18.5 sz/d (range 0.07–70). In CBZ group the mean seizure frequency was 0.6 ± 0.6 sz/day (range 0.04–1.63, median 0.46). However it did not differ from that in the VEH group. Mean behavioural seizure severity did not differ between groups. The mean seizure duration was longer in the CBZ group compared to that in the VEH group (77 ± 16 sec vs. 68 ± 19 sec, p < 0.05). Conclusions: The three AEDs with different mechanisms of action, LEV, CBZ, and VPA were not antiepileptogenic when started at 24 h after SE induction and continued for 7 days. (Supported by NIH/NINDS R21 NS 049525.) 1 Joseph A. Oommen, 1 Alli Kraus, and 1 Robert S. Fisher ( 1 Neurology, Stanford, Stanford, CA ) Rationale: Since CSF drug delivery is clinically useful for antibiotics and antineoplastics, direct delivery of anti‐epileptic medication to the CSF may also be useful and improve the tolerability and efficacy of these drugs. We performed a proof‐in‐principle study of intraventricular gabapentin (GPN) in the rat. Methods: Anesthetized Sprague‐Dawley rats were implanted with two Alzet model 2001 osmotic pumps, each delivering 1.0 μL/hour for one week. Pumps were connected to left and right intraventricular catheters, with tips in relation to Bregma at: 1.4 mm lateral and 0.8 mm posterior on the right, 2 mm lateral and 1.3 mm posterior on the left, and 2.5 mm below the top of cortex. GPN 80 μg/μL was installed into each pump and delivered at 3.8 mg per day. Half of the animals had pumps filled with isotonic normal saline (control group), and half with GBP. Seizure theshold was tested after five days, by dripping the convulsant liquid anesthetic flurothyl onto tissue paper at 10 μL/minute in a sealed 3.2 L container. A single observer, blinded as to treatment group, recorded the time to first myoclonic twitch, first partial seizure and first tonic‐clonic seizure. Seizures were judged behaviorally; no EEG was used as part of this protocol. Results: Of 60 rats used, 6 were lost due to mortality or pump/catheter malfunction. For the surviving 54, time to first tonic‐clonic seizure was 295.8 ± 58.8 s (n = 28) for control group, versus 338.0 ± 89.9 s (n = 26) for the rats with GBP in the pump (p = 0.049 by two‐tailed t‐test). The mean time to onset of first myoclonic jerk was 158.7 ± 20.8 vs. 164.6 ± 33.5 s (p = n.s.). Effects of flurothyl can be influenced by animal weight; however, regression of time to tonic‐clonic seizure versus weight yielded no significant relationship. Serum gabapentin levels obtained immediately after testing the seizure threshold were undetectable (<1 μg/ml) in every animal. We did not perform quantitative studies of GPN distribution through brain, but methylene blue dye in the pumps distributed into periventricular white matter and also over cortex, especially ipsilaterally. Conclusions: Intraventricular delivery of gabapentin elevates the threshold for generalized tonic‐clonic seizures in the rat flurothyl model, but not time to the first myoclonic jerk or first partial seizure. Interpretation is limited, as the degree of penetration and distribution of the GPN into brain cannot be detailed. Unmeasurable serum levels suggest the effect is not systemic. Intraventricular delivery of antiepileptic drugs would be useful for drugs with poor oral absorption or which do not penetrate the blood‐brain barrier. Intraventricular administration of drugs that do penetrate into brain after oral ingestion (e.g. GPN) may or may not be useful, depending upon achievable improvements in the therapeutic/toxic ratio and in steady brain levels of drug. (Supported by Medtronic, and by the Maslah Saul MD Chair, the James and Carrie Anderson Epilepsy Laboratory, and the Susan Horngren Fund.) 1 Mansi B. Parekh, 2 Lan B. Hoang‐Minh, 1 Hector Sepulveda, 3 Angela Handlock, 1 Wendy Norman, 1 Justin C. Sanchez, 1 William L. Ditto, 1 Paul R. Carney, and 4 Thomas H. Mareci ( 1 Neuroscience, University of Florida, Gainesville, FL ; 2 Biomedical Engineering, University of Florida, Gainesville, FL ; 3 Pediatrics, University of Florida, Gainesville, FL ; and 4 Biochemistry and Molecular Biology, University of Florida, Gainesville, FL ) Rationale: In this study, we used MR diffusion tensor imaging (DTI) to determine the temporal hippocampal and parahippocampal structural changes in vivo at 11.1 Tesla (T) during the latent period of epileptogenesis and to examine these same regions in the excised intact brain at 17.6 T. The mesial temporal lobe model was also used to determine if chronic stimulating/recording microelectrodes distort MR images collected at 11.1 T. Methods: Fifty micron gold plated tungsten wires (2) were implanted in the ventral hippocampus in Sprague Dawley rats (n = 6). Spontaneously seizing rats were obtained following experimentally inducted status epilepticus (SE). Rats were video recorded to capture spontaneous seizures. Injured rat brains were examined in vivo at 11.1 T pre/post wire implantation and after SE at 3, 5, 7, 10, 20, 40, 60 days. The intact excised brain of 2 rats were imaged at 17.6 T. Diffusion weighted images were acquired using gradient directions specified by the tessellations of an icosahedron on the hemisphere. The diffusion weighted images were fit to a rank‐2 tensor model of diffusion. Results: Five of the 6 rats developed seizures. The DT images of the rat, without seizures, looked identical to controls. In the other 5 rats, diffusivity increased in the hippocampus and fimbria in vivo. In 2 of these rats, diffusivity also increased in the entorhinal cortex. Increased diffusivity was seen as early as day 7 post‐stimulation. We examined the excised brains of a seizing rat and one that did not. In the seizing rat, the diffusivity increased bilaterally in the hippocampal region, identified with mossy fiber sprouting. The hippocampus showed atrophy ipsilateral to the simulation. Decreased fractional anisotropy indicated a loss of structure in the dentate gyrus and CA2/3 region of the pyramidal cells (Fig.1). Conclusions: The induced SE produces bilateral structural changes, visible in detail with DTI, in the hippocampal and parahippocampal structures. Stereotaxic images allowed repeated measures of the structural changes during epileptogenesis and the implanted wires produced minimal distortions. Future work will relate the physiological and structural changes, during epileptogenesis, using electrophysiology and DTI. (figure 1) (Supported by NIH grants, R01 EB004752 and R01 NS42075, Wilder Epilepsy RC, and the UF Alumni Foundation.) 1 Alexia Paucard, 2 Akitaka Yamamoto, and 1 David C. Henshall ( 1 Physiology and Medical Physics, Royal College of Surgeons, Dublin, Ireland ; and 2 RS Dow Neurobiology Laboratories, Legacy Research, Portland, OR ) Rationale: Hippocampal sclerosis is the commonest lesion identified in temporal lobe epilepsy and has been proposed as both cause and effect of seizures. Additionally, an emerging concept in epilepsy research is that the propagation of paroxysmal activity is necessary and sufficient to transform a naïve structure into one that is capable of generating spontaneous (epileptic) seizures. This transformation underlies the formation of an epileptogenic focus that can entrain the generation of seizures in interconnected structures. However, whether seizure‐injury occurs within such communicating fibre tracts is largely unknown but recent reports document the presence of lesions within white matter tracts in epilepsy patients. According, we evaluated white matter tract injury within the corpus callosum in a novel seizure model in mice. Methods: Adult male C57Bl/6 mice underwent focally‐evoked seizures induced by intraamygdala kainic acid microinjection with continuous electroencephalography. Seizures were terminated by intravenous lorazepam 40 (n = 10) or 50 minutes (n = 7) following kainate/vehicle (n = 3) injection and brains obtained 24 hours later. Coronal sections were processed for detection of DNA fragmentation using the TUNEL technique. To identify the cell populations of the corpus callosum, the slides were labelled for oligodendrocytes with antibodies against 2,3‐cyclic‐nucleotide 3‐ phosphodiesterase (CNP) or myelin basic protein (MBP) and for astrocytes with anti‐glial fibrillary acidic protein (GFAP). Results: Seizures in mice terminated after 40 min resulted predominantly in unilateral TUNEL staining within ipsilateral CA3 and CA1. In contrast, when seizure activity was extended to 50 min contralateral CA1 and CA3 hippocampal damage was more frequently observed. Examination of the ipsilateral and contralateral corpus callosum revealed the presence of cell death. Our data also revealed significant correlations between numbers of degenerating cells in contralateral (r = 0.460, p < 0.05) corpus callosum and the severity of neuronal cell death in CA3 hippocampus ipsilateral. Seizures induced oligodendrocyte death in the corpus callosum as revealed by the immunofluorescence colocalization of TUNEL/MBP or CNP. In contrast, the presence of a gliogenesis is shown around the corpus callosum. Conclusions: The present data establish that prolonged focal limbic seizures in mice induce damage to non‐neuronal elements within the corpus callosum and show the degree of this injury is a function of hippocampal damage. Our data indicate seizures injure white matter tracts and offer a potential mechanism for seizure generation and spread to contralateral and extrahippocampal structures. (Supported by Marie Curie Grant.) 1 Challon R. Perry, 1 Minah Suh, 1 Hongtao Ma, 1 Mingrui Zhao, and 1 Theodore H. Schwartz ( 1 Department of Neurological Surgery, Weill Medical College of Cornell University, New York, NY ) Rationale: Cortical stimulation is playing an increasing role in clinical neuroscience in the treatment of epilepsy, stroke and pain. The neuromodulatory effects of cortical stimulation include both excitation and inhibition, the balance of which varies with the stimulation parameters. As a result, cortical stimulation also influences cerebral hemodynamics, including perfusion and oxygenation, which may also vary depending on the stimulation parameters, although little data exists on this subject. In this study, we investigate the effects of stimulation frequency on the hemodynamic response using optical recording of intrinsic signals (ORIS). Methods: Using bipolar tungsten electrodes, the rat sensory neocortex was stimulated at variable frequencies in a pseudo‐random fashion at 5 Hz, 10 Hz, 50 Hz, 100 Hz, and 250 Hz at a constant charge (CC). The amplitude remained constant at 1 mA. CC was maintained by delivering stimulation at 100 spikes at each frequency and varying the duration of stimulation. Simultaneous field potential (f.p.) recording and ORIS was performed at 570 nm and 610 nm to quantitatively measure total hemoglobin (Hbt), deoxygenated hemoglobin (Hbr) and oxygenated hemoglobin (HbO2) using a pathlength correction. Optical data was averaged over 10 trials at each frequency and each trial lasted 100 sec so the optical signal could return to baseline. Trials with afterdischarges were eliminated. Results: A large increase in Hbr was recorded at all stimulation frequencies but the maximum occurred at 10 Hz (2.12 ± 0.20%, n = 3). The relationship between the increase in Hbr and stimulation frequency (in log scale) could be characterized by the equation y =−1.1164 × 2 + 3.1836x – 0.2979 (R2 = 0.74). There was a significant correlation between the duration of stimulation and the duration of the increase in Hbr (r = .86, p < 0.05). The largest increase in Hbt occurred at 50 Hz (4.04 ± 1.08%, n = 3) and the relationship between the increase in Hbt and stimulation frequency (in log scale) could be characterized by the equation y =−1.3816 × 2 + 5.3414x – 1.2291 (R2 = 0.98, p < 0.05). The duration of the increase in Hbt was generally longer than the duration of stimulation by 80 seconds, regardless of frequency. Conclusions: These results demonstrate that the duration of stimulation is more important than the frequency in determining the duration of the dip and that the relationship between frequency and both Hbt and Hbr is not linear, with a plateau occurring between 10–50 Hz. (Supported by NIH.) 1 David Revesz, 1 Mikael Tjernstrom, 1 Thorleif Thorlin, and 1 Elinor Ben‐Menachem ( 1 Department of Clinical Neuroscience and Rehabilitation, Institute of Neuroscience and Physiology, the Sahlgrenska Academy at Gothenburg University, Gothenburg, Sweden ) Rationale: The birth of new neurons from neuronal stem cells, the process called neurogenesis, has been seen in adult brains from both rodents and humans. The process is especially prominent in the hippocampus, which is involved in mood and memory functions and epilepsy. A new hypothesis for the development of depression in humans has been formed over the last few years, which includes neurogenesis as a factor of importance in the depressive disease: Reduction of neurogenesis in the hippocampus is a causality factor in the generation of depression and stimulated neurogenesis is part of the recovery process from the depressive state. The same applies for epilepsy. VNS therapy, which has been used successfully for many years in the treatment of epilepsy, was early on in its development noted to have mood elevation effects in VNS treated patients, with effects on depressive symptoms. This anti‐depressive effect from VNS has been proven in clinical studies, and VNS is now approved by the FDA for the treatment of severe depression in humans as well as epilepsy. The mechanisms behind the antiepileptic and anti‐depressive effects from VNS are, however, not fully understood. We hypothesize that a VNS‐induced stimulation of the neurogenesis is part of the mechanism behind the observed clinical effects from VNS. Methods: The degree of stem cell proliferation within the dentate gyrus of the hippocampus was studied. The number of new cells were measured by BrdU (bromodeoxy‐uridine) incorporation. Immunohistochemistry with antibodies against BrdU were used for the detection of the newly divided cells. Eight rats implanted with VNS‐leads and generators were stimulated for two days with the following parameters: 0.75 mA output current, 250 microseconds pulse width, 20 Hz frequency, 30 seconds on and 5 minutes off. These rats were compared with eight control rats that had VNS surgery for implantation of the VNS equipment, but with devices constantly in the off position. Results: There was a statistically significant difference between the two groups with a 50% increase of stem cell proliferation at the 0.03 significance level in the VNS‐stimulated rats as compared to sham operated animals. Conclusions: These results indicate that VNS treatment in the rat affects the stem cell proliferation in the hippocampus. In the light of the prevailing theory of a link between hippocampal neurogenesis, epilepsy and depression, this finding might represent an interesting path in the search for mechanisms behind the actions of VNS. (Supported by Cyberonics, Houston, TX, USA; Vastra Gotalandsregionen, Sweden.) 1 Alexander Rotenberg, 1 Khanh Hoang, 2 Alvaro Pascual‐Leone, 1 James J. Riviello, and 1 Frances E. Jensen ( 1 Neurology, Children's Hospital, Boston, MA ; and 2 Neurology, Beth Israel Deaconess Medical Center, Boston, MA ) Rationale: rTMS is a noninvasive method for cortical stimulation that is emerging as a new therapeutic tool in epilepsy. The technique is based on the principle of electromagnetic induction where small intracranial electrical currents are induced by a strong rapidly‐changing extracranial magnetic field. Limited human trials and animal experiments in our laboratory show acute suppression and reduction in frequency of partial seizures with EEG‐guided single pulse and 0.5 Hz rTMS. However, the capacity of rTMS to suppress generalized (absence) seizures has not been thoroughly examined. Accordingly, we sought to determine whether whether 0.5 Hz rTMS can suppress frequent PTZ‐induced generalized seizures in a rat absence epilepsy model. Methods: Generalized (absence) seizures were triggered in adult male Long Evans rats (n = 10) with PTZ (50 mg/kg, i.p.). Immediately after PTZ injection, and thereafter at 30‐minute intervals, one group (n = 4) was treated with 0.5 Hz repetitive TMS for 10 minutes (300 stimuli). A second group (n = 6) remained as untreated controls. Continuous EEG was recorded with subdermal wires electrodes for 120 minutes after PTZ injection, and then analyzed for seizure frequency during periods between TMS sessions. Results: (1) Frequent brief absence seizures accompanied by generalized spike‐wave discharges on surface EEG were triggered by PTZ injection in all rats. (2) 3/6 untreated rats died within the monitoring period – 0/4 rats died in the 0.5 Hz rTMS group. (3) Average cumulative seizure frequency was lower in the treated group after 60 min (t = 2.65; p < 0.033), 90 min (t = 2.98; p < 0.021) and 120 min (t = 2.85; p < 0.030) of monitoring. Conclusions: 0.5 Hz rTMS is effective in suppression of generalized seizures in the PTZ absence epilepsy rat model. Or finding may be a step toward eventual treatment of generalized epilepsy with low frequency rTMS. (Supported by Citizens United for Research in Epilepsy (CURE).) 1,2 Raman Sankar, 1,3 Stephane Auvin, 1 JoAnne Nakagawa, and 1 Don Shin ( 1 Pediatrics, David Geffen School of Medicine at UCLA, Los Angeles, CA ; 2 Neurology, David Geffen School of Medicine at UCLA, Los Angeles, CA ; and 3 Pediatric Neurology, EA 1046 School of Medicine, Lille, France ) Rationale: Lithium pilocarpine status epilepticus (LiPC‐SE) causes prolonged seizures and subsequent widespread neuronal injury. In the developing rat brain, untreated LiPC‐SE results in age dependent hippocampal injury as well as widespread extrahippocampal neuronal damage. We examined whether brief durations of SE are sufficient to induce injury in the immature brain. Methods: EEG data was collected using acquisition software obtained from wireless transponders implanted s.c. in two week old wistar pups. Rats were pretreated with LiCl (3 mEq) on the day of surgery and then subjected to SE 16–20 hr later with pilocarpine (60 mg/kg) at P14. Diazepam (5 mg/kg) and phenobarbital (25 mg/kg) were given after varying durations of SE using EEG recordings to determine both the delay of onset after PC injection, as well as the total duration of SE. 24 hr after the onset, rats were given pentobarbital (100 mg/kg) followed by perfusion/fixation with 4% paraformaldehyde. Brains were embedded in paraffin and cut at 8 μ intervals. Sections were stained with Fluoro‐Jade B (F‐J B) as a marker of neuronal injury, and analyzed using a semiquantative scale (0 = no injury, 1 = trace, 2 = < 10%, 3 = 10–25%, 4 = 26–50%, 5 = > 50% F‐J B labeled neurons) bilaterally in three sections containing the dorsal hippocampus. Results: SE was established in over 90% of the P14 rats. Behavioral manifestations of seizures appeared almost immediately after PC injection, but synchronized video monitoring revealed a delay of 9.8 ± 1.6 min before the appearance of continuous, rhythmic, high frequency electrographic activity used to mark the onset of SE. Prolonged recordings showed that the combination of diazepam/phenobarbital was sufficient to prevent the return of seizures over the following 24 hr period. While 30 min of SE was not sufficient to induce hippocampal injury with the exception of trace dentate granule cells, injury was seen in the amygdala (2.3 ± 0.1) and cortex (1.8 ± 0.1). No brightly labeled neurons were seen in the thalamus, but some lightly stained, irregularly shaped neurons were detected. Conclusions: Previous studies have shown that P14 rats subjected to LiPC experience several hours of SE resulting in extensive neuronal injury, particularly in the CA1 hippocampal subfield. While early pharmacological intervention was sufficient to protect those neurons from injury, extrahippocampal structures remain vulnerable after only 30 min of SE. Studies in adult rats suggest protection of the Ammon's horn alone is not sufficient to prevent epilepsy. 25% of untreated P14 rats undergoing LiPC‐SE go on to demonstrate spontaneous seizures. Whether the neuronal injury that results after only 30 min is sufficient to induce epilepsy is under investigation. (Supported by NS046516 & DAPA Foundation (RS), AEAC Association (SA).) 1 Matthew R. Sarkisian, 1 Christopher M. Bartley, 2 Hongbo Chi, 3 Fumihiko Nakamura, 1 Kazue Hashimoto‐Torii, 1 Masaaki Torii, 2 Richard A. Flavell, and 1 Pasko Rakic ( 1 Neurobiology, Yale University School of Medicine, New Haven, CT ; 2 Immunobiology, Yale University School of Medicine, New Haven, CT ; and 3 Medicine, Harvard Medical School, Boston, MA ) Rationale: Periventricular nodular heterotopia (PVNH) is a congenital malformation of human cerebral cortex frequently associated with mutations in the Filamin‐A (FLN‐A) gene but the pathogenetic mechanisms remain unclear. Both loss and gain of FLN‐A function have been proposed to disrupt the migration initiation of cortical neurons. Here we show that MEKK4 (a MAP3K member of the mitogen‐activated protein kinase pathway) is involved in Fln‐A regulation and PVNH formation. Methods: We performed histological analyses on the brains of MEKK4‐deficient mice. Additionally, we designed short interference RNA (siRNA) constructs targeting MEKK4, incorporated these constructs into the developing cerebral cortex by in utero electroporation and analyzed the fates of transfected cells. Immunocytochemistry, in situ hybridization, electron microscopy, GST‐pulldown and western blot analyses were also performed Results: MEKK4−/− mice develop bilateral PVNH in the cerebral cortex that arise from breaches in ventricular surface lining. Using neuronal markers, we show that cells within PVNH have differentiated and failed to migrate into the cortical plate. In utero electoporation of MEKK4 siRNA at E14.5 and analysis at P0 revealed dramatic disruption of migration‐initiation of cortical neurons compared to control. Analysis of Fln‐A expression in MEKK4−/− forebrain revealed elevated and ectopic Fln‐A, while MEKK4 RNAi increased Fln‐A phosphorylation in vitro. Finally, recombinant‐MKK4/SEK1 protein precipitated a complex containing MEKK4 and Fln‐A suggesting that MKK4/SEK1 bridges these molecules during development. Conclusions: Disrupted migration initiation in MEKK4‐deficient cells may be related to both defects of the ventricular surface lining and dysregulated Fln‐A. Our data demonstrate an essential role for MEKK4 in the migration of developing cortical neurons and provide insight into the pathogenesis of human PVNH. (Supported by the Epilepsy Foundation of America postdoctoral fellowship (to M.R.S), an under‐represented minority fellowship (to C.M.B) on U.S. Public Health Service grants (to P.R.), a Child Health Research Grant from the Charles H. Hood Foundation, Inc. (Boston) (to H.C.). R.A.F. is an Investigator of the Howard Hughes Medical Institute.) 1 Steven C. Schachter, 2 H. Steve White, 3,4 Lauren Murphree, and 4 James Stables ( 1 Osher Institute, Harvard Medical School, Boston, MA ; 2 Pharmacology and Toxicology, University of Utah, Salt Lake City, UT ; 3 SRA International, Fairfax, VA ; and 4 Anticonvulsant Screening Project, NINDS, Bethesda, MD ) Rationale: Huperzine A (HupA), a sesquiterpene alkaloid derived from Chinese club moss (Huperzia serrata), has traditionally been used in China for swelling, fever and inflammation. HupA reversibly inhibits acetylcholinesterase and reduces glutamate‐induced neuronal death as an NMDA receptor antagonist. We previously reported that huperzine potently protects mice against clonic seizures induced by s.c. pentylenetetrazol at 1 mg/kg (White, 2005) with a TD50 of 0.83 mg/kg. We now report results from the 6‐Hz model of psychomotor seizures. Methods: Anesthetic/electrolyte solution (0.5% tetracaine hydrochloride in 0.9% saline) was applied to the eyes of male, 18–25 g CF1 mice (Charles River Laboratories, Wilmington, MA) prior to placement of corneal electrodes. Seizures induced by 6‐Hz corneal stimulation (Barton, 2001) using three levels of stimulation – 22, 32 and 44 mA – (3 sec duration) were assessed at various times after i.p. administration of 0.09 – 1.1 mg/kg HupA. Median effective doses (ED50) and 95% confidence intervals were calculated for each the three current stimulations evaluated. Results: ED50 values for i.p. HupA were 0.28, 0.34 and 0.78 mg/kg for 22, 32, and 44 mA, respectively. The following results were observed at the highest doses tested at each stimulation level: at 22 mA, 7/8 animals were protected with a dosage of 0.7 mg/kg; at 32 mA, 8/8 animals were protected with 0.83 mg/kg; and at 44 mA, 6/8 animals were protected with 1.1 mg/kg. Conclusions: HupA produced a potent anticonvulsant effect in the 6‐Hz model of psychomotor seizures at all stimulation levels tested and with ED50 values less than the TD50, suggesting a possible advantage over phenytoin, carbamazepine, lamotrigine and topiramate, each of which display limited efficacy in this model at doses devoid of behavioral toxicity. Further, the less than 2‐fold ratio of dosages effective across the range of stimulations suggests a possible advantage over other drugs active in this model such as levetiracetam. These findings and the known action of huperzine as an NMDA receptor antagonist warrant further pre‐clinical and clinical evaluation. References Barton ME et al. Epilepsy Res 2001;47:217–27. White HS et al. Epilepsia 2005;46(suppl 8):220. (Supported by: The investigators gratefully acknowledge the support of the NINDS Anticonvulsant Screening Project and a grant from The Epilepsy Project (S.S.).) 1 Melodie R. Winawer, 2 Nicholas Makarenko, 2 Tana M. Hintz, 2 Sameh M. Kamel, 2 Daniel P. McCloskey, 2 Nandini Nair, 3 Abraham A. Palmer, and 2,4 Helen E. Scharfman ( 1 Neurol. & Sergievsky Center, Columbia University, New York, NY ; 2 CNRRR, Helen Hayes Hospital, West Haverstraw, NY ; 3 Human Genetics, University of Chicago, Chicago, IL ; and 4 Pharmacol. & Neurol., Columbia University, New York, NY ) Rationale: To identify new genes involved in temporal lobe epilepsy (TLE) using quantitative trait locus (QTL) mapping, initial studies were conducted to compare the pilocarpine model of TLE in A/J vs. DBA/2J mouse strains, two strains that are fully sequenced. The incidence and latency to status epilepticus were compared using pilocarpine doses that spanned the dose‐response curve. Animals that had status were also compared after recurrent, spontaneous seizures developed, to examine structural changes that often accompany chronic seizures: hippocampal neuronal loss, mossy fiber sprouting, mossy fiber neuropeptide Y (NPY) expression, and hilar ectopic granule cells (EGCs). Methods: Male mice (10 wks old) were administered atropine methylbromide (5 mg/kg, s.c.) 30’ before pilocarpine hydrochloride (200, 220, 250, or 300 mg/kg, i.p.), and 1 hr after the onset of status, diazepam was administered (5 mg/kg, i.p.). Mice were transcardially‐perfused > 4 wks after status, and immunocytochemistry was conducted using antibodies to NeuN, a neuronal marker (1:5,000, Chemicon), NPY (1:30,000, Peninsula), or PROX1 to label granule cells (1:60,000, Covance). PROX1‐labeled EGCs were quantified using stereology (StereoInvestigator software). Results: A/J mice had a higher incidence of status, if animals which died after a tonic‐clonic seizure, prior to clear signs of status, were excluded. However, the incidence of these tonic‐clonic seizures was higher for A/J mice than DBA mice, and the tonic‐clonic event could have been the start of status, so an incidence comparison is complex. When status occurred, the mean (±sem) latency to status was longer for A/J mice at all doses (A/J: 132.6 ± 6.5min, n = 26; DBA: 30.4 ± 1.7min, n = 24; p < 0.05). Recordings in A/J mice confirmed that the onset of behavioral status reflected the onset of electrographic status (n = 4). All DBA mice demonstrated neuronal loss in CA3 (n = 6/6), but all A/J mice did not (n = 3/6). Neuronal loss was detected in area CA1 in DBA (n = 3/6) but not A/J mice (n = 0/6). Both strains demonstrated NPY expression in mossy fibers (A/J: n = 5/6; DBA: n = 4/5). There were more EGCs in DBA mice (A/J: 3029.9 ± 1053.4 cells/hippocampus, n = 4; DBA: 9590.2 ± 1979.2, n = 6; p < 0.05). Conclusions: A/J and DBA/2J mice show differences in acute behavioral and long‐term anatomical changes with the pilocarpine model. Differences suggest that use of one variable to define susceptibility can be misleading. The results suggest guidelines for future studies of strain‐dependent seizure susceptibility, and characterize an infrequently‐studied strain for QTL mapping of seizure susceptibility. (Supported by NINDS R01 41490, K02 NS050429, K23 NS02211.) 1 Irene Schlifke, 2 Biljana Georgievska, 2 Deniz Kirik, and 1 Merab Kokaia ( 1 Experimental Epilepsy Group, Wallenberg Neuroscience Center, BMC A‐11, Lund, Sweden ; and 2 Neurobiology, Wallenberg Neuroscience Center, BMC A‐11, Lund, Sweden ) Rationale: The glial cell line derived neurotrophic factor (GDNF) has recently been indicated as an endogeneous anti‐epileptic agent, as well as it, due to its trophic action, could rescue neurons from excocytotic cell death and promote hippocampal neurogenesis. In the present study, we aimed to evaluate the ability of viral vector‐mediated hippocampal overexpression of GDNF to supress seizures and alter seizure‐induced cell death and neurogenesis. Methods: A recombinant adeno‐associated viral vector (rAAV) encoding for GDNF or GFP (control) was injected in multiple sites of the hippocampal formation of rats. A recording/stimulating electrode was placed in the ventral hippocampus, and animals were stimulated either for 1 s once daily, inducing kindling epileptogenesis, or for 1 h continously to induce self‐sustained status epilepticus (SE). Kindled animals were re‐stimulated (i.e. re‐kindled) four weeks following termination of kindling to assess the sustainability of hyperexcitability. Behavioural seizure activity (graded according to the scale of Racine) and EEG were recorded throughout the experiments. All experiments were conducted according to international ethical guidelines for animal care and were approved by the local ethical comittee. Results: Using the kindling model, we found that transfection with rAAV‐GDNF decreased the number of generalized seizures, prevented an increase of the duration of behavioural convulsions seen in re‐kindled control animals, and increased the threshold for induction of re‐kindled seizures. In the SE‐model of epilepsy, animals overexpressing GDNF developed fewer generalized stage 5 seizures and exhibited a significantly better survival rate following SE. We further found that in GDNF‐overexpressing animals, a higher proportion of new‐born (BrdU+) cells acquired a neuronal fate, as determined by staining for DCX and Neun, compared to control animals. The SE‐induced damage in hippocampal CA1/CA3 and dentate gyrus, as assessed by fluorojade staining, was unchanged by rAAV‐GDNF transfection. Conclusions: Taken together, these data suggest that rAAV‐mediated gene transfer is a feasible method for long‐term delivery of GDNF to the seizure focus, and that such GDNF overexpression is able to suppress seizures in different animal models of epilepsy. (Supported by Swedish Research Council (Vetenskapsradet), the Segerfalk Foundation, the Craaford Foundation and the Kock's Foundation.) 1 Hector Sepulveda, 1 Lan Hoang‐Minh, 2 Mansi B. Parekh, 3 Angela Hadlock, 3 Wendy Norman, 3 Justin C. Sanchez, 1 William L. Ditto, 3 Paul R. Carney, and 4 Thomas H. Mareci ( 1 Biomedical Engineering, University of Florida, Gainesville, FL ; 2 Neuroscience, University of Florida, Gainesville, FL ; 3 Pediatrics, University of Florida, Gainesville, FL ; and 4 Biochemistry and Molecular Biology, University of Florida, Gainesville, FL ) Rationale: We used an 11.1 Tesla (T) magnet to determine in vivo temporal hippocampal and parahippocampal structural changes in the latent period of epileptogenesis in a rat model of mesial temporal lobe epilepsy. We also determined if chronic microelectrodes distort MR images to assess the feasibility of MRI with electrophysiological arrays. Methods: Fifty micron gold plated tungsten wires (2) were implanted in the ventral hippocampus in Sprague Dawley rats (n = 4). Spontaneously seizing rats were obtained following experimentally inducted status epilepticus (SE). Rats were video recorded to capture spontaneous seizures. An MRI compatible stereotaxic frame was developed, to allow repeatable positioning under isoflurane in the 11.1T/40 cm bore magnet, along with a 3 cm, 150 degree arc, linear surface MR coil. Rats were imaged pre/post wire implantation and after SE at 3, 5, 7, 10, 20, 40, 60 days. T2‐weighted images were collected and quantified. Results: Brain images and wire locations were obtained with the stereotaxic frame, which provided a repeatable platform. Significantly increased T2 values were observed in the amygdala, entorhinal and piriform cortices, particularly on the contralateral side to stimulation, in three out of four rats. T2‐weighted images of the first rat showed a contralaterally growing volume of hyper‐intense tissue in the hippocampus and amygdala (Fig. 1). The brain of the second rat did not change; the wires were medial to the hippocampus near the alveus. The third rat showed hyper‐intensity in the contralateral amygdala, with wires in the ventral hippocampus. The fourth rat was hyper‐intense in both sides of the hippocampus and amygdala by day 3, but slightly decreased in time, with wires in the hippocampus near the alveus. Conclusions: Unilateral electrically induced SE produces bilateral structural changes in the hippocampal and parahippocampal structures. Stereotaxic images allowed repeated measures of the structural changes during epileptogenesis. The implanted wires produced minimal distortions. Their positioning in the ventral hippocampus may play a role in which structures show changes. Future work will relate the physiological and structural changes during epileptogenesis using electrophysiology and high resolution MRI. (figure 1) (Supported by NIH grant R01 EB004752, Wilder ERC, UF Alumni Foundation.) 1 Margaret N. Shouse, 1 John C. Scordato, and 1 Paul R. Farber ( 1 Sleep Disturbance Research, VA GLAHS, North Hills, CA ) Rationale: To describe some similarities and dissimilarities between feline epilepsy models, Landau‐Kleffner Syndrome (LKS) and Electrical Status Epilepticus in Slow‐sleep (ESES) with respect to sleep state modulation, neurobehavioral consequences and early intervention. Methods: The sleep‐wake state distribution of seizures and neurobehavioral consequences were quantified in relation to monoamine concentrations (microdialysis) as well as before and after lesions, systemic and localized drug administration and/or photic stimulation in feline epilepsy models (n = 86), mostly in the amygdala kindling epilepsy model. 1 Similarities and Dissimilarities of age‐related amygdala kindling in kittens to Electrical status epilepticus (ESES) and/or Landau‐Kleffner Syndrome (LKS) SIMILARITIES DISSIMILARITIES AGE DEPENDENCY Critical period for onset GTCs can last to adulthood in kittens SPONTANEOUS SEIZURE‐RELATED SYMPTOMATOLOGY: Seizure types (GTCs, partial motor, atypical absence, complex‐partial); IID types & cortical “foci;” Foci change over time ESES rare in kittens;Clinical seizures start 1st kittens & ESES vs. LKS; Fronto‐central foci in ESES vs. Centro‐ temporal foci in LKS and kittens SLEEP PATTERNS: NREM onset seizures; ↑ IIDs & spread in NREM & ↓ IIDs & spread in REM. Occasional ESES symptoms in kittens; Sleep deficits in kittens DEVELOPMENTAL/NEUROBEHAVIORAL DEFICITS: ↑ sensitivity & IIDs to sound; Social isolation; Mood change, Stereotypies, Hyperactive, atonic/dystonic aggressive/placid Type/extent of residual deficits depend on early onset, site and duration of CSWS in humans vs.GTCs in kittens PROGNOSIS & AED RESPONSE Good for seizures; ESES self‐limited; Residual neurobehavioral symptoms Same AEDs untested ENCEPHALOPATHY; PATHOPHYSIOLOGY Localization‐related with subtle/minor cell loss/pathology at foci ? GTCs = generalized tonic‐clonic convulsions; ↑= increased ↓= decreased; CSWS = Continuous slow spike wave during sleep (Supported by the Department of Veterans Affairs.) Results: 1) Table 1 compares the kindling model to LKS and ESES with respect to sleep state modulation, spontaneous neurobehavioral consequences and early intervention. 2) Neural generators of synchronous EEG oscillations (tonic background slow waves and sleep EEG transients), can combine to promote electrographic seizure propagation during NREM; antigravity muscle tone permits seizure‐related movement. 3) Neural generators of asynchronous neuronal discharge patterns can reduce electrographic seizures during REM sleep; skeletal motor paralysis blocks seizure‐related movement during REM. 4) Neurobehavioral consequences depend on the site of the focus, regardless of overt histopathological change in the young. 5) The prognosis is best with early intervention. Conclusions: The kindling process is thought to have trans‐synaptic effects on cells distal to the kindled focus. Age‐related changes could underlie the progression of epileptic and non‐epileptic symptoms in kindled kittens, LKS and ESES. 1 Ajay K. Srivastava, and 1 H. Steve White ( 1 Pharmacology and Toxicology, University of Utah, Salt Lake City, UT ) Rationale: Recently we have characterized the pharmacology of antiepileptic drugs (AEDs) in the lamotrigine (LTG)‐resistant pentylenetetrazlole kindled (Epilepsia 2003; 44,42) and LTG‐resistant amygdala kindled rat (Epilepsia 2004; 45(7):12). In this model of pharmacoresistant epilepsy, CBZ, PHT, and TPM were all inactive at non‐toxic doses. In contrast, VPA and retigabine (Epilepsia 2005;46(8):217) were both effective. The broad‐spectrum AED felbamate (FBM) displays a favorable preclinical profile in various models of epilepsy including the phenytoin‐resistant amygdala kindled rat (Neuropharmacology 2000;39(10):1893–903). The present study aimed to evaluate the efficacy of FBM in the LTG‐resistant amygdala kindled rat model of refractory epilepsy. Methods: Two groups of male Sprague‐Dawley rats were kindled via basolateral amygdala stimulation according to the method described by Srivastava et al., (Epilepsia 2004; 45(7):12). One hour before each kindling stimulation, rats in the control group received 0.5% methylcellulose and rats in the experimental group received LTG (5 mg/kg, i.p.). Seizure severity and afterdischarge duration was recorded after each kindling stimulation. Treatment was continued until all rats displayed 4 consecutive Stage 5 seizures (Clin Neurophysiol 1972; 32:281–294). Two days later, both groups were challenged with 15 mg/kg, i.p LTG to confirm LTG‐sensitivity (control) and LTG‐resistance (experimental). The efficacy of FBM (100, 150 and 200, mg/kg administered 90 min prior to kindling stimulation) was then evaluated. The ED50 and 95% confidence intervals were calulated using Probit analysis. Results: Felbamate dose‐dependently reduced the seizure severity and decreased the ADD in the fully kindled LTG‐resistant rat. At the highest dose tested (200 mg/kg), FBM decreased the seizure score from 5 ± 0 to 2.25 ± 0.6 and the ADD from 57 ± 5 sec. to 22 ± 30 sec. The calculated ED50 and (95% confidence intervals) for FBM against Stage 4 and 5 seizures was 169 and (131 – 206) mg/kg, respectively. Conclusions: In an animal model of pharmacoresistant epilepsy, the broad‐spectrum AED FBM was found to effectively block the fully kindled secondarily generalized kindled seizure. In this particular model, FBM is as effective as VPA and the investigational AED retigabine (albeit less potent than retigabine). These findings demonstrated that the LTG‐resistant kindled rat can be utilized to differentiate the AED profile of novel AEDs and confirm that FBM possesses a unique profile relative to CBZ, PHT, LTG, and TPM. Ongoing studies continue to evaluate the mechanism underlying the development of pharmaco‐resistance in this model. (Supported by NINDS grant 1 R21‐NS049624–01.) 1,2 Kimberly Statler, 1 Seth Swank, and 2 H. Steve White ( 1 Pediatrics, University of Utah, Salt Lake City, UT ; and 2 Pharmacology and Toxicology, University of Utah, Salt Lake City, UT ) Rationale: Pediatric traumatic brain injury (TBI) is a common cause of childhood death and disability. Post‐traumatic epilepsy (PTE) and cognitive disabilities are common sequelae that may manifest years after TBI. PTE afflicts 10% of children after severe TBI but remains poorly understood. To test the hypothesis that TBI during brain development increases seizure susceptibility later in life, we used standard electrical stimulation techniques to assess seizure thresholds in a rat TBI model. Methods: Immature, male Sprague‐Dawley rats (n = 20/grp) underwent TBI by controlled cortical impact (6‐mm rounded tip, 4 m/sec velocity, 2‐mm deformation, 100 msec duration) to left parietal cortex using isoflurane anesthesia (1%) on post‐natal day (PND) 16–18. Results were compared to those obtained from age‐matched sham (craniotomy only using isoflurane) and naive rats (n = 20/grp). Seizure thresholds were assessed during adolescence and adulthood for tonic (PND 34 and 60), clonic (PND 37 and 63) and limbic (PND 40) seizures. Full convulsive current (CC) curves were generated by staircase procedure and the median CC (CC50) was calculated using Probit analysis. A p‐value < 0.05 was considered significant. Results: Tonic and clonic seizure thresholds increased with age in naive rats (p < 0.05). As shown in Table 1, seizure responses during adolescence (PND 34–40) were similar among groups for tonic and clonic seizures; however, limbic seizure responses showed a trend toward lower thresholds after TBI. In adults (PND 60–63), tonic seizure thresholds were decreased in both sham and TBI (vs. naive) groups. Conversely, clonic seizure thresholds were similar among sham and naive rats but decreased after TBI (vs. either naive or sham). Conclusions: TBI to left parietal cortex during brain development attenuates normal maturational increases in clonic seizure thresholds and may increase susceptibility to limbic seizures. Given the involvement of frontal cortex and limbic system in learning and decision‐making functions, this finding may have implications not only for PTE, but also for post‐injury learning and behavioral deficits. Delineation of the incidence of PTE and cognitive deficits, as well as further characterization of longitudinal maturational changes in seizure thresholds post‐TBI are ongoing. 1 Seizure Thresholds Tonic Clonic Limbic adolescent naive 27.7 (25.1, 30.5) 12.3 (10.4, 13.9) 89.0 (87.6, 90.5) sham 30.4 (29.0, 33.0) 13.7 (12.0, 15.5) 87.8 (86.4, 89.2) TBI 26.9 (23.7, 29.6) 11.6 (9.7, 13.5) 87.2 (85.4, 88.9)# adult naive 36.5 (35.1, 37.8) 15.0 (13.7, 16.3) sham 34.5 (33.8, 35.8)* 14.8 (13.5, 16.1) TBI 33.7 (32.1, 35.3)* 12.0 (10.3, 13.7)*∧ Data presented as CC50 (95% CI) and compared by Probit analysis; *p < 0.05 vs. naive, ∧p < 0.05 vs. sham; #p < 0.1 vs. naive (Supported by NIH K12‐HD 01410; PCMC Foundation; University of Utah Child Health Research Center.) 1 Kerry‐Ann A. Stewart, 2 Robert S. Fujinami, 1 Karen S. Wilcox, and 1 H. Steve White ( 1 Pharmacology & Toxicology, University of Utah, Salt Lake City, UT ; and 2 Neurology, University of Utah School of Medicine, Salt Lake City, UT ) Rationale: Viral infections of the CNS are associated with a significantly increased risk for seizures and epileptogenesis. Up to 50% of adult‐onset epilepsy cases in developing countries are associated with CNS infection. To our knowledge, no available animal model accurately reproduces the pathology of human encephalitis‐induced epilepsy. We have established an animal model where infection of C57BL/6 (B6) mice with Theiler's murine encephalomyelitis virus (TMEV) leads to the development of seizures during the acute encephalomyelitis phase of infection. Stage 4/5 seizures (Racine, Electroenceph. Clin. Neurophys. 32: 281–94, 1972) occur in approximately 50% the mice between 3–10 days post infection and appear to remit as the infection clears. The present study tested the hypothesis that seized mice have altered seizure thresholds post inoculation, which increases their risk for the development of spontaneous seizures. Methods: B6 mice were infected intracerebrally with the Daniel's strain of TMEV (TMEV‐DA). Animals were monitored for the development of seizures and separated into two groups: seized versus non‐seized. Two months after infection with 3 × 105 PFU TMEV‐DA, minimal clonic and maximal tonic hindlimb extension (THE) seizure thresholds of seized and non‐seized mice were assessed by electroconvulsive seizure testing (EST). Additionally, the rate of corneal kindling acquisition (number of stimulations required to reach fully kindled state, i.e., 3 consecutive Stage 5 seizures) was assessed to determine whether TMEV inoculation leads to an increased rate of kindling. Results: At two months post infection, seized mice had a significantly lower minimal clonic seizure threshold compared to the non‐seized mice (seized CC50= 6.7 mA, non‐seized CC50= 7.8 mA, p < 0.01). In contrast, there was no significant difference in maximal THE seizure thresholds between the seized and non‐seized mice (seized and non‐seized CC50= 13.5 mA, p < 0.99). In a preliminary study to assess acquisition rate of corneal kindling, 8/8 seized mice displayed fully generalized Stage 5 seizures, versus 5/8 non‐seized mice. Furthermore, seized mice had much shorter latencies to generalized seizures than non‐seized mice. Conclusions: The results obtained in the present study demonstrate altered seizure susceptibility following recovery from seizures associated with TMEV‐ induced encephalitis. Furthermore, they suggest that these animals may be more susceptible to epileptogenesis following subsequent challenges. Although preliminary, these results further support this as a model of encephalitis‐ induced hyperexcitability. (Supported by Robert and Joyce Rice Epilepsy Fellowship (KAS); Max Abrams Memorial Fellowship from Epilepsy Foundation of America (KAS).) 1 Nikolaus Sucher, 2 Steven C. Schachter, 3 Nam‐In Baek, 1 Bombi Lee, 2 Jongbae Park, 1 Frances Jensen, 4 H. Steve White, 5,6 Lauren Murphree, and 6 James Stables ( 1 Department of Neurology, Children's Hospital, Boston, MA ; 2 Osher Institute, Harvard Medical School, Boston, MA ; 3 Graduate School of Biotechnology & Plant Metabolism Research Center, Kyung Hee University, Seoul, Korea ; 4 Pharmacology and Toxicology, University of Utah, Salt Lake City, UT ; 5 SRA International, Fairfax, VA ; and 6 Anticonvulsant Screening Project, NINDS, Bethesda, MD ) Rationale: Herbal medicines for epilepsy have a long tradition and continue to be used widely. An international team has been assembled to identify Asian herbal medicines used for epilepsy, isolate their constituent compounds, and test these compounds in established animal models of epilepsy and in vitro assays to determine their potential mechanisms of action. We report the preliminary findings from the first herbal medicine to be evaluated. Methods: Five pure compounds were isolated from an Asian herbal extract commonly used for epilepsy and were designated ADD 377008, 377009, 377010, 377011, and 377012 by the NIH/NINDS Anticonvulsant Screening Project. Compounds were assessed in the mouse maximal electroshock and s.c. Metrazol (pentylenetetrazol; s.c. Met) models of generalized tonic‐clonic and myoclonic seizures, respectively, following p.o. administration to Swiss‐Webster mice. Minimal motor impairment was evaluated in the rotarod test. Four of the compounds were tested for effects against NMDA receptor‐mediated excitotoxicity in primary neuronal cultures derived from rat cortex (Sun et al., NeuroSignals 2003;12:31–8). Results: ADD 377010 protected 5/5 animals after a dose of 300 mg/kg (i.p.) in the scMET model and had no effect in the MES model. No other compound was active up to 300 mg/kg in either model, and none caused motor impairment. All four compounds tested prevented NMDA receptor‐mediated cytotoxicity when co‐applied with NMDA. Conclusions: Compounds isolated from an Asian herbal medicine demonstrate promising activity in the NMDA assay, and one also effectively blocked seizures in the scMET model, suggesting it may be responsible for the anticonvulsant properties of its parent extract. Further evaluations of compounds isolated from herbal medicines used for epilepsy are warranted. (Supported by an American Epilepsy Society Research Infrastructure Award. The investigators gratefully acknowledge the support of the NINDS Anticonvulsant Screening Project.) 1,2 Lucie Suchomelova, 1 Roger Baldwin, and 1,2 Claude G. Wasterlain ( 1 Epilepsy Research, VA GLAHS, Los Angeles, CA ; and 2 Neurology, UCLA, Los Angeles, CA ) Rationale: The role of febrile seizures (FS) in epileptogenesis and the contribution of fever to the long‐term consequences of prolonged FS and febrile status epilepticus (SE) are poorly understood. Controversy remains regarding any causal relationship between prolonged FS and temporal lobe epilepsy. This study investigated the contribution of hyperthermia per se to SE‐induced epileptogenesis, using the lithium‐pilocarpine model of SE in immature rats, and carefully controlling for brain temperature and seizure severity. Methods: SE was induced in rats at postnatal day 10 (P10) by administration of lithium (3mEq/kg, i.p.), and 24 hours later by pilocarpine (60 mg/kg, s.c.). Body temperature was maintained during 30 min of SE at 39 ± 1°C (hyperthermic animals) or at 35 ± 1°C (normothermic animals). SE was terminated by diazepam (0.5 mg/kg, i.p.). Control animals received lithium and diazepam in presence of hyperthermia or normothermia. The course of SE was recorded by video/EEG for 24 hours. The severity of SE was assessed by measuring the following parameters: duration of SE (time of last seizure), and total seizure time (time spent in seizures subtracting interictal time) using Harmonie Software. Separate groups of animals, subjected to SE at P10 without surgery, were monitored four months later for the occurrence of spontaneous recurrent seizures (SRS) by telemetry/videotape continuously for one month. Results: We were able to control brain and body temperature during and after SE in both study groups: hyperthermic (39 ± 1°C) and normothermic (35 ± 1°C). Terminating SE by diazepam was effective in both groups (10.2 ± 1.5 min in hyperthermic animals and 12.3 ± 2.2 min in normothermic animals). Total seizure time (54.5 ± 15.2 min after SE in hyperthermic animals and 56.7 ± 12.1 min after SE in normothermic animals) and duration of SE (87.2 ± 23.2 min after SE in hyperthermic animals vs. 95.2 ± 28.2 min after SE in normothermic animals) were similar between groups. Only hyperthermic animals developed convulsive seizures (50%). Both groups of animals that experienced SE at P10 developed electrographic seizures 4 months later (83 vs 55%). However, the frequency and seizure severity were significantly higher in hyperthermic animals (12.5 ± 3.5 vs 4.2 ± 2.0 SRS/day). Conclusions: Since the precise role of hyperthermia during SE in long‐term consequences such as epileptogenesis and neuronal injury has never been adequately defined, this study tightly controlled seizure duration and seizure severity in order to isolate hyperthermia as the main variable and to study its consequences. In this model, hyperthermia greatly increased the epileptogenicity of SE, leading to subsequent behavioral seizures not seen in the normothermic group, and increased the severity and incidence of spontaneous electrographic seizures. (Supported by Grant NS13515 from NINDS, National Institutes of Health, and by the Research Service of the Veterans Health Administration.) 1 Jessica R. Sudbury, and 1,2 Massimo Avoli ( 1 Neurology and Neurosurgery, Montreal Neurological Institute, McGill University, Montreal, QC, Canada ; and 2 Human Physiology and Pharmacology, La Sapienza University, Rome, Italy ) Rationale: The participation of the insular cortex (IC) in temporal lobe epilepsy (TLE) has long been suggested from studies of human epileptic patients; however, its network interactions with the perirhinal cortex (PC) and hippocampus remain uninvestigated. Here, we used a hippocampus‐parahippocampus‐IC rat brain slice to provide the first characterization of network epileptiform interactions among these structures during application of the convulsant 4‐aminopyridine (4‐AP). Methods: Brain slices (450 μm) were prepared from adult male Sprague‐Dawley rats. Field potential recordings were acquired using ACSF‐filled glass electrodes positioned within area CA3 of the hippocampus, the PC and the IC. 4‐AP (50 μM) was bath applied to induce epileptiform synchronous network activity. Results: 4‐AP induced three types of epileptiform activity: (i) fast interictal events (duration = 0.1–0.2 s; interval of occurrence = 0.7–1.2 s) that were mainly observed in CA3 as well as (ii) slow interictal (duration = 0.5–2.0 s; interval of occurrence = 4.0–21.7 s) and (iii) prolonged ictal‐like (duration = 11.3–48.7 s; interval of occurrence = 17.9–332 s) that occurred synchronously among the PC and the IC and occasionally entered the hippocampus. Only slow interictal discharges were present in PC and IC in 15/30 slices. Moreover, slow interictal and ictal events were generated by the IC following surgical isolation. Ictal discharges were abolished by NMDA receptor antagonism while interictal activity persisted. Further antagonism of non‐NMDA glutamatergic receptors abolished interictal discharges and disclosed slow field potentials that recurred at intervals of 11.4–28.0 s in all structures. Bath application of the μ‐opioid receptor agonist (D‐Ala2,N‐Me‐Phe4‐,Gly‐ol5)enkephalin (DAGO, 10μM) significantly reduced the occurrence of these slow potentials in all structures (n = 5); this effect was reversible with naloxone methiodide (20μM). Application of the GABAA receptor antagonist picrotoxin (50μM) abolished these field potentials. Conclusions: Our results demonstrate that epileptiform activity occurs synchronously among PC and IC networks, while analogous activity can be generated independently within the isolated IC. These observations suggest that the network characteristics necessary and sufficient to initiate epileptiform network events are present within the IC. (Supported by: Funded by CIHR (Grant 8109), CURE and the Savoy Foundation.) 1 T. Sutula, 1 J. Ockuly, 1 C. Stafstrom, and 1 A. Roopra ( 1 Department of Neurology, University of Wisconsin, Madison, WI ) Rationale: During investigation of mechanisms of the ketogenic diet, the glycolysis inhibitor 2DG had unexpected anticonvulsant and antiepileptic effects in hippocampal slices and against seizures evoked by kindling of the perforant path. Anticonvulsant and antiepileptic properties of 2DG were further characterized by kindling of olfactory bulb and in screening models in cooperation with the NINDS Antiepileptic Screening Program (ASP). Cognitive and systemic toxicity was examined in rats receiving 2DG for as long as 6 months. Methods: Rats received olfactory bulb or perforant path kindling stimulation twice daily with 1‐sec trains of 60‐Hz 1‐ms pulses at the lowest intensity evoking ADs by a standardized protocol. Interictal and ictal burst frequencies were examined in rat hippocampal slices in 7.5 mM (K+)o. 2DG was evaluated in NINDS ASP screening models in rats and mice including maximal electroshock, subcutaneous metrazol, 6‐Hz stimulation, and Frings audiogenic mice. Cognitive effects were evaluated in rats in the Morris water maze. Results: Application of 2DG (10mM) to hippocampal slices in 7.5 mM (K+)o reduced CA3 interictal bursts by ∼40% and ictal bursts by 60%. 2DG (250 mg/kg) 30 min before kindling of perforant path in rats induced a 2‐fold increase in AD threshold (an anticonvulsant effect) and a 2‐fold increase in the number of ADs required to evoke Class V seizures (an antiepileptic effect). In contrast, 2DG prior to olfactory bulb kindling also resulted in 2‐fold slowing of kindling progression, but had no effect on AD threshold. Initial evidence of anticonvulsant activity was observed in preliminary screening against seizures evoked by subcutaneous metrazol in rats, by 6‐Hz stimulation in mice, and in Fring's audiogenic mice, but not against seizures evoked by maximal electroshock. Additional testing is underway to further characterize the activity found in these models. 2DG at 2 gm/kg daily for 2 weeks in rats had no effect on Morris water maze performance, and body weights of rats receiving 0.5 gm/kg daily for 6 months did not differ from controls. Conclusions: 2DG has novel anticonvulsant and antiepileptic activity against experimental seizures compared to currently available anticonvulsants and a favorable preclinical toxicity profile. Acute anticonvulsant action was observed in vitro against both interictal and ictal discharges. Robust chronic antiepileptic action against kindling progression was independent of the seizure initiation site, but anticonvulsant action against AD threshold was stimulation site specific. The in vivo chronic effects on kindled seizures have been associated with a novel mechanism of antiglycolytic metabolic suppression of seizure‐induced BDNF and trkB expression by the NADH sensor CtBP and the transcription factor NRSF, which regulates neuronal gene expression. (Supported by NINDS R01 25020, the Wisconsin Alumni Research Foundation, and the NINDS Antiepileptic Screening Program.) 1,2 Janos Szabadics, 2 Gabor Tamas, and 1 Ivan Soltesz ( 1 Department of Anatomy & Neurobiology, University of California, Irvine ; and 2 Department of Comparative Physiology, University of Szeged, Hungary ) Rationale: A distinguishing feature of neurogliaform cells is their extremly dense axonal arborization. Moreover, neurogliaform cells are the only known cortical cell type that can reliably activate postsynaptic GABAA and GABAB receptors after discharging a single action potential. Using three‐dimensional light microscopic reconstructions and correlated electron microscopic investigations, we found that neurogliaform cells formed synapses (62 ± 28 μm) closer to the soma of layer 2/3 pyramidal cells than Martinotti cells (102 ± 48 μm) that evoked faster inhibitory responses. Thus, the slow kinetic properties of GABAA responses evoked by neurogliaform cells can not be explained exclusively by the dendritic localization of the synapses. However, there is a twofold difference in unitary axon densities between neurogliaform cells and other interneurons (0.954 ± 0.299 and 0.453 ± 0.015 μm axon in 1000 μm3).We propose that neurogliaform cells release GABA all over their dense axonal field to activate GABA receptors on dendrites present in the axonal cloud. Methods: To test this hypothesis, we studied the physiological properties of responses between neurogliaform cells, other identified interneurons and layer 2/3 pyramidal cells using whole cell patch clamp recording in neocortical slices with the AMPA, NMDA and GABAB receptors blocked. Results: Neurogliaform cells evoked slower GABAA currents (10–90% rise time: 4.5 ± 1.6 ms, weigthed decay time constant: 42 ± 9.7 ms, amplitude: 97 ± 69 pA) than basket cells (10–90% rise time: 0.7 ± 0.2 ms, weigthed decay time constant: 8.1 ± 1.9 ms, amplitude: 130 ± 109 pA). The variability in amplitudes of neurogliaform cell‐evoked IPSCs (0.14 ± 0.04, S.D./mean) was much less than in the basket cell‐evoked responses (0.29 ± 0.1). Furthermore, gradual changing of extracellular calcium resulted in continuous increases in the amplitude of the responses. Blocking of GABA transporter 1 by NO‐711 dramatically increased the decay time constant of neurogliaform cell‐evoked IPSCs (by 203 ± 110%) and did not affect basket cell‐evoked IPSCs. The low affinity, competitive antagonist of GABAA receptors, TPMPA had a larger effect (58 ± 17%) on the amplitude the neurogliaform cells effect compared to basket cells (35 ± 17%). Alfa1 subunit specific GABAA agonists (zolpidem, Cl218,872) had significant effets on the neurogliaform cell‐evoked responses. However, the properties of this modulatory action was different than on basket cell‐evoked IPSCs. Conclusions: These results indicate that a spill‐over‐like mechanism of GABA release takes place following single action potentials in neurogliaform cells, in contrast to the more conventional basket cell synapses. These data suggest that neurogliaform play highly specialized role in regulating neuronal activity in the cortical networks. (Supported by NIH NS35915 to IS & GT; OTKA T049535 and HHMI 55005625 to GT; George E. Hewitt Foundation and Boehringer Fonds to JS.) 1 C. Akos Szabo, 2 Shalini Narayana, 2 Peter V. Kochunov, 2 Crystal Franklin, 1 Koyle Knape, 2 M. Duff Davis, 2 Peter T. Fox, 3 M. Michelle Leland, and 3,4 Jeff T. Williams ( 1 South Texas Comprehensive Epilepsy Center, UTHSCSA, San Antonio, TX ; 2 Research Imaging Center, UTHSCSA, San Antonio, TX ; 3 Southwest National Primate Research Center, San Antonio, TX ; and 4 Department of Genetics, Southwest Foundation for Biomedical Research, San Antonio, TX ) Rationale: The baboon (Papio hamadryas spp) offers a natural primate animal model of photosensitive generalized epilepsy. This study compared changes in cerebral blood flow (CBF) during intermittent light stimulation (ILS) between photosensitive and asymptomatic baboons. Methods: Six photosensitive and four asymptomatic baboons, matched for age, gender and weight, were selected based upon previous scalp EEG evaluation. Continuous intravenous ketamine (5–13 mg/kg) was used for sedation. Subjects underwent five sequential blood flow PET studies within 60 minutes using 20 mCi 15O‐labelled water. Images were acquired in 3D mode (CTI/Siemens HR+ scanner, 63 contiguous slices, 2.4mm thickness). Three resting scans were alternated with two activation scans. ILS was performed at 25 Hz for 60 seconds before to 60 seconds after the start of an activation scan. PET images were coregistered with MRI (3T Siemens Trio, T1‐weighted 3D sequence, TE = 6, TR = 14, TI = 800msec, flip angle 19°, NEX = 6). PET scans were reviewed and corrected for motion artifact. Resting scans were contrasted with activation scans and averaged independently for both groups. Results: The controls showed greatest ILS‐induced activation in the left middle frontal and inferior temporal gyri, left brainstem structures and right postcentral gyrus, and bilaterally in the occipital lobes, posterior cingulate gyrus and cerebellum. In contrast, the photosensitive animals showed strongest ILS activation in the right anterior cingulate and medial orbital gyri, amygdala, globus pallidum, left inferior and superior temporal gyri, and the pineal gland. A striking finding was the absence of occipital activation during ILS in the photosensitive animals. Deactivations were noted in the right orbitofrontal and anterior cingulate cortices in the controls, and in the posterior cingulate gyrus, brainstem and cerebellum of the photosensitive animals. Conclusions: The patterns of ILS‐induced CBF changes differed between control and photosensitive groups. These differences of activations and inhibitions suggest involvement of specific cortical‐subcortical networks in photosensitivity. (Supported by National Institutes of Health (P51 RR13986 through the SFBR to CAS, 1 R01 NS047755 to JTW, and by facilities constructed with support from Research Facilities Improvement Program Grants C06 RR15456 and C06 RR014578 from the National Center for Research Resources).) 1 Patricia Szot, 1 Sylvia S. White, 2 Krista Gilby, and 2 Dan C. McIntyre ( 1 MIRECC, VA Puget Sound Health Care System, Seattle, WA ; and 2 Department of Psychology, Carleton University, Ottawa, ON, Canada ) Rationale: Kindling is a progressive increase in seizure activity following the daily application of focal, low‐intensity electrical stimulation of different forebrain regions. Two strains of rats were developed to exhibit different rates of amygdala kindling, and were called Fast and Slow kindlers. These animals also show different response profiles to α2‐adrenoreceptor (AR) agonists. Therefore, the objective of this work was to examine α2‐AR binding sites and subtype mRNA expression in the CNS of naive Fast and Slow rats. Methods:α2‐AR binding sites were measured in naive Fast (n = 8) and Slow (n = 8) kindlers with the α2‐AR antagonist (3H)RX821002. α2‐AR binding sites were measured in the locus coeruleus (LC), neocortex, bed nucleus of the stria terminalis, thalamus, septum, amygdala, habenula and hippocampus. α2‐AR is composed of three different subtypes. α2A‐ and α2C‐AR mRNA were determined in the CNS of Fast and Slow kindlers by in situ hybridization with oligonucleotides. α2A‐AR mRNA was measured in the LC, cortex, septum, hypothalamus and amygdala, and α2C‐AR mRNA was measured in the LC, cortex, hippocampus, amygdala, cerebellum, caudate and nucleus accumbens. Results:α2‐AR binding sites, as determined by (3H)RX821002, were not statistically different between Fast and Slow kindlers in all regions of the CNS. α2A‐AR mRNA expression, the subtype responsible for (3H)RX821002 binding, therefore, was not different. However, α2C‐AR mRNA expression was different between Fast and Slow kindlers in many regions of the CNS, including nucleus accumbens, dentate gyrus granule cell layer, cingulate gyrus, cerebellum and neocortex layer II: in these cases, α2C‐AR mRNA expression was significantly elevated in the Slow compared to Fast kindlers. Conclusions: These data indicate a signifcant increase in several brain structures in the α2C‐AR mRNA expression in Slow versus the Fast kindlers. These differences in α2C‐AR mRNA may explain the different α2 adrenergic pharmacological profiles between the two strains, and speak additionally to adrenergic involvement in epileptogenesis. (Supported by VA Puget Sound health Care Center and CIHR.) 1 Mayuko Takaki, 2 Yuto Ueda, 3 Akira Nakajima, 4 Yoshiya Murashima, 2 Taku Doi, 2 Keiko Nagatomo, and 5 James Willmore ( 1 Section of Integrative Physiology, Department of Medical Sciences, Faculty of Medicine, University of Miyazaki, Miyazaki, Japan ; 2 Department of Psychiatry, Faculty of Medicine, University of Miyazaki, Miyazaki, Japan ; 3 Section of Chemistry, Department of Medical Science, Faculty of Medicine, University of Miyazaki, Miyazaki, Japan ; 4 Department of Neural Plasticity, Tokyo Institute of Psychiatry, Tokyo, Japan ; and 5 Department of Pharmacology and Physiology, Saint Louis University School of Medicine, St. Louis, MO ) Rationale: The EL mouse is an inbred mutant strain that is used as an animal model of secondarily generalized seizures. Epileptogenesis in EL mice is thought to be caused by collapse of redox state induced by either decreased antioxidant protection or excessive free radical formation. However, measurement of reductants in brain tissue extracts is problematic because synergistic actions of antioxidants such as ascorbic acid, alpha tocopherol or glutathione are effective in scavenging free radicals. Using in vivo methods we developed, we evaluated hippocampal antioxidant ability in EL mice with X‐band Electron Paramagnetic Resonance (EPR) spectroscopy and microdialysis. Methods: Using sodium pentobarbital anesthesia, each mouse underwent stereotactic implantation of guide cannula into the ventral hippocampus. EL mice were injected i.p. with the nitroxide radical 3‐methoxycarbonyl‐2,2,5,5‐tetramethylpyrrolidine‐1‐oxyl (PCAM) at 677 micromol/kg (3.3 ml saline/kg). Following PCAM injection sequential measurements with EPR were used to monitor levels of nitroxide radicals in the extracellular fluid of the hippocampus. In vivo antioxidant efficacy was estimated based on the principle that PCAM loses paramagnetism and that the decay rate of the EPR signal expressed as half life will reflect tissue antioxidant efficacy. Animals used included experimental groups of 6 five week old and 6 eight week old EL mice; all animals were interictal. Mice from the originating strain of EL mice, DDY mice of similar age, were used as controls. Results: EPR half‐life in the five week old mice was equal to control. However, by eight weeks there was significant prolongation in EPR half‐life in EL mice when compared to both control groups of DDY mice. None of the EL mice used in this experiment were subjected to seizure‐inducing stimulation and they remained in the interictal state. Conclusions: Since five week EL mice do not seize in response to tossing stimulus while by eight weeks EL mice are subject to seizures, we concluded that the decreased in vivo antioxidant ability is a fundamental change related to development of seizures. (Supported by Grant‐in‐Aid for Scientific Research (C) (2) (18591297) from the Ministry of Education, Science, Sport and Culture, Japan (to Y.U.).) 1 Nobuhiro Tanaka, 1 Hirochika Imoto, 1 Masami Fujii, 2 Jouji Uchiyama, 1 Hirosuke Fujisawa, 2 Kimihiko Nakano, 3 Ichiro Kunitsugu, 2 Takashi Saito, and 1 Michiyasu Suzuki ( 1 Neurosurgery, School of Medicine Yamaguchi University, Ube, Yamaguchi, Japan ; 2 Applied Medical Engineering Science, Graduate School of Medicine Yamaguchi University, Ube, Yamaguchi, Japan ; and 3 Public Health, School of Medicine Yamaguchi University, Ube, Yamaguchi, Japan ) Rationale: Local cortical cooling for the elimination of epileptic discharges (EDs) has recently become a focus of research. However, the effect of focal cooling on seizures arising from the temporal lobe has yet to be thoroughly investigated. We devised an apparatus for the cooling of the hippocampus in order to examine the effect of focal cooling on experimental hippocampal seizures. Methods: We developed a cooling needle (measuring 1 mm diameter and 6mm length) made of copper which was attached to a thermoelectric (Peltier) chip. Next, experiments were performed on adult male Sprague‐Dawley rats under halothane anesthesia. The cooling needle, a thermocouple and a needle electrode for the EEG recording were then inserted into the right hippocampus after performing a craniotomy. Kainic acid (KA) (1 mg/ml, 1 μl) was injected into the left hippocampus to provoke epileptiform discharges (EDs). Thereafter, focal cooling was started and it was continued for 5 minutes after confirming the appearance of EDs in the right hippocampus which had been provoked by the KA in the contralateral hippocampus. Results: The temperatures of the cooled site in the hippocampus were 9°C in the center, while they were under 25°C within 2.4 mm in radius from the cooling center in phantom experiments using the cooling device.The EDs began to significantly decrease in amplitude from 90 seconds after the start of cooling (p < 0.01). The mean amplitudes of the EDs were suppressed to 68.1 ± 4.8% at the end of the cooling and then stayed low values during the subsequent period. The temperatures of the cooling site were maintained at 33.1 ± 0.7°C before cooling and then they decreased to 14.5 ± 1.3°C finally. Conclusions: We confirmed that hippocampal cooling suppressed the EDs provoked by KA injection in the contralateral hippocampus. Our results therefore seem to suggest that hippocampal cooling may thus be a potentially useful minimally invasive therapy, which might thus be able to replace the need to perform a temporal lobectomy. 1 Kristina Surdova, 1 Erik Taubøll, 2 Anette Krogenæs, 2 Ellen Dahl, 2 Jan L. Lyche, 2 Irma C. Oskam, and 2 Erik Ropstad ( 1 Department of Neurology, Division of Clinical Neuroscience, Rikshospitalet‐Radiumhospitalet, University of Oslo, Oslo, Norway ; and 2 Department of Production Animal Sciences, Norwegian School of Veterinary Science, University of Oslo, Oslo, Norway ) Rationale: It is well known that valproate (VPA) affects endocrine function. There are, however, contradictory reports in the literature as to whether or not VPA affects the hypothalamic‐pituitary‐adrenal axis. Such effects, if present, are important since this hormonal axis is crucial in the homeostasis of fundamental bodily functions and in regulation of stress responses. In the present study the effects of VPA exposure on basal and stress‐induced cortisol was investigated in male adolescent goats. Methods: Twenty‐three bucks aged 2 months at the start of experiment were given VPA mixture (n = 16) or control solution (n = 7) perorally twice daily for 8 months. VPA mixture was given in single doses up to 62.5 mg/kg to sustain therapeutic plasma concentrations (300–600 μmol/l). In Experiment 1, blood samples were analyzed for cortisol weekly from weeks 35 to 50 of the year. In experiment 2, bucks aged 9 months were exposed to a period of mild stress associated with blood sampling from indwelling jugular vein catheters every 15 minutes for 7 hours. The effects of VPA on plasma cortisol concentration were studied in repeated measured mixed model analyses. Results: VPA treated bucks in experiment 1 had significantly lower basal cortisol concentrations than control animals during the study period (F1,14= 6.22, p = 0.026). A significant time effect (F7,98 = 5.92, p < 0.0001) was found with a decrease in basal cortisol plasma concentrations towards the end of the study period. A significant time/treatment interaction (F7,98= 2.63, p = 0.016) was found with a more pronounced decrease in plasma cortisol in the control group than in the treated group. In experiment 2 the cortisol response to mild stress was different in the VPA group compared to control. Overall, there was a lower cortisol response in the VPA group compared with controls. Mean basal plasma cortisol concentration was 2 ng/ml in both groups. During the first period of mild stress plasma cortisol concentration in the control group was 4–8 ng/ml compared to 3–6 ng/ml in the VPA group. In both groups cortisol decreased after 1.5 hours of blood sampling. The group difference was significant when only main effects were included in the model (F1,14= 47.2, p < 0.0001). Conclusions: VPA exposure was associated with a significant reduction in both basal and stress‐induced plasma cortisol concentration in male adolescent goats. The observed effects can be a potential problem with long‐term VPA treatment. On the other hand, it may be speculated that the reduction in cortisol levels and reduced response to mild stress could be related to the mood stabilizing effect of VPA. 1 Michael R. Taylor, 1 Matthew T. Dinday, 2 Sally Chege, and 1,2 Scott C. Baraban ( 1 Neurological Surgery, UCSF, San Francisco, CA ; and 2 Graduate Program In Neuroscience, UCSF, San Francisco, CA ) Rationale: Studies from humans, rodents, and fruit flies have identified dozens of single gene mutations that result in heritable forms of epilepsy or increased seizure susceptibility. Although this genetic basis for epilepsy and seizure susceptibility is well established, there have been no efforts to identify gene mutations conferring seizure resistance. Large‐scale mutagenesis screening in simple vertebrates offers a powerful approach to further studies of the genetic basis of epilepsy. Here we describe the initial results of the first forward‐genetic screen to identify seizure resistant zebrafish mutants. Methods: Approximately 5,000 F1 adult fish were generated from matings between ENU‐treated males and wild‐type females. From these founders 1,896 F2 families were generated and ∼500,000 F3 larvae were used for screening. Using a behavioral assay, we isolated several seizure‐resistant (SR) familes based on their ability to survive prolonged exposure to the convulsant pentylenetetrazol (PTZ). All SR mutants were further characterized by examining resistance to another convulsant (pilocarpine), tracking behavioral responses, and recording electrical activity. In addition, mutants were genetically mapped by bulked‐segregant analysis, and recombination analysis with microsatellite markers was used for fine‐mapping. Results: Six seizure‐resistant mutant families were identified in a forward‐genetic screen over a four‐year period. One mutant, s198, was also shown to be resistant in a pilocarpine survival assay, suggesting multi‐convulsant resistance. Three of the mutants exhibited reduced behavioral activity upon acute PTZ exposure (s139, s333 & s387), and one of these also showed decreased c‐Fos expression by RT‐PCR (s333). Electrophysiology data on mutant s334 revealed an inability to generate long‐duration ictal‐like discharge in response to PTZ or 4‐aminopyridine. Mutant s198 was randomly chosen for mapping studies. Analysis of 192 microsatellite markers, evenly distributed throughout the zebrafish genome, placed the mutated gene on linkage group 24. Recombination analysis with 658 meioses localized the mutation within a 1.1‐centimorgan region (approximately 800 kb). Conclusions: In conclusion, SR zebrafish mutants were successfully identified in a large‐scale forward‐genetic screen. Mutants were isolated with multi‐convulsant resistance, reduced seizure behaviors, and no ictal‐like discharge. Linkage studies on one mutant, s198, have been completed. Further identification and characterization of the defective genes in these mutants will present a more complete understanding of the genetics of seizures, and provide a unique opportunity for developing a genetically‐based cure for epilepsy. (Supported by Epilepsy Foundation of America/Milken Family (M.R.T.) and Klingenstein Fund (S.C.B.).) 1 Michael A. Treiman, 1 Jie Wu, 1 Steven T. Marsh, 1,2 Levi B. Good, 1 Norman Wang, 1 Ryoichi Kimura, 1 Kevin Ellsworth, and 1,2 David M. Treiman ( 1 Neurology Research, Barrow Neurological Institute, Phoenix, AZ ; and 2 Harrington Department of Bioengineering, Arizona State University, Tempe, AZ ) Rationale: Treiman et al. (Epilepsy Res 5:49–60, 1990) described five stages of progressive electroencephalographic (EEG) changes during clinical and experimental status epilepticus (SE), which reflect the underlying severity of the episode of SE. Longer duration of SE and later EEG stages translate to increased problematic cessation of seizures, perhaps due to progressive attenuation of GABA‐mediated inhibition. We hypothesized that marked electrophysiologic changes in GABAA receptor function correlate with a specific EEG stage during experimental SE in rats. Methods: Ten‐week‐old Sprague‐Dawley rats were injected (i.p.) with 3 mM LiCl followed by 30 mg/kg pilocarpine 24 h later to induce SE. Animals were sacrificed before initiation of SE (control), or during EEG Stages I, III, or V, and brains were rapidly removed, placed in cold (4°C) artificial cerebrospinal fluid (ACSF) for 1 min, sliced using a vibratome, and then oxygenated for 1 h in ACSF at room temperature. Paired‐pulse responses with interpulse intervals (IPI) of 30, 50, 100 and 150 msec were induced by stimulation of Schaffer collaterals and recorded from the cell‐body layer of the CA1 region of the hippocampus. γ oscillations (30–80 Hz) were then elicited by tetanic stimulation at a voltage twice that of threshold and were also recorded from CA1. Some Stage V hippocampal slices were perfused with 100 μM GABA or 1 μM diazepam following initial electrophysiological recordings. Results: Paired‐pulse stimulation resulted in increased inhibition in Stage I hippocampal slices compared to control slices (p < 0.05, IPI = 50 msec) but decreased inhibition in Stage V slices compared to Stage I slices (p < 0.05, IPI = 50 msec). γ oscillations were progressively attenuated in slices from later EEG stages. Amplitude (p < 0.01), duration (p < 0.05) and spike count (p < 0.05) were all reduced in Stage V slices compared to control slices. Perfusion with GABA, and to a lesser extent with diazepam, resulted in recovery of γ oscillations in Stage V slices. Conclusions: These results are consistent with other reports of attenuation of GABA‐mediated inhibition during prolonged SE, which may explain the increased refractoriness of later stages of SE to treatment. In addition, our results indicate a possible impairment of presynaptic release of GABA occurred in this model. This observation suggests GABA reuptake inhibitors could be useful for the treatment of refractory status epilepticus. We also report the new finding that paired‐pulse inhibition is greater during Stage I compared to control slices and speculate that this may represent an initial compensatory response to excessive excitation before GABA‐mediated inhibition begins to deteriorate as SE progresses. (Supported by Barrow Neurological Foundation.) 1,2 David M. Treiman, 1 Norman Wang, 1,2 Levi B. Good, 1 Steven T. Marsh, and 1 Kevin J. Garvey ( 1 Neurology Research, Barrow Neurological Institute, Phoenix, AZ ; and 2 Harrington Department of Bioengineering, Arizona State University, Tempe, AZ ) Rationale: There is a need for more effective drugs for the treatment of status epilepticus (SE) but testing potential SE drugs in human patients is difficult. The cobalt‐homocysteine model of generalized convulsive status epilepticus (GCSE) (Epilepsy Res 2:79–86, 1988) can be used to predict the median effective dose (ED50) and median effective concentration (EC50) of drugs against SE. We used this experimental model to study the potential efficacy of levetiracetam in the treatment of GCSE. Methods: 135 male 200–300 gm male Sprague‐Dawley rats were implanted with 4 stainless steel epidural electrodes for EEG recording. 2.5 mg powered cobalt was placed under the left frontal electrode. When the rat exhibited interictal polyspikes and/or right UE twitching, SE was induced by injecting 8 mM homocysteine thiolactone IP. Immediately after the 2nd GTC seizure, levetiracetam (LTA) was administered IP and was considered effective if no further seizures occurred 10–30 minutes after injection. Animals were euthanized at 30 minutes after onset of SE and blood and brains collected for LTA determination by HPLC. Because LTA was not effective alone, its ED50 was determined after administration of 0.1 mg/kg diazepam (ED50 of DZM alone =∼ 5 mg/kg), using doses from 0 to 1000 mg/kg. The neuroprotective potential of LTA was determined by comparing visual‐spatial performance in a Morris Water maze at various EEG stages of SE with and without LTA (500 mg/kg). Results: LTA alone at doses of 10 to 6000 mg/kg was not effective at stopping GCSE in the cobalt‐homocysteine model. The ED50 of LTA, when given in combination with 0.1 mg/kg DZM, was 550 mg/kg (CI 385–800 mg/kg), based on doses from 0–800 mg/kg. Paradoxically, at 1000 mg/kg only 14% of the rats stopped GCSE. 500 mg/kg LTA had no neuroprotective effect on water maze performance in rats in which SE was stopped at EEG stage V, whether given during SE (Stage III), or when SE was stopped at Stage V. Serum and brain LTA concentrations 30 minutes after administration will be shown. Conclusions: LTA is not effective alone in the cobalt‐homocysteine model of GCSE. Its efficacy is enhanced by prior‐ or co‐administration of a subtherapeutic dose of diazepam, but the ED50 is still ∼ 550 mg/kg. Whether or not LTA should be considered for use in human SE will at least partially depend on the serum concentration that corresponds to the ED50 of 550 mg/kg in this model. If the serum concentration is in a range safe for human administration, it can serve as a target concentration in further study of the potential of LTA in the treatment of human SE. It may be that LTA will be an effective second treatment that should be used to provide long‐term protection against relapse in human GCSE after initial use of a benzodiazepine such as diazepam or lorazepam. (Supported by UCB and the Barrow Neurological Foundation.) 1 Andrew K. Tryba, and 2 Charles J. Marcuccilli ( 1 Physiology, Medical College of Wisconsin, Milwaukee, WI ; and 2 Pediatric Neurology, Children's Hospital of Wisconsin, Milwaukee, WI ) Rationale: The neurotrophin, brain derived neurotrophic factor (BDNF), plays a critical role in neural development, migration, plasticity, recovery from injury and learning. In epileptogenic brain areas, such as the temporal lobe, BDNF is up‐regulated during repeated seizures, whereby it is proposed to enhance glutamatergic synaptic transmission and decrease GABAergic inhibition, enhancing the overall excitability of the network (Neuroscientist. 2005 11(4):282–7). BDNF additionally enhances intrinsic neuronal membrane excitability (Nature 1999;401 (6756):918–21). Thus, BDNF has been proposed to promote epileptiform activity (Trends Neurosci. 2001; 24(1):47–53). Although, BDNF over‐expression and exogenous application lead to reduced seizure threshold in experimental models where inhibition is reduced, no one has before shown that BDNF alone induces cortical epileptiform activity. Here, we test the hypothesis that BDNF application can trigger epileptiform activity. Methods: We used brain slice preparations containing temporal lobe cortex of CD‐1 mice (age P10‐P15), where we simultaneously recorded network population activity and from individual cortical neurons using whole cell current clamp techniques. Results: BDNF application (50ng/mL) triggered paroxysmal depolarization shift (PDS) bursting (n = 9/10); PDS bursts were NMDA‐dependent and blocked by the NMDA antagonist, CPP. BDNF enhanced bursting properties in intrinsic bursting (IB) neurons (n = 3), but not in non‐rhythmogenic neurons (n = 7). Application of BDNF also induced seizure‐like population activity in some (n = 8) preparations. Conclusions: Our data support our working hypothesis that BDNF can preferentially enhance bursting properties of rhythmogenic neurons that may trigger SLA. Additionally, BDNF can trigger PDS bursting, one of the hallmarks of epilepsy. Thus, our results suggest that BDNF signaling may be involved in the genesis of epilepsy. (Supported by NIH RO1 HL079294–02 (AKT).) 1,2,3 Pamela A. Valentine, 1 Elizabeth J. Jensen, and 2 Lana J. Ozen ( 1 Department of Clinical Neurosciences, University of Calgary, Calgary, AB, Canada ; 2 Department of Psychology, University of Calgary, Calgary, AB, Canada ; and 3 Neuroconnections Program, Hotchkiss Brain Institute, Calgary, AB, Canada ) Rationale: Deep brain stimulation (DBS) has a long history of therapeutic use in the treatment of a number of neurological disorders, perhaps most effectively for movement disorders such as Parkinson's disease. Neural stimulation may also provide a promising innovative technique for the treatment of pharmacoresistant seizure disorders. While there is experimental and clinical evidence that direct electrical stimulation of deep brain structures can prevent or decrease seizure activity, fundamental questions remain to be resolved including the identification of the most effective brain sites and stimulation parameters, long‐term consequences and potential side effects. The mechanism underlying deep brain stimulation is as yet unknown, although it is thought that the application of high‐frequency stimulation results in a net inhibitory effect that serves to desynchronize activity, effectively abolishing pathologically synchronized epileptiform activity. This may explain why stimulation of a number of structures have been shown to have anticonvulsant properties. Methods: In this study we examined the effect of high frequency stimulation (applied to a number of subcortical sites) on seizure thresholds of electrically kindled temporal lobe seizures. The DBS consisted of unipolar, biphasic square wave pulses, with pulse durations of 300 μs, delivered at a frequency of 100 Hz and an intensity of 200 μA. The DBS was applied 1 second prior to the application of the kindling stimulation to ensure effective activation prior to the seizure inducing stimulus. The seizure activity was scored with respect to the afterdischarge (AD) duration and seizure severity. Results: Preliminary data revealed DBS applied to the entopeduncular nucleus resulted in a strong anticonvulsant effect on the seizure threshold and AD characteristics. We observed the increase in seizure threshold to occur over time which was observed to commence at about the 12th kindling session. The nominal progression in AD characteristics typically observed during kindling were largely absent, with the AD duration becoming shorter, the spike frequency lower and the spike complexity more simplistic. Conclusions: We conclude there may be strong anticonvulsant effects of DBS on temporal cortex seizures that are dependent on the site of stimulation. (Supported by Natural Sciences and Engineering Council of Canada (NSERC), Faculty of Medicine, Hotchkiss Brain Institute, University of Calgary.) 1,2 Erwin A. Van Vliet, 1 Silvana da Costa Araujo, 3 Sandra Redeker, 1 Rosalinde Van Schaik, 3 Eleonora Aronica, and 1,2 Jan A. Gorter ( 1 Center for Neuroscience, Swammerdam Institute for Life Sciences, Amsterdam, Netherlands ; 2 Epilepsy Institute of the Netherlands, Heemstede, Netherlands ; and 3 Department of (Neuro) Pathology, Academic Medical Center, Amsterdam, Netherlands ) Rationale: Disruption of the blood‐brain barrier (BBB) has been found to be associated with various neurological disorders, including the acquired epilepsies such as temporal lobe epilepsy (TLE). However, it is not known to what extent the BBB integrity changes during epileptogenesis and whether alterations in BBB permeability may contribute to the development of epilepsy. To get more insight into this, we determined BBB permeability in epileptic rat and human and studied the possible consequences of BBB opening for the subsequent progression of TLE. Methods: BBB permeability was investigated in the rat at different time points after electrically induced SE using various tracers that bind to albumin. The cellular localization of albumin was further investigated using immunocytochemistry in rat and human epileptic brain specimens. In addition, we investigated whether BBB opening was related to seizure progression. Results: In a rat model for TLE, in which epilepsy develops after an electrically induced status epilepticus (SE), the BBB was most affected 1 day after SE in limbic brain regions. Although a partial recovery took place during the latent period, the BBB was still permeable in the chronic epileptic phase. Tracer deposits (albumin) were found in microglia, astrocytes and neurons. Similarly, albumin extravasation in human was most prominent directly after a fatal SE in astrocytes and neurons, and to lesser extent in hippocampi of temporal lobe epilepsy patients. BBB permeability in chronic epileptic rats was positively correlated to seizure frequency. Artificial opening of the BBB by mannitol in the chronic epileptic phase induced a persistent increase in the number of seizures in the majority of rats. Conclusions: These findings indicate that BBB leakage occurs during the latent and chronic epileptic phase and suggest that this permanent disruption can contribute to the progression of epilepsy. (Supported by the Epilepsy Institute of The Netherlands.) 1 Lia C. Liefaard, 2 Bart A. Ploeger, 3 Adriaan A. Lammertsma, 1 Meindert Danhof, and 1,4 Rob A. Voskuyl ( 1 Division of Pharmacology, LACDR, Leiden, Netherlands ; 2 LAP&P Consultants BV, Leiden, Netherlands ; 3 Department of Nuclear Medicine & PET Research, VU Medical Centre, Amsterdam, Netherlands ; and 4 Epilepsy Institute of the Netherlands (SEIN), Heemstede, Netherlands ) Rationale: The GABAA‐receptor plays an important role in epileptogenesis. A key question is whether downregulation or changes in receptor properties are involved in reduced efficacy of antiepileptic drugs, in particular in pharmacoresistance. Recently, a full saturation approach was developed, in which the whole range of receptor occupancies, obtained from a single PET‐experiment per animal, was used to calculate both GABAA‐receptor density (Bmax) and affinity (KD). This method was used to examine the influence of kindling on Bmax and KD. Methods: Sprague Dawley rats were amygdala kindled up to 5 consecutive stage V seizures (n = 12). Implanted but unstimulated rats served as controls (n = 22). The PET experiment was performed 10–14 days after the last seizure. After injection of an excess amount of flumazenil (FMZ) to fully saturate the receptors, the concentration‐time curves of FMZ were measured in blood with HPLC‐UV and in brain with PET. Next, Bmax and KD were estimated using population pharmacokinetic (PK) modelling. The model consists of a blood, a tissue and 2 brain (“free” and “bound”) compartments. The total FMZ concentration in brain, as measured by PET, reflects the sum of “free” and “bound” concentrations. In the model, exchange between blood, tissue and “brain free” compartments is described by first order rate constants, and specific binding depends on both the concentration of free ligand and the concentration of receptors available for binding. Population PK modelling allows for simultaneous analysis of data from all animals, taking interindividual parameter variability into account. Results: In control rats, all model parameters, including specific binding of FMZ in the brain, as characterized by Bmax (14.5 ± 3.7 ng/ml) and KD (4.68 ± 1.5 ng/ml), could be estimated adequately. Kindling did not affect KD, but Bmax decreased to 64 ± 16% of control. This finding corresponds with a reduction in maximal effect of the allosteric modulator midazolam to 72% of control, as reported by Cleton et al. In addition, the volume of distribution of the brain was increased to 180 ± 27% of control, indicating that transport from the brain was decreased. Conclusions: This study shows that the full saturation technique, using a single injection of FMZ, allows detection of moderate reductions in GABAA‐receptor density and changes in brain PK. The decreased Bmax may underlie the reduced efficacy of midazolam in kindled rats. In addition, the study shows the feasibility to conduct (longitudinal) studies on epileptogenesis and pharmacoresistance development in chronic epilepsy models. (Supported by Christelijke Vereniging voor de Verpleging van Lijders aan Epilepsie and National Epilepsy Fund, #02–06.) 1 Yue Wang, 3 Hitten P. Zaveri, 1 Arko Ghosh, 2 Henning Beckstrom, 1 Nihal C. De Lanerolle, and 1 Tore Eid ( 1 Department of Neurosurgery, Yale University School of Medicine, New Haven, CT ; 2 Department of Anatomy, University of Oslo, Oslo, Norway ; and 3 Department of Neurology, Yale University School of Medicine, New Haven, CT ) Rationale: Glutamine synthetase (GS), which converts glutamate to glutamine, is deficient in specific areas of the sclerotic hippocampus in patients with mesial temporal lobe epilepsy (MTLE). A key question is whether hippocampal GS deficiency is causally related to the generation of seizures in MTLE. Methods: The GS inhibitor methionine sulfoximine (MSO) was infused chronically into the hippocampus of rats at the following doses: 1.0 mg/ml (n = 4), 2.5 mg/ml (n = 18), 5.0 mg/ml (n = 11), and 10 mg/ml (n = 39). Saline infused animals (n = 30) were used as controls. The animals underwent long‐term intracranial EEG and video monitoring for up to 83 days; the activity of GS was assessed biochemically; and the brain pathology was explored by Nissl‐ and silver‐staining of histological sections. Results: Infusion of MSO, but not saline, caused significant (p < 0.01) inhibition of brain glutamine synthetase. A dose of ≥ 2.5 mg/ml MSO consistently produced spontaneously recurrent seizures (stages I – IV) after a clinically silent interval of 5–9 days, whereas no seizures were observed in saline treated animals. Hippocampal gliosis and neurodegeneration were present in all MSO‐treated animals in a dose‐dependant manner (i.e. increasing severity in pathology change with higher dose of MSO). Conclusions: These studies suggest that the deficiency of hippocampal GS is a causative factor in the generation of spontaneously recurring seizures in MTLE, and that GS may represent a novel therapeutic target for this disease. 1 Andrew M. White, 1 Philip A. Williams, 1 Jennifer L. Hellier, 2 F. Edward Dudek, and 1 Kevin J. Staley ( 1 Department of Pediatrics, University of Colorado, Denver, CO ; and 2 Department of Physiology, University of Utah, Salt Lake City, UT ) Rationale: Following severe head trauma, 10 – 50% of survivors will develop epilepsy. Identifying the subset of individuals that will go on to have chronic seizures would greatly improve their clinical care. We used the repeated low‐dose kainate model of temporal lobe epilepsy to investigate the spectral content of the EEG before and after treatment (pre‐ictal) to determine if there are EEG predictors of future epilepsy. Methods: We used implantable radiotelemetry units to continuously monitor the EEG in freely behaving rats. Three sets of rats were observed: (1) rats treated until motor status (MSE) (n = 5), (2) rats treated until electrical status (ESE) (n = 5), and (3) saline treated rats (n = 3). An average Fast Fourier Transform was calculated using the Welch method and a rectangular window. Each window contained 65,536 points (262 sec). The results were averaged over an entire day (329 windows). Three power sectra were compared: (1) prior to kainate, (2) 3 days after kainate, and (3) 8 days after kainate. Results: All MSE rats developed chronic epilepsy. Their power spectrum did change after kainate administration. In 4/5 of the rats considered, the total power increased. The increases corresponded mostly to the lower frequencies, below 5 Hz. These increases persisted in 3/5 of the rats 8 days after treatment. By comparing the spectral density when frequent spikes are present (this was present only with MSE rats) with that when they were absent, we note that there is an increase in both high frequency (due to the sharpness of the spike) and low frequency (due to the periodicity of the spiking) power. We note that the results were invariant to the type of window used (Hamming, Hann, rectangular). None of the ESE rats developed chronic seizures. In these rats, we note that the total power decreased after kainate treatment in 4/5 of the rats. The fact that the power does not increase is not surprising because visually there was no change in the signal (no spikes). We also do not see the low frequency increase in the power spectrum noted in the MSE rats. For control rats the average EEG total power changed by only 10% from before to after kainate treatment. Conclusions: Kainate‐injured rats that develop epilepsy exhibit spectral changes (increases in low‐frequency power) in the EEG that appear prior to the onset of frequent motor seizures. Some of these spectral changes are due to the appearance of spikes (either individually or in clusters). If similar changes are found to be present in humans, it could be possible to predict whether an individual is likely to go on to develop chronic epilepsy following a brain injury. This will have a large positive influence on the care of patients with brain injury and could focus treatment trials on those at highest risk. (Supported by Funding was provided by the NIH.) 1 Dawn N. Wilson, 1 Isabel Chung, and 1 Sookyong Koh ( 1 Neurology, Children's Memorial Hospital, Chicago, IL ) Rationale: Prolonged seizures during childhood can lead to neuropsychiatric consequences including memory impairment, learning disability, and behavioral problems. No effective post‐seizure treatment exists to prevent this neurological insult. Using kainic acid (KA)‐induced seizures, a rat model of temporal lobe epilepsy, we have demonstrated a reversal of seizure‐induced decrease in exploratory behavior in open field testing of young rats by exposure to an enriched environment for 7–10 days after KA. These behavioral changes were accompanied by transcriptional alterations of genes involved in proliferation, long‐term potentiation, and memory consolidation including Arc, Homer1a, and Egr1. The expression of these genes were decreased after KA‐induced seizure but significantly increased after post‐seizure exposure to an enriched environment. To demonstrate a causal role for these genes in alleviating seizure‐induced behavioral deficits, we have delivered recombinant adeno‐associated virus (rAAV2) containing Arc into the hippocampus of young rats after KA‐induced seizure and assessed the subsequent behavioral effects. Methods: Arc, Homer1a, and Egr1 were cloned into AAV2 shuttle vectors and high‐titer rAAV2 vectors were produced by the CMRC Viral Vector Core. Long‐Evans male rats were injected either with PBS or KA at postnatal day 21 (P21) and divided into 5 groups (KA‐ARC, PBS‐ARC, KA‐GFP, PBS‐GFP, PBS‐No Surgery). One day later, rAAV2‐Arc or rAAV2‐GFP was stereotactically injected into the hippocampus of anesthetized rats in the surgical groups. Rats underwent the open field test at P28, 35, and 42. Animals were sacrificed at P44 and brains were immediately flash‐frozen. 40 μm cryostat sections were cut and processed for quantitative real‐time RT‐PCR (qRT‐PCR), in situ hybridization (ISH), immunocytochemistry (ICC), and western blot. Results: In the open field test, KA‐ARC animals explored significantly more than KA surgical controls (two‐way ANOVA, p = 0.0196, n = 8). By the third week of open field testing, a 1.6‐fold increase in exploratory behavior of KA‐ARC rats compared to KA controls was paralleled by a >2.5‐fold increase in Arc mRNA (qRT‐PCR, p = 0.0013, n = 8) and a 2‐fold increase of Arc protein in western blot with Arc antibody. Shown by ISH and ICC, Arc mRNA and protein were primarily localized to the dentate granule cells of the hippocampus. Exploratory behavior significantly correlated with Arc mRNA expression seen in qRT‐PCR (r = 0.701, p = 0.0012, n = 18). Conclusions: Expression of Arc mRNA and protein was increased by injection of rAAV2‐Arc into rat brain, which led to an improvement in exploratory behavior after KA‐induced seizure. These results demonstrate that Arc expression can reverse seizure‐induced behavioral impairments. Future in vivo studies injecting viral vectors containing Egr1 and Homer1a into rat brain will elucidate whether these genes will also afford protection against behavioral deficits after seizure. (Supported by Child Neurology Foundation, K02NS048237.) 1 Melodie Winawer, 3 Rachel Kuperman, 1 Martin Niethammer, 4 Steven Sherman, 5 Daniel Rabinowitz, 5 Irene Plana‐Guell, 7 Christine Ponder, 6 Helen Scharfman, and 2 Abraham Palmer ( 1 Neurology & G.H. Sergievsky Center, Columbia University, New York, NY ; 2 Human Genetics, University of Chicago, Chicago, IL ; 3 Pediatric Neurology, Columbia University, New York, NY ; 4 Epidemiology, Columbia University, New York, NY ; 5 Statistics, Columbia University, New York, NY ; 6 Pharmacology and Neurology Columbia University, Helen Hayes Hospital, West Haverstraw, NY ; and 7 Genetics and Development, Columbia University, New York, NY ) Rationale: Inbred mouse strains are valuable to identify genes contributing to complex genetic diseases such as epilepsy. Chromosome substitution strains (CSS) are strains in which a single chromosome from one inbred strain (donor) has been transferred onto a second strain (host) by repeated backcrossing. A panel of CSS in which each chromosome is represented may be used to identify quantitative trait loci (QTLs) that contribute to a phenotype. Pilocarpine hydrochloride is a muscarinic cholinergic agonist that causes limbic seizures in mice, modeling human temporal lobe epilepsy. QTLs for susceptibility to pilocarpine‐induced seizures have not been reported. C57BL/6J (B6) mice have been shown to be resistant to seizures; the A/J strain has been less studied, but appears more susceptible to seizures. Here, we report QTLs identified using a B6 (host) X A/J (donor) CSS panel to localize genes involved in susceptibility to pilocarpine‐induced seizures. Methods: 340 adult male mice–at least 10 of each CSS, 50 B6, and 39 A/J–were tested for susceptibility to pilocarpine‐induced seizures between 10 and 12 weeks of age. Mice were pretreated with atropine (5 mg/kg), followed in 30 minutes by pilocarpine (250 mg/kg), then continuously observed for 3 hours. Using a seizure staging system adapted from established (Racine) rodent seizure scales, highest stage reached, latency to each stage, and latency between stages were recorded for all mice. Significance of results in the CSS panel screen was determined by nonparametric comparison of mean highest stage reached (Mann‐Whitney), and cox proportional hazard regression was used to generate hazard ratios and p‐values for latency variables. Results: B6 mice were more resistant to seizures and were slower to reach each stage compared to A/J mice (p < .01). The CSS for chromosomes 10 and 18 progressed to the most severe seizure stages, diverging dramatically from the B6 seizure‐resistant phenotype (p < .01). Latency to all seizure stages and between stages was also significantly shorter for CH10 and CH18 mice, with hazard rates as high as 25 times that of B6 mice. Conclusions: Results of CSS mapping suggest the presence of seizure susceptibility loci on mouse chromosomes 10 and 18. This approach can provide a framework for identifying potentially novel homologous candidate genes for human temporal lobe epilepsy. (Supported by NINDS K23 NS02211, K02 NS050429, R01 NS36319, R0141490.) 1 Yan Yang, 1 Connie L. Mahaffey, 1 Terry P. Maddatu, 1 Gregory A. Cox, 1 Joel H. Graber, and 1 Wayne N. Frankel ( 1 Neurogenetics, The Jackson Laboratory, Bar Harbor, ME ) Rationale: BRUNOL4 (Bruno like 4) is a brain specific RNA binding protein, the loss of which leads to both convulsive and absence seizures in mice. The expression of Brunol4 is concentrated in neurons of the brain regions controlling synchronization and oscillation such as cerebral cortex, hippocampus and thalamus, suggesting that BRUNOL4 may play important roles in the maintenance of normal neuronal rhythmic activities. We hypothesize that BRUNOL4 is involved in the processing of mRNAs encoding proteins critical to neuronal excitation. Thus, BRUNOL4 deficiency would lead to altered expression of several molecules which in turn would result in seizures. This study aims to identify differentially expressed genes in the Brunol4 mutant brains and to determine how BRUNOL4 regulates gene expression. Methods: RNAs were extracted from mutant and control brains before the onset of overt seizures. The transcriptomes between mutants and controls were compared by microarrays, followed by northern and western analysis to confirm the expression difference. A potential BRUNOL4 recognition site in the target RNA transcripts was identified by multi‐species sequence comparison. Direct binding between BRUNOL4 and its RNA targets was demonstrated by RNA‐immunoprecipitation in neurons. The consequences of the BRUNOL4‐RNA interactions were investigated through reporter assays and RNA decay studies in cultured neurons. Results: Multiple potential BRUNOL4 target RNAs with altered expression were identified in mutants. Subsequent work confirmed the reduced expression of four RNAs encoding proteins critical to neuronal excitation in the mutant brain. The RNAs encode N‐ethylmaleimide‐sensitive factor, serotonin receptor 2c, synapsin II and α‐synuclein, some of which have been implicated in seizures while others may represent novel pathways in epilepsy. Sequence analysis revealed a highly conserved U/G rich motif in the 3′ untranslated region (UTR) of all four genes. RNA‐ immunoprecipitation showed that BRUNOL4 interacts with the 3′ UTRs of the four RNA transcripts through binding to the conserved U/G rich region. In luciferase reporter assays, the expression of chimeric reporters containing 3′ UTRs of the target RNAs was down‐regulated in the Brunol4 mutant neurons compared with wildtype neurons, a defect that can be rescued by adding back BRUNOL4 through transfection. Preliminary studies suggest that the mRNA decay rates of the four RNA transcripts are shorter in mutant neurons. Conclusions: BRUNOL4 appears to modulate neuronal excitability at the RNA level through fine‐tuning the stability of RNA transcripts encoding proteins critical to neuronal excitation. Brunol4 mutant mice represent the first animal model where compromised mRNA processing leads to epilepsy, thus providing a novel disease mechanism for seizure disorders. (Supported by a research award from Citizens United for Research in Epilepsy (YY) and a grant from NINDS (NS31348 to WNF).) 1 Guo Yin, 1 Peter I. Jukkola, 1 Hong Xu, and 1,2 Kevin M. Kelly ( 1 Neurology, Center for Neuroscience Research, Allegheny‐Singer Research Institute, Pittsburgh, PA ; and 2 Neurology, Drexel University College of Medicine, Philadelphia, PA ) Rationale: Macroscopic properties of the EEG are frequently correlated with abnormal behaviors. We observed an apparent correlation of increased generalized spike wave discharge (SWD, absence seizure) amplitude to abnormal behaviors in aged F344 rats during long‐term video‐EEG recordings. We sought to validate quantitatively whether SWD amplitude could be used as a reliable marker to predict myoclonic arousals and limbic seizures in these animals. Methods: SWD amplitude recorded from six EEG channels during the initial 5 hours of a 24 hour recording was observed, and 7 schema were established to predict two major types of abnormal behaviors (myoclonic arousal and limbic seizures) in the following 19 hours of recording. The 7 schema consisted of: 1) event occurrence frequency; 2) myoclonic arousal; 3) head nodding (class II); 4) forelimb or hindlimb clonus (class III); 5) running and rearing (class IV); 6) falling and jumping (class V); and 7) fatality. Five episodes of SWDs were chosen randomly in the first 5 hours of recording and an average amplitude was generated from 6 EEG derivations: F3‐C3, C3‐P3, F3‐P3, C4‐P4, F3‐F4, and C3‐C4. Quantitative prediction for different abnormal behaviors were determined based on the amplitude variation of the 6 channels. For example, it would be predicted that there would be no myoclonic arousal when the amplitude of F3‐C3 was greater than that of F3‐P3 by more than 5%, whereas there would be 1 to 2 events of myoclonic arousal in the subsequent 19 hours when the F3‐P3 amplitude was over 10% greater than F3‐C3 while the amplitude of C3‐P3 was less than or equal to C4‐P4. For purposes of simplicity, a standard scheme of quantitative comparison was designed: 1 class III event = 2 class II events, 1 class IV event = 4 class II events. Accuracy was judged by comparing the number of observed events to the predicted results. For example, based on SWD amplitude, the range of actual class II events is predicted between 5 and 8 times. Prediction failure occurs when the actual class II event number is 3 or 10. Results: Twenty two EEG files of 20 month old F344 rats (n = 9) were analyzed and 3 prediction categories included: 1) frequency of occurrence of myoclonic arousal; 2) event frequency; and 3) limbic event class and fatality. Analysis indicated prediction accuracy of 80% for myoclonic arousals, 70% for event frequency, and 75% for limbic events ± fatalities. Taken together, prediction accuracy was ∼70%. Conclusions: These findings suggest that macroscopic properties of electrocerebral activity in aged F344 rats may be used as predictors of abnormal ictal behaviors in these animals. These findings also suggest the possibility that specific differences in EEG spectral power may be quantifiable and equally predictive of these behaviors. (Supported by NIH R01NS046015 to KMK.) 1 Junli Zhou, and 1 Gregory L. Holmes ( 1 Neuroloscience Center at Dartmouth, Dartmouth Medical School, Hanover, NH ) Rationale: Patients with epilepsy suffer from varying degrees of postictal impairment including confusion and amnesia. This impairment adds substantially to the disease burden of epilepsy. However, the mechanism responsible for postictal cognitive impairment is unclear. In this study we monitored place cells in the CA1 region of the hippocampus before and after spontaneous seizures in epileptic rats. Place cells fire action potentials when the animal is in a specific location in space, the so‐called place field. Place cell function correlates well with performance in tasks of visual‐spatial memory and are an excellent measure of spatial memory. Methods: Male Sprague‐Dawley rats at P70 were given lithium chloride (127 mg/kg) given intraperitoneally (IP) 18 hours before an IP injection of pilocarpine (34 mg/kg) to induce SE. One month later the rats were implanted with eight tetrodes and place cells and interneurons recorded. We compared place cell and interneuron firing patterns before and after spontaneous seizures. For the pyramidal neurons we measured coherence, information content, AP firing, and total number of APs. Results: A total of 12 spontaneous seizures were recorded from six rats. Of 87 cells recorded in these six rats, 20 met the requirements for a place cell and were detected during all four recording sessions (two before and two after the seizure). In these cells there were significant decreases in coherence, information content, firing rate and total spikes following the seizure (Figure). In addition, eight cells that did not have place firing prior to the seizure developed firing fields following the seizure. Nine cells that had firing fields on both recording sessions prior to the seizure could not be recorded following the seizure. Seven interneurons were recorded from four rats were recorded. The seizures had no effect on firing frequency of the interneurons. Conclusions: Following seizures there was a dramatic decrease in firing rate of action potentials. In addition, firing fields were aberrant with reduced coherence and information content. In addition, to postictal suppression of firing patterns, seizures lead to the emergence of previously silent cells. Interneuron firing was not affected by seizures. The concept that postical impairment is due to simple neuronal exhaustion is no longer tenable. Our findings demonstrate that the effects of seizures are complex, affecting neuronal function in a cell specific manner. (figure 1) (Supported by Western Massachusetts Epilepsy Awareness Fund, Friends of Shannon McDermott, the Sara fund, and grants from NINDS (Grants: NS27984 and NS44295).) 1 Jacqueline J. Ardesch, 1 Laetitia J.J.C. Wagener‐Schimmel, 2,3 Hans E. Van Der Aa, 1 Gerard Hageman, and 3 Hendrik P.J. Buschman ( 1 Neurology, Medisch Spectrum Twente, Enschede, Netherlands ; 2 Neurosurgery, Medisch Spectrum Twente, Enschede, Netherlands ; and 3 Twente Institute for Neuromodulation, Medisch Spectrum Twente, Enschede, Netherlands ) Rationale: Few adverse events on heart rate have been reported with vagus nerve stimulation (VNS) for refractory epilepsy. This effect may be caused by electrical stimulation of the efferent nerve fibers innervating the heart. We describe three cases with intraoperative bradycardia during device testing. Methods: From 1999 until 2006 approximately 90 patients with medically refractory epilepsy have been implanted with a VNS system in our centre. We review 3 patients who showed a bradycardia during intraoperative device testing (Lead Test) and their VNS‐therapy follow‐up outcome. Results: Patient 1, a 32 year old female with complex partial epilepsy, experienced a bradycardia of 43 beats/min during the first Lead Test, and 25 beats/min during a second test. Ten days after the operation the VNS system was activated under ECG‐monitoring. No cardiac events occurred. Patient 2, a 52 year old male with partial epilepsy, experienced bradycardia of 40 beats/min. Ten days after implantation no cardiac effects were recorded when VNS therapy started. Patient 3, a 59 year old female with tonic clonic seizures, had a change in heart rate from 63 beats/min to 54 beats/min. No change in cardiac rhythm occurred when VNS therapy started. All three have had 4–6 years of chronic stimulation without cardiac side effects. The effect on epilepsy varied between 30%– > 50% reduction in seizure frequency. They also had less severe seizures and improved postictal recovery. Patient 1 and 2 have received a new pulse generator when battery was depleted. During the subsequent intraoperative Lead Tests no cardiac changes on ECG occurred. Conclusions: Our results suggest that bradycardia during intraoperative device testing does not have to be a reason for aborting the operation. However, postoperative VNS‐therapy onset should be done carefully and under ECG monitoring. Furthermore, long‐term VNS in these patients seems safe in respect to heart rate. 1 Edward Bromfield, 1 Barbara Dworetzky, 2 Shelley Hurwitz, 1 Zina Eluri, 1 Lara O’Brien, 1 Sonia Replansky, and 3 David Mostofsky ( 1 Neurology, Brigham and Women's Hospital, Boston, MA ; 2 Medicine, Brigham and Women's Hospital, Boston, MA ; and 3 Psychology, Boston University, Boston, MA ) Rationale: Fatty acids reduce seizures in animal models, but there is little human data. In a recent randomized trial (Epilepsy and Behavior 2005;7:253–8), patients with intractable epilepsy showed only a transient response. We are reporting the second such trial. Methods: Adults with uncontrolled epilepsy (at least 4 seizures/mo) were randomized 1:1, double‐blinded, to PUFA (eicosapentanoic acid (EPA) plus docosahexanoic acid (DHA), 2.2 mg/d in a 3:2 ratio) or placebo (mineral oil). Following a 4‐wk prospective baseline and 1‐wk titration phase, subjects entered a 12‐wk treatment period. Quality of life was assessed (QOLIE‐31) before and after treatment. Upon completion, subjects had the opportunity to participate in a 1‐mo open‐label PUFA trial. Nonparametric statistics were used. Results: 27 subjects were randomized; 4 decided not to participate before entering baseline (3 had been randomized to placebo and 1 to PUFA). 2 other subjects, both randomized to placebo, were dropped during baseline, 1 because of noncompliance and the other because of increased seizures. Completers included 12 PUFA subjects (7F, 5M; ages 25–55 (mean 37)), vs 9 on placebo (5F, 4M; ages 22–62 (mean 39)). 10 on PUFA and 6 on placebo had focal epilepsy, while 2 and 3 had generalized epilepsy (cryptogenic/symptomatic in 2 and 1, idiopathic in 2 on placebo). Subjects took from 1 to 3 antiepileptic drugs, with no significant group differences. 2 on placebo vs. 0 on PUFA had >50% decrease in seizure frequency from baseline during the 12‐week treatment (p = 0.17). Seizure frequency increased 6% on PUFA and decreased 12% on placebo (p = 0.21). On the other hand, of 19 subjects completing the 4‐wk open‐label PUFA trial, 15 experienced fewer seizures than during baseline (p = 0.02), 5 by at least 50%; 4 of these 5 had previously been on placebo. QOLIE‐31 scores increased an average of 1 point among PUFA subjects, and decreased 6 points among placebo subjects (p = 0.23). PUFA subjects during the randomized trial and all subjects during the open‐label phase showed no differences in serum drug concentrations other than lamotrigine, which declined by a mean of 17% among 6 patients. Conclusions: In this trial, PUFA was not superior to placebo as a treatment for intractable epilepsy. Given PUFA's absence of adverse effects and efficacy in animal models, additional trials of different doses or EPA:DHA ratios could be justified, particularly a longer, larger trial including children as well as adults. Adequate controls will be needed to account for what can be a strong placebo effect. (Supported by CURE Foundation; The General Clinical Research Center of Brigham and Women's Hospital, supported by the National Center for Research Resources (5M01 RR‐02635); Carlson Laboratories, suppliers of the treatment and placebo capsules.) 1,2 Hendrik P.J. Buschman, 2 Jasper R. Sikken, 3 Willem Kersing, 2 Peter H. Veltink, and 1,4 Hans E. Van Der Aa ( 1 Institute for Neuromodulation (TWIN), Medisch Spectrum Twente, Enschede, Netherlands ; 2 Biomedical Signals & Systems Group, Department Electrical Engineering, University of Twente, Enschede, Netherlands ; 3 Otolaryngology and Phoniatrics, Medisch Spectrum Twente, Enschede, Netherlands ; and 4 Neurosurgery, Medisch Spectrum Twente, Enschede, Netherlands ) Rationale: VNS is an effective treatment for patients with medically refractory epilepsy. It is generally accepted that VNS owes its antiepileptic effect from afferent nerve fiber stimulation. Co activation of efferent fibers can cause side‐effects. One of the most frequent reported side‐effects is hoarseness due to stimulation of the recurrent laryngeal nerve. At high stimulation amplitudes even spastic contractions of ipsilateral intrinsic laryngeal muscles can occur. For a group of VNS‐therapy users this side‐effect interferes severely with their daily activities. We investigated the recruitment properties of the recurrent laryngeal nerve and analyzed electromyographic and morphometric alterations on the vocal folds during VNS. Methods: Vocal fold EMG experiments were conducted intra‐operatively during the implantation of a VNS system. When the patient was anesthetized the endotracheal EMG tube (XOmed) was inserted, and the recording wires connected to the Monitor. During surgery the VNS‐therapy stimulation lead and pulse generator were implanted following normal procedure. The pulse generator was then programmed to stimulate in magnet mode for 14 seconds at 2 Hz, and one of the following pulse durations: 130, 250, 500, 750, or 1000 micros. At each pulse width the EMG‐threshold current was determined by electrical stimulation of the VN with increasing stimulation currents. When EMG signal amplitude saturated upon stimulation a second pulse duration was randomly selected and the above procedure repeated. Laryngostroboscopic examination was performed 1 and 6 months after surgery. Special attention was given to the effects of spontaneous stimulation on the larynx. If no effect was noticed a higher intensity stimulation cycle was given by means of ‘magnet‐activation’. Results: The vocal fold EMG and morphodynamic changes in the larynx have been analyzed in 5 patients. In all patients left vocal fold EMG threshold was between 0.25 and 0.50 mA. Pulse duration had little influence on the EMG threshold level. Vocal fold EMG saturation levels were reached between 0.75 and 1.00 mA. Videostroboscopic monitoring at therapeutic levels (1.25 – 2.25 mA) showed that VNS is well tolerated. ‘Magnet’ stimulation induced an adductory spasm of either the ipsilateral vocal fold or the vestibular fold, and was present remarkably irrespective of the presence of hoarseness. Conclusions: VNS causes pronounced effects on the vocal folds even at low stimulation amplitudes. At therapeutic levels the effect on the vocal folds is maximal even at the lowest stimulation pulse durations. The vocal fold contractions, however, do not necessarily give audible effects (hoarseness). 1 Robert S. Fisher, and SANTE Study Group ( 1 Neurology, Stanford Medical Center, Stanford, CA; and See Listing on Epilepsycontrol.com ) Rationale: We present an interim report for the SANTE (Stimulation of the Anterior Nucleus of the Thalamus for Epilepsy) trial, to illustrate the structure of the trial. Methods: The subject pool consists of patients 18–65 years old, inclusive, with partial onset seizures with or without generalization, and with observable clinical seizure manifestations. Seizures must be documented by prior video‐EEG monitoring, and they must average at least six per month. Patients must have failed at least three AEDs, with dosing stable for 30 days prior to enrollment. Women must be using a reliable method of birth control. Excluded are patients with IQ < 70 or inability to complete neuropsychological testing, progressive lesions or diseases, or psychiatric hospitalization, suicide attempt or psychosis (excluding postictal psychosis) within five years. If present, a VNS is to be explanted at time of thalamic implantation. After a three‐month baseline, eligible subjects receive bilateral anterior thalamic nuclei implantation of a dual‐channel Medtronic Model 7428 Kinetra Neurostimulation System with Model 3387 DBS Brain Leads. One month later, stimulation is initiated at either 0 (placebo) or 5 (active treatment) V, with pulse width 90 us, 145 pulses per second, on for 1 minute and off for 5. The double‐blind period continued for 3 months, after which there is 9 months of systematic variation of certain stimulation parameters. Results: As of 3/27/06, 98 patients were enrolled and 58 implanted at 15 US medical centers. Of these, 25 discontinued prior to implantation. Reasons included withdrawal of consent, inability to maintain stable antiepileptic drugs, status epilepticus, insufficient numbers of seizures in the baseline, and miscellaneous other reasons. Among 85 patients completing the baseline visit, age (mean ± SD, range) was 35.4 ± 10.9 (18.5–60.9) years, 45 male, 40 female. Epilepsy duration was 22.2 ± 13.1 (2–60) years. AEDs numbered 1 in 9 patients (11.3%), 2 in 37 (46.3%), 3 in 32 (40.0%), and 4 in 2 (2.5%). A vagus nerve stimulator previously had been implanted in 36 (42.4%); 17 (20.0%) had prior epilepsy surgery and 12 had both epilepsy surgery and VNS implantation. Conclusions: The SANTE trial is proceeding satisfactorily. Efficacy and safety data will be analyzed and presented after 102 patients have completed the blinded phase. (Supported by Medtronic.) 1 Eric H. Kossoff, 2 Hannah M. Rowley, 3 Saurabh R. Sinha, and 1 Eileen P. Vining ( 1 Epilepsy, Johns Hopkins Medical Institutions, Baltimore, MD ; 2 Nutrition, Johns Hopkins Medical Institutions, Baltimore, MD ; and 3 Neurology, Sinai Hospital, Baltimore, MD ) Rationale: The ketogenic diet is not typically offered to adults with epilepsy due to the significant lifestyle alterations needed for its use. The modified Atkins diet has been recently demonstrated to be therapeutic for children with intractable epilepsy without the need for admission, weighing foods, and fluid, calorie, and protein restrictions. We hypothesized the modified Atkins diet would be similarly well tolerated and effective for adults. Methods: Study inclusion criteria included age over 18, and at least weekly seizures and 2 attempted anticonvulsants. Carbohydrates were initially restricted to 15 grams/day, fats encouraged, and fluids, protein, and calories were ad lib. Medications were unchanged for the first 3 months. Patients were examined and laboratory results obtained at baseline, 1, 3, and 6 months on the diet. Results: Thirty patients, median age 36 years (range, 18 to 53), 19 (63%) female, consented and 29 started the diet. The median number of prior medications was 8 (range, 2–14); seizures were occurring 10 times per week (range, 1–140). Twelve patients had vagus nerve stimulators, 8 had failed resective surgery. Using an intent‐to‐treat analysis, approximately 40% had a >50% seizure reduction after 1 and 3 months on the diet. In those with seizure reduction, the median time to improvement was 2 weeks (range, 1 week to 2 months). Improvement was slightly higher in those with moderate to large urinary ketosis at the 1‐month visit, p = 0.08. The mean weight loss was 6.9 kg; mean BMI decreased from 29.0 to 26.5, p = 0.001. Cholesterol increased from a mean of 182 to 201 mg/dl after 1–3 months, p = 0.02. Compliance was problematic; 11 (37%) discontinued the diet prior to 3 months. Conclusions: A modified Atkins diet appears to be an effective therapy for adults with intractable epilepsy and can also induce weight loss. Considering the high dropout rate, a 2‐month trial appears to be adequate. (Supported by General Clinical Research Center, NIH/National Center for Research Resources grant M01‐RR00052.) 1 Kevin Novak, 2 JoAnn Hoeppner, 1 Ruzica K. Ristanovic, 1 Lawrence P. Bernstein, 1 Jesse Taber, and 3 Jeffrey Cozzens ( 1 Neurology, Evanston Northwestern Healthcare; Northwestern University, Evanston, IL ; 2 Pediatrics ; and 3 Neurosurgery ) Rationale: Patients receiving VNS therapy for refractory epilepsy often appear on follow‐up to be more alert, attentive, and energetic, independent of the effects of VNS on seizure frequency. This effect has not been fully demonstrated, and potential mechanisms of action have not been explored. We hypothesized that the apparent improvement in attention and energy results from a reduction in IEDs, which may interfere with cognitive processes. Methods: A heterogeneous group of 12 subjects with medically refractory, localization related epilepsy was evaluated at baseline, and at 6 months after VNS implantation surgery. Subjects had different epilepsy syndromes, baseline seizure and IED rates, and baseline IQs. During a 6 hour EEG recording session, an extensive battery of neuropsychological tests was administered to test for global cognitive function in several domains, with particular emphasis on attention. EEG spikes were detected with Persyst software, while custom software was used for EEG noise reduction and analysis of power spectrum, coherence, and statistics. Quality of life was examined with the QOLIE‐31 questionnaire. Seizure frequency was recorded by subject and caregiver report. The paired Student's t test or Wilcoxon signed rank test was used for statistical analysis (p < 0.05 considered significant). Results: The group had nonsignificant reductions in seizure frequency (mean 10.0 to 2.4/month, p = 0.07) and IED rate (64.1 to 7.7 spikes/hr, p = 0.08). There were no significant changes in relative EEG power acutely (VNS‐on vs ‐off) or chronically (baseline vs VNS‐off), although there was a trend of decreased delta and increased alpha power after VNS. Coherence did not change acutely with stimulation, but 9 of 12 subjects had significant chronic changes. Quality of life trended toward improvement (QOLIE‐31 raw score 52.5 to 61.9, p = 0.17). No changes were seen in any of the cognitive domains examined in standard neuropsychological testing, but the longer lasting of the two computerized tests of sustained attention/vigilance trended toward improvement (Steer clear driving: 7.02 vs. 4.95 cows/min hit, p = 0.11). The longer duration sustained attention driving test performance was correlated with baseline IED rate (r2 = 0.27, p = 0.05), and improved performance correlated with decreased IED rate after VNS (r2 = 0.31, p = 0.04). Conclusions: Although nonsignificant, 6 months of VNS in a small number of subjects resulted in a trend of reduced seizure frequency and IED rate, but had no measurable effect on cognition. The trend of reduced IED rates significantly correlated with improved ability to sustain attention in a longer duration task. Perhaps a larger study on a more homogeneous group of subjects selected for high IED rates and low to moderate baseline cognitive function would show larger changes in cognition. (Supported by Cyberonics, Inc.) 1,2 Marianna V. Spanaki, 3 David Greene, 1,2 Brien J. Smith, 1 David Burdette, 3 Martha Morrell, and 1,2 Gregory L. Barkley ( 1 Neurology, Henry Ford Hospital, Detroit, MI ; 2 Neurology, Wayne State University, Detroit, MI ; and 3 NeuroPace, Inc., Mountain View, CA ) Rationale: A significant percentage of women with epilepsy (WWE) have reported increased seizure frequency during the perimenstrual period (catamenial epilepsy). We first documented catamenial exacerbation of epileptiform activity using the Responsive Neurostimulator System (RNS) in 2005 (M V Spanaki, et al. Epilepsia 2005;46(Suppl 8):222). In this study we continuously measured epileptiform activity in two WWE over a total of 17 menstrual cycles. Methods: As part of a FDA‐approved multicenter feasibility study investigating the safety and potential efficacy of the RNS for epilepsy, we have implanted two WWE who were not good surgical candidates. The patients kept diaries of their menstrual periods. After seizure onset patterns were identified for both patients, detectors were programmed to identify electrocorticogram (ECOG) patterns that occurred prior to clinical seizures. These patterns appear epileptiform and occur frequently, but the vast majority of these bursts do not result in clinical seizures. Detection data from the RNS were then analyzed in multi‐day windows around the date of menses onset and compared to detection rates on days exclusive of these perimenstrual windows to determine if the daily detection rate differed between the menstrual and non‐menstrual intervals. Results: Data was gathered for 9 and 8 menstrual cycles for patients H and M, respectively. The detection rates for the different windows are shown in the graph. (figure 1)The detection data show statistical difference for both patients at window sizes of ± 3 days relative to the onset of the patient's menstrual cycle at the alpha = 0.05 level (95% confidence). Conclusions: The RNS has detection and diagnostic capabilities that allow objective measurement of epileptiform activity over prolonged periods of time in ambulatory patients. Use of RNS has confirmed the existence of perimenstrually exacerbated seizure activity. (Supported by: The RNS is limited by U.S. law to investigational use only. This study was sponsored by NeuroPace, Inc.) 1 Alan W.C. Yuen, 1 Josemir W.A. Sander, 1 Dominique Fluegel, 1 Philip N. Patsalos, 2 Lucy Browning, 1 Gail S. Bell, 3 Tony L. Johnson, and 1 Matthias M. Koepp ( 1 Department of Clinical and Experimental Epilepsy, UCL Institute of Neurology, London, United Kingdom ; 2 MRC Human Nutrition Research, Cambridge, Cambridge, United Kingdom ; and 3 MRC Biostatistics Unit, Cambridge, United Kingdom ) Rationale: Fatty acids (FA) have important roles in determining optimal structural and functional properties of neuronal membranes and may have an important impact on epilepsy. Pharmacology studies have shown that the FA, eicosapentaenoic acid (EPA, an omega‐3 FA), docosahexaenoic acid (DHA, an omega‐3 FA) and arachidonic acid (AA, an omega‐6 FA) can increase seizure thresholds. One study showed that the ketogenic diet increased plasma concentrations of EPA, DHA and AA. This increase may be a factor in the diet's anti‐seizure effects. In a double blind study examining effects of omega‐3 FA supplementation in people with epilepsy, we also examined FA profiles prior to and following supplementation. Methods: Patients with chronic epilepsy were randomised to receive capsules providing EPA (1g) and DHA (0.7g) daily or placebo (mixed vegetable oils) in a 12‐week double blind study. Seizure counts, adverse events, AEDs and 28 different red blood cell (RBC) FA concentrations were assessed during the study. The effect of concomitant AEDs on baseline FA concentrations was assessed by comparing those on AEDs with those not on the AEDs, using the Wilcoxon rank sum test. This study was approved by the local ethics committee. Results: The RBC FA profile obtained in 56 subjects prior to the start of the study showed mean% of AA, EPA and DHA of 13.1, 0.9 and 3.9. Carbamazepine (CBZ) and oxcarbazepine (OXC) appear to affect baseline FA concentrations, so analysis was done comparing those on CBZ (26) and those on OXC (20) with those taking neither CBZ nor OXC (10). Those taking CBZ showed 22% lower mean DHA (p < 0.05), 20% lower long chain omega‐3 FA (p < 0.05) and 5% lower monosaturated FA (p < 0.01). Those taking OXC showed non‐significantly higher EPA (23%), DHA (14%) and AA (6%) levels, which together contributed to a higher total polyunsaturated FA of 8% (p < 0.05). Like CBZ, OXC was also associated with lower (8%) monosaturated FA (p < 0.001). Following 12 weeks of supplementation, the 29 subjects in supplement group showed a 13% reduction in AA (p < 0.001), 216% increase in EPA (p < 0.001) and a 55% increase in DHA (p < 0.001). Caution needs to be taken in interpreting these results as no correction factor was applied for the multiple analyses undertaken. Conclusions: CBZ appears to be associated with lower DHA and long chain omega‐3 FA whilst OXC appears to be associated with higher total polyunsaturated FA. Hence OXC appears to be associated with a potentially more favourable FA profile. These findings need to be confirmed in subjects starting on these drugs as monotherapy. This study also showed that EPA and DHA concentrations are increased with omega‐3 FA supplementation. (Supported by Seven Seas Ltd. UK provided study medication.) 1 Ann Bergin, 2 Karen Costas, 2 Vanessa Ludlow, 1 Francie Mandel, and 3 Virginia Kimonis ( 1 Neurology, Childrens Hospital ; 2 GI/Nutrition ; and 3 Genetics, Childrens Hospital, Boston, MA ) Rationale: Angelman syndrome (AS) is associated with deletion of the maternal chromosome 15q11.2–15q13 in >70% of cases. Uniparental disomy, mutations of the UBE3A gene or of an imprinting center, cause an additional 15% of cases. Refractory epilepsy is particularly common in deletion‐associated cases. Multiple seizure types occur including atypical absences and prolonged non‐convulsive status epilepticus (NCSE). Anticonvulsant drugs (AEDs), often in combination, are incompletely effective. Each of the molecular mechanisms known to be associated with AS causes dysfunction of the UBE3A gene product, the E3 ubiquitin‐ligase protein. In deletion cases other genes, including the gene encoding g‐aminobutyric acid (GABAA) β3 subunit is deleted. A model for AS, mice lacking the β3 subunit, show myoclonic jerks, abnormal EEG and seizures with motor and learning defects. The β3 subunit deficit may result in abnormal thalamocortical functioning, favoring atypical absence seizures and NCSE. The ketogenic diet (KD) is an effective treatment for refractory epilepsy. The mechanisms by which the KD exerts its anticonvulsant effect are not known. Proposed molecular targets include mitochondrial uncoupling proteins, potassium channels and, of interest for deletion‐positive AS patients, GABA‐ergic neurotransmission favoring inhibitory effect. In this context we present 3 patients with Angelman syndrome and refractory epilepsy who were successfully treated with the ketogenic diet. Methods: A modified ketogenic diet (brief fast, full hydration, acid‐base correction) was initiated in hospital after neurological, nutritional and neurometabolic testing revealed no contraindication to the KD. Final KD ratio was determined based on efficacy and level of ketosis. Results: A: 3 year old girl, deletion positive AS and refractory NCSE, on 4 AEDs. Seizure free without complications on KD. Improved alertness, motor control, respiratory health, wellbeing. B: 2 year old boy, deletion positive AS, on 2 AEDs. Seizure free off AEDs on KD. More alert, more progress. C: 14 year old boy, one of 2 affected sibs with linkage to 15q11–13 but no mutation, with refractory status epilepticus, on 5 AEDs. Decrease in admissions for status epilepticus from 5/year to 1/3 years. Continued rarer seizures. Improved alertness, wellbeing. No adverse effects. Conclusions: Epilepsy in 2 deletion‐positive AS cases was completely controlled, and a useful response occurred in an undeleted case. Based on these initial observations we recommend consideration of KD treatment early in the course of refractory epilepsy in AS individuals. Different molecular pathology may underlie the differential response seen in these cases. There are now 2 animal models of AS available with different molecular characteristics. Further clinical and in vivo studies of the KD in AS may clarify the role of the KD and offer insights into the pathophysiology of epilepsy in AS and into KD mechanisms. 1 A.G. Christina Bergqvist, 2 Joan I. Schall, 1 Erica L. Richard, 3 Paul R. Gallagher, and 2 Virginia A. Stallings ( 1 Division of Neurology, The Children's Hospital of Philadelphia, Philadelphia, PA ; 2 Division of Gastroenterology, Hepatology, and Nutrition, The Children's Hospital of Philadelphia, Philadelphia, PA ; and 3 Division of Biostatistics, The Children's Hospital of Philadelphia, Philadelphia, PA ) Rationale: The purpose of this study was to determine if hypoglycemia or hyperglycemia predicts clinical ketogenic diet (KD) response by evaluating first morning blood glucose during the initial 21 days after initiation of the KD in a cohort of children with intractable epilepsy. In addition, the correlation between change in weight status and blood glucose was explored. Methods: First morning fasting whole blood glucose was measured daily for the first 21 days after KD initiation. Moderate hypoglycemia was defined as a glucose < 60 mg/dL at any time during the 21 days; severe hypoglycemia as < 45 mg/dL; and hyperglycemia as > 100 mg/dL. Z‐scores for weight (WAZ) and height (HAZ) were obtained at baseline, the day of discharge, 0.5 and 1 month of full KD therapy. Associations between clinical response to the KD, responder defined as > 50% reduction of seizure frequency at 3 months, hypoglycemia, hyperglycemia, style of KD initiation (fasting or gradual) and weight status were evaluated using Fisher's exact test and longitudinal mixed effects models. Results: Forty‐five subjects ages 1 to 12 years (mean age 5.3 ± 2.7 y) participated. Growth status was suboptimal before KD initiation (WAZ −0.3 ± 1.6 and HAZ −0.4 ± 2.0). Weight status declined over 1 month of KD therapy (to −0.6 ± 1.7 for WAZ). KD responder status was not associated with low or elevated blood glucose. A fasting or gradual KD initiation was also not associated. Children who had declining weight status during KD initiation were more likely to be hypoglycemic during full KD therapy, but this was unrelated to response to KD. Conclusions: Low blood glucose during KD therapy was not necessary for response to KD with clinically significant seizure reduction. Hypoglycemia was related to declining weight status. An effective KD can be provided in a manner to minimize side‐effects and maximize efficacy. Low and elevated blood glucose by ketogenic diet responder status Number Subjects All Responders Non‐responders p‐value Total Number of Subjects 45 30 15 Ever any glucose > 100 13 10 3 0.285 Ever any glucose < 60 28 19 9 0.539 Ever any glucose < 45 4 2 2 0.407 (Supported by: This study was supported in part by RRK‐23 16074 and General Clinical Research Center (MO1RR00240), the Nutrition Center of the Children's Hospital of Philadelphia, and the Catharine Brown Foundation.) 1 Thomas A. Buzzerd ( 1 Clinical, Associates In Behavioral Health, Tunkhannock, PA ) Rationale: Roughly 500,000 children and adolescents across the U.S. suffer from seizure activity. Currently, medication management programs are the primary intervention with only varying success at controlling seizure activity. Increased attention and study needs to be focused on non‐pharmacological intervention strategies like stem cell therapy. This case study evaluated the effectiveness of human, umbilical cord stem cell (hUCSC) therapy over a ten month period of time in an eight year old male with cerebral palsy and epilepsy. Methods: Direct observations and indirect measures (such as standardized assessment tools) were conducted and administered; interviews with support persons were also conducted. Seizure activity, motor skills, communication skills and academic performance baseline were evaluated for change following two stem cell injections (1.8–2.0 and 2.8–3.0 million stem cells) administered six months apart. A single subject research design was used to collect baseline and post‐injection data. Results: A t‐test analysis of data groups presented statistical significance (at p = .05) in seizure activity. The post injection data represented the subject as exhibiting approximately a 90% reduction in both seizure frequency and duration during the period of this study. Conclusions: Although a single case study does not serve as “proof” that stem cell therapy is an effective treatment. Studies into this new area of science must critically evaluate all outcomes that may be discovered. (Supported by Private Funder.) 1 Maria G. Dahlin, 2 Lena Hjelte, and 1 Per E. Amark ( 1 Neuropediatrics, Astrid Lindgrens Childrens Hospital, Karolinska Hospital, Stockholm, Sweden ; and 2 Cystic Fibrosis, Astrid Lindgrens Childrens Hospital, Karolinska Hospital, Stockholm, Sweden ) Rationale: The ketogenic diet (KD) is a high‐fat, low‐protein, very low‐carbohydrate diet used in the treatment of medically refractory epilepsy in children. Its anticonvulsant mechanism is still unclear but alterations of fatty acid (FA) levels has been proposed as a mechanism of action. We examined the influence of the KD on FA levels and seizure control. Methods: The levels of seventeen FAs in plasma phospholipids were determined before and 1, 6, and 12 months after initiation of the KD in 25 children (mean age 6.3 years) with severe epilepsy. The majority were on a KD ratio of 4:1 (fats:proteins and carbohydrates). After 1 month, the KD was supplemented with polyunsaturated omega‐3 FAs by adding fish‐oil (4–8 g/day). Seizures were quantified. Results: Highly significant changes in the levels of several FAs were found in relation to the KD. After 1 month on KD significant changes were seen in the levels of 8 FAs, after 6 months on KD in 10 FAs and after 12 months on KD in 12 FAs. Changes were found in the levels of polyunsaturated, monounsaturated as well as saturated FAs. Among others, increases were seen in linoleic acid (18:2 n6) and EPA (20:5 n3) and decreases in arachidonic acid (20:4 n6). The ratio of omega‐6 to omega‐3 FAs decreased from 7.0 before diet to 4.9 at 12 months after diet start. However, no correlation of FA levels to seizure response was found. Conclusions: The ketogenic diet induced prominent changes in several major plasma fatty acid levels but a correlation to seizure control was not found. A decreased ratio of omega‐6 to omega‐3 fatty acids was observed which may be of benefit to reduce risks of cardiovascular disease. Further studies are needed on how different diet compositions, in terms of type and amount of various fatty acids, affect blood fatty acid profile, in order to define an optimal diet regime for maximal seizure control as well as minimal cardiovascular risks. (Supported by the Margaretahemmet Association Sunnerdahl Handicap Fund.) 1 Eric Dinnerstein, 1 Karen R. Richardson, and 1 Barbara C. Jobst ( 1 Section of Neurology, Dartmouth Hitchcock Medical Center, Lebanon, NH) Methods of terminating seizures Patient Epileptic focus Praying Talking Swallowing Listening Writing Rubbing 1 Right + + + 2 Right + 3 Right + 4 Right + 5 Right + 6 Left + + + 7 Left + + Rationale: Patients often report that they can terminate seizures with certain cognitive tasks. Case reports decribe the ability of cognitive tasks to suppress/interrupt seizure activity. A systematic review of such experiences has not yet been documented in the literature. Is it not known whether cognitive activity arresting seizures is hemispheric specific, and whether it correlates with side of seizure onset. Methods: 20 adult patients with mesial temporal epilepsy were retrospectivley interviewed by telephone or in writing. All patients had unilateral temporal lobe seizure onset and underwent temporal lobectomy. Patients were asked to describe, whether they developed a cognitive method/technique to terminate seizures, when they were preceded by an aura. In addition a list of cognitive activities specific to each hemisphere was presented to all patients. Results: Seven patients reported cognitive acitivities that terminated seizures after an aura occured.Five patients had right mesial temporal seizure onset, two had left mesial temporal seizure onset. Patients with right mesial temporal onset seemed to control their seizures with praying, while patients with a left mesial temporal onset controlled their seizures via rubbing or tapping the contralateral affected limb (see table). The small number of patients precluded statistical analysis. Conclusions: Despite being a limited study, this survey provides an interesting observation. Some patients with right sided mesial temporal seizure onset can terminate their seizures with right hemispheric activities such as praying. Some select patients with a left mesial temporal seizure onset are able to terminate their seizures using left hemispheric tasks, such as talking, writing, tapping and rubbing the right side. Neurocognitive tasks represented in close proximity to the seizure focus could influence electrical seizure activity. Neurocognitive therapy could evolve as treatment option for epilepsy. These findings need to be confirmed in a larger prospective study with objective validation. 2 Celina C. Martinez, 1 Adam L. Hartman, and 1 Eric H. Kossoff ( 1 Pediatric Epilepsy, Johns Hopkins Medical Institutions, Baltimore, MD ; and 2 School of Medicine, Johns Hopkins Medical Institutions, Baltimore, MD ) Rationale: Approximately 10% of children on the ketogenic diet become seizure‐free after its initiation. The diet typically is discontinued after 2 years, but there is no information regarding the likelihood and severity of seizure recurrence. Methods: A retrospective chart review examined children who became seizure‐free on the ketogenic diet at our institution over a 12‐year period; the indication for discontinuing the diet was freedom from seizures. All children had at least 6 months of seizure freedom and were on the diet for that period prior to discontinuation. Results: Sixty‐six children discontinued the diet after a median of 2.1 years due to seizure freedom; 92% were medication‐free. Twelve (18%) had recurrence of their seizures at a median of 3.0 years after the diet was stopped. Seven (58%) became seizure‐free a second time, 2 after reinstitution of the diet. Four patients restarted the diet but remain intractable to date. The presence of recent electrographic epileptiform activity and an abnormal MRI each significantly increased the likelihood of recurrence (45% and 38%, respectively); age, seizure type, seizure frequency, medication freedom, and time to initial diet response did not. Three patients with recurrent seizures had tuberous sclerosis complex. Conclusions: The risk of recurrent seizures subsequent to discontinuing the diet after achieving seizure freedom was approximately 1 in 5, and more than half became seizure‐free again, usually with anticonvulsants alone. The risk of recurrent intractability was three times less; however the diet, when restarted in these patients, was not as successful. Obtaining a recent EEG prior to discontinuing the diet in this situation may be advisable in counseling families. 1 Hrisimir Kostov, 1 Pål G. Larsson, and 2 Geir K. Røste ( 1 Department of Neurodiagnostics, The National Center for Epilepsy, Rikshospitalet University Hospital, Sandvika, Norway ; and 2 Department of Neurosurgery, Rikshospitalet University Hospital, Oslo, Norway ) Rationale: Vagus nerve stimulation (VNS) is accepted as an effective and safe treatment for refractory partial epilepsy. Its value for treating patients with refractory idiopathic generalized epilepsy (IGE) is not well documented. Methods: All patients with medically refractory IGE, with EEG/video documented seizures, implanted with VNS in our epilepsy surgery program between 1995 and 2006 were included. A total of 12 patients (2 M, 10 F) from 11 to 48 years (mean age: 30.7 ± 14.4 years) were identified. Mean follow‐up period was 18 months (±14; range, 3 – 48). Patients were treated with an average of 2.25 AEDs/patient at implantation time. Wilcoxon's paired signed rank test was used for statistical analysis. Results: There was 62% seizure reduction for generalized tonic‐clonic seizures (p = 0.0020), 58% for absence seizures (p = 0.0003) and 40% for myoclonic seizures (p = 0.0156). Eight patients were considered responders (≥ 50% seizure reduction); two of these patients became seizure‐free. Two patients had < 50% seizure reduction, and two patients showed no change in seizure frequency. At the last follow‐up visit, the patients were receiving an average of 1.67 AEDs (−0.58 AEDs/patient; p = 0.0625). Two patients (both responders) are currently being treated with VNS therapy only. Most commonly used stimulations parameters were: pulse width 250 μs, frequency 20 Hz, 30 seconds on and 5 minutes off. Median output current at the last follow‐up was 1.5 mA. Due to lack of response in four patients, parameters were changed to rapid cycling (7 sec. on and 0.2/0.3 min off); two of these patients became then responders. An improvement in alertness (“better” or “much better”) was observed in 6 of the patients. According to 5 patients, the magnet was always or most times effective. Nine patients reported some side effects, which were mild and tended to decrease over time. Conclusions: In this patient population, VNS therapy was effective and safe for the treatment of medically refractory IGE. Rapid cycling seems to be effective for some of the non‐responders. There was an improvement in alertness in half of the patients, and a trend towards a reduction in the number of AEDs/patient. 1,2 Elizabeth A. Leleszi, 1,2 Eishi Asano, and 1,2 Gyula Acsadi ( 1 Neurology, Children's Hospital of Michigan, Detroit, MI ; and 2 Pediatrics, Children's Hospital of Michigan, Detroit, MI ) Rationale: Celiac disease can be associated with a wide spectrum of neurological manifestations including ataxia, epilepsy, dementia, neuropathy, myopathy, cerebral calcification, cerebellar atrophy and leukoencephalopathy, which are mostly related to the metabolic consequences of severe malabsorption. We describe a four year‐old child, presenting with myoclonic epilepsy described as head drops and occasional upper extremity clonus. The seizures continued to happen up to thirty times a day for several months even on multiple antiepileptic medications. An ictal EEG showed generalized high amplitude spike and slow wave activity with duration lasting 1–2 seconds. A brain MRI was normal. As part of an allergy workup, high antigluten antibodies were found and the HLA typing (HLA‐DQB1*0302, 0301) was also consistent with celiac disease. Unfortunately, a jejunal biopsy was not preformed. After the patient was started on a gluten‐free diet, the seizures resolved completely within three days and the EEG was normalized. The anti‐epileptic medications were gradually discontinued without any relapse for at least four more months. Conclusions: Gluten from dietary sources may evoke an immune response of the gastrointestinal cells via gliadin and tissue transglutaminase in susceptible individuals and lead to celiac disease with multiorgan manifestations. However, the clinical and immunological consequences of asymptomatic gluten intolerance are not well established. This is the first case report describing a complete resolution of seizure activity with a gluten‐free diet alone. Our case suggests a possibility that the epilepsy was caused by a so far unknown immunological process directed to the central nervous system (e.g. antibodies to some ion channels.) Further investigations are necessary to determine a relationship and pathomechanism between intractable epilepsy and gluten intolerance. 1 Jullie W. Pan, 2 C.J. Segal‐Issacson, and 2 Brandy Cowell ( 1 Neurosurgery, Yale University School of Medicine, New Haven, CT ; and 2 Epidemiology and Social Medicine, Albert Einstein College of Medicine, Bronx, NY ) Rationale: A number of reports have found that low carbohydrate (e.g., Atkins) or low glycemic index diets may help in reducing seizure frequency. Given the relatively lower ketosis achieved in such diets as compared to the ketogenic diet, it is possible that the effect of seizure control is not via direct ketone (or ketone derivatives) levels in the brain, but through a physiological mechanism. We examined for such a mechanism by studying the effects of 4 weeks of an Atkins diet on cerebral energetics in n = 7 healthy overweight control adult volunteers. Methods: We used high field 31P whole brain MRSI to study healthy overweight (n = 7, mean age 34.7 ± 10.0years, 4F) volunteers, examining them prior to initiating a low carbohydrate diet (Atkins Phase 1) and after 4 weeks of sustaining this diet. For optimum diet control, meals from a licensed caterer were provided to the subjects, and the macronutrient distribution of the diet was 65% fat: 30% protein: 5% carbohydrate. The mean body mass index at baseline was 34.0 ± 2.7, and after 4 weeks of the diet was 31.2 ± 2.2. Weight loss was achieved by all volunteers, averaging 6.7 ± 3.2 kg. 31P MRSI was performed at 4Tesla. Including scout imaging and calibrations, the duration of the 31P study was ∼75min. The 31P acquisition has an effective sampling radius of 1.4cm; previous studies have shown the reproducibility to be 10%. Quantification was performed accounting for tissue volume, coil loading and relaxation. Plasma samples were frozen at –80C until analysis; b‐hydroxybutyrate concentrations were spectrophotometrically assayed after each study using a COBAS‐FARA analyzer. Results: A mild ketosis was achieved in all subjects, with plasma BHB rising from 0.24 ± 0.05 to 1.64 ± 0.93. Example data (position and spectra) are shown in Fig. 1, showing PCr, ATP and inorganic phosphate Pi. Data and loci of spectral analysis are shown in the Table. From all volunteers, the ATP and PCr concentrations in the hippocampus and thalamus rose significantly between D0 to D28, while there was no significant change in occipital cortical tissue. Conclusions: Cerebral energetics changed with use of a low carbohydrate diet, with increased ATP and PCr concentrations in the thalamus and hippocampus. The data are consistent with a view that the change in fuel type and availability is characterized by an alteration in energetics particularly in subcortical brain structures. The increased availability of high energy phosphates in key limbic structures may be an important factor in how a low carbohydrate diet affects seizure control. (figure 1) (Supported by NIH R01NS40550.) 1 Roberto Poma, 2 John Ives, 3 Alexander Rotenberg, and 3 Alvaro Pascual‐Leone ( 1 Clinical Studies, Ontario Veterinary College—University of Guelph, Guelph, ON, Canada ; 2 Clinical Neuroscience, University of Western Ontario, London, ON, Canada ; 3 Division of Epilepsy, Harvard Medical School, Boston, MA ; and 4 Department of Neurology, Beath Israel Deaconess Medical Center, Boston, MA ) Rationale: The purpose of this study was to investigate the role of repetitive transcranial magnetic stimulation (rTMS) as an alternative treatment for 3 dogs suffering from refractory idiopathic epilepsy. Methods: Three client‐owned epileptic dogs refractory to AED treatment were studied. All dogs were diagnosed with idiopathic epilepsy based on the history of generalized seizures, normal neurological examination, unremarkable brain MRI and EEG features. Affected dogs were followed for at least 60 days before and after rTMS. A calendar of the seizures was kept to evaluate the interval between seizures before and after rTMS. Each dog received multiple AED treatment including phenobarbital, potassium bromide and levetiracetam. No dosage change was made in the two months prior to and following the study. A total amount of five treatments were delivered to each dog over five consecutive days (1 treatment per day). The technique of stimulation consisted in a 9 cm focal point coil located over the skull in proximity of the vertex. Different rates of stimulation were used between dog 1 and dogs 2 and 3. In dog 1, each treatment consisted in low‐rate rTMS delivered in 20 consecutive trains of 90 pulses each at 1 Hz frequency and 60% of the maximal coil output. The inter‐train interval (ITI) was 120 seconds. Dog 2 and 3 received a treatment protocol consisting of two consecutive sessions at different rates. The first session consisted of priming the cerebral cortex with high‐rate rTMS characterized by 20 trains (40 seconds ITI) of 60 pulses each at 6 Hz frequency and 55% of the maximal coil output. The second session of stimulation consisted in low‐rate rTMS characterized by 10 trains (60 seconds ITI) of 60 pulses each at 1 Hz frequency and 60% of the maximal coil output. The longest seizure‐free interval during the pre‐rTMS period was compared with the seizure‐free interval immediately post‐rTMS. Results: Dog 1 and 3 had a short‐term improvement of the seizure interval in the immediate post‐rTMS period (27 days and 14 days respectively) as compared to the longest seizure‐free interval of the pre‐rTMS period (20 days and 12 days respectively). Dog 2 had an equal seizure‐free interval between the immediate post‐rTMS period (19 days) and the longest seizure‐free interval of the pre‐rTMS period (19 days). Conclusions: Regardless of the technique of stimulation used in this study (low‐rate rTMS versus high‐rTMS followed by low‐rate rTMS), an equal or prolonged seizure‐free interval in the immediate post‐rTMS period was observed in all dogs suggesting the possible short‐term beneficial effect of rTMS in canine epilepsy. (Supported by Ontario Veterinary College Pet Trust Foundation.) 1 Steven M. Rothman, 1 Xiao‐Feng Yang, and 2 Brigitte F. Schmidt ( 1 Neurology and Hope Center for Neurological Disorders, Washington University School of Medicine, St. Louis, MO ; and 2 Molecular Biosensor and Imaging Center, Carnegie Mellon University, Pittsburgh, PA ) Rationale: Caged compounds are inert prodrugs that are capable of releasing small bioactive molecules when excited by light at the appropriate wavelength. They were initially developed to facilitate rapid, step increases in the local concentration of neurotransmitters and are typically activated with a very brief pulse from a mercury bulb or laser. We wondered whether we could use low levels of light from an ultraviolet diode (UV LED) to release GABA from a new caged analog and modulate seizure‐like activity in cultured neurons. Methods: We used cultured hippocampal neurons after about 2 weeks in vitro and recorded in whole cell bridge or voltage clamp mode. GABA was applied by whole bath perfusion or by uncaging 4‐(((2H‐1‐benzopyran‐2‐one‐7‐amino‐4‐methoxy)carbonyl)amino)butanoic acid (BC204). A UV LED (Nicchia 365 nm/100 mW) was activated by a custom fabricated driver for 4 seconds in the uncaging experiments. Results: We first verified that our UV LED was capable of generating sufficiently high GABA concentrations to activate GABAA receptors in our cultures. We found that driver currents between 25 and 250 mA produced readily detectable currents in the presence of BC204 (30 μM). These currents are well below the maximum capacity of the UV LED (700 mA). The current elicited by 150 mA corresponded to a 10 μM GABA current, when normalized to a GABA dose response curve. The currents produced by uncaging BC204 were reduced by 89% in the presence of picrotoxin (100 μM). BC204 itself at 30 μM is a very weak GABAA blocker, reducing 3 μM GABA currents by about 5%. When cultures were exposed to medium lacking magnesium, they generated increased synaptic activity, paroxysmal depolarization shifts, and often “seizure like” discharges that were abolished by perfusion with GABA as low as 3 μM. In the presence of BC204 (30 μM), a 4 second illumination with 100–200 mA also eliminated spontaneous activity. The figure below shows the effect of a 4 second light pulse on bursting. (figure 1)Illumination in the absence of BC204 produced no detectable current. Conclusions: These results suggest that it may be possible to utilize a compact, low power UV LED in combination with locally applied caged GABA to activate tonic GABAA and GABAB receptors and modulate paroxysmal activity. These receptors can be activated by GABA concentrations in the low micromolar range, which it may be possible to achieve in vivo. (Supported by Alafi Family Foundation, the NIH (R01 NS 42936 to SMR), and NSF (MCB‐8920118 to BFS).) 1 Christina A.G. Bergqvist, 2 Joan I. Schall, and 2 Virginia A. Stallings ( 1 Pediatrics/Neurology, The Children's Hospital of Philadelphia, University of Pennsylvania, Philadelphia, PA ; and 2 Pediatrics/Division of GI, Hepatology and Nutrition, The Children's Hospital of Philadelphia, Philadelphia, PA ) Rationale: The aims of this study were to describe vitamin D status including dietary intake in a contemporary cohort of children with intractable epilepsy prescribed newer anti‐epileptic drugs (AED), and to determine the effects of the ketogenic diet (KD) on serum vitamin D status over 15 months. Methods: Prepubertal children 1–14 years with intractable epilepsy were eligible. Serum vitamin D (25‐OHD and 1,25‐OHD) and parathyroid hormone (PTH) were obtained prior to initiation of KD therapy and every three months thereafter. Three‐day weighed dietary records were obtained at baseline and one month on KD therapy, and included vitamin and mineral supplementation. Results: 45 children with intractable epilepsy (age 5.1 ± 2.7 years) enrolled in the study. At baseline 4% had deficient and 51% had insufficient serum 25‐OHD levels. Vitamin D intake was inadequate; 46% consuming less than the recommended intake. Adequate vitamin D intake, fewer AED and generalized seizures were associated with higher serum 25‐OHD levels (p < 0.01). After 3 months on the KD, vitamin D status improved, 25‐OHD levels increased (p < 0.001), and PTH declined (p < 0.001). Over the next 12 months of KD exposure, the 25‐OHD levels steadily declined (p < 0.001), while 1,25‐OHD and PTH were not significantly changed. Conclusions: Children with intractable epilepsy treated with newer AED had poor vitamin D intake and blood levels. KD therapy with vitamin D supplementation improved vitamin D status; however, over the next 12 months of KD therapy a decline in 25‐OHD was observed. (figure 1) (Supported by: Supported in part by RRK‐23 16074 and General Clinical Research Center (MO1RR00240), the Nutrition Center of the Children's Hospital of Philadelphia, and the Catharine Brown Foundation.) 1 Christie Snively, 1 Paul R. Carney, 1 Zhao Liu, 1 Debbie Ringdahl, 1 Kelly Winger, 1 Crystal Jackson, 2 Thameshwarie Singh, 2 Lauren E. Little, and 2 Peggy R. Borum ( 1 Pediatric Neurology, University of Florida, Gainesville, FL ; and 2 Food Science and Human Nutrition, University of Florida, Gainesville, FL ) Rationale: Ketogenic Therapy (KT) is a treatment for intractable epilepsy that is designed to mimic metabolism during starvation. Although KT has been practiced for many decades, the guidelines for practice have often been communicated orally and based on personal experience. Subjective measurements of seizure increase or decrease and of antiepileptic drug (AED) requirements are usually used. The goal of our project was to develop an evidence based KT program for treatment of epilepsy. Methods: A review of the literature on KT over the past decade was conducted and an evidence grading system was used to document current practice. A retrospective database was developed to evaluate KT provided at our center for the past decade. There are currently 184 current and non‐current patients in the retrospective database of which 95 are males and 89 are females. The average age at initiation was 7.1 ± 4.9 years old and the average time on diet from non‐current patients (n = 133) was 17.7 ± 18.9 months. Patient tracking forms for pre‐initiation, initiation, post‐initiation, and termination of KT were created. Systems for quantifying and tracking seizures and medications were created to evaluate patients' progress on KT. This includes Clinic Seizure Score (CSS), Daily Diary Seizure Score (DDSS), and a Medication Score (MS). The CSS is dependant upon the caregiver's recollection at clinic of seizure activity since the last clinic. The DDSS is based on daily seizure records of the caregiver and used to track the total amount of seizure activity between clinic visits. The MS allows us to evaluate both the total number and dosage of all the AEDs that a patient is taking based on their age and weight. Results: The evidence based guidelines have been created by the University of Florida Pediatric Comprehensive Epilepsy Center. For initiation, pre‐therapy records are collected and the patient comes in after an overnight fast and starts on a 1.5:1 ratio at full calories which is then increased by 0.5 every other meal for oral feeders and every day for tube feeders. During post‐initiation, parents document daily ketones, specific gravity, and seizure activity as well as dietary intake for 24 hours per week. Patients are monitored using the CSS, DDSS, MS as well as the following lab parameters: comprehensive metabolic panel, complete blood count, lipid profile, serum beta‐hydroxybutyrate, lactate, and acylcarnitine profile. Data from the patient tracking forms are entered in a prospective database on a weekly basis. Conclusions: The data obtained from the databases will allow us to evaluate and expand the evidence based KG guidelines for optimal administration and monitoring of patients throughout KT. 1 Jessica A. Yankura, 1 Amy French, 1 Michael A. Rogawski, 2 Adam Hartman, and 1 Maciej Gasior ( 1 Epilepsy Research Section, NINDS, National Institutes of Health, Bethesda, MD ; and 2 Pediatric Epilepsy Center, John Hopkins Hospital, Baltimore, MD ) Rationale: Anticonvulsant effects of the ketogenic diet seem dependent on high fat and low carbohydrate availability that results in the development of ketosis along with a modest reduction in serum glucose levels. However, it remains unclear if low glucose levels alone can decrease neuronal excitability and have anticonvulsant effects. Glucose depletion attenuates seizure activity in in vitro models whereas both proconvulsant and anticonvulsant effects have been reported in in vivo models (1,2). In the present study we sought to clarify the influence of glucose in regulating seizure susceptibility in vivo by simulating glucose deprivation with 2‐deoxyglucose (2‐DG), a structural analog of glucose that does not undergo glycolysis. Methods: Anticonvulsant effects of 2‐DG were assessed in amygdala‐kindled rats and in Swiss mice in the intravenous pentylenetetrazol (PTZ) and kainic acid models, and in the electroconvulsive threshold (ECT) and 6 Hz models. 2‐DG (125 – 500 mg/kg) was administered intraperitoneally 30 min before seizure testing. Results: In amygdala‐kindled rats, 2‐DG had no effect on afterdischarge threshold, afterdischarge duration, behavioral seizure score and behavioral seizure duration. In mice, administration of 2‐DG had no effect on seizure threshold in the PTZ test and was slightly proconvulsant in the kainic acid model. It had no effect on ECT. Only in the 6 Hz seizure test did 2‐DG cause a small, yet significant, dose‐dependent elevation in seizure threshold. Conclusions: In contrast to a prior study using olfactory bulb and perforant path kindling (2), we did not find 2‐DG to have potent anticonvulsant efficacy in the various in vivo seizure models we studied. Our findings suggest that low glucose availability plays a limited role in the seizure protection conferred by the ketogenic diet. 1 Kirchner, A., Veliskova, J. & Velisek, L. Differential effects of low glucose concentrations on seizures and epileptiform activity in vivo and in vitro. Eur. J. Neurosci. 23:1512–1522, 2006. 2 Stafstrom, C., Kriegler, S., Valley, M., Ockuly, J., Roopra, A. & Sutula, T. 2‐Deoxyglucose exerts anticonvulsant and antiepileptic actions in experimental epilepsy models. Epilepsia 46 (Suppl. 8), 268–269, 2005. (Supported by NINDS, NIH.) 1 Mahmoud M. Abu‐Ata, 1 Bassel Abou‐Khalil, 1 Kevin Haas, 1 Patrick Lavin, and 1 Peter Konrad ( 1 Neurology, Vanderbilt University Medical Center, Nashville, TN ) Rationale: To evaluate the visual field defects in patients who underwent SAH (selective amygdalohippocampectomy) for intractable epilepsy using automated static visual field testing. Background: Visual field defects, specifically contralateral upper quadrant homonymous quadrantanopsia, are well known complications after conventional temporal lobectomy, however, few studies have evaluated visual field defects after SAH using automated static visual field perimetry. Methods: We evaluated 6 adult patients with history of intractable temporal lobe epilepsy who had SAH within the last 3 years using neuro‐ophthalmologic evaluation and automated Humphrey visual field analysis post‐operatively. All patients were diagnosed with mesial temporal lobe sclerosis and had their surgery at our institution using the same image guided surgical technique, in which an incision is made in the middle temporal gyrus to locate the temporal horn, then the amygdala and 3 cm of hippocampus are removed using the microscope. After obtaining written consents, all patients were questioned about visual complaints and were examined for visual fields by confrontation using finger counting, visual acuity ocular motility, pupil size and reactivity and fundoscopy. Humphrey automated visual field analysis was then performed using the 24–2 program that tests 54 points that straddle the horizontal and vertical meridians. The visual field for each eye was divided into 4 quadrants. The depth of any identified defect was determined by calculating the algebraic difference between measured value and expected value and was displayed on a grid corresponding to the location of the test points creating a map that showed the location and extent of the visual field defects. The pattern deviation was then averaged in the quadrant contralateral to surgery, separately for the ipsilateral and contralateral eye. Deviations less than −5 were displayed (Figure). Results: 4 patients had left and 2 had right SAH. None had identifiable visual field defects by history or confrontation visual field exam. All had identifiable visual field defects in the contralateral superior quadrant by perimetry. 4 had incomplete contralateral superior quadrantonopsia and 2 had only small peripheral defects in the contralateral superior quadrants. The defect affected mostly the medial sector near the vertical meridian and was steeper in the ipsilateral eye. Conclusions: Visual field defects are an expected complication of SAH. The expected defects are predominantly incomplete contralateral superior quadrantonopsias, similiar to those previously reported with conventional anterior temporal lobectomy. (figure 1) 1 Indrani E. Acosta, 1 William O. Tatum, 1 Fernando L. Vale, and 1 Selim R. Benbadis ( 1 Department of Neurology & NeuroSurgery, Tampa General Hospital, Tampa, FL ) Rationale: Temporal Lobectomy is the most common performed epilepsy surgery. Specialized surgical centers infrequently perform temporal lobectomies on patients 60 years of age or older. We reviewed our experience on patients in this age group who underwent a temporal lobectomy at our Comprehensive Epilepsy Program. Methods: We reviewed our surgery database and identified patients who underwent a temporal lobectomy for intractable seizures at the age of 60 years or older. All patients underwent a comprehensive presurgical evaluation, including EEG‐video monitoring, MRI, SPECT or PET, neuropsychologicla testing, and Wada test. Post surgical outcome was measured using Engel classification. Follow up ranged from 5 months up to 20 months. Results: Over a 5 year period, 7 patients 60 years or older underwent a temporal lobectomy at our Epilepsy Center. This represents less than 5% of resective epilepsy surgeries performed. Patient's ages ranged from 60 to 76 years of age (mean 63.4 years). Duration of epilepsy ranged from 1 year to 53 years (mean 34.8 years). Etiology was mesiotemporal sclerosis in 4 patients and unclear in 3. Invasive EEG was performed in 1 patient. Post‐surgical outcome was class I in 4 patients, class II in 2 patient. No patients fell into Class III or Class IV. One patient developed psychogenic seizures after surgery. Conclusions: Temporal lobectomies are rarely performed after the age of 60, but should be considered and seem so have a comparable outcome to that of younger patients. 1 Katherine Bayless, 1 Jana E. Jones, 2 Lincoln Ramirez, and 1 Bruce P. Hermann ( 1 Neurology, University of Wisconsin, Madison, WI ; and 2 Neurological Surgery, University of Wisconsin, Madison, WI ) Rationale: This study examined the long‐term (10 years+) psychosocial outcomes of a consecutive series of patients who underwent anterior temporal lobectomy (ATL) compared with a group of medically managed patients. These same individuals were initially assessed in 1999 (mean 5 years follow‐up) (Jones et al., 2002). This second follow‐up interview occurred 6 years later in 2005. The primary focus of this study was to examine the changes maintained in actual life performance after surgery in the following areas: employment, independent living, driving and financial independence. Self‐reported quality of life was also assessed. Additionally, during this second follow‐up interview questions pertaining to the presence or absence of mood and anxiety disorders were included. Methods: The sample consisted of 57 adults who underwent ATL and 18 controls that were evaluated for surgery during the same time period as those who had surgery. However, the medically managed group did not have surgery and continued to receive medical management. Participants were interviewed 8–15 years after surgery or presurgical evaluation. A structured clinical interview was conducted to obtain information regarding seizure frequency and psychosocial status. Seizure frequency was evaluated for the number seizures in the year prior to the interview and number of seizures since the initial follow‐up interview in 1999. Results: As indicated in the initial follow‐up evaluation in 1999, significant differences in psychosocial outcomes were maintained between the surgical and medical management groups (p < .001). Favorable psychosocial outcomes were significantly more common in surgical participants (50.9%) than medically managed patients (11.1%). Individuals in the surgery group were more likely to be employed (70.2% v. 44.4%), driving (77.2% v.33.3%), living independently (80.7% v. 44.4%) and financially independent (80.7% v. 44.4%). Interestingly, there were no significant differences between the two groups in terms of self reported quality of life. In terms of seizure frequency, 71.9% of the surgery group was seizure free in the past year compared to 27.8% in the medically managed group. Since 1999, 50.4% of participants in the surgery group have remained seizure free. Among the two groups 15.8% of the surgery participants reported a diagnosis of a mood or anxiety disorder in the past year compared to 17.6% in the medically managed group. There were no significant differences between the two groups. Conclusions: Surgery appears to be associated with the long‐term maintenance of favorable psychosocial outcomes at a very long‐term follow‐up of 8–15 years after surgery. Rates of employment, driving, independent living and financial independence appear to be positively influenced by surgery. Seizure freedom and reduced seizure frequency appear to be a significant long term outcome of surgery. 1 Christine M. Bower, 1 Eric M. Cheng, 2 Susan S. Spencer, 1 Stefanie Vassar, and 1 Barbara G. Vickrey ( 1 Department of Neurology, University of California Los Angeles, Los Angeles, CA ; 2 Department of Neurology, Greater Los Angeles VA Medical Center, Los Angeles, CA ; and 3 Department of Neurology, Yale University School of Medicine, New Haven, CT ) Rationale: Although the clinical goal of resective epilepsy surgery is seizure freedom, patients may have a wide set of expectations for this invasive procedure. This study's goals were to assess the nature, range, and frequency of preoperative expectations for resective epilepsy surgery, and to explore whether expectations vary across patient sociodemographic and clinical characteristics. Methods: 396 adults and adolescents with refractory epilepsy were enrolled in a seven‐center cohort study and underwent resective surgery. During baseline interviews, patients responded to open‐ended questions about expectations for surgery. Investigators reviewed text responses and identified a preliminary set of expectation themes. Each text response was placed on a card, then “sorted” into piles representing distinct themes (expectations) by consensus of three investigators. Differences in the frequencies for which each expectation theme was identified were explored across gender, ethnicity, education, whether temporal or extratemporal lobe epilepsy, and side of surgery, using chi‐square. Results: Among 391 respondents, nine unique expectations themes were identified by >15% of the sample: driving (n = 241, 61.6%); job/school (n = 166, 42.5%); independence (n = 115, 29.4%); seizure cessation (n = 100, 25.6%); social functioning (n = 90, 23.0%); quality of life (n = 82, 21%); medication discontinuance (n = 78, 20.0%); physical activities (n = 71, 18.2%); and cognition (n = 67, 17.1%). Non‐whites and whites differed on 3 of 9 expectation themes: non‐whites (n = 72, 18% of sample) endorsed job/school and cognition more frequently and driving less frequently than did whites (p < 0.03). Respondents with a higher level of education (n = 95, 24% of sample with bachelor's degree or higher) endorsed cognition more frequently than those with less education (p < 0.003). Subjects with a right‐sided resection (n = 190, 49% of sample) endorsed social functioning more frequently than those with a left‐sided resection (p < 0.05). There were no differences in expectations by gender or temporal versus extratemporal lobe epilepsy. Conclusions: Patients awaiting resective epilepsy surgery have definable expectations for surgery, the most predominant of which are driving and employment/educational attainment. While the majority of these expectations were similar across a range of patient characteristics, there were differences on one‐third of the most frequently reported expectations by ethnicity. Future studies are needed to explore these potential sociodemographic differences, and to understand how expectations may aide in the clinical decision‐making process of resective epilepsy surgery. (Supported by Grant #RO1 NS 32375 from the National Institute of Neurological Disorders and Stroke.) 1 Mar Carreño, 1 Juan Luis Becerra, 1 Joaquin Castillo, 1 Antonio Donaire, and 1 Iratxe Maestro ( 1 Neurology, Hospital Clinic, Barcelona, Barcelona, Spain ) Rationale: Little is known about the long term outcome of patients with medically refractory epilepsy who undergo presurgical evaluation but are not considered good surgical candidates or reject surgery. A previous study with a small number of patients found out that up to 20% of those patients may eventually become seizure free (Selwa et al, 2003). This information is essential to counsel patients attended in a surgical programme. Methods: We performed a retrospective chart review and telephone survey on all patients who had undergone presurgical evaluation at the Epilepsy Monitoring Unit (EMU) of Hospital Clinic (Barcelona, Spain) from 1995 until 2004, but did not undergo epilepsy surgery afterwards, either because they were not considered to be good surgical candidates or because the patient declined surgery. We inquired about current seizure frequency, antiepileptic drug treatment, overall impression of change, quality of life, marital and laboral situation. Results: From a total of 167 telephone calls, we finally contacted 80 patients who agreed to answer the survey. Of them, only 7 patients (9.3%) were currently seizure free, with a mean seizure free period of 3 years (2–6). Seizure freedom was associated with use of new AEDs in 2/7 patients. Five patients (6.2%) had died, with death being possibly related to seizures (SUDEP or CRA during prolonged seizures) in all of them. Among those patients who were not seizure free, the vast majority (89%) reported that their seizure frequency was either the same or lower than at the time of presurgical evaluation. No significant differences were found between seizure free and not seizure free patients respect to the type of epilepsy, etiology, presence of a lesion on the MRI and reason not to have surgery. Overall, 54% of all patients interviewed reported feeling better and more satisfied with their lives than when they were admitted to the EMU. Although most patients did not report changes in their marital or laboral status, a higher proportion of patients in the seizure free group (71%) were actively working, compared to 32% of patients in the group with continued seizures. Conclusions: A small percentage (9.3%) of patients with refractory epilepsy who undergo presurgical evaluation but not epilepsy surgery may eventually become seizure free on drugs. Up to 6% of patients may die as a consequence of their epilepsy. In patients who do not become seizure free, seizure frequency will be, in general, similar or lower with time, and up to 50% of patients will report feeling better and more satisfied with their lives than at the moment of presurgical evaluation. This information may be used to counsel patients attending epilepsy surgery programmes. 1 Shiro Chitoku, 1 Yshihiro Sumi, 2 Masahiro Mizobuchi, 1 Toshiaki Osato, 1 Takehiko Sasaki, 1 Jouji Nakagawara, and 1 Hirohiko Nakamura ( 1 Department of Neurosurgery, Nakamura Memorial Hospital, Sapporo, Hokkaido, Japan ; and 2 Department of Neurology, Nakamura Memorial Hospital, Sapporo, Hokkaido, Japan ) Rationale: To analyze characteristics and risk factors of seizures following ruptured aneurysmal subarachnoid hemorrhage (SAH). Methods: We studied a total of 544 patients with acute SAH treated with clipping (444), detachable coil embolization (37), and conservative treatment (67) at Nakamura memorial hospital between 1998 and 2004. The age ranged from 26 to 92 years old (mean 60 years old). We characterized seizures following SAH and compared risk factors of seizures between seizure group and non‐seizure group. Results: Sixty‐seven (12%) patients had seizures (seizure group) 58 (87%) in clipping, 7 (10%) in embolism, and 2(3%) in conservative treatment. Other 477 (88%) patients had no seizure history for 7 years follow‐up (non‐seizure group). Thirty (44%) of 67 patients presented seizure within 24 hours following SAH (acute onset), and 37 patients (56%) after 24 hours (late onset). Sudden generalized convulsions occurred in 27 (90%) patients in the acute onset group. Focal seizures occurred in 31 (84%) patients in the late onset group. Two (7%) patients had following epileptic seizure disorders in the acute onset group and 11 (30%) in the late onset group. Thirty‐one (46%) of 67 patients in seizure group resulted in good prognosis (modified Rankin Scale0–2) and 320 (65%) of 477 patients in non‐seizure group (p < 0.05) Seizure group showed more than three times higher rate of symptomatic hydrocephalus, 34(51%):76(16%); infarction due to vasospasm, 18(27%):41(9%); and other co‐morbidities, 34(51%):48(10%); compared with those of non‐seizure group (p < 0.05). Conclusions: Seizure group revealed worse prognosis and more correlated with hydrocephalus, vasospasms and co‐morbidities than non‐seizure group. Seizures following SAH characterized generalized convulsions at acute onset, and focal seizures at late onset. The late onset seizure group tended to develop epileptic seizure disorders compared with acute onset seizure group. 1,2,3 Luciana L. D´Alessio, 3 Juan José J.J. López‐Costa, 1 Hector H. Konopka, 1 Damián D. Consalvo, 1 Eduardo E. Seoane, 2 Laura L. Guelman, 1,3 Silvia S. Kochen, and 2 Luis María L.M. Zieher ( 1 Epilepsy Center, Ramos Mejía Hospital, Buenos Aires, Capital Federal, Argentina ; 2 Pharmacology Department, Medicine, Buenos Aires University, Buenos Aires, Capital Federal, Argentina ; and 3 CONICET, CONICET, Buenos Aires, Capital Federal, Argentina ) Rationale: Nitric oxide (NO) is a short lived free radical with diverse functions as a biological messenger molecule, and has been implicated in numerous aspects of the physiology and pathology of the CNS including epilepsy. NO can acts as anticonvulsant or proconvulsant depending on the experimental model and the pharmacological parameters used. There is little information about changes in NO levels in human epileptic brains. Neurons containing nicotinamide adenine dinucleotide phosphate‐diaphorase (NADPH‐d) synthesizes NO. The aim of this study is to investigate the anatomical distribution, morphology, optical density and cell sizes of NADPH‐d neurons in the temporal cortex of patients with hippocampal sclerosis who underwent anterior temporal lobectomy for intractable temporal lobe seizures. Methods: Brain samples of 7 × 5 × 5 mm from temporal cortex were obtained from 7 patients was compared with 5 postmortem controls. Samples were fixed and incubated at 37°C in 1 mg/ml of B‐NADPH and 0,2 mg/ml of nitroblue tetrazolio. NADPH d activity was measured as distribution of formazan deposits (endproduct of the histochemical reaction). Neuronal area and neuropil staining optical density was determined. Results: Sprouting (dendritic arborization) and larger NADPH‐d positive neurons, with an increase in staining intensity were found in epileptic´s temporal cortex (p < 0.05). Conclusions: The increased area and the observed morphological changes in NADPH‐d reactive cells found in this study indicate plastic changes and suggest an up‐regulation in NO system in the neocortex. These changes could contribute to abnormal cortical excitability however, could also be involved in adaptative neuroprotective function. 1,2 Alaa Hassan, 1 Heinz Pannek, 1 Friedrich Behne, 1 Reinhard Schulz, 2 Mohammed Nayel, 2 Ahmed Issa, and 1 Alois Ebner ( 1 Preoperative Evaluation, Epilepsy Center Bethel, Bielefeld, NRW, Germany ; and 2 Neurosurgery, Cairo University, Cairo, Cairo, Egypt ) Rationale: To evaluate the long‐term seizure outcome of 15 years of extratemporal neocortical epilepsy surgery of adults in the Bethel Epilepsy Center comparing the outcome between patients operated in three consecutive 5‐year periods. Methods: Between 1991 and 2006, 221 adult patients with extratemporal drug resistant epilepsies were operated on. From 221 patients 28 patients (12.6%) had more than one operation, five patients (2.6%) received VNS. We retrospectively reviewed the outcome of all patients. Patients with Rasmussen's encephalitis and those who only had a biopsy were excluded. Data was gathered by reviewing hospital files, outpatient reports, two‐year in‐patient postoperative follow‐up results, and using mail and telephone questionnaires. Three patients were not available for review, six patients died, and in ten patients data was incomplete. Patients were divided into three groups, depending on the time of operation: group 1 (1991–1995), group 2 (1995–2000), and group 3 (2000–2005). Results: Group 1: (54 patients, 24.4%) mean age of patients at seizure onset in years 10.6 ± 8.8 (range 1–41), mean age at surgery 28.13 ± 10.3 (range 16–57), mean duration of epilepsy 17.4 ± 9.7 (range 1 to 38), mean duration of follow‐up 11.4 ± 2.09 (range 1 to 14); outcome class 1 at the 6‐month follow‐up (47.0%), after 1 year (48.5%), 2 years (45.5%), 5 years (46.6%). Group 2: (90 patients, 40.7%), mean age at seizure onset 13.4 ± 10.4 (range 0.8–52), mean age at surgery 28.02 ± 10.1 (range 16–59), mean duration of epilepsy 15.6 ± 9.2 (range 0–44), mean duration of follow‐up 7.6 ± 1.6 (range 2–10 yrs); outcome class 1 at the follow‐up after 6 months (53.4%), 1 year (50.0%), 2 years (50.0%), 5 years (53.7%). Group 3: (77 patients, 34.4%) mean age at seizure onset 16.8 ± 15.0 (range 0.4–65), mean age at surgery 31.1 ± 13.1 (range 16–69), mean duration of epilepsy 14.2 ± 13.2 (range 0–64), mean duration of follow‐up 2.6 ± 1.5 (range 0.5–5 yrs); outcome class 1 at the follow‐up after 6 months (63.5%), 1 year (57.1%), 2 years (61.9%), 5 years (57.9%). A statistically significant difference was only found in the the 2‐year outcome group (p = 0.002). Conclusions: The comparison of postoperative seizure outcome of extratemporal epilepsy surgery in adults over the last 15 years shows mixed results. A statistically significant improvement was seen only in the two‐year postoperative outcome whereas at other points in time there was a tendency for improvement. 1,3 Frederick A. Schmitt, 2 Curtis A. Given, 1 Meriem K. Bensalem‐Owen, 1 Michelle L. Mattingly, and 1 Toufic A. Fakhoury ( 1 Department of Neurology, University of Kentucky Medical Center, Lexington, KY ; 2 Department of Radiology, University of Kentucky Medical Center, Lexington, KY ; and 3 Sanders Brown Center on Aging, University of Kentucky Medical Center, Lexington, KY ) Rationale: Presurgical evaluation of language and memory using the internal carotid WADA (ICAW) procedure is a standard approach for patients with medically intractable epilepsy. However, ICAW results often provide equivocal memory data, especially in cases with a dominant hemisphere seizure focus, as a result of aphasia with amytal injection. Therefore, selective preoperative intracarotid amytal procedures (IAP) have been developed to target hippocampal functions. Several variations of the selective IAP have been reported and include selective injections of amytal into: (1) anterior choroidal artery (AChA) and (2) posterior cerebral artery (PCA). This report focuses on a clinical series of patients who had selective amytal injections through the internal carotid artery (ICA) with temporary balloon catheter occlusion of the distal ICA beyond the AchA origin. Methods: Twenty‐seven patients with intractable epilepsy were evaluated using the distal ICA occlusion approach. Three patients underwent selective procedures for both hemispheres while the remaining 24 had selective injections targeting the hippocampus in the affected hemisphere along with a standard IAP of the contralateral hemisphere. Of these patients, 7 (26%) had previously ‘failed’ standard IAP injections given generalized anesthetic effects on the dominant hemisphere. Memory and language were evaluated in all patients using a modified McGill protocol that required learning of 3 visual objects and 3 verbal items after selective amytal injection. Subsequent free and cued recall of these items along with recognition was completed after resolution of the patient's homonymous hemianopsia. Results: All patients demonstrated homonymous hemianopsia with the selective injection and 78% showed brief temporal slowing on EEG ipsilateral to the injected hemisphere. Evaluation of memory showed encoding specific effects on memory items for 78% of cases where selective left‐sided injection resulted in proportionally poorer recall of verbally encoded items. In the 3 right‐sided selective IAPs, no memory encoding effects were observed. This ‘selective’ memory effect was independent of the presence of temporal slowing on EEG. Conclusions: Selective WADA procedures targeting the hippocampus through temporary occlusion of the supra‐clinoidal ICA and injection through the AchA can be an effective tool in the presurgical evaluation of patients when evaluating information on memory preservation post surgical resection of a seizure focus. Further, selective WADA procedures generally show memory effects that are consistent with the verbal or visual presentation mode of the memory items used in IAP assessments. 1 Georges A. Ghacibeh, 1 Jeffrey Smith, 1 Keyur Patel, 1 Sandeep Konka, and 1 Stephan Eisenschenk ( 1 Department of Neurology, University of Florida, Gainesville, FL ) Rationale: Patients with mesiotemporal epilepsy have a better chance of achieving seizure freedom after epilepsy surgery than those with neocortical epilepsy. The risk of seizure recurrence depends on several factors. Previous studies revealed that interictal epileptiform activity (IEA) on post‐operative EEG did not predict the risk of seizure recurrence. However, it is not known whether there is a difference based on the type of surgery. The aim of this study was to evaluate whether the relationship between IEA on post‐operative EEG and seizure recurrence after epilepsy surgery is different in patients who underwent neocortical and mesiotemporal resections. Methods: We reviewed retrospectively the electronic medical records of patients who underwent epilepsy surgery at the Comprehensive Epilepsy Program of the University of Florida between January 1997 and December 2002. Post‐operative follow‐up clinic notes were reviewed up to the last clinic note to determine the recurrence of seizures. EEG reports of the post‐operative EEG were reviewed to assess for the presence of IEA. Epileptiform abnormalities were defined as frank spikes or sharp waves or electrographic seizures. Results: A total of 94 patients who had both adequate clinic follow‐up and post‐operative EEGs were identified. EEGs were performed on average 106 days after surgery and patients were followed in clinic on average 32 months after surgery. Seventy patients had mesiotemporal resections. Of those, 16 (23%) had IEA on their EEGs and 22 (31%) had seizure recurrence. A chi‐square test revealed no significant relationship between seizure recurrence and the presence of IEA. Twenty four patients had neocortical resections. Of those, 14 (58%) had IEA and 12 (50%) had seizure recurrence. A chi‐square test revealed a significant relationship between the presence of IEA and seizure recurrence (p = 0.013). However, 14 out of the 22 mesiotemporal and 8 out of the 12 neocortical surgery patients had seizure recurrence within 3 months after surgery, prior to the post‐operative EEG. Conclusions: This study revealed that epileptiform activity on early post‐operative EEG correlate with seizure recurrence in neocortical but not mesiotemporal surgeries. However, since most patients have seizure recurrence within three months after surgery, evaluation of the utility of earlier EEG is needed. 1 Ryder P. Gwinn, 1 John Morgan, 1 Lisa Caylor, 1 Alan Haltiner, 1 Laura Allen, and 1 Michael Doherty ( 1 Epilepsy Center, Seattle Neuroscience Institute, Seattle, WA ) Rationale: Hippocampal depth electrodes are useful in lateralizing and localizing medial temporal lobe seizures. Historically these have been placed either orthogonally or transversely using frame based or frameless stereotaxy. This method can provide quality electrical recordings of hippocampal activity, but is a blind procedure with inherent risks and limitations including bleeding, hippocampal injury, inconsistent hippocampal orientation, and incomplete sampling. We report the placement of ventricular depth electrodes under framless stereotactic and neuro‐endoscopic guidance using direct visualization to assure safe and consistent placement of hippocampal recording electrodes, with quality recordings leading to successful seizure localization. Methods: Frameless stereotactic navigation (Steath Treon Plus, Medtronic) was used to guide occipito‐parietal burr hole placement and endoscope trajectory. A neuro‐endoscope (Channel™ Neuroendoscope, Medtronic) was placed into the atria of the lateral ventricle, then guided down the temporal horn under direct visualization. A 10 contact, 5 mm spacing depth electrode (Ad‐tech, Racine WI) was passed through the endoscope with 1 – 2 contacts placed into the temporal pole. The endoscope was then removed using flouroscopy to verify stable electrode position. Electrode position was verified by co‐registration of post‐operative CT scan to pre‐operative MRI in all patients. Results: Eleven ventricular electrodes were placed in 6 patients undergoing surgery for intracranial localization of partial seizures. Successful navigation of the temporal horn was achieved in all patients, and electrodes were verified to be touching the hippocampus in all patients visually and by post‐operative imaging. No infection, neurologic injury, or post‐operative hemorrhage was seen. Quality recordings were obtained from all electrodes, and seizure onsets were documented in 7/11 electrodes, with reversal potentials seen along the hippocampus using a bipolar montage. Two patients underwent anteromedial temporal resection 6 weeks post‐implantation, with pathology showing gliosis adjacent to the electrode location. Conclusions: Electrocorticography may be performed using depth electrodes placed in the temporal horn of the lateral ventricle. Placement of temporal horn electrodes under direct neuro‐endoscopic visualization can be performed safely, and provide high quality recordings of hippocampal activity without significant injury to the structure. Temporal horn electrodes are situated in a predictable anatomical relationship to hippocampal structures for reproducible recordings from the same hippocampal subfields between patients. Neuro‐endoscopically placed ventricular electrodes may offer a safe and effective alternative to orthogonal and longitudinal hippocampal electrodes for localization of medial temporal lobe seizures. 1 Jeffery A. Hall ( 1 Neurosurgery, Montreal Neurological Hospital, Montreal, QC, Canada ) Rationale: Proponents of tissue‐sparing selective temporal resections for patients with mesial temporal sclerosis and intractable epilepsy begrudgingly acknowledge the lack of neuropsychological data demonstrating any related savings with these procedures. Seizure outcome in well‐selected patients appears to be equivalent between the more “selective” and larger traditional temporal lobectomies. One interpretation of why no obvious savings follows from selective resection is that the tissue is somehow rendered dysfunctional by the operation. Alternatively, it may be that we are not adequately probing the function of the spared temporal tissue. Methods: Ten consecutive patients who underwent transcortical (second temporal gyrus) transventricular selective amygdalohippocampectomy for mesial temporal sclerosis and intractable epilepsy were examined post‐operatively with formal visual field testing. Each patient also had post‐operative MRI to demonstrate the extent of resection and exclude complications. These patients were compared to a reference case of intractable non‐lesional temporal lobe epilepsy operated by the same surgeon. This patient underwent a more radical traditional temporal corticoamygdalohippocampectomy. Pre and post operative visual fields and MRI are available for this patient also. Results: The patient who underwent traditional temporal lobectomy demonstrated the expected homonymous superior quadrantopsia. Patients who underwent the selective procedure demonstrated some evidence of visual field loss, albeit to a lesser extent than that of the reference case. Conclusions: This finding has several important implications the most of salient of which suggests that the selective procedure results in the sparing of viable and functionally important brain tissue. Additionally, these data support the notion from recent studies of MR tractography, that the fibers of Meyer's Loop course further anterior in the temporal lobe than is generally acknowledged. (Supported by Department of Neurology and Neurosurgery McGill University.) 1 Lizbeth Hernandez‐Ronquillo, 1 Jose F. Tellez‐Zenteno, 1 Samuel Wiebe, and 1 Nathalie Jette ( 1 Department of Clinical Neurosciences, University of Calgary, Calgary, AB, Canada ) Rationale: In recent years, advances in surgical techniques have allowed patients without identifiable neuroimaging abnormalities to undergo epilepsy surgery. However, there is a notion that patients with normal MRI or histopathology may have worse surgical outcomes than those with lesional epilepsy. The objective of this study is to provide evidence‐based estimates of the surgical outcome in patients with non‐lesional epilepsy and to identify sources of results variation between published studies Methods: Medline, Index Medicus, bibliographies of reviews, original articles, and book chapters were searched to identify published articles since 1991, describing outcomes in ≥5 patients of any age who underwent resective or non‐resective epilepsy surgery without an identifiable structural lesion by imaging or histopathology. Two reviewers independently assessed study eligibility and extracted the data. Disagreements were resolved through discussion. Results: Twenty‐nine articles fulfilled eligibility criteria and described outcomes in 655 patients with non‐lesional epilepsy surgery. Non‐lesional status was defined by MRI in 56% of studies and by histopathology in 44%. Overall, 24% of patients described were non‐lesional. The median proportion of seizure‐free patients was 44% (95% CI 39–49) in non‐lesional cases vs. 61% (95% CI 58–64) in those with lesions. The corresponding proportion was 46% (95% CI 39–53) if non‐lesional status was defined by MRI, versus 42% (95% CI 35–49) if defined by histopathology. In non‐lesional patients with temporal lobe surgery, 40% (95% CI 33–46) were seizure free. We explore aspects related to the validity of these results. Conclusions: Overall, non‐lesional epilepsy surgery has inferior results compared with lesional cases (44% vs 61% respectively). The response after surgery was the same whether lesional status was defined using MRI or histopathology. Sources of heterogeneity are explored. 1 Hyeon‐Seob Byeon, 1 Seung‐Chul Hong, 1 Jong‐Su Kim, 1 Jung‐Il Lee, 1 Jong‐Hyun Kim, 1 Do‐Hyun Nam, and 2 Dae Won Seo ( 1 Neurosurgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea ; and 2 Neurology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea ) Rationale: To compare seizure outcome in the patients with cavernous angioma treated by lesionectomy alone or further resection and to analyze the factors influencing on the surgical outcome. Methods: Fifty‐nine patients (27 male) with supratentorial cavernous angioma associate with seizure were operated in our institute. Clinical, radiological and electrophysiologic data were collected retrospectively. The mean age was 32 years (ranged 2–67) and the mean follow up duration after surgery was 34.3 months (ranged 12–103). Results: Thirty‐two patients underwent lesionectomy and remaining 27 patients underwent standard temporal lobectomy with tumor removal. Seizure outcome was compared using the Engel classification scheme between two kinds of technique. At the last follow up, 33 patients were involved in Engel class I, 6 patients in class II, 4 patients in class III and 2 patients were class IV. Estimated survival in patients who had shorter seizure histories and fewer preoperative seizures did not differ between the two surgical techniques. The important factors influencing the outcome were duration of seizure, location of tumor, multiple epileptic foci on electroencephalogram and surgical technique. Conclusions: These findings indicated that shorter seizure histories and fewer preoperative seizures can be effectively treated by lesionectomy alone, whereas those with longer histories and more seizures are not effectively treated by this procedure and may require epileptic surgery. And early surgical intervention is important for good treatment outcome. 1 Hariprasad Kunhiveedu, 1 Tobias Loddenkemper, 1 Stephan U. Schüle, and 1 Richard C. Burgess ( 1 Department of Neurology, The Cleveland Clinic, Cleveland, OH ) Rationale: Prediction of ictal onset zone by subdural interictal spike pattern has rarely been analyzed. The objective of our study was to analyze the relationship between interictal spikes and ictal onset zones on subdural EEG recordings. Methods: Fifteen patients (aged 25 to 61 years) undergoing prolonged subdural intracranial EEG recordings were included. On average 92 electrodes (range 42 to 119) were implanted. Representative interictal and ictal EEG recordings were reviewed. Interictal spike populations were identified by voltage from baseline, earliest onset, frequency of the spike population, presence of an aftergoing slow wave and duration. Distribution of the interictal activity was defined as 80% fall‐off from the maximum. The electrodes with the earliest rhythmic evolution determined the ictal onset. The relationship between interictal EEG and ictal onset electrodes was analyzed. Results: Interictal spikes were seen from a discrete single focus in five patients, from two different foci in six patients and were multifocal or diffuse in four patients. Ictal onset was seen from circumscribed areas in 14 patients and was diffuse in one patient. Cases with more than one spiking pattern were not associated with a higher number of implanted electrodes. Single EEG seizure patterns were recorded in eight patients. In five patients two different EEG seizure patterns were recorded, one patient had three different EEG seizure patterns and one additional patient had five different EEG seizure patterns. Interestingly, cases with more than one subdural EEG seizure pattern were associated with a higher number of implanted electrodes (p < 0.05). In eight patients (53%) the most frequent spike focus was concordant with a subdural EEG seizure onset zone. Six out of these eight patients presented with only one subdural EEG seizure type. One of these eight patients presented with very broad spikes, but their maximum was concordant with the ictal onset zone. In three additional patients (20%) the most frequent or second most frequent spike population occurred one or two electrodes (1–2 cm) adjacent to a seizure onset zone. In three patients diffuse spiking was seen whereas the seizure onset was focal. One patient had diffuse seizure onset with focal spikes. Conclusions: Ictal onset zones on subdural EEG recordings were related to the most frequent subdural interictal spike focus in 53% of cases. After inclusion of the second most frequent spike focus and surrounding electrodes, an ictal onset zones could be predicted in 73% by interictal spiking. However, not all ictal onset zones were predicted by the interictal spike focus, in particular if a patient presented with more than one ictal onset zone. Interestingly, a higher number of implanted electrodes and a higher number of recorded ictal onset zones may be related. Analysis of interictal subdural EEG may be helpful in the prediction of the ictal onset zone. 1 Addolorata Mascia, 1 Antonio Sparano, 1 Vincenzo Esposito, 1 Sergio Paolini, 1 Roberta Morace, 1 Giancarlo Di Gennaro, and 1 Pier Paolo Quarato ( 1 Epilepsy Surgery Unit, IRCCS Neuromed, Pozzilli, Isernia, Italy; Epilepsy Surgery Unit, IRCCS Neuromed, Pozzilli, Isernia, Italy; Epilepsy Surgery Unit; Department of Neurosurgery, IRCCS Neuromed; La Sapienza University, Rome, Italy, Pozzilli, Isernia, Italy; Epilepsy Surgery Unit; Department of Neurosurgery, IRCCS Neuromed; La Sapienza University, Rome, Italy, Pozzilli, Isernia, Italy; Epilepsy Surgery Unit; Department of Neurosurgery, IRCCS Neuromed; La Sapienza University, Rome, Italy, Pozzilli, Isernia, Italy; Epilepsy Surgery Unit, IRCCS Neuromed, Pozzilli, Isernia, Italy; and Epilepsy Surgery Unit, IRCCS Neuromed, Pozzilli, Isernia, Italy ) Rationale: To determine the potential effectiveness and tolerability of levitiracetam (Lev) versus oxacarbazepine (Oxc) versus placebo in preventing early postoperative seizures in patients undergoing brain surgery for mass lesions. Methods: 30 patients requiring elective craniotomy for supratentorial brain mass lesions were randomized to three groups of equal size, with a prospective, open‐label, controlled design. One group receveid Lev (13.3–40 mg/kg/die) the other group receveid Oxc (18–20 mg/kg/die), the third group receveid placebo. In all the three groups the drug administration were oral with a titration performed within the 30 days preceding surgery. Early postoperative seizures (seizures occurring in the 7 days following surgery) were considered for the purpose of the study. Patients having chronic epilepsy were not included. On the contrary, heralding seizures were not considered as esclusion criterion Results: Early post‐operative seizures were observed in all three groups. Patients taking Lev or Oxc showed less seizures than patients taking placebo. Conclusions: In our study Lev and Oxc seemed to be similarly effective in preventing early seizures after brain surgery. Both the drugs showed a good profile of tolerability and all patients remaining on each treatment during all the duration of the study. 1 Shearwood McClelland III, 2 Rebeca E. Garcia, 3 Daniel M. Peraza, 4 Tina T. Shih, 5 Lawrence J. Hirsch, 3 Joy Hirsch, and 2 Robert R. Goodman ( 1 Neurosurgery, University of Minnesota, Minneapolis, MN ; 2 Neurological Surgery, Columbia University College of Physicians and Surgeons, New York, NY ; 3 Radiology, Columbia University College of Physicians and Surgeons, New York, NY ; 4 Neurology, University of California at San Francisco, San Francisco, CA ; and 5 Neurology, Columbia University College of Physicians and Surgeons, New York, NY ) Rationale: The nondominant amygdala is crucial for processing facial expression and emotional recognition (ER) of visual stimuli, particularly in facial fear recognition. Patients with temporal lobe epilepsy (TLE) associated with mesial temporal sclerosis (MTS) often incur amygdalar and hippocampal damage. Previous studies have shown that patients with right‐sided (nondominant) MTS experience impaired ER if TLE onset occurred before age six. This finding has resulted in the hypothesis that early right mesiotemporal insult impairs plasticity, resulting in ER deficits, while damage later in life (seizure onset after age five) results in no deficit. However, this hypothesis has not been tested in a uniformly seizure‐free postsurgical population. This study was performed to examine this issue in late‐onset postsurgical patients. Methods: Controls (n = 10) and late‐onset patients (n = 5) were recruited. All patients underwent nondominant anteromedial temporal lobectomy (AMTL), had TLE onset after age five, Wada‐confirmed left‐hemisphere language dominance and memory support, MTS on both pre‐operative MRI and biopsy, and were Engel class I five years after surgery. Using a standardized (Ekman and Friesen) human face series depicting neutrality and the six basic emotions conveyed by facial expression (happiness, sadness, fear, disgust, anger and surprise), subjects were asked to match the affect of one of two faces to that of a simultaneously presented target face. Target faces expressed fear, anger, or happiness. Results: Statistical analysis (t‐test) revealed that the late‐onset group had ER (as measured by percentage of faces correct and reaction time) for fear (p = 0.871), anger (p = 0.256), and happiness (p = 0.608) comparable to controls. All 95% confidence intervals included zero. Conclusions: Following curative AMTL in nondominant MTS patients, TLE onset after age five does not predict significant ER impairment for facial expressions. Future studies are needed to examine whether early TLE onset continues to impair ER following AMTL. 1 Pradeep Modur, and 2 Warren Milteer ( 1 Neurology, University of Louisville, Louisville, KY ; and 2 Hazelwood Center, Louisville, KY ) Rationale: To investigate the impact of vagus nerve stimulation (VNS) therapy in institutionalized patients with Lennox‐Gastaut Syndrome (LGS) with respect to emergency room (ER) utilization for seizures. Methods: We identified patients with LGS from a retrospective chart review of 98 institutionalized patients with epilepsy. LGS was defined when all of the following criteria were met: mental retardation; multiple seizure types; frequent seizures refractory to antiepileptic drugs (AEDs); and prolonged video‐EEG monitoring showing 2 of these 5 findings – slow spike‐wave discharges, generalized fast activity, multifocal (≥3 sites) epileptiform discharges, tonic seizures and atypical absence seizures. The institution's protocol specified administration of intrarectal diazepam (Diastat) for seizures or seizure clusters lasting >3 minutes, and transfer to an ER if seizures persisted >10 minutes. Because of frequent seizure clusters in this population, the number of days with documented seizures was counted instead of the actual number of seizures. The comparator group consisted of patients without VNS (non‐VNS group) followed over a period of 18 months. The intervention group consisted of those with VNS, and were analyzed in two periods: 6 months before and 6 months after implantation. The primary outcome measure was the number of seizure‐related ER visits in the VNS group before and after implantation. Secondary outcome measures were the number of seizure days and the number of Diastat administrations in the VNS group during the same periods. Fixed effects poisson model and poisson regression were used for statistical analysis. Results: There were 9 patients with LGS, and 4 of them had been implanted with VNS. In the non‐VNS group, there were 5 patients (3 males), aged 24–53 years. In the VNS group of 4 patients (2 males) aged 25–56 years, the age at implantation was 22–53 years. Compared to non‐VNS patients, the VNS patients before implantation had significantly higher Diastat administrations (p < 0.0005) and ER visits (p < 0.0005) but not seizure days (p = 0.06). In the VNS group, during 6 months of follow‐up after implantation, the ER visits reduced significantly (10 versus 2, p = 0.04), the number of seizure days reduced minimally (107 versus 99, p = 0.58) while the number of Diastat administrations increased slightly (36 versus 50, p = 0.13). Conclusions: In a group of institutionalized LGS patients more prone to ER visits, VNS therapy significantly reduced the number of ER visits. The observed trend towards an increase in the administration of Diastat after VNS implantation suggests that the favorable outcome with respect to ER visits may be related to the combined effect of VNS and Diastat in reducing the duration of seizures or seizure clusters. Prospective studies are needed to confirm this and evaluate other potential factors such as VNS parameters and concomitant AEDs influencing this outcome. (Supported by Grant from the state of Kentucky.) 1 Paula Pergami, 1 Christian A. Keller, 1 Adriana E. Palade, 1 John F. Brick, 1 Warren Boling, 1 Alexander Torres‐Trejo, and 1 Daunice Lohr ( 1 Neurology, West Virginia University, Morgantown, WV; and Neurosurgery, West Virginia University, Morgantown, WV ) Rationale: The last few years have seen an increased interest in mechanisms regulating body weight. At hippocampal level, obesity can be viewed as the result of an imbalance between the tendency to seek a reward (food) and learned inhibitory processes. Our objective was to investigate the relation between removal of different brain structures and the development of postoperative eating disorder resulting in weight gain. Methods: In order to identify significant differences in weight changes, two groups of patients with intractable epilepsy were studied: patients with medically refractory partial seizures status post amygdalohippocampectomy (n = 25) and patients with extratemporal epilepsy who underwent surgery for the same condition (n = 18). BMI was calculated at baseline and at 6, 12, 18 and 24 months following surgery. Differences between the two groups were analyzed. The two groups were comparable for control of postoperative seizure, medications, age, gender and co‐morbidities. Results: Postoperatively, a significant difference in weight, consisting of weight gain was observed in patients with temporal lobe epilepsy following the amygdalectomy compared to patients undergoing extratemporal surgery. Conclusions: Our results further strengthen the hypothesis that the amygdala plays a role in eating behavior. This might involve complex neuronal circuits controlling behaviors like rewards, learning and memory formation. This finding supports a possible role for amygdala in eating disorders and challenges the traditional idea of eating disorders associated with hypothalamic disturbances. 1 Dimitris G. Placantonakis, 1 Saadat Shariff, 2 Douglas Labar, 2 Cynthia Harden, 2 Syed Hosain, 3 Neil Schaul, 3 Dimitrius Kolesnik, and 1 Theodore H. Schwartz ( 1 Neurological Surgery, Weill Cornell Medical College, New York, NY ; 2 Neurology, Weill Cornell Medical College, New York, NY ; and 3 Neurology, New York Hospital of Queens, Flushing, NY ) Rationale: Medically refractory epilepsy is amenable to neurosurgical intervention if the epileptogenic focus can be accurately localized. In some situations, the scalp EEG and MRI are discordant and/or non‐lateralizing, yet the clinician is still highly suspicious that a single focus exists, either based on a dominant interictal focus or stereotypical semiology. In these situations, a bilateral intracranial survey can be helpful to lateralize the ictal onset zone. The risks and success of such bilateral surveys have not been well described. Methods: We retrospectively reviewed 20 patients with refractory seizures treated over a 4‐year period. Patients underwent bilateral placement of subdural strip electrodes and, in 50% of cases, stereotactic implantation of depth electrodes into the mesial temporal lobes. In all cases the MRI scans and scalp EEG were not sufficiently concordant to lateralize seizure onsets. The mean age was 35.1 years and 75% of patients were male. Results: Patients were electrophysiologically monitored for an average of 7.4 days. Thirty‐five percent (n = 7) subsequently underwent unilateral implantation of electrodes to more accurately localize the seizure focus. Sixty percent (n = 12) eventually underwent a therapeutic procedure. Of those, 58.3% underwent resection of epileptogenic foci without unilateral invasive mapping. Reasons for no further surgery (n = 8) included failing the WADA (37.5%), multifocal onsets (25%), refusal of further treatment (12.5%) and negative intraoperative ECoG (12.5%). There was 1 complication, involving a retained electrode fragment that was removed in a separate minor procedure. Of the surgically treated patients, 58.3% have no seizures currently, 16.7% have improved and 25% remain unchanged. Conclusions: Bilateral placement of subdural strip and depth electrodes to survey the cortex in patients with non‐lateralizing scalp video‐EEG and MRI but clinical suspicion for localized epilepsy is both safe and effective. Of those patients who go on to a therapeutic procedure, 75% are either improved or seizure‐free. 1 Fabio Sebastiano, 1 Pier Paolo Quarato, 1 Vincenzo Esposito, 1 Angelo Picardi, 1 Antonio Sparano, 1 Addolorata Mascia, 1 Mario Manfredi, and 1 Giancarlo Di Gennaro ( 1 Epilepsy Surgery Unit, IRCCS Neuromed, Pozzilli, Isernia, Italy; Epilepsy Surgery Unit, IRCCS Neuromed, Pozzilli, Isernia, Italy; Epilepsy Surgery Unit; Department of Neurosurgery, IRCCS Neuromed; La Sapienza University, Pozzilli, Isernia, Rome, Italy; Center of Epidemiology and Health Surveillance and Promotion, Italian National Institute of Health, Rome, Italy, Rome, Lazio, Italy; Epilepsy Surgery Unit, IRCCS Neuromed, Pozzilli, Isernia, Italy; Epilepsy Surgery Unit, IRCCS Neuromed, Pozzilli, Isernia, Italy; Epilepsy Surgery Unit; Department of Neurology, IRCCS Neuromed; La Sapienza University, Pozzilli, Isernia, Rome, Italy; and Epilepsy Surgery Unit, IRCCS Neuromed, Pozzilli, Isernia, Italy ) Rationale: To evaluate a novel method for localization of subdural electrodes in presurgical assessment of patients with drug‐resistant focal epilepsy. Basic demographic, clinical, and neuroradiological characteristics of patients Patient N. Sex Age (years) Age at epilepsy onset (years) Epilepsy duration (years) Neurological examination Number of seizures per month MRI 1 f 37 19 18 Normal 20 r‐T dysplasia 2 m 21 11 10 Normal 20 r‐ and l‐TPO dysplasia 3 m 36 9 27 Normal 60 r‐P dysplasia 4 m 47 14 33 l‐upper quadranopia 16 r‐ and l‐TPO glyosis 5 m 28 7 21 Normal 20 Normal 6 f 12 2 10 l‐hemisoma sensory‐motor deficit 60 Normal 7 m 21 18 3 l‐upper quadranopia 10 Normal 8 f 31 31 19 Normal 20 r‐T tumour Methods: We studied eight consecutive patients with posterior epilepsy in whom subdural electrodes were implanted for presurgical evaluation. Electrodes were detected on post‐implantation brain CT scans through a semiautomated procedure based on a MATLAB routine. Then, post‐implantation CT scans were fused with pre‐implantation MRI to localize the electrodes in relation to the underlying cortical structures. Two independent raters tested the reliability of this procedure by comparing 3D‐rendered MR images of the electrodes with electrode position as determined by intraoperative digital photography. Results: In each patient, all electrodes could be correctly localized and visualized in a stereotactic space, thus allowing optimal surgery planning. For both raters, the agreement between the procedure‐generated images and the digital photographs was good in all patients. Conclusions: While our findings need confirmation on larger samples including patients with anterior epilepsy, this procedure allowed to localize subdural electrodes and to establish the spatial relationship of each electrode to the underlying brain structure, either normal or damaged, on brain convessity, basal and medial cortex. (figure 1) 1,2 Marianna V. Spanaki, 3 Kost V. Elisevich, and 1 Brien J. Smith ( 1 Neurology, Henry Ford Hospital, Detroit, MI ; 2 Neurology, Wayne State University, Detroit, MI ; and 3 Neurosurgery, Henry Ford Hospital, Detroit, MI ) Rationale: Epilepsy surgery is considered to be the most effective treatment in refractory epilepsy with seizure free outcomes from 45% to 67.9% (Engel J. Surgery for seizures. N Engl J Med 1996;334:647–652). However, little data exists on the number of older patients who are treated surgically, the post surgical complications and outcome rates in this patient population. Methods: We identified all patients who were 50 years and older and had epilepsy surgery for refractory epilepsy at Henry Ford Hospital (HFH) Epilepsy Program from 1994 to 2005. Data on age at surgery, duration of epilepsy, risk factors, noninvasive and invasive video/EEG monitoring, surgery site, pathology, outcome, and medical conditions at surgery as well as complications were collected through search of the Epilepsy Monitoring Unit admission database and the HFH electronic database (CarePlus). Patients who had tumors that presented with seizures and surgery was performed in the context of tumor treatment were not included. Descriptive statistics were applied. Results: Thirty five patients from 50 to 70 years old (median age = 55; 19 females) underwent epilepsy surgery at HFH. Duration of epilepsy was from 6 months to 50 years (median = 35 years). Nineteen patients (54%) with inconsistent localization during scalp continuous video EEG monitoring had invasive monitoring with grids, strips and/or depth electrodes. Thirty one patients had temporal resections (16 had left anterior temporal resection), 3 extratemporal and one subpial transection. Pathology revealed mesial temporal sclerosis in 14 patients, heterotopia in 2, dual pathology in 2, cavernous angioma in 2, dysembryoplastic neuroepithelial tumor in 1. In the remainder pathology was consistent with various degrees of gliosis or non diagnostic. Duration of follow up ranged from 1 month to 13 years (median follow up = 2 years). One patient was lost to follow up. Twenty five patients (25/34; 73.5%) were seizure free after surgery, 4 of which remained seizure free off antiepileptic medication. Medical history prior to surgery was consistent with hypertension (HTN) in 6 patients, hypercholesterenemia and HTN in 3, diabetes and HTN in 2 and coronary artery disease in 3 (2 of which had previous angioplasty). One patient had right MCA infarct one day after surgery that resulted in mild left hemiparesis. Following surgery, 1 patient developed transient angina and another one pulmonary embolism that was successfully treated. Conclusions: Surgical treatment of refractory epilepsy in older patient population is a safe treatment and yields seizure freedom outcomes that are comparable to those reported in young patients. Co‐morbid conditions and advanced age should not prevent physicians from offering epilepsy surgery to appropriate candidates. 1 K. Meng Tan, 1 Elson L. So, 1 Gregory D. Cascino, 2 Fredric B. Meyer, 2 W. Richard Marsh, and 1 Gregory A. Worrell ( 1 Department of Neurology, Mayo Clinic ; and 2 Department of Neurosurgery, Mayo Clinic, Rochester, MN ) Rationale: Patients with medically refractory temporal lobe epilepsy may be excellent candidates for epilepsy surgery. However, lateralization of the ictal onset zone is not always clear from non‐invasive studies. In such cases, bilateral temporal depth electrode implantation may be undertaken prior to considering temporal lobe resection. Although intraoperative pharmacologic activation is known to increase interictal epileptiform discharges on electrocorticography, the clinical relevance of this procedure has not been established. Methods: We performed a retrospective review of all epilepsy patients who underwent bilateral temporal depth electrode implantation during the period 1997–2005 and received intraoperative alfentanil activation. We identified ten patients and characterized the data provided by EEG (interictal and ictal), MRI, intraoperative electrocorticography (before and after pharmacologic activation) and subsequent prolonged intracranial EEG monitoring, with respect to localization of the ictal onset zone. We focused on the correlation between pharmacologic activation and lateralization of the epileptogenic temporal lobe as determined by intracranial EEG monitoring. Results: 6 females and 4 males meeting inclusion criteria were identified. Administration of alfentanil intraoperatively caused increased activation of epileptiform discharges bilaterally in 5 patients, and no definite activation in another 3 patients. In the remaining 2 patients, the side of pharmacologic activation did not correspond with the epileptogenic temporal lobe determined on subsequent prolonged intracranial monitoring. Conclusions: Intraoperative pharmacologic activation does not appear to be useful in patients with temporal lobe epilepsy of unclear lateralization. 1 Denise S. Taylor, 1 Heber L. Varela, 1 William O. Tatum, 1 Fernando L. Vale, and 1 Selim R. Benbadis ( 1 Neurology & Neurosurgery, University of South Florida & Tampa General Hospital, Tampa, FL ) Rationale: Lateralization of temporal lobe epilepsy (TLE) is sometimes unclear based on the non‐invasive presurgical evaluation. In this situation, invasive EEG‐video monitoring is performed, using either depth or subdural electrodes. We reviewed our experience with bitemporal subdural strips for lateralization of TLE. Methods: We reviewed the data on all patients who received bitemporal subdural strips for lateralization of TLE at our center from December 2000 to March 2006. We only analyzed patients whose sole question was lateralization of TLE; if non‐temporal coverage was performed, patients were excluded. Outcome used Engel's classification. Results: Over a 5‐year period, 21 patients underwent bitemporal subdural strips. Out of 1,356 patients monitored, this represents 1.5% of patients. The results for the 21 patients with bitemporal subdural strips evaluation was as follows: • 11 patients (52%) had clear lateralization (100% of seizures on one side) resulting in a temporal lobectomy: 8 class I, 2 class II, and 1 class III. • 5 patients had a predominance but less than 100% lateralization. These were offered a temporal lobectomy and presented before surgery as an unlikely seizure free outcome but a probable improvement (“palliative” interventions). 4 patients underwent a lobectomy (2 class III, 2 class IV), and 1 declined. • 1 patient underwent resection of a large arachnoid cyst that was contralateral to seizure onset (outcome class II). • 4 patients (19%) had true bitemporal epilepsy without predominance and no resection was performed. Conclusions: Subdural strips achieved lateralization of TLE in about half of cases where the non‐invasive evaluation failed to do so. When lateralization was 100% to one side, postoperative outcome was excellent. About 19% had true bitemporal epilepsy. 3 Fuyuki Hirashima, 1 Eric Dinnerstein, 1 Karen L. Gilbert, 2 David W. Roberts, 1 Vijay M. Thadani, and 1 Peter D. Williamson ( 1 Section of Neurology, Dartmouth Hitchcock Medical Center, Lebanon, NH ; 2 Section of Neurosurgery, Dartmouth Hitchcock Medical Center, Lebanon, NH ; and 3 Dartmouth Medical School, Hanover, NH ) Rationale: We wished to study variables that influence outcome, in terms of seizure control, among patients with mesial temporal sclerosis who undergo surgery. We compared two surgical techniques. Some patients had a standard temporal lobectomy, and others had a selective amygdalo‐hippocampectomy. In this retrospective study, we compared the outcomes of these surgical techniques with respect to seizure control. Methods: 61 patients with medically refractory epilepsy and mesial temporal sclerosis (seen on MRI and subsequent pathological examination) were evaluated. The two groups were very similar with regard to age, age at surgery, duration of epilepsy, handedness, risk factors such as febrile seizures, and side of resection. The number of anti‐epileptic medications, before and after surgery, was the same in the two groups. 30 patients had a standard temporal lobectomy, and 31 had the more selective procedure. The choice of procedure was arbitrary. Follow‐up information was collected at 6, 12, 18 and 24 months, and yearly thereafter. The mean duration of follow‐up was 4.38 ± 1.02 years. Outcome was assessed using the Engel classification. Patients were classified as Class I outcomes if they were seizure‐free for the previous year. To simplify the statistics, patients in Classes II, III, and IV were pooled. Neuropsychological testing was done before surgery, and 6–12 months after surgery. Results: Class I outcomes ranged from 63% to 96%. A trend towards a difference between the two surgical groups, with more Class I outcomes among those who had the standard procedure, was detected 4 years after surgery (89% vs. 67%), and a statistically significant difference was apparent 5 years after surgery (96% vs. 63%, p < .01). However, if the data were analyzed with Class I defined as seizure‐free since surgery, there was no difference between the two groups. Neuropsychological testing did not show a significant difference between the two groups before or after surgery. Conclusions: These data suggest that in patients with mesial temporal sclerosis a better outcome with respect to seizure control may be seen with a wider resection of the anterior‐lateral temporal lobe. The difference, if there is one, depends on how a Class I outcome is defined. Roughly the same% of patients in both groups has post‐operative seizures; but after several years, a higher% of those who had the wider resection has been seizure‐free for at least a year. The clinical significance of this difference is uncertain. 1 Alejandro Torres‐Trejo, 1 Adriana Palade, 1 Warren Boling, and 1 John F. Brick ( 1 Neurology, West Virginia Univerisity, Morgantown, WV ) Rationale: After epilepsy surgery for cortical dysplasia, a high percentage of patients may remain seizure free. The clinician for various medical reasons may decide to discontinue one or more antiepileptic drugs. It is reasonable to consider discontinuation of the antiepileptic drugs (AED) to perhaps decrease some of the side effects associated with these medications. We would like to identify whether patients will continue to remain seizure free after discontinuation of AED's. Methods: We are reporting five patients who underwent epilepsy surgery for resection of epileptic focus with pathology reported as cortical dysplasia. The seizure focus was localized electrographically. Structures resected include right posterior temporal region, right central sulcus, amygdalohippocampectomy, right frontal lobectomy, and a left parietal region resection. For various medical resons all of the patients had their AED's discontinued between six months to three years. Results: All of these patients remained seizure‐free between six months to three years after surgery, but had recurrent seizures after withdrawal of medications was attempted. Conclusions: Patients that had successful epilepsy surgery resecting the epileptic focus due to cortical dysplasia remained seizure free while on AED's but had recurrence of seizures when the medications were discontinued. This suggests that patients with cortical dysplasia may fail AED discontinuation despite having good seizure control postoperatively if the AED's are discontinued before three years. 1 K. Upchurch, 2 J. Stern, 2 R. Staba, 2 C.L. Wilson, 2 J. Engel Jr., and 1 I. Fried ( 1 Neurosurgery, David Geffen School of Medicine, UCLA, Los Angeles, CA ; and 2 Neurology, David Geffen School of Medicine, UCLA, Los Angeles, CA ) Rationale: For cases of pharmacologically refractory epilepsy in which data from noninvasive diagnostic studies (Phase I) are discordant with respect to the epileptogenic zone, a hypothesis concerning the location of ictal onset is formulated. On the basis of this hypothesis, patients may undergo invasive tests such as ictal EEG recordings via intracerebral depth electrodes (Phase II). The purpose of this study was to assess clinical outcomes (diagnostic and therapeutic) of invasive chronic depth electrode recording for patients with medically intractable epilepsy in a series of 100 patients who underwent operation by a single surgeon at a single institution. Methods: Using the UCLA Seizure Disorders Center surgical database, the history, operative course, and clinical outcomes of 100 patients with medically intractable epilepsy who underwent implantation of intracerebral depth electrodes at UCLA Medical Center between January 1993 and August 2005 were retrospectively reviewed. Results: Of the 100 patients reviewed, 54 were males; mean age was 33.2 years; and age range was 12–50 years. All surgeries were performed stereotactically by the same surgeon (I.F.) with pre‐operative targeting based on stereotactic cerebral angiography and magnetic resonance imaging. The average number of depth electrodes implanted for each patient was 10 (range 6 to 12). In 60/100 patients, a single epileptogenic zone was localized via the first Phase II study; 10/100 underwent a second Phase II study with subdural grids/strips; and 59/100 underwent Phase III resection. Morbidity secondary to intracerebral depth electrode implantation consisted of: (1) small intracerebral hematoma incidentally revealed on post‐implant or post‐deplant imaging in three asymptomatic patients; (2) contaminant bacteria on either Gram stain or culture on deplanted electrodes in two asymptomatic patients who were clinically without other evidence of infection; and one episode of fever and cerebrospinal fluid pleiocytosis during implantation in one patient whose deplanted electrodes and cerebrospinal fluid showed negative bacterial cultures. Mortality was zero. Conclusions: These data suggest that chronically implanted intracerebral depth electrodes for electrophysiologic recording are safe and continue to be important in patient selection for Phase III epilepsy surgery, despite the emergence of sophisticated noninvasive diagnostic modalities. (Supported by Epilepsy Foundation Post‐Doctoral Research Fellowship.) 1,3 Hans E. Van Der Aa, 2 Jacqueline J. Ardesch, 2 Gerard Hageman, and 3 Hendrik P.J. Buschman ( 1 Neurosurgery, Medisch Spectrum Twente, Enschede, Netherlands ; 2 Neurology, Medisch Spectrum Twente, Enschede, Netherlands ; and 3 Institute for Neuromodulation (TWIN), Medisch Spectrum Twente, Enschede, Netherlands ) Rationale: VNS‐therapy for treatment of patients with medically refractory epilepsy in the Netherlands is part of a continuous quality improvement system. This encompasses the organisatory structure, responsibilities, processes and facilities for carrying out quality care. One of the basic and core elements of a quality system is the systematic monitoring of the process of care via the quality circle. The first outcome results of VNS‐therapy in patients with epilepsy within this setting in the Netherlands are presented. Methods: The quality system was developed in a close collaboration with medical specialists from 5 neurosurgery and 5 epilepsy centers. Monitoring of the VNS‐quality system is performed in three phases (intake, implant, and follow‐up). The main indicators are reduction of epileptic seizures and changes in seizure severity, quality of life (QOLIE‐89) and adverse events. The effect of VNS during follow‐up was recorded after 1, 3, 6, 9 and 12 months. Results: Between 2002 and 2006, 122 patients were included following the national protocol for VNS‐therapy. During the first 2 years patient referral was low (∼12 patients p.y.). Thereafter ∼30 patients per year were included. Of these patients 9 were excluded. Of the remaining patients, at the time of analysis, seventy had received an implant, 22 were on the waiting list for implantation and 21 were in the intake phase. The mean age at intake was 40 y (SE 1.5 y). The onset of epilepsy in these patients was at 12.7 y (SE 1.1 y), and had a history with medically refractory epilepsy of 27.7 y (SE 1.5 y). They had 60.0 (SE 11.1) epileptic seizures per month. Most patients had localized partial epilepsy (59% symptomatic and 31% cryptogenic). A medically significant (>40%) decrease on seizure frequency during VNS was observed in 35% of patients after 1 month, and 49%, 61%, 42% and 64%, after respectively 3, 6, 9 and 12 months. Seizure severity was markedly reduced in 46%, 46%, 52%, 62% and 55% of patients after respectively 1, 3, 6, 9 and 12 months. About 80% of patients used ‘magnet activation’ to prevent or halt a seizure. 52% of patients experience a better or much better post‐ictal recovery period with VNS. Side‐effects were mainly voice related (hoarseness: 68%) and throat pain (32%). Quality of life showed statistically significant improvement in “change in health” and “general health”. Conclusions: Monitoring the effect of VNS‐therapy in a national quality‐system setting guaranties high quality care. Moreover it provides an effective means to review the outcome results of the participating centers. This national data‐base may eventually help identify response predictors to further improve the effectiveness of VNS‐therapy, which will subsequently be implemented in the national VNS‐therapy protocol. 1 Kenou Van Rijckevorsel, 2 Alexandra Volckaert, 1 Marianne De Tourtchaninoff, 1 Geraldo Vaz, and 1 Christian Raftopoulos ( 1 Reference Center for Refractory Epilepsy, Université Catholique de Louvain, Brussels, Belgium ; and 2 Psychology Faculty, Université Catholique de Louvain, Brussels, Belgium ) Rationale: Epilepsy surgery is an effective treatment for seizure control, especially in temporal lobe epilepsy (TLE). Morbidity is low, often limited to visual field defect and memory problems. However, long‐term cognitive and emotional effects are poorly documented, especially in selective amygdalo‐hippocampectomy (Sel‐AH). This paper studied memory (hippocampus,) emotions (amygdala) and quality of life (QOL) in 21 TLE patients after Sel‐AH. Methods: Out of 31 Sel‐AH patients operated between 2001 and 2005, 21 adults (5M, mean age 35 (19–53) years, 15 left) completed the tests; 4 children or teenagers were excluded, 4 refused or did not answer, 2 died of possible SUDEP more than 2 years after AH. Memory (immediate, work and general) was tested subjectively (QAM) and objectively (MEM‐III Wechsler 3rd ed). QOL was appreciated with QOLIE‐31‐P questionary and emotions were analyzed by presenting computed pictures of fear, sadness and joy (Mann‐Witney and Greenhouse‐Geisser tests). Results were compared to matched controls or control population. Results: 15 (71.43%) are seizure free since surgery. 4 complained spontaneously from memory and 4 were chronically depressed, already before AH. For memory, patients' performances are inferior to controls' in QAM for reading and movie, especially in left AH. With MEM‐III test, patients are less competent than controls, but work memory is the most preserved or the best recovered without side inference of AH. There is no correlation between QAM and MEM‐III results. For QOL, patients are less anxious than unoperated epilepsy controls regarding seizures and medications. However, patients are more anxious than controls in general regarding the ongoing (test) situation. In emotional testing, there are significant differences between controls and patients, for control situations (human‐animal and male‐female) and in emotional testings: patients made more errors and more omissions with a slightly increased reaction time (RT). However, these differences were minimal without side inference of AH. Conclusions: Memory, QOL and emotions have been tested in 21 adult Sel‐AH patients, compared to matched controls or control population. Memory is less performant in patients compared to controls, but not deteriorated by surgery. QOL is improved in surgery patients compared to unoperated epilepsy patients regarding seizures and medications. For emotions, patients differ significantly from controls by making more errors and omissions with a slightly increased RT. However, these results are the same for test and emotional situations and are probably unrelated to surgery. 1 Marcos Velasco, 1 Ana L. Velasco, and 1 Francisco Velasco ( 1 Stereotactic and Functional Neurosurgery, Hospital General de México, Mexico City, Mexico ) Rationale: Resective surgery of the epileptic focus yields very good results, nevertheless there are cases that escape this surgical possibility, i.e. patients with bilateral hippocampal foci or patients with epileptic focus located nearby eloquent areas for speech and memory (usually the left side). These latter patients cannot be operated on because it would mean having severe neurological impairment particularly related to short term memory. These patients are candidates for neuromodulating procedures. The present study evaluates the long term efficacy of ESH in the treatment of complex partial seizures in patients with intractable mesial temporal lobe epilepsy. Methods: Nine patients were studied. They had at least a 6 month basal period during which a careful seizure record was obtained. They all underwent bilateral hippocampal electrode implantation to establish focus was laterality and location. 3 patients had bilateral and 6 had unilateral foci. Diagnostic electrodes were explanted and definitive Medtronic electrodes were implanted directed to the hippocampal foci. Position was confirmed with MRI and afterwards the DBS system internalized. Patients signed the informed consent approved by the Hospital's Ethics Committee and started a double blind stimulation protocol. Patients attended every 3 months for seizure count and neuropsychological tests. Results: Follow up went from at least 18 months to 5 years. Patients were divided in two groups: 5 had normal MRIs and 4 had hippocampal sclerosis. Patients with normal MRI had a seizure reduction of >95%. Patients with hippocampal sclerosis had a seizure reduction of 30–50%. None of the patients showed deterioration of neuropsychological function, patients who had >95% seizure reduction, showed improvement in the neurophsychological tests. No patient showed adverse effects. 1 patient had to be explanted after 2 years due to skin erosion in the trajectory of the system. 1 patient had skin erosion which was corrected with plastic surgery. Conclusions: ESH provides a non lesional method that improves seizure outcome without deterioration of memory in patients with unilateral or bilateral hippocampal epileptic focus. 1 Clarissa L. Yasuda, 1 Leonardo Bonilha, 1 Chris Rorden, 1 Helder Tedeschi, 1 Evandro Oliveira, 1 Li M. Li, and 1 Fernando Cendes ( 1 Neurology, State University of Campinas (UNICAMP), Brazil; Neuropsychiatry, University of South Carolina; and Communication Sciences and Disorders, University of South Carolina ) Rationale: Resection of hippocampus (HIP)has been the standard procedure for patients with drug resistant (DR) medial temporal lobe epilepsy (MTLE) (1), with a success rate of approximately 75%(1,2). The reasons for the failure of almost 25% of the operated patients remain unknown. We aimed to investigate the influence of resection of medial temporal lobe (MTL) structures in addition to HIP in the surgical outcome(SO) of these patients. Methods: We studied 43 patients with DR, unilateral MTLE who underwent anteromedial temporal resection in the Epilepsy Service at UNICAMP, between 1998 and 2004. We selected patients according to our standard surgical investigation protocol (2) and included only patients with hippocampal atrophy supporting both clinical and electrophysiological findings. This study was approved by the ethics committee of our institution. All patients underwent a routine MRI scan at least six months after the surgery, with the following parameters: T1weighted with 1 mm isotropic voxels acquired on a 2 Tesla scanner. We used the Engel's postoperative outcome scale (1) and conducted a voxel lesion symptom mapping (VLSM) (3) of their post‐operative MRI to discriminate the contribution of distinct segments of the MTL cortex to the SO (4). We computed linear regressions and applied a cost function transformation of the resection maps for each patient to a common set of spatial coordinates. Results: Patients underwent surgery with the mean age of 37 (range 17–56) years, and the mean follow up period was 40 (range 11–99) months. The SO classification was: 76% as Engel I,14% as Engel II,5% as Engel III and 5% as Engel IV. We observed that the extent of hippocampal removal was associated with better outcomes (t = 2.371, p = 0.023). However, when the resection of the HIP was associated with resection of the entorhinal cortex there was a greater likelihood of seizure control after surgery (t = 3.286, p = 0.002) (Fig. 1). Figure 1. VLSM maps are overlaid in a normal T1 template. Colors that are associated with low p‐values (green, yellow and red) are the most significantly associated with a good SO and correspond to the entorhinal cortex and HIP. (figure 1) Conclusions: Based upon this finding, the efficiency of the surgical treatment of MTLE can be improved by including the entorhinal cortex to the hippocampal resection. References 1. N Engl J Med 2001;345:311–318. 2. Seizure 2006;15:35–40. 3. Nat Neurosci 2003;6:448–450. 4. Hum Brain Mapp 2004;22:145–154. (Supported by FAPESP:05/59258.) 1,2 Arthur Cukiert, 1,2 Jose A. Burattini, 1,2 Pedro P. Mariani, 1,2 Lauro Seda, 1,2 Cristine M. Baldauf, 1,2 Meire Argentoni‐Baldochi, 1,2 Carla Baise‐Zung, 1,2 Cassio R. Forster, and 1 Valeria A. Mello ( 1 Neurology & Neurosurgery, Hospital Brigadeiro, Sao Paulo, Sao Paulo, Brazil ; and 2 Neurology & Neurosurgery, Clinica de Epilepsia de Sao Paulo, Sao Paulo, Sao Paulo, Brazil ) Rationale: Children with refractory epilepsy often present with delayed growth. We studied the effect of epilepsy surgery on linear growth of kids with refractory focal or generalized epilepsy. Methods: Eleven consecutive children with refractory epilepsy were evaluated as for their linear growth after epilepsy surgery. Six kids were submitted to focal cortical resection and five to extensive callosal section (90%). Age ranged from 3 months to 8 years (mean = 4.1 years). All patients had daily seizures. Four children had cortical dysplasia on MRI, two had gliotic lesions and five diffuse atrophy (those submitted to callosal section). Antiepileptic medications were kept at the same regimen as preoperatively, at least for one postoperative year. Follow up time ranged from 1.1 to 5 years (mean = 1.8 years). Results: Preoperatively, one patient was growing under the 2.5 growth percentile, 8 within the 2.5 and 10 percentile and one within the 10 and 25 percentile intervals. All patients submitted to cortical resection have been rendered seizure‐free after surgery; all kids submitted to callosal section had a 90% decrease of the generalized seizure frequency, but none was seizure‐free. All patients improved their linear growth after surgery and all of them were growing at least two percentile intervals above the preoperative level. Although not statistically significant, this “catch‐up” phenomenon was clinically more noticeable in those kids submitted to callosal section. Conclusions: Children with refractory epilepsy that underwent successful epilepsy surgery caught‐up with linear growth after surgery. Although this could be seen as recovery from a chronic illness in the kids with focal resections that were seizure‐free postoperatively, it is unlikely that an isolated response to improved health would be responsible for the additional growth in patients submitted to callosal section, who were not seizure‐free postoperatively. In this pediatric population, an inhibitory effect on growth from the epileptic activity might explain this postoperative catch‐up phenomenon, especially in children with generalized epilepsy. (Supported by Sao Paulo Secretary of Health.) 1 Jason S. Doescher, 1 Frank J. Ritter, 1 Patricia E. Penovich, 1 Deanna L. Dickens, 2 Mary Beth Dunn, and 1 Michael D. Frost ( 1 Minnesota Epilepsy Group, PA, Minnesota Epilepsy Group, PA, St. Paul, MN ; and 2 United Neurosurgery Associates, United Neurosurgery Associates, St. Paul, MN ) Rationale: Many clinical variables are considered when selecting pediatric epilepsy surgery candidates with the objective of predicting beneficial outcome. Methods: We analyzed all pediatric patients who underwent initial resective surgery at Minnesota Epilepsy Group from Jan 2000‐Dec 2002. Beneficial outcome was defined as Engel class I or II. Beneficial outcomes were compared to many diagnostic variables: attainment of early childhood milestones, presence of nocturnal seizures, history of febrile seizures, history of status epilepticus, symptomatic versus cryptogenic etiology, temporal vs. extra‐temporal lobe seizures, complete vs. partial resection, and presence of multiple conflicting diagnostic variables. Outcome measurements were done at 24‐months. Statistical method utilized was Fisher's Exact Test. Results: 46 consecutive subjects were identified and reviewed; 23 (50%) underwent temporal lobe only resection; extra‐temporal resections included 10 (22%) frontal, 1 (2%) parietal, and 12 (26%) multi‐lobe. In subjects with available follow‐up data, 19/33 (58%) were seizure free at 24‐months. Achievement of early childhood milestones was not associated with beneficial seizure outcome. The clinical history of nocturnal seizures, febrile seizures, or status epilepticus did not statistically separate beneficial outcomes groups. Etiology was not a statistically significant influence on beneficial outcome at 24‐months (symptomatic 23/30: 77%; cryptogenic 1/3: 33%; p < 0.174); however, this is may be influenced by low numbers within the cryptogenic group. Temporal lobe surgery was statistically more successful in achieving a beneficial outcome of Engel I‐II in 15/17 (88%) subjects compared to 9/16 (56%) of the extra‐temporal group (p < 0.046). Complete resection of the epileptogenic zone was associated with beneficial outcome in 20/24 (83%) subjects compared to partial resections 4/9 subjects (44%) (p < 0.039). The presence of multiple conflicting diagnostic variables was the strongest predictor of beneficial outcome as 24/27 (89%) subjects achieved Engel I‐II compared to none of the 6 subjects with >1 contralateral diagnostic variable (Fisher's Exact test p < 0.001). Conclusions: Favorable diagnostic variables associated with an Engel class I‐II outcome following epilepsy surgery include temporal lobe resections, complete resections, and absence of multiple contralateral diagnostic variables. In this retrospective review, the absence of multiple contralateral variables was the strongest predictor. 1 Georg Dorfmüller, 1 Emidio Procaccini, 1 Martine Fohlen, 1 Christine Bulteau, 1 Claude Jalin, and 1 Olivier Delalande ( 1 Pediatric Neurosurgery, Fondation Rothschild, Paris, France ) Rationale: To present our single‐center experience with a frameless robot‐guided endoscopic system for the transventricular disconnection of hypothalamic hamartomas (HH) associated with drug‐resistant epilepsy. Methods: Fourty‐three patients (9 m to 34 y, median age: 10 y) were operated on either through a pterional approach (8) or by means of transventricular endoscopy (21) or subsequently by both techniques (14), in order to disconnect the HH below or at the wall of the third ventricle. For the endoscopic approach we used a frameless stereotactic robot‐guided navigation (Neuromate®, Schaerer‐Mayfield). The trajectory of the endoscopic approach was defined by the hamartomas plane of attachment on the hypothalamus, for which we have proposed an anatomo‐topographic classification. Results: In terms of seizure outcome, we achieved the best results in HH with a vertical broad plane of attachment, extending into the third ventricle, enabling us to disconnect it through the endoscopic route. In this subtype (n = 12), 83% of the patients became seizure‐free, as compared to 43% in the other three types (horizontal attachment, attachment in multiple planes and giant HH). Other parameters, such as the patients age, seizure type and frequency or the duration of illness, were not significantly related with the outcome. In all, of 42 patients followed up, 21 (50%) became seizure‐free and 2 had rare events, while 17 (41%) had a worthwhile reduction in their seizure frequency, and 2 patients remained unchanged. Furthermore, improvement in several patients with disease‐related cognitive and behavioural disturbances could be objectified postoperatively. In terms of surgery‐related morbidity, the endoscopic approach had a significantly lower complication rate. Conclusions: Our results confirm the safety and efficacy of the disconnection of HH extending into the third ventricle with our robot‐guided endoscopic approach. 1 Eliana Garzon, 2 Ajay Gupta, 2 Jennifer Haut, 2 Deepak Lachhwani, 2 Prakash Kotagal, 2 Elaine Wyllie, 3 William Bingaman, 1 Americo Sakamoto, and 2 Hans Luders ( 1 Neurology and Neurosurgery, Universidade Federal de Sao Paulo, Sao Paulo, Sao Paulo, Brazil ; 2 Neurology, Cleveland Clinic Foudation, Cleveland, OH ; and 3 Neurosurgery, Cleveland Clinic Foundation, Cleveland, OH ) Rationale: In infants and young children with congenital or early onset acquired lesions, interictal scalp EEG may show generalized or multiregional abnormalities but ictal EEG concordant with semiology and brain MRI usually clarifies the epileptogenic zone for surgical option. We studied 6 children with catastrophic epilepsy whose ictal scalp EEG was classified as maximum over the unaffect hemisphere contralateral to the brain lesion. Factors leading to paradoxical lateralization of ictal scalp EEG over the unaffected hemisphere were evaluated. Methods: Of 209 patients in the pediatric epilepsy surgery database of Cleveland Clinic Foundation, clinical, EEG, brain imaging, surgery and postoperative records of 6 patients, whose EEG was classified as maximum in the unaffected hemisphere were reviewed. Digital records were restored and analyzed independently. Paradoxical lateralization was defined as scalp EEG seizure pattern that developed eiher exclusively or bilaterally but maximally contralateral to the to the side of the lesion. Results: Six patients, 4 females, 3–14 years in age at the time of the pre‐surgical evaluation, were identified. Despite the paradoxical lateralization of ictal scalp EEG during recorded seizures, lateralizing signs (>1 sign in 3 patients) of focal tonic (4 patients), head versive (2 patients), hemiclonic (1 patient) and asymmetric epileptic spasm (1 patient) were noted contralateral (concordant) to the side of lesion. Interictal EEG showed generalized or lateralized and multiregional epileptiform discharges with predominance (> 65%) over the affected hemisphere in 3/6, generalized but maximum over the unaffected hemisphre in 2/6, and generalized with no focal predominance in 1 patient. Brain MRI showed extensive destructive (encephaloclastic) lesions such as stroke in 3 and encephalomalacia (post status epilepticus and infectious) in 3 patients. All patients underwent hemispherectomy and 5/6 patients were seizure free after 5 months to 3 years of follow‐up. Conclusions: Extensive congenital or early onset acquired encephaloclastic lesions may produce an ictal scalp EEG pattern that appear maximum over the unaffected hemisphere. Interictal EEG in these patients may also show bilateral and multiregional abnormalities with or without any predominant region of abnormality. These scalp EEG patterns should be recognized and they do not preclude benefit from epilepsy surgery if the neurological deficits, semiology, and imaging implicate the abnormal hemisphere. Early spread to the normal hemisphere, the presence of encephaloclastic lesion modifying scalp EEG patterns, and secondary epileptogenesis may explain these findings. (Supported by CAPES (Brazil).) 1 Tiziana Granata, 1 Elena Freri, 2 Carlo Marras, 3 Alessandro Pincherle, 4 Marina Casazza, 1 Francesca Ragona, 2 Paolo Ferroli, 3 Flavio Villani, 3 Roberto Spreafico, and 2 Giovanni Broggi ( 1 Department of Child Neurology, Neurological Institute C.Besta, Milan, Italy ; 2 Department of Neurosurgery, Neurological Institute C.Besta, Milan, Italy ; 3 Department of Clinical Epileptology, Neurological Institute C.Besta, Italy ; and 4 Department of Clinical Neurophisiology, Neurological Institute C.Besta, Italy ) Rationale: We report on the outcome of 13 patients who underwent hemispherotomy, for the relief of drug resistant seizures, at the Neurological Institute of Milan, between 2000 and 2005. Methods: The 13 patients were affected by static or progressive encephalopathies, involving the right hemisphere in 7 cases, and the left in 6. Diagnosis was made by MRI and confirmed by pathological study. Motor, mental, and language evaluation have been performed before surgery and at least yearly during a follow‐up lasting 6–72 months (median 35.5 months). All the patient underwent peri‐insular hemispherotomy. Results: Four patients had malformation of cortical development (2 hemimegalencepahly, 2 cortical dysplasia), 3 patients suffered of acquired perinatal brain damage and 6 patients were affected by Rasmussen encephalitis (RE). Age at the first seizure was 1 month – 11 years (median 4.4 years), age at surgery was 6 months‐18 years (median 7.3 years), the mean delay between the onset of seizures and surgery being 3.3 years (range 4 months‐9 years). In all the patients motor deficit contralateral to the affected hemisphere, mental retardation or cognitive decline, of different severity, were present at the time of treatment. Post‐op course was complicated in 2 cases (chemical meningitis, subdural hematoma). There were no life‐threatening complications. Among the 10 patients followed‐up for at least 12 months, surgery was judged as successful in 8 cases. Six patients became seizure‐free (3/4 cases with RE, 2/2 cases with acquired brain damage, 1/2 cases with diffuse cortical dysplasia), whereas 2 children (1 with RE, 1 with cortical dysplasia) showed >50% reduction of seizures. Antiepileptic treatment was withdrawn in 3 patients, and simplified in 4. In all the 8 patients neurological condition improved: motor deficits remained unchanged in 3 patients and improved in 5; a significant cognitive and behavioural improvement was observed in 5 patients, no patient with left hemispherotomy experienced worsening in language. No improvement in seizure frequency, nor in motor and cognitive skills was, by contrast, obtained in the 2 children affected by hemimegalencephaly. Conclusions: Our data support the efficacy of peri‐insular hemispherotomy in unilateral encephalopathies. As already reported, the seizure outcome is related to the underlying pathology, with the best results in acquired brain damage and in RE, and the worst in hemimegalencephaly. Our data also confirm that control of seizures is often associated with improvement of motor and mental abilities when surgery is performed in childhood. 1 Jin Sook Lee, 1 Jee Yoon Park, 1 Jin Hwa Moon, 1 Hee Hwang, 1 Ki Joong Kim, 1 Yong Seung Hwang, 2 Seung Ki Kim, 2 Byung Kyu Cho, and 3 Ho Jin Park ( 1 Pediatrics, Seoul National University Hospital, Seoul, Republic of Korea ; 2 Pediatric Neurosurgery, Seoul National University Hospital, Seoul, Republic of Korea ; and 3 Pediatrics, Eulji University Hospital, Daejeon, Republic of Korea ) Rationale: Dysembryoplastic neuroepithelial tumor (DNT) is a benign brain tumor and often occurs in association with medically intractable partial seizures. The authors report our experience of DNT patients treated with epilepsy surgery. Methods: A retrospective review and analysis of medical records were performed on brain MRI findings, interictal and ictal EEG features pathological findings, other clinical characteristics, and surgical outcome of 14 patients with refractory epilepsy secondary to DNT. Results: All patients showed cortical or subcortical lesions with T1 low and T2 high signal intensity, and only one case showed focal small enhancing portion within the tumor. Other MRI findings included cystic formation (6), calcification (2) and hippocampal atrophy (1). Temporal lobe was involved in 9 cases (64.3%), while frontal, parietal or occipital lobes were involved in 2, 3, and 1 respectively. Mean age at seizure onset was 7.2 years (range 0.3 – 13.8 years). All patients presented with complex partial seizures with or without secondary generalization, however, one patient had previous history of infantile spasms. Ten of the 12 patients revealed abnormal findings on interictal EEG. Long‐term video‐EEG monitoring (VEM) was done in 5 patients. The epileptogenic lobe was correctly localized by VEM in 4 patients who had lesions in the temporal lobe. Intracranial recording was performed in one patient, in whom lesion was located in the parietal lobe but scalp VEM failed to localize ictal onset zone. Lesionectomy was done in 8 cases (57%) and temporal lobectomy with amygdalohippocampectomy was done in 6 cases (43%). Pathology revealed associated cortical dysplasia (CD) in 4 patients (28.6%), and one of them showed oligodendroglioma component as well as CD. Mean follow‐up period after surgery was 2.8 years (range 0.2 – 5.9 years). All had an Engel Class IA outcome (3 patients still have been on antiepileptic drugs) without recurrence of tumor. Second operation was performed in one patient with temporal lobe involvement. Postoperative morbidity (visual field defect and hemiparesis) was observed in 2 cases who had lesions in the temporal lobe (21.4%). Follow‐up MRI could not exclude the possibility of residual tumor in 3 cases, however, no further seizure was reported so far. Six of the 9 patients showed normal postoperative EEG. Conclusions: DNT most frequently involved the temporal lobe and was often associated with CD. Gross total removal is thought to ensure favorable prognosis in most cases regardless of type of surgery. However, careful resection is necessary to reduce morbidity especially in case of temporal lobe involvement. 1 Nicoletta Longoni, 1 Warren W. Boling, 3 Adriana Palade, 4 Angel Wabulya, 2 Frederick Andermann, 1 Priscah Mujuru, 4 Warf Benjamin, 6 Donald Gross, 5 Richard Bittar, 7 Gavin Fabinyi, and 8 Richard Byrne ( 1 Neurosurgery, West Virginia University, Morgantown, WV ; 2 Neurology and Neurosurgery, Montreal Neurological Institute, Montrea, QC, Canada ; 3 Neurology, West Virginia University, Morgantown, WV ; 4 Community Medicine, West Virginia University, Morgantown, WV ; 5 CURE Childrens Hospital of Uganda, Mbale, Uganda ; 6 Neurosurgery, Monash University, Melbourne, Victoria, Australia ; 7 Neurology, University of Alberta, Edmonton, AB, Canada ; 8 Neurosurgery, University of Melbourne, Melbourne, Victoria, Australia ; and 9 Neurosurgery, Rush University Medical Center, Chicago, IL ) Rationale: Since May 2005 an Epilepsy Monitoring Unit (EMU) has been in place at the CURE Children's Hospital of Uganda (CHU) in Mbale, Uganda, East Africa. Children are pre screened in regionalized clinics prior to evaluation at CHU comprehensive epilepsy program, the first such program in East Africa. The program is dedicated to the evaluation and treatment of pharmacoresistant temporal lobe epilepsy. Methods: Prescreening relies on identifying the stereotypical signs and symptoms of temporal lobe epilepsy. Treatment candidates must have intractable epilepsy, seizure frequency over 1 per month, and a history of two of the following: ‐ febrile convulsions ‐ typical aura ‐ typical semiology of stare and/or automatisms Patients less than 6 years of age or multiple extra‐temporal intracranial lesions were excluded. EEG and CTimaging are done at CHU. The data are transferred for analysis to WVU. Epilepsy surgery is performed at CHU. Results: 30 patients have been evaluated at CHU. Seizure frequency ranged between 30/day to 2/month with a mean of 2.2 seizures/day. 25 patients presented an aura. Essentially all patients had a history of malaria. 27 had febrile convulsions. To date, 19 patients have completed an evaluation in the comprehensive program. EEG recording has shown: Left temporal (5), right temporal (2), frontal (2), bitemporal (2), multifocal (3), generalized (4), and non‐epileptic (1). CT showed scar in 4 cases. Four patients had convincing temporal lobe epilepsy and underwent surgery at CHU. Conclusions: Epilepsy prevalence has been estimated to be 20 per 1000 in one region of Eastern Africa, 10 times the prevalence in developed countries making intractable epilepsy a significant public health problem in the region. Malaria was essentially universal in the population screened and we suspect cerebral malaria as an important cause of intractable epilepsy. One case of “non‐epileptic spell” was identified (5% in the population studied). (Supported by USAID, American Schools and Hospitals Abroad.) 1 Angel W. Hernandez, 1 Saleem I. Malik, and 1 David Donahue ( 1 Comprehensive Epilepsy Program, Cook Children's Medical Center, Fort Worth, TX ) Rationale: Patients undergoing brain surgeries are at risk of developing bacterial infections post‐operatively, but herpes simplex virus (HSV) reactivation is a rare complication. Few reports are available on patients that develop reactivation herpes virus encephalitis following epilepsy surgery. We report a case of HSV reactivation post epilepsy surgery to raise awareness of this risk and to consider treatment pre and post surgery on patient with a history of prior HSV infection. Methods: We report one case of reactivation of herpes simplex virus after epilepsy surgery on a 12 year old with pharmaco‐resistant localization‐related epilepsy of left temporo‐parietal onset. This patient had a history of negative cerebrospinal fluid (CSF) polymerase chain reaction (PCR) for HSV as an infant. Four days after resection of the epileptogenic focus the patient developed acute repetitive seizures, meningeal signs, confusion, lethargy and high fever. PCR in the CSF was positive for HSV. Results: The patient's seizures stopped 48 hours after starting treatment with Acyclovir. The confusion and lethargy resolved 48–72 hours after initiating treatment. There were no sequelae to the infection. The patient is now 1.5 years seizure free and off anti‐convulsants. Conclusions: Though early recognition and treatment of HSV encephalitis is essential, we believe that prophylactic treatment with acyclovir pre and post brain surgery is warranted in patients with prior HSV exposure to decrease the chances of this complication. 1 Manuel Marrufo, 1 Yong D. Park, 1 Ki Hyeong Lee, 2 Mark R. Lee, 2 Joseph R. Smith, and 2 Scott Y. Rahimi ( 1 Neurology, Medical College of Georgia, Augusta, GA ; and 2 Neurosurgery, Medical College of Georgia, Augusta, GA ) Rationale: Corpus callosotomy is a treatment option for epilepsy patients who are not resective surgery candidate. EEG as predictor or outcome for corpus callosotomy cases has been studied exclusively among adult population. Methods: Patient selection: A retrospective anlysis included all patients under 18 yrs who underwent corpus callosotomy from July 02 to April 06. Data collection: The demographic data, seizure characteristics and seizure outcome was obtained from the medical records, as well as direct phone interviews at least 6 month following the surgery. The EEG findings were reviewed by two authors (YDP and MM) blinded to the outcome. Seizure and EEG patterns classification: Seizure type was classified and recorded in to two main categories: Category A for drop attacks and tonic seizures; Category B for GTCS, atypical absence, adverse postures, CPS and myoclonus. Ictal EEG was classified into two different patterns: Type A for generalized desynchronization, spike and wave, and diffuse beta activity; and Type B for focal onset confined to one hemisphere or with secondary generalization. Outcome evaluation: Two seizure outcomes were defined by the reduction of seizure: Class I ≥ 90% reduction, and Class II < 90% or less. Statistical Methods: Chi‐Square test run with SPSS 13 package. Results: There were a total of 35 patients followed for 11mo (6–12mo). The average age at surgery and the average age of seizure onset were 7.4 years (range 1 – 17 yrs) and 2.36 yrs (range 0–10 yrs) respectively. Sex ratio was 20:15 (M:F). In terms of seizure characteristic, 24 out of 35 (68%) patients belonged to category A while 11/35 (32%) to category B. When correlated with outcome of surgery, patients with Category A seizures showed a tendency to do better than Category B (Category A 85% (12/14) vs. Category B 15% (2/14), p‐value 0.077). Ictal EEG pattern did not predict the outcome: Type A 47.6% (10/21) vs Type B 30.8% (4/13), p‐value 0.272. Other demographic or pathology characteristics were not different between the two outcome groups. Conclusions: Our data suggest that ictal EEG pattern does not predict outcome following corpus callosotomy among the pediatric population unlike the published data on adult. However, consistent with previous reports, “drop” seizures tended to respond better to this surgical procedure than other seizure types in our population. Careful selection of candidate regarding seizure types, not necessarily ictal EEG patterns, are recommended for corpus callosotomy. 1 Bruno Maton, 1 Prasanna Jayakar, 1 Trevor Resnick, 1 Glenn Morrison, 1 John Ragheb, 1 Catalina Dunoyer, 1 Pat Dean, and 1 Michael Duchowny ( 1 Comprehensive Epilepsy Center, Brain Institute, Miami Children's Hospital, Miami, FL ) Rationale: Temporal lobe epilepsy (TLE) that begins in early life is often a catastrophic disorder with pharmaco‐resistant seizures and secondary neurological deterioration. Few data are available regarding epilepsy surgery performed during infancy and early childhood and no prior study has focused on TLE. Methods: We analyzed the results of temporal resection, performed for epilepsy as the primary indication, between 1979 and 2003, in children less than age 5 years at time of surgery who had at least 2 years of follow‐up. Results: 20 children (14 males) were identified with a mean age at surgery of 26 months and a mean age at seizure onset of 12 months. Clinical presentation was age‐dependent. Typical psychomotor seizures (n = 4; mean age at surgery 37 months) may be followed by prominent motor changes (n = 7; 30 months) that can be isolated (n = 3; 23 months). Epileptic spasms were noted in 6 patients and were frequently associated with partial seizures. The interictal EEG was lateralizing in 15 patients and ictal EEG was lateralizing in 18 patients. Brain MRI had good predictive value in 16 patients; ictal SPECT was concordant in 4/8 patients. Invasive EEG was employed in 6 cases. At mean follow‐up of 5.5 years, 65% of the children were seizure‐free and 15% had > 90% seizure reduction. Morbidity included infection and hydrocephalus in one case and stroke‐related hemiparesis in two cases. Cortical dysplasia was identified in 8 children, tumors in 8 including 2 DNET, 2 ganglioglioma, and 4 malignant tumors. Hippocampal sclerosis was present in 4 cases as dual pathology. Conclusions: Surgery for TLE can be performed during early life with similar results as in older children and in adults. 2 Ann‐Christine Duhaime, 3 Barbara C. Jobst, and 1 Richard P. Morse ( 1 Pediatrics and Neurology, Dartmouth‐Hitchcock Medical Center, Lebanon, NH ; 2 Pediatric Neurosurgery, Dartmouth‐Hitchcock Medical Center, Lebanon, NH ; and 3 Neurology, Dartmouth‐Hitchcock Medical Center, Lebanon, NH ) Rationale: Supplementary Motor Area (SMA) seizures are relatively common in children but there are small numbers reported in terms of surgical approach and outcome. Despite the fact that SMA epilepsy typically begins in childhood and is often refractory to AED treatment, surgical approaches to its management most often are delayed until adulthood. We report on the clinical features, evaluative test results (SPECT, invasive EEG) and surgical outcome in a series of 5 non‐lesional children and adolescents, 3 of whom have had resective epilepsy surgery. Methods: Five children were characterized by means of seizure description, V‐EEG, SISCOM, fMRI, and invasive intracranial recording with brain mapping, and seizures were localized to the SMA. Results: Three of the five patients have had resective surgery, with mixed outcomes. None were initially seizure‐free: one has had a few seizures due to non‐compliance but otherwise has gone months at a time without seizures, one required a second resection in the primary motor cortex and is now seizure‐free, and one has had a >50% reduction in seizures. Of the non‐operated patients, one declined surgery and one has been moderately well‐controlled with AED therapy. All patients were non‐lesional on MRI imaging; pathology showed cortical dysplasia in one and non‐specific changes in the other two patients. The patient who required a second operation has a mild L hemiparesis, but had no deficits after the initial resection limited to the SMA, and the other two patients have no neurologic deficits. Conclusions: SMA seizures are not uncommon in children, with characteristic clinical features and seizure semiology like those described in adults. The seizures can be approached surgically with acceptable results and without deficits. Identification and characterization of SMA in children may allow for earlier successful surgical intervention, especially in cases involving focal cortical dysplasia as the underlying cause of the epilepsy. Use of adjunctive methods to localize, including SISCOM and invasive monitoring with stimulation and mapping, is essential for non‐lesional cases. A common confounder appears to be the rapid propagation to the SMA from adjacent cortical areas (primary motor) or simultaneous onset, making surgery more challenging. 1 Joo Hee Seo, 2 Su Jeong You, 2 Hoon Chul Kang, 1 Heung Dong Kim, 3 Tae‐Sung Ko, 4 Yong Soon Hwang, 5 Dong Suk Kim, 6 Jung‐Kyo Lee, and 4 Sang Keun Park ( 1 Department of Pediatrics, Severance Hospital, Handicapped Children's Research Institute, Brain Research Institute, Yonsei University College of Medicine, Seoul, Korea ; 2 Department of Pediatrics, Epilepsy Center, Inje University College of Medicine, Sang‐gye Paik Hospital, Seoul, Korea ; 3 Department of Pediatrics, Asan Medical Center, Ulsan University College of Medicine, Seoul, Korea ; 4 Department of Neurosurgery, Epilepsy Center, Inje University College of Medicine, Sang‐gye Paik Hospital, Seoul, Korea ; 5 Department of Neurosurgery, Severance Hospital, Handicapped Children's Research Institute, Brain Research Institute, Yonsei University College of Medicine, Seoul, Korea ; and 6 Department of Neurosurgery, Ulsan University College of Medicine, Asan Medical Center, Seoul, Korea ) Rationale: We sought to evaluate the long‐term efficacy of vagus nerve stimulation (VNS) in refractory pediatric epilepsy. Methods: Twenty eight patients could be followed over 12 months in three tertiary care referral epilepsy centers. Data including seizure frequencies, quality of life measures and adverse events had been prospectively filed up through 5 years. Results: VNS resulted in a >50% reduction in seizure frequency in 53.6% (15/28) of children with 28.6% (8/28) patients achieving a >90% reduction. When compared a seizure reduction rate according to seizure types (generalized vs paritial seizure), etiologies (symptomatic vs cryptogenic), and seizure duration before VNS implantation, we could not find any significant difference of it. However, a seizure reduction rate had a tendency to be in inverse proportion to seizure duration. In addition, of 11 patients who had partial seizure disorders, 7 patients (63.6%) obtained a reduction of seizure frequency with >50% and then again, of 17 patients who had generalized seizure disorders, 8 patients (47.1%) did. Quality of life was also measured to be enhanced as follows, memory in 50.0% (8/16) patients, mood in 62.5% (10/16) patients, behavior in 68.8% (11/16) patients, alertness in 68.8% (11/16) patients, achievement in 37.5% (6/16) and verbal skill in 43.8% (7/16) patients. Adverse events included hoarseness in 7 patients, dyspnea at sleep in 2 patients, wound infection in 1 patient but were transient and successfully managed by careful follow‐up and an adjustment of parameters. Conclusions: Our data show that VNS roles as one of alternative therapy for refractory pediatric epilepsy without any significant adverse events. 1 Nicolas Specchio, 1 Lucia Fusco, 1 Giuseppe Gobbi, 1 Federico Vigevano, and 1 Tiziana Granata ( 1 Division of Neuroscience, Bambino Gesù Children's Hospital, Rome, Rome, Italy; Division of Neuroscience, Bambino Gesù Children's Hospital, Rome, Rome, Italy; Division of Neuropediatrics, Ospedale Maggiore Cà Pizzardi, Bologna, Bologna, Italy; Division of Neuroscience, Bambino Gesù Children's Hospital, Roma, Roma, Italy; and Division of Child Neurology, Istituto Nazionale Neurologico C. Besta, Milano, Milano, Italy ) Rationale: Rasmussen encephalitis (RE) is an acquired neurological disease characterized by intractable partial seizures and symptoms of progressive dysfunction of one cerebral hemisphere. Laboratory findings suggested the possibility of a immune‐mediate component in the pathogenesis of the disease implying. However, surgical treatment still remains the preferred treatment. We report our experience on the surgical treatment of 10 RE patients collected in the context of an Italian collaborative study. Methods: The series included 10 patients with RE followed in the centers of Bambino Gesù Children's Hopsital (Roma), Neurological Institute “C. Besta” (Milano) and Ospedale Maggiore (Bologna). Pre‐ and post‐operative evaluations included: EEG recordings with video monitoring, MRI, assessment of motor function and cognitive evaluation based on clinical observation and standardized tests. The surgical procedures was based on functional disconnection of the affected hemisphere (hemispherotomy), performed according to vertex or peri‐insular approach. Results: The series includes 10 patients (6 M,4 F), aged between 10 and 23 years (16.3 ± 3.4) observed between 1993 and 2000. The age at onset ranged between 3 and 11 years (6.5 ± 2.8); surgery was performed 8 months to 9 years from disease onset (4.1 ± 2.7). Post‐surgical follow‐up ranged between 5 and 10 years (6.9 ± 1.5). The disease onset was marked by partial seizures, epilepsia partialis continua and unilateral limb paresis followed by status epilepticus. Neurologic deficits appeared in all patients within 24 months. EEG recordings consistently showed slow and epileptic abnormalities over the affected hemisphere. At the time of surgery patients were treated with three or more AED. All, but one, patients became seizure‐free after surgery, AE treatment has been stopped in 7 patients, and reduced in 3. Patients still treated are given a mean of 1.5 ± 1 drugs. The need for neurological evaluations and EEG monitoring reduced from monthly to yearly. Conclusions: After surgery seizures ceased in all patients, thus demonstrating the successfulness of the operative procedures. It is noteworthy that postural control improved in all patients and regained unassisted gait. Mental abilities, as evaluated by standardized scales, improved in all cases. None of them performed anymore habitual monthly EEG and neurological evaluations. Surgical treatment of RE appeared in our series as a gold standard because of its effectiveness in seizure control, dramatic improvement in cognitive evaluations and reduction in the general healthcare assistance. 1 Marino M. Bianchin, 1 Vera C. Terra‐Bustamante, 1 Tonicarlo R. Velasco, 2 Neife H. Deghaide, 1 Hélio R. Machado, 2 Eduardo A. Donadi, and 1 Américo C. Sakamoto ( 1 Department of Neurology, Psychiatry, and Medical Psychology, Ribeirão Preto School of Medicine, University of São Paulo, Ribeirão Preto, São Paulo, Brazil ; and 2 Division of Clinical Immunology, Ribeirão Preto School of Medicine, University of São Paulo, Ribeirão Preto, São Paulo, Brazil ) Rationale: Rasmussen encephalitis (RE) is a rare disorder of unknown etiology. The disease has been attributed to possible viral infections or autoimmune mechanisms. HLA is the landmark of autoimmune disorders and specific alleles have been associated with different neurological diseases. Here we report the results of HLA Class‐Ia study in Brazilian RE patients. Methods: After approval by the Ethic Committee of our institution and after informed consent was obtained from the patients, DNA samples were obtained from 13 Brazilian patients with pathologically confirmed RE. For the study of allele prevalence, data were compared with those for ethnically matched control population. The Fischer test was used to determine possible differences. The level of significance was established at p < 0.05. Results: No specific HLA Class‐Ia was found to be strongly associated with RE in our patients. HLA class Ia genes in Rasmussen Encephalitis HLA A Rasmussen Control p Positive Negative Positive Negative A1 2 24 229 1961 0.89 A2 5 21 264 1926 0.42 A3 3 23 226 1964 0.84 A23 2 24 133 2057 0.73 A24 3 23 227 1963 0.96 A25 2 24 41 2149 0.15 A26 3 23 67 2123 0.16 A30 2 24 124 2066 0.98 A32 1 25 71 2119 0.86 A33 2 24 79 2111 0.56 A69 1 25 5 2185 0.06 HLA B Positive Negative Positive Negative p B7 2 24 153 2161 0.83 B8 2 24 128 2186 0.96 B13 1 25 37 2277 0.90 B14 3 23 117 2197 0.30 B15 1 25 179 2135 0.71 B18 1 25 120 2194 0.76 B35 6 20 274 2040 0.15 B38 2 24 42 2272 0.14 B39 2 24 66 2248 0.38 B44 1 25 269 2045 0.35 B45 1 25 45 2269 0.49 B49 1 25 71 2243 0.82 B51 1 21 200 2114 0.61 B52 2 24 36 2278 0.09 Conclusions: HLA‐Class‐Ia molecules are expressed in most somatic cells including certain neurons, making some neural population potentially susceptible to autoimmune attacks under triggering stimuli from cytotoxic T‐cells. Additionally, HLA Class‐Ia molecules might have other roles in the CNS in terms of neuroplasticity and neuroprotection, or might be involved in normal brain development. In fact, HLA Class‐Ia molecules have been associated with some neurological disorders like schizophrenia, autism, or dyslexia. RE is an autoimmune disorder observed mostly during the years of brain development and associated with neuroplastic alterations that leads to an intriguing form of epilepsia partialis continua. However, in spite of multiple potential roles for HLA molecules in RE, in this group of patients no specific HLA Class‐Ia was found to be strongly associated with RE. We conclude that none of the specific HLA Class‐Ia alleles studied here is particularly involved in RE. However, larger samples involving multicentric studies are necessary to confirm our findings. (Supported by FAPESP.) 1 Mary L. Zupanc, 2 Ellianne Dos Santos Rubio, 1 Rhonda Roell‐Werner, 1 Michael J. Schwabe, 3 Wade Mueller, 3 Sean Lew, 1 Kathy Eggener, and 1 Maria Chico ( 1 Department of Neurology, Medical College of Wisconsin, Milwaukee, WI ; 2 Department of Pediatric Neurology, Erasmus Medical College, Sophia Children's Hospital, Rotterdam, Netherlands ; and 3 Department of Neurosurgery, Medical College of Wisconsin, Milwaukee, WI ) Rationale: To study seizure outcome and determine the correlation between seizure outcome and quality of life measurements in children after epilepsy surgery. Methods: A consecutive, retrospective analysis of seizure outcome and quality of life assessment was conducted in 83 pediatric patients who underwent epilepsy surgery for medically refractory epilepsy at Children's Hospital of Wisconsin between 2002–2005. Data was collected via a chart review and examination of post‐operative patients. Data collection included seizure outcome measurements as described in the Engel classification system. Quality of life assessment was performed by sending quality of life questionnaires to caregivers of our patients (QOLCE and QOLIE‐AD‐48). Results: The seizure outcomes for all surgery types were Class I‐68.7%; Class II‐12%; Class III‐19.3%. If corpus callosotomies were excluded, Class I outcome was 72%. Seizure free outcome was highest in temporal lobectomies (84.2%) followed by hemispherectomies (76.2%), focal resections (61.2%), and finally corpus callosotomies (25%). Temporal lobectomies were more common among older patients. Seizure free outcome was 50% in infants; 67.9% in young children, and 78.4% in adolescents. Cortical dysplasia was predictive of a less favorable seizure free outcome. 62.7% of the QOLCE surveys and 45.8% of the QOLIE‐AD‐48 surveys were returned. The quality of life measures for the surveys paralleled seizure outcome. The measures of physical activity, cognition, social activity, general health, and overall quality of life were significantly better in children with a seizure free outcome than children who were not seizure free. Conclusions: Epilepsy surgery in medically refractory epilepsy in children significantly improves quality of life as well as seizure control. 1 Warren W. Boling ( 1 Neurosurgery, West Virginia University, Morgantown, WV ) Rationale: The author describes a novel keyhole approach to selective amygdalohippocampectomy (SAH). The surgical technique has proven to be successful in the treatment of medically intractable mesial temporal lobe epilepsy (MTLE). Image guidance is required in the keyhole approach to orient the cranial opening, verify cortical landmarks for positioning the corticectomy, and assess the extent of hippocampal disconnection. The restricted surgical exposure has not limited the ability to perform SAH with favorable results on reducing the seizure tendency. Methods: All patients with one year or more follow‐up after keyhole SAH at West Virginia University (WVU) were included in this report. Forty‐eight surgeries for temporal lobe epilepsy have been performed since 2002 by the author with over one year follow‐up. Ten had keyhole SAH performed and are the subjects reported here in. Results of surgery in keyhole SAH subjects were contrasted with individuals undergoing SAH at WVU using standard craniotomy approach. All individuals since 2002 who underwent standard SAH surgery by the author for epilepsy and with over 1 year of postoperative follow‐up were included in the evaluation. Results: Ten subjects had keyhole approach for SAH performed. Seven (70%) had an excellent outcome from surgery becoming seizure free (Class I). One subject had over 90% seizure frequency reduction (Class III), and 2 had less than 90% reduction in seizure frequency (Class IV). Three subjects had no stereoelectroencephalography (SEEG) performed prior to keyhole SAH (all seizure free), and the remainder had SEEG performed for further confirmation of MTLE. There were no surgical complications encountered. Twenty‐One surgeries were performed for SAH using a standard craniotomy approach. 81% were seizure free after surgery. Excellent outcomes compared favorably between keyhole and non‐keyhole approaches (p = 0.54). Hospital stay after surgery was determined for the subjects who underwent keyhole SAH at WVU and an age matched control of patients who had SAH performed via a non‐keyhole (pterional craniotomy) approach at WVU. In keyhole SAH subjects the mean length of hospital stay was 3 days (range 2 – 4 days). Non‐keyhole SAH patients stayed in the hospital a mean of 4 days (range 2 – 6 days) after surgery representing a trend to shorter hospital stays after keyhole SAH (p = 0.11). Conclusions: The keyhole approach to SAH is a strictly image guided procedure. The corticectomy is positioned at the upper bank of T2 anterior to the precentral suclus in the dominant hemisphere and in front of the central sulcus in the non‐dominant hemisphere. These cortical landmarks can only be confirmed with image guidance. The results of surgery on the seizure tendency depend on an accurate diagnosis of MTLE, and compare favorably with non‐keyhole SAH and corticoamygdalohippocampectomy for temporal lobe epilepsy. The keyhole SAH is an effective approach in patients undergoing surgery for MTLE. 1 Richard C. Burgess ( 1 Neurology, Cleveland Clinic, Cleveland, OH ) Rationale: The Cleveland Clinic has a 25 year history as a tertiary referral center for patients with epilepsy. Rigorously kept records of seizure/epilepsy classification and neurophysiologic results are saved electronically. With the advent of a hospital‐wide electronic medical record, picture archiving and communication system, digtal video recordings, and other scattered electronic repositories, data about our patients is entered electronically in many areas, but is not available in an integrated fashion when needed to generate reports to referring doctors or to carry out sophisticated research queries. We sought to put all of this information at the neurologist's fingertips for clinical and research purposes, by developing a system that provides access via standard web browser. Methods: The database contains both text and graphics. Data is imported from other servers, as well as maintained by pointers. For access we employed the Faces framework deployed on a J2EE application server. Because of its deployment in a very busy clinical center, careful attention was paid to real‐time performance issues, e.g. by employing asynchronous Javascript and XML to handle page display simultaneous with data retrieval. In addition to the carefully classified neurophysiologic and epilepsy information, a range of graphical and tabular information has been organized for incorporation into the database. This includes pictorial information, formerly available ad hoc from a variety of sources, such as: grid placement maps, depth placement maps, stimulation maps, ictal and nterictal discharge maps, MRI & PET segments, ictal SPECT, results of volume reconstructions, coregistrations showing grids, intra‐operative photos, video clips, evoked potentials, pathology slides, etc. Results: The application is a complex mix of Java files, JSP files, JSP fragments, Javascript files, configuration files, etc. — over 2000 files and approximately14 megabytes. In addition to ongoing data input the database has been back‐filled with data on more than 65,000 patients from our previous codified textual database. Data output is provided in standardized formats that are familiar to users, easy to work with, and aesthetically pleasing, like HTML and PDF. It is possible to create and view the inpatient evaluation reports in their entirety. New classes of reports (such as for patient management conferences) can be derived from the fundamental information by abstracting and using a built‐in templating engine. Conclusions: Learning from expeience in previous large database development projects, we focused our efforts during this software design on a) obtaining rapid results, quickly visible to and useable by the users, and b) employing modern web‐based tools to speed development and performance. This philosophy has yielded a clinical workflow aid for communication with many disparate sources and for manipulation of data in order to flexibly present results. It has also been designed to include powerful data aggregating tools for research purposes. 1 Hélène Catenoix, 2 Marc Guénot, 1 François Mauguière, 2 Marc Sindou, and 1 Jean Isnard ( 1 Functional Neurology and Epileptology Department, Neurologic Hospital, Lyon, Rhône, France ; and 2 Functional Neurosurgery Department, Neurologic Hospital, Lyon, Rhône, France ) Rationale: Intracranial EEG recordings in presurgical epilepsy assessment are done in our unit using stereotactically implanted depth electrodes. Five years ago, we explored use of these electrodes to produce RFTC lesions inside the epileptogenic area. We report the results in 42 patients. Methods: 42 consecutive patients explored by video‐SEEG recordings were enrolled in this study between June 2001 and February 2006. RFTCs were produced by inducing a Joule effect between electrode contacts where discharge onsets had been recorded. Four to 31 lesions (mean, 12.5) were performed per patient. The epileptic focus (EF) was located in the frontal (8 patients), occipital (7), insular (4), parietal (3) and temporal (20) lobe. Several etiologies were found: cortical dysplasia (CD) (21), hippocampal sclerosis (4), periventricular heteropia (PH) (3), posttraumatic focal atrophy (1) and cryptogenic epilepsy (13). The results were evaluated on the decrease in the number of seizures: good outcome (GO, >70%) and using a self‐administered questionnaire, each patient estimated the presence (a) or absence (b) of improvement in their quality of life (QL). Results: The GO rate 1 month after the intervention (57%) decreased to 35% at the 6‐month evaluation and then stabilized. The QL evaluation showed comparable mean results (a = 38%) but rose to 66% on the subgroup of patients who were inoperable after SEEG exploration. The GO rate was positively correlated with: 1) anterior location of the EF: 75, 37 and 33% of GO for insular, frontal and temporal lobe epilepsy, respectively, with no change for parietal and occipital epilepsy. 2) the etiology, with the best results for symptomatic epilepsy (37 and 33% GO in CD and PH, respectively) vs 23% GO in cryptogenic forms. No general or neurological complication occurred during the procedures other than two transient deficits (paresthetic sensations in the mouth and mild apraxia of the hand). 20 unsatisfied patients had a secondary classical intervention (19 now in Engel class 1 and 1 in class 2) Conclusions: This study reports the results at 5 years of RFTC treatment given to 42 epileptic patients during SEEG exploration. This procedure is: 1) sure: complications are rare, minor and always reversible despite treatment often targeting high‐value functional cortical and/or inaccessible areas 2) effective: benefit in terms of seizure control concerns 1/3+ patients with GO maintained long term from the 6th month after treatment 3) offers improvement in quality of life, especially for the 2/3 of patients considered inoperable after invasive exploration 4) no surgical burden on prognosis: 19/20 patients operated after RFTC failure are in Engel class 1 These results changed our SEEG exploration strategy by combining diagnosis and therapeutic potential until now unknown with this technique. (Supported by HCL.) 1 Stephan Chabardes, 2 Lorella Minotti, 1 Dominique Hoffmann, 1 Eric Seigneuret, 2 Philippe Kahane, and 1 Alim L. Benabid ( 1 Neurosurgery Department, CHU Michallon, Grenoble, France ; and 2 Neurology Department, CHU Michallon, Grenoble, France ) Rationale: Temporal lobe epilepsy (TLE) is the most common form of medically intractable partial epilepsy in adults, and surgery (anterior temporal lobectomy or a selective amygdalo hippocampectomy) is effective in the majority of patients. We have developed since 1998 a new surgical technic to treat non lesional TLE which consists in disconnecting the whole temporal lobe from the insula and surrounding structures. The aim of this study was to assess the safety and results of temporal surgical disconnection of patients suffering from non lesional TLE. Methods: Since 1998, 62 patients suffering from TLE, or extra TLE have been treated by surgical disconnection of the epileptogenic zone. Among this group, 45 patients (mean age: 35 y ± 10) suffered from TLE and underwent temporal disconnection (left and right disconnection in 20 and 25 patients respectively). There was 16 males and 29 females, 12 were left handed and 33 were right handed. All patients underwent a comprehensive pre‐surgical evaluation with MRI, chronic video‐EEG recording and neuropsychological assessment. 16 patients (35%) underwent pre‐surgical evaluation with depth electrodes (SEEG). All patients underwent a post operative MRI, neuropsychological testing, repeated scalp EEG and visual field examination. Temporal disconnection was performed under neuronavigation and consisted in disconnecting the lateral temporal neocortex, the whole hippocampus, para‐hippocampus and the amygdala. Veins and arteries were spared in order to avoid post‐operative temporal lobe ischemia and subsequent temporal lobe swelling. Results: At two years follow‐up, 84.4% of patients were Engel's I (58% were Ia). Four patients were class II and 3 class IV. Post‐operative clinical morbidity was as follow: persistent mild hemiparesia (n = 1) or mild facial paresia (n = 1), definitive quadranopsia (n = 23), hemianopsia (n = 1). In dominant temporal lobe, verbal memory was worsened in 69%. Radiological post‐operative morbidity was as follow: 2 thalamic or pallidal limited ischemia without clinical consequences, 2 temporal horn cystic dilatation (one requested reintervention without clinical consequences). At 6 months and one year follow up, scalp EEG showed temporal spikes respectively in 27% and 19% of non cured patients compared to 8.5% and 6%of patients seizure free. Conclusions: Temporal disconnection was effective in 85% of patients (Engel's Class I). Post –operative mild contro‐lateral motor deficit was found in about 4%, quadranopsia in 51% and verbal memory deficit in dominant hemisphere in about 69%. Presence of spikes in post‐operative scalp EEG was less frequent in patients seizure free. Comparative studies are required to evaluate temporal disconnection as an alternative to lobectomy in non lesional TLE. 1 Anita Datta, 1 Matt B. Wheatley, 1 Tom Snyder, 1 S.N. Ahmed, 1 Donald Gross, and 1 David B. Sinclair ( 1 Comprehensive Epilepsy Program, University of Alberta, Edmonton, AB, Canada ) Rationale: The objective of the study was to review our experience and compare the results of the selective amygdalohippocampectomy in children versus adults with refractory temporal lobe epilepsy. Methods: A retrospective case series was used in the setting of a tertiary care hospital which provides care to both children and adults. All patients underwent a selective amygdalohippocampectomy performed by a single neurosurgeon and had at least one year of follow up. Patient charts were reviewed for pertinent clinical information and the data was compared between children and adults. Outcome measures included seizure frequency, neuropsychological assessment, EEG, neuro‐imaging and anti‐epileptic drugs pre and post‐operatively. Results: Twenty three patients, 9 children and 14 adults were studied. Age at surgery varied from 6 to 58 years. Follow‐up ranged from 1 to 4 years. Surgical outcome was variable between the two groups. Amongst the younger cohort (6–18 years of age), 5 patients had good outcome (Engel Class I or II), two patients had a poor outcome (Engel class III) and two patients had refractory seizures (Engel Class IV) that required returning to the operating room for a standard anterior temporal lobectomy. This differed from the adults, who all had a good outcome. Ten patients were seizure‐free (Engel Class I) and the remainder had rare seizures (Engel Class II). Non‐MTS pathology was more common in children. Thirteen of 14 (93%) had MTS in the adult population, while only 4 of 9 children (44%) had MTS. The pathology at surgery in this group included mesial temporal sclerosis (MTS), brain tumors, focal encephalitis and normal pathology. Conclusions: Selective amygdalohippocampectomy can lead to favorable seizure and neuropsychological outcome in adults with refractory temporal lobe epilepsy. However, preliminary results show less favorable results in children. The difference is probably related to the different pathology between the two groups. Adults most commonly have MTS, whereas children more commonly have non‐MTS temporal lobe pathology. Clinicians should be aware of this difference when considering surgical management in children with refractory temporal lobe epilepsy. 1 Gena R. Ghearing, 1 Drahomira Sencakova, 2 Barbara Schauble, 1 Gregory D. Cascino, 1 Elson L. So, 3 Clifford R. Jack Jr., and 1 Gregory A. Worrell ( 1 Neurology, Mayo Clinic, Rochester, MN ; 2 Section of EEG and Epileptology, Department of Neurology, University Hospital of Berne, Inselspital, Berne, Switzerland ; and 3 Radiology, Mayo Clinic, Rochester, MN ) Rationale: Unilateral hippocampal atrophy is associated with an excellent surgical outcome after temporal lobectomy. The management of patients with bitemporal disease is more controversial and many undergo intracranial EEG (iEEG) with bitemporal depth electrodes. The purpose of this study is to evaluate the usefulness of intracranial monitoring for predicting surgical outcomes in patients with both asymmetric and symmetric hippocampal volumes. Methods: Volumetric hippocampal analysis was performed on the MRIs of all consecutive intractable temporal lobe epilepsy patients between the ages of 16 and 45 years who underwent bitemporal depth electrode recording at our institution over a seven year period. Symmetric hippocampal volumes were defined by a difference between the right and left hippocampus between the range – 0.1 and + 0.3 cm3. The hippocampal volumes were compared to age matched controls, and significant atrophy defined as greater than two standard deviation difference from control. The localization of ictal and interictal discharges on scalp and iEEG were reviewed, and surgical outcome determined for each patient. Results: Of the 27 patients, 14 had asymmetric and 13 had symmetric hippocampal volumes. The surgical outcome was available for 22 patients (four did not have surgery and one was lost to follow up) with an average follow up of 51 months. Ten of 12 surgical patients with asymmetric atrophy (83%) had an Engel class 1 outcome. Nine of these patients with a favorable outcome had significant atrophy on the resected side. The predominant side of seizure onset during iEEG was concordant with the smaller hippocampus in all patients who had significant atrophy. Four out of ten surgical patients with symmetric atrophy (40%) had a favorable outcome. The degree of atrophy did not predict surgical outcome. Two of the patients with a favorable outcome (50%) and three of the patients with an unfavorable outcome (50%) had significant bilateral atrophy. Of the four patients who had a favorable surgical outcome, three had a predominance of seizures arising from the resected side during iEEG, and three of the patients who had unfavorable outcomes also had a predominance of seizures from the resected side. Conclusions: In this limited retrospective study, patients with asymmetric hippocampal atrophy had a high probability (83%) of seizure free outcome. The smaller hippocampus was concordant with the predominant side of iEEG seizure onset in all patients with significant asymmetric atrophy. Conversely, only 40% of patients with symmetric hippocampal atrophy had favorable surgical outcomes, and iEEG did not prove helpful for predicting outcome. 1 Jorge A. Gonzalez, 2 Dileep Nair, and 1 William Bingaman ( 1 Neurological Surgery, Cleveland Clinic, Cleveland, OH ; 2 Neurology, Cleveland Clinic, Cleveland, OH ; and 3 University Heights, OH ) Rationale: Treatment of patients who fail epilepsy surgery is problematic. Selected patients may be candidates for further surgery, potentially leading to a significant decrease in seizures frequency and severity. The purpose of this manuscript is to present our long‐term outcome series of highly investigated patients who failed resective epilepsy surgery and subsequently underwent re‐operative resective procedures. Additionally, our intention is to identify possible factors associated with unsatisfactory seizure outcome, improving selection criteria for re‐operation in failed epilepsy surgery. Methods: We performed a retrospective consecutive analysis of patients who underwent re‐operative procedures due to medically intractable epilepsy at our institution from 1990 to 2001. Seventy patients underwent re‐operative epilepsy surgery with 57 patients having a minimum follow‐up of 2 years. We assessed the relationship between seizure outcome and categorical variables using Chi‐square and Fisher's exact tests, and the relationship between outcome and continuous variables using Wilcoxon rank‐sum test. For all analyses, statistical significance was set at a p value of 0.05. Results: Of the 57 patients, there were 29 males and 28 females. Age of seizure onset ranged from 3 months to 39 years (mean 10.7 ± 10.3 years, median 7 years). The mean age at re‐operation was 24.7 ± 12 years (range 4–50 years). Follow‐up ranged from 24 to 228 months (mean: 128 months, mode: 132 months). Seizure outcome were classified according to Engel's classification and divided in two groups: Favorable outcome (Engel's classes I and II) and non‐favorable outcome (Engel's classes III and IV). Fifty two percent of patients had favorable outcome. Patients with tumors as initial pathology had better outcome in comparison with malformations of cortical development (MCD) and mesial temporal sclerosis (MTS) (p < 0.05). Conclusions: Re‐operation in selected patients failing epilepsy resective surgery should be considered as approximately 50% of patients may have benefit. Patients with MCD and MTS are less likely to improve after re‐operation. 1 Robert E. Gross, 1 Edward K. Sung, 1 David LaBorde, 1 Michele Johnson, and 1 Klaus Mewes ( 1 Department of Neurosurgery, Emory University School of Medicine, Atlanta, GA ) Rationale: Frameless image‐guidance techniques for implantation of depth electrodes for intracranial monitoring are increasingly used, but few reports have assessed the accuracy and precision of these techniques. Methods: Frameless image‐guidance using the Stealth® (Medtronic) neuronavigational hardware and Cranial® software was used exclusively for the implantation of depth electrodes since 1998. Electrodes were implanted through burr holes using an orthogonal (trans‐temporal) approach (n = 32 patients), typically into the amygdala and 2 – 3 electrodes in the hippocampus or parahippocampal gyrus on each side, or occipital approach (n = 3 patients). We analyzed the implantation of all electrodes for which the intra‐operative planning trajectories (saved on the pre‐operative MRI scan by the Stealth) and post‐operative MR imaging were available (n = 155 electrodes: 150 orthogonal, 5 occipital). The pre‐ and post‐operative image sets were registered using the navigational software, and a vector was constructed from the “intended” target to the “actual” location of the electrode tip (taking into account imaging artifact around the electrode). Results: The discrepancy between the “intended” and “actual” location of the orthogonal electrode tips was 5.6 ± 2.9 mm. There was no significant difference between accuracy of electrodes implanted on the first vs. second side as might be expected from brain shift, nor on the left (n = 80) vs. right (n = 70) side. The direction of error could be characterized as “too deep” for 93 electrodes as opposed to “too shallow” for 50 electrodes probably due to difficulty in securing the electrode precisely during placement. Five electrodes implanted via an occipital approach had an error of 7.7 ± 8.0 mm, reflecting lower accuracy and greater variability likely due to the longer passage. No patients implanted via an orthogonal approach required reoperation in contrast to 1 of 3 patients implanted occipitally. No complications definitively related to post‐operative MR imaging were detected over the 150 electrodes and 32 patients, although 2 patients recently implanted experienced transient neurologic deficits which in one patient may have been infectious in etiology but in the other was uncertain. Conclusions: Frameless image‐guidance implantation of depth electrodes via an orthogonal approach is an effective technique with sufficient accuracy for intracranial monitoring. 1 Marc Guenot, 2 Jean Isnard, 2 Philippe Ryvlin, 2 Helene Catenoix, 2 Francois Mauguiere, and 1 Marc Sindou ( 1 Department of Functional Neurosurgery, Hop. P. Wertheimer, Hospices Civils de Lyon, Lyon, France ; and 2 Department of Functional Neurology, Hop. P. Wertheimer, Hospices Civils de Lyon, Lyon, France ) Rationale: Except in case of Mesial Temporal Lobe Epilepsy Syndrom (MTLE), and despite their more recent refinement, the noninvasive techniques still often fail to clearly localize the epileptogenic zone in a large number of patients presenting with drug‐resistant temporal lobe epilepsy. Different techniques of intracranial recordings are currently available. StereoElectroEncephaloGraphy (SEEG), which consists in the stereotactic orthogonal implantation of depth electrodes (5 to 15, 11 on average) into the brain, is the only one allowing to record the mesial as well as the lateral aspects of the temporal lobe, including the depth of the sulci. The object of this paper is to display the curent usefulness of SEEG in terms of surgical strategy related to temporal resection, in 100 consecutive procedures. Methods: 100 patients, suffering from drug‐resistant epilepsy arising from a temporal onset, underwent a SEEG procedure between 2000 and 2005. For each of these patients, the non invasive investigation, which included clinical and neuropsychological examination, long‐term video‐EEG recording, functional imaging, and MRI scan, were not in favor of a MTLE syndrom, and were not congruent enough to clearly identify the epileptogenic focus. Consequently, the sites of implantation of the electrodes were chosen in order to precise: either the side of the onset of seizures, or their uni or multilobar feature, or a possible strictly neocortical or insular onset of the seizures, and also, using direct electrode stimulation, the proximity of speech area. Results: Complications occurred in 3 patients (1 superficial infection, and 2 breakages of electrodes). ‐ SEEG was helpful in most (91%) of the 100 patients to confirm, or cancel, surgical indication, and to adjust the extent of the resection: ‐ In 33% of cases, SEEG‐recordings led to an indication of a strictly extra‐limbic lobectomy, i.e. a tailored (temporo‐basal or lateral) neocortectomy sparing the temporo‐mesial structures. ‐ In 20% of cases, SEEG led to a limited limbic resection, which included the amygdala, as well as the pole, while sparing most of the (non‐atrophic) hippocampus. ‐ In 25% of cases, SEEG allowed to plan an enlarged temporal resection, despite the fact that it showed a wide, or even multiple, epileptogenic zone. ‐ In 13% af cases, SEEG allowed to definitly exclude the patients from surgery. ‐ In 9% of cases, SEEG confirmed the indication of a conventional limbic resection (anterior temporal lobectomy). Conclusions: SEEG proved to be a safe and a very useful method for complex cases candidate for temporal lobe epilepsy surgery. It should be considered as soon as the non invasive investigation do not display a clear MTLE syndrome. 1 Hirochika Imoto, 1 Masami Fujii, 2 Jouji Uchiyama, 1 Nobuhiro Tanaka, 3 Kimihiko Nakano, 1 Sadahiro Nomura, 1 Hirosuke Fujisawa, 2 Takashi Saito, and 1 Michiyasu Suzuki ( 1 Neurosurgery, Yamaguchi University School of Medicine, Minamikogushi, Ube, Yamaguchi, Japan ; 2 Applied Medical Engineering Science, Graduate School of Medicine Yamaguchi University, Ube, Yamaguchi, Japan ; and 3 Mechanical Engineering, Faculty of Engineering Yamaguchi University, Ube, Yamaguchi, Japan ) Rationale: Local brain cooling using an implantable cooling system is an attractive prophylactic technique for seizures since it has a potential to be an alternative to a surgical resection of the epileptogenic foci. In this report, we applied the newly devised local brain cooling system to intractable epilepsy in humans. Methods: We utilized a Peltier chip as the cooling device and therefore applied this cooling device to two patients with medically intractable epilepsy. During surgery, cooling was performed for two minutes in the human cortex where the epileptiform discharges (EDs) were recorded and therefore it had to be resected. Electrocorticograms and the temperature just beneath the cooling site were recorded before and during cooling. Changes of EDs between before and during cooling were quantitatively analyzed. Results: The first case was a 12‐year‐old girl with temporal lobe epilepsy, while the second case was a 2‐year‐old boy with parietal lobe epilepsy caused by tuberous sclerosis. In both cases, EDs diminished during the cooling process when the temperature of the brain surface reached less than 25°C. Conclusions: This is the first report demonstrating the effectiveness of the cooling device using a Peltier chip for the treatment of human epilepsy. Owing to recent advances in the precision machinery industry, an implantable local cooling system in humans is therefore expected to become a reality in the near future. 1 Floor E. Jansen, 1 Alexander C. Van Huffelen, 1 Ale Algra, and 1 Onno Van Nieuwenhuizen ( 1 Neurology, Pediatric Neurology, Clinical Neurophysiology and Clinical Epidemiology Julius Center for Health Sciences and Primary Care, University Medical Center, Utrecht, Netherlands ) Rationale: Cortical tubers are associated with epilepsy and intellectual disability in the majority of patients with tuberous sclerosis. Seizures are often drug resistant. Uncertainties exist concerning the selection of patients suitable for epilepsy surgery. We performed a systematic review of the available literature to assess the overall outcome of epilepsy surgery and to identify factors related to good seizure outcome. Methods: A comprehensive literature search of Medline, Embase, and bibliographies of reviews and book chapters published from 1960 until May 2006 was performed. Studies were included if they at least reported quantitative seizure outcome, and a description of the type of surgery. Seizure outcome was analysed both as seizure freedom and good seizure outcome (seizure reduction of >90%) versus poor seizure outcome. Results: Twentyfive studies, including 177 patients, were found to be eligible. Seizure freedom was achieved in 101 patients (57%). Seizure frequency was reduced by > 90% in another 32 patients (18%). Moderate or severe intellectual disability (IQ < 70) (RR 1.8; 95% CI 1.2–2.8), the presence of tonic seizures (RR 1.7; 95% CI 1.2–2.4) multifocal abnormality on SPECT examination (RR 7.0; 95% CI 1.1–43) and corpus callosotomy (corpus callosotomy versus resective surgery RR 2.5; 95% CI 2.1–3.0) predicted an unfavourable outcome. Conclusions: A good outcome of surgery (seizures reduction of > 90% or complete cure) was achieved in 75% of patients. Resective epilepsy surgery should be considered in all drug resistant TSC patient, as a relation between multiple seizure types with early onset, multiple cortical tubers and multifocal or generalized epileptogenicity and a poor outcome is not supported. (Supported by the Epilepsy Fund of the Netherlands (grant number 02–13, F.E.J.).) 1 Andres M. Kanner, 2 Richard Byrne, 1 Ada V. Chicharro, 1 Marlis Frey, and 1 Michael C. Smith ( 1 Neurological Sciences, Rush University Medical Center, Chicago, IL ; and 2 Neurosurgery, Rush University Medical Center, Chicago, IL ) Rationale: Following an antero‐temporal lobectomy, one third to one half of patients who become free of any disabling seizures continue experiencing auras. The cause of temporal lobe epilepsy (TLE), duration of the seizure disorder and the presence of generalized tonic‐clonic (GTC) seizures have been identified as predictors of post‐surgical outcome folloiwng a temporal lobectomy. A comorbid psychiatric history has also been associated with a worse post‐surgical seizure outcome in these patients. Predictors of persistence of post‐surgical auras in patients free of disabling seizures are yet to be identified. Methods: 97 consecutive patients, 58 men and 39 women with a mean age of 30.6 ± 10.7 years who underwent an antero‐temporal lobectomy at the Rush Epilepsy Center were included in the study. The cause of TLE was: mesial temporal sclerosis in 60 (62%), lesional TLE in 18 (18.5%) and idiopathic TLE in 19 (19.5%). The mean duration of the seizure disorder was 19.7 ± 8.8 years. All patients had undergone a psychiatric evaluation as part of their presurgical evaluation (not out of suspicion of psychiatric disorder). All patients had to have a minimal post‐surgical follow‐up period of 2 years before being included in the study (mean post‐surgical follow‐up: 6.9 ± 3.0 years). The folloiwng variables were entered into a logistic regression model to identify the predictors of achieving a seizure‐free state without auras: cause of TLE, duration of seizure disorder, occurrence of GTC seizures and life‐time history of depression. Results: Among the 97 patients, 37 (38%) were free of any disabling seizures and auras since the time of surgery (Class IA of Engel). An additional 43 patients (44.3%) were free of disabling seizures but had auras (Class I‐B). A life time prevalence of depression of 48.5% (n = 47) was identified among the 97 patients. The absence of a life‐time history of depression was the only variable that predicted the achievemnet of a seizure‐free outcome without auras (p < 0.0001). This model correctly classified 76% the data. A lifetime history of depression continued to be a predictor, but to a lesser degree when the post‐surgical outcomes consisted of freedom of disabling seizures with or without auras (p = 0.001) and freedom of disabling seizures in the last two years (p = 0.06). On the other hand the cause of TLE became a predictor in the setting of these two outcomes (p = 0.04 and p = 0.009, respectively). Conclusions: A presurgical history of depression appears to predict the peristence of auras after an antero‐temporal lobectomy in patients that become free of diabling seizures. The pathogenic mechanisms that mediate thsi phenomenon are yet to be identified. 1 Keith W. MacDougall, 1 Jorge G. Burneo, 1 Richard S. McLachlan, 1 Andrew G. Parrent, and 1 David A. Steven ( 1 Department of Clinical Neurological Sciences, University of Western Ontario, London, ON, Canada ) Rationale: In many patients being considered for epilepsy surgery, a standard pre‐surgical evaluation consisting of a scalp electroencephalogram (EEG), magnetic resonance imaging (MRI), functional neuro‐imaging and a neuropsychological assessment is sufficient to identify the epileptogenic zone. When this evaluation cannot identify the epileptogenic focus, invasive intracranial electrodes are often required. Although there is considerable data about the outcome of patients undergoing epilepsy surgery, most of these investigations examine all patients (with and without invasive recordings) together. This purpose of this study was to specifically examine patients who underwent invasive EEG and to determine how many of these patients eventually underwent resective surgery. In those who had surgery, we wished to examine the eventual outcome. Methods: We retrospectively reviewed charts for all patients admitted to the London health Sciences Centre Epilepsy Monitoring Unit between January 2000 and July 2004 for intracranial electrode monitoring. Surgical outcomes were reported using the Engel classification. We used a multivariate analysis to determine which factors if any were associated with successful surgery. Results: There were 103 subdural electrode implantations performed by our two epilepsy surgeons in 100 patients. Of these, 78 patients went on to have resective surgery. Temporal lobectomy was performed in 39 (50%) patients. One‐year follow‐up was available on 66 of the 78 patients (85%). Thirty patients (45%) realized an Engel I outcome at 1 year. An additional 20 (30%) had their seizure frequency significantly reduced. Predictors of an Engel I outcome included younger age (p < 0.0001) and having a temporal lobectomy (adjusted OR = 4.17; CI = 1.12 – 12.8). Factors that did not significantly affect outcome included the duration, area, and location of electrode coverage, the presence of a pre‐operative lesion or the final pathology. Conclusions: Excellent results from eventual resective surgery can be achieved in the subset of patients needing invasive recordings. Younger patients with temporal lobe epilepsy seem to have the highest likelihood of seizure freedom. The need for bilateral or multilobar electrodes did not negatively affect prognosis. 1 Taketoshi Maehara, 1 Tadash Nariai, 1 Toshiya Momose, 1 Chihiro Hosoda, 2 Kenji Ishii, 2 Kiich Ishiwata, and 1 Kikuo Ohno ( 1 Neurosurgery, Tokyo Medical and Dental Universuty, Tokyo, Japan ; and 2 Positron Medical Cente, Tokyo Metropolitan Institute of Gerontology, Tokyo, Japan ) Rationale: With the advent of intraoperative navigation systems, structural lesions disclosed by MRI can now be removed more easily and accurately. But in cases receiving epilepsy surgery, the seizures can only be stopped by removing the surrounding epileptogenic areas together with the lesions themselves. To accomplish this, we need to functionally detect peri‐lesional epileptogenic areas during surgery. Our group recently introduced a method of F18‐fluorodeoxyglucose (FDG) and C11‐flumazenil (FMZ) PET co‐registration as a reference for the detection of epileptogenic areas during image‐guided epilepsy surgery for patients with organic lesions. In this study we investigated the clinical role of multimodal navigation epilepsy surgery using FDG‐PET and FMZ‐PET. Methods: We studied a series of 10 patients who suffered from intractable epilepsy and were found to have organic lesions on MRI. All of the subjects underwent FDG‐PET and FMZ‐PET studies before their operations. Surgical specimens demonstrated cavernous angiomas in 2 patients and benign tumors in 8 (3 low‐grade astrocytomas, 2 DNTs, 2 gangliogliomas, and 1 monomorphous angiocentric glioma). The lesions were located in the mesial temporal lobe (mT) in 6 patients and in the extra‐mesial‐temporal lobe (extra‐mT) in 4. Our first steps were to examine the relationships between the organic lesions and hypometabolic areas on FDG‐PET, and then to compare the FDG‐hypometabolic areas and low‐uptake areas on FMZ‐PET. Next, we performed intraoperative electrocorticography (ECoG) in and around the hypometabolic areas to examine the epileptogenicity of the hypometabolic areas on FDG‐PET and the low‐uptake areas on FMZ‐PET. Results: 1) The hypometabolic areas on FDG‐PET were much wider than the low‐uptake areas on FMZ‐PET in 8 cases, and slightly wider in 3 cases. 2) Spikes on intraoperative ECoGs were recorded in the hypometabolic areas in all 6 of the mT cases. 3) The hypometabolic areas were much wider than the ECoG‐spiking areas in 2 of the 4 extra‐mT patients, and slightly wider in the other 2. ECoG‐spiking areas were wider than the low‐uptake areas on FMZ. 4) Four of the 6 mT patients underwent anterior temporal lobectomy and the other 2 underwent lesionectomy along with anterior temporal resection. The extra‐mT patients underwent lesionectomy along with resection of any ECoG‐spiking areas which exhibited hypometabolism on FDG‐PET and low‐uptake on FMZ‐PET. 5) All 10 patients became seizure‐free during follow‐up periods ranging from 6 to 24 months. Conclusions: The use of multi‐modal navigation surgery by FDG‐PET and FMZ‐PET in combination with intraoperative ECoG may become an important method for focus resection in epileptic patients with organic lesions, especially in patients with extra‐mesial temporal lobe epilepsy. 1 Michiharu Morino, 1 Yumiko Urano, 1 Masaki Yoshimura, and 1 Kenji Ohata ( 1 Neurosurgery, Osaka City University Graduate School of Medicine, Osaka, Japan ) Rationale: Although selective amygdalohippocampectomy was developed in epilepsy surgery to spare unaffected brain tissue minimizing the cognitive consequences of temporal lobe surgery, several previous studies have shown that it was equivocal whether selective amygdalohippocampectomy for intractable epilepsy could lead to a better memory outcome or not. The aim of the present study was to investigate the effects of selective surgery to short‐term memory outcome in patients of intractable mesial temporal lobe epilepsy who underwent transsylvian selective amygdalohippocampampectomy (TSA). Methods: The 51 patients received comprehensive neuropsychological testing of verbal and nonverbal memory, attension, and delayed recall before and 1 months after unilateral TSA. Results: Significant improvements in verbal memory and delayed recall were found particularly for the right resected group. For right TSA, the outcome of verbal memory was not related to hippocampal pathology, however, the risk for a nonverbal memory decline was increased when patients had non hippocampal sclerosis. After left TSA, there were no significant decline and it indicated that all aspects of memory were preserved after surgery although the risk for a verbal memory decline was increased of patients without hippocampal sclerosis. Memory before and 1 month after TSA Right TSA p values Left TSA p values Preop. 1MP Preop. 1MP Memory Verbal 74.9 (14.5) 87.1 (18.5) p < 0.001 77.8 (18.5) 73.8 (16.1) NS Non verbal 81.0 (17.5) 87.2 (16.1) NS 90.8 (17.6) 92.1 (21.1) NS General 73.3 (14.8) 84.6 (18.1) p < 0.001 78.9 (18.6) 76.2 (17.3) NS Delayed Recall 72.0 (15.9) 78.8 (20.0) p = 0.0015 75.2 (19.8) 72.4 (17.6) NS Attension 83.7 (16.6) 86.8 (16.7) NS 91.2 (16.4) 90.8 (20.8) NS Cells provide the mean standardized scores with the standard deviations in parentheses. NS: not significant (p>0.05). MP: month postoperative. Conclusions: The results indicate that left TSA can lead to preserve the memory function, particularly right TSA can lead to a significant improvements in memory functions. 1 William Murphy, 1 Terry Myles, 1 Walter Hader, 1 Neelan Pillay, 1 Lorie Hamiwka, 1 Elaine Wirrell, 1 Paolo Federico, 1 Nathalie Jette, and 1 Samuel Wiebe ( 1 Clinical Neurosciences, University of Calgary, Calgary, AB, Canada ) Rationale: Reports on the diagnostic accuracy of structural MRI in detecting putative lesions in patients undergoing epilepsy surgery vary substantially, and are probably influenced by MRI protocols, types of lesions, and expertise. Using histopathology as the gold standard, we assess the overall sensitivity and specificity of structural MRI in an unselected cohort of patients undergoing epilepsy surgery, and provide post‐tests probabilities for negative and positive studies. Methods: In a retrospective cohort of consecutive cases undergoing resective epilepsy surgery, we identified those who had histopathological tissue examination,ie., the gold standard. MRIs protocols contained a combination of general and epilepsy‐specific sequences, and experts and non‐experts performed MRI interpretation. The presence or absence of a putative lesion on structural MRI, using epilepsy specific protocols, was compared with the presence or absence of a histopathological lesion. We obtained sensitivity (true positive rate), specificity (true negative rate), positive and negative likelihood ratios, and post‐test probabilities using a range of plausible pre‐test probabilities. We also obtained 95% confidence intervals around each estimate. Factors influencing the diagnostic accuracy were assessed when the sample size allowed. Results: Of 135 consecutive patients evaluated for epilepsy surgery, 117 had MRI and histopathological analyses. Histopathological abnormalities occurred in 104 patients (prevalence of 89%), of whom the MRI found a lesion in 93, yielding a false negative rate of 11%, and a sensitivity of 89% (95% CI = 83%, 95%). Of 13 patients with normal histopathology, the MRI showed a lesion in eight, with a false positive rate of 62%, and a specificity of only 38% (95% CI = 12%, 64%). The likelihood ratios for a positive and a negative MRI were 1.44 (low) and 0.3 (medium‐high), respectively. Indicating that in this analysis, the MRI seems to be more useful when negative, because its false positive rate was too high. When the chance of finding a lesion is 50%, a positive MRI increases the probability to only 59%, and a negative MRI decreases is to 22%. Conclusions: In unselected patients undergoing epilepsy surgery Structural MRI has a sensitivity of almost 90% and a specificity of only 38%. The causes for the high false positive rate of MRI in our cohort are explored. Reference Von OJ, Urbach H, Jungbluth S, et al. Standard magnetic resonance imaging is inadequate for patients with refractory focal epilepsy. J Neurol Neurosurg Psychiatry 2002;73:643–647. 1 Yu‐tze Ng, 2 Norman C. Wang, 2 Steve C. Chung, 3 Erin C. Prenger, 1 Harold L. Rekate, 2 Iman Feiz‐Irfan, and 1 John F. Kerrigan ( 1 Pediatric Neuroscience, Barrow Neurological Institute/St. Joseph's Hospital, Phoenix, AZ ; 2 Neurology, Barrow Neurological Institute, Phoenix, AZ ; and 3 Neuroradiology, Barrow Neurological Institute, Phoenix, AZ ) Rationale: Hypothalamic hamartomas (HHs) are rare developmental abnormalities of the inferior hypothalamus that often cause refractory, symptomatic, mixed epilepsy. Multiple surgical approaches have been described. We present a series of 39 patients where a transcortical, transventricular, endoscopic resection was performed. Methods: Forty‐two patients with refractory epilepsy underwent surgical resection of HH between October 2003 and April 2005 at our institution, i.e. all patients with more than 12 month follow‐up. All patients underwent endoscopic resection and were prospectively studied. Neurosurgery was performed with the endoscope held by an articulated pneumatic arm and tracked with a frameless stereotactic neuronavigation system. Data collection and follow‐up was performed by personal interview. Three patients were excluded because they underwent combined surgical approaches (subfrontal and transcallosal). Results: The average age of the patients was 15.8 years (range 8 months to 55 years). There were 24 males. All 39 patients had very frequent seizures, usually multiple types. The average number of each patient's total daily seizures was 18.2 (range 0.1–192) and an average of 2.2 seizure types each. 33 (85%) patients had trademark gelastic seizures at the time of surgery. The patients were on an average of 2.0 antiepileptic drugs (AEDs). 25 patients had either cognitive and/or behavioral impairment. Postoperative MRI demonstrated 100% resection and disconnection of the HH from the hypothalamus in 12 patients. There were no mortalities. Significant morbidity was limited to one patient with residual hemiparesis. Two patients were lost to follow‐up. At latest follow‐up (average 20.1 months (range 13 to 28 months)), 18/37 (48.6%) patients were seizure‐free, 26/37 (70.3%) patients had >90% reduction in their seizures and another 8/37 (21.6%) of patients had 50–90% seizure reduction. Fourteen patients were off all AEDs. Twenty‐five (67.6%) patients reported either improved cognition and/or behavior. The percentage of disconnection trended toward seizure freedom (p = 0.069). The percentage of resection did not significantly correlate with seizure freedom (p = 0.11). In an endoscopic resection group of 33 patients, the average length of post‐surgical hospital stay was 5.8 days (SD = 3.1) as compared to 18 patients who underwent transcallosal resection and stayed for 8.7 days (SD = 4.7). This overall difference i.e. shorter hospital stay amongst the endoscopically resected patients was statistically significant (p < 0.03). Conclusions: Endoscopic resection of HH is safe and effective for the treatment of seizures. Its efficacy appears to be comparable to transcallosal resection of HH, however with significantly shorter patient recovery time post‐operatively. 1 Tomonori Ono, 1 Keisuke Toda, 1 Shiro Baba, 1 Hiroshi Baba, and 2 Kenji Ono ( 1 Neurosurgery, National Nagasaki Medical Center, Omura, Nagasaki, Japan ; and 2 Yokoo Hospital, Isahaya, Nagasaki, Japan ) Rationale: Surgery for bilateral temporal lobe epilepsy (Bi‐TLE), herein defined as seizures arising independently from both temporal lobes, has been demonstrated to have poor efficacy. However, lateralized MRI abnormality and a preponderance of interictal spikes on EEG have been identified as factors for good surgical outcome. Ictal EEG findings are also a decisive factor for determining the resection side. Predominant ictal onset side (PIOS) is considered to be a direct measure of laterality of ictogenisity, however, this has yet to be quantitative analyzed. In this study, we reviewed our surgical series of Bi‐TLE patients and determined the correlation between quantitatively assessed PIOS and postoperative outcome. Methods: We analyzed data from 23 patients with Bi‐TLE who had undergone anterior temporal lobectomy. Patients' ages at the time of the operation ranged from 14 to 55 years (mean 31.2 years). Diagnoses of Bi‐TLE were confirmed using subdural electrode recordings, and the side of resection was determined by PIOS. Postoperative seizure outcome was assessed using Engel's outcome classification between 9 and 161 months (mean 80 months) after surgery. Percentages of ictal onset side calculated on the basis of all recorded seizures and other clinical characteristics were correlated with postoperative seizure outcome. Results: Regarding postoperative seizure outcome, 9 patients (39.1%) were class I, 8 patients (34.8%) were class II, and 6 patients (26.1%) were class III. The following items were found to be statistically significant factors for good surgical outcome: 1) no history of severe encephalitis (p = 0.03), 2) higher percentage of PIOS (class I vs. class II‐III, 87.6 ± 3.1% vs. 73.7 ± 3.3%, p = 0.009), and 3) lateralized lesion (p = 0.01), or absence of bilateral abnormalities on MRI (p = 0.02). Furthermore, multivariate analyses showed that factors 1) and 2) were independent predictors of good outcome. Conclusions: Surgery for Bi‐TLE results in a greater reduction of seizures if preoperative evaluations show good predictive factors. In particular, when the percentage of PIOS is greater than 85%, a more favorable outcome can be obtained. 1 Andre Palmini, 1 Eliseu Paglioli, 1 Jaderson Costa, 1 Caroline Torres, 1 Jose Victor Martinez, 1 Eduardo Portela, 1 Helio F. Silva‐Filho, 1 Eduardo Paglioli, 1 Mirna Portuguez, 1 Jose‐Rubiao Hoeffel‐Filho, 1 Ligia Coutinho, 1 François Dubeau, and 1 Frederick Andermann ( 1 Department of Neurology and Neurosurgery, Pontificia Universidade Católica do RS, Porto Alegre, RS, Brazil; and Department of Neurology and Neurosurgery, McGill University, Montreal, QC, Canada ) Rationale: Incomplete resections are not uncommon in surgery for epilepsy due to FCD, however, the dynamics of long‐term surgical outcome following these resections have not been investigated in detail. We report on the variable patterns of seizure recurrence following incomplete resections for FCD and relate these to the histopathologic type. Methods: Forty‐two children and adult patients with FCD identified through MRI and/or confirmed on histopathology were followed for a mean of 54 months. In 25, a complete resection of the dysplastic tissue was not possible in a first operation. They all underwent preoperative evaluation and surgery was guided by acute ECoG and cortical electrical stimulation. Residual lesion was identified either at the end of operation or at post‐op MRI. The degree of seizure control was evaluated periodically, and classified according to Engel's outcome scale. FCD lesions were classified histopathologically in types I, IIA, and IIB. Results: Three recurrence patterns were identified: (i) in 6 patients, seizures recurred within 7 days of operation, with higher frequency and intensity than in the preoperative period. Three needed ICU admission for status epilepticus; (ii) 9 presented recurrent seizures during the first post‐op month and continued seizing with similar frequency as before surgery over the years or until reoperation. Finally (iii) 10 patients had a post‐op seizure free interval between 3 months and 2 years, following which seizures recurred. Histopathologic types IIA and IIB were similarly represented in all 3 recurrence patterns. Conclusions: Lasting seizure freedom was incompatible with incomplete lesion resections in FCD types II A and B. Interference with inhibitory networks associated with dysplastic lesions may explain the patterns of running up of seizure frequency and intensity, particularly early post‐resection severe deterioration leading to status epilepticus. (Supported by FAPERGS.) 1 Tarek A. Radwan, 1 David W. Roberts, 1 Vijay M. Thadani, 1 Peter D. Williamson, 1 Richard P. Morse, 1 Ann‐Christine Duhaime, and 1 Barbara C. Jobst ( 1 Neurosurgery & Neurology, Dartmouth‐Hitchcock Medical Center, Lebanon, NH ) Rationale: Interhemispheric electrodes (IHE's) are increasingly used in epilepsy surgery, but the use of IHE's has been associated with heightened concern about risk of bridging vein disruption & haemorrhage. We hypothesize that the use of IHE is safe and indispensable in localizing epilepsy originating from the mesial surface of the brain. Methods: We performed a retrospective review of complication rate, type of resection, EEG findings and surgical outcome of all patients who had undergone interhemispheric intracranial electrode monitoring. The interhemispheric location was anterior in 22 (69%), and posterior in 10 (31%). Results: 21 patients had IHE implanted. 17 had interhemispheric grids (curvilinear‐doublesided Adtech 2 × 8 or 3 × 8 grids), four had 1 × 8 or 1 × 4 strip electrodes placed interhemispherically. One patient had two interhemispheric 3 × 8 grids implanted. The interhemispheric location used was anterior in 22 (69%), and posterior in 10 (31%). Complications as a result of IHE's implantation consisted of temporary leg weakness in two patients which resolved over several days. No hemorrhagic complications or other complication related to electrode implantation were observed. Of all the patients implanted with IHE's, 20 (95%) underwent resective surgery, with 12 (57%) undergoing mesial resection. Out of those 12 patients, 7 (58%) had normal MRI, while 5 (41%) had a lesion identified on MRI. In the seven patients with normal MRI, localization and resection was only possible due corticography obtained from IHE. In the remaining patients with abnormal MRI IHE determined extent of resection. Four patients had not clearly defined developmental abnormalities on MRI and in one patient with a tumor IHE clearly localized seizure onset zone. Of the 9 patients that did not localize mesially 8 (89%) underwent resection after localizing to other cortical areas. In those patients IHE were helpful in excluding mesial seizure onset. One patient (11%) did not localize and did not undergo any resection. Outcome by Engel Class for all patients that had IHE's implanted and underwent resection was: Class I‐ 13 (Ia‐12/Ib‐1) (62%); Class II‐ 3 (14%); Class III‐ 2 (10%); Class IV‐ 3 (14%) Conclusions: Interhemispheric intracranial monitoring electrodes are a safe and effective tool in epilepsy surgery. 1 Fortino Salazar, 2 Imad Najm, 3 Prakash Kotagal, 1 Ann Warbel, and 1 William Bingaman ( 1 Neurosurgery/Epilepsy Surgery, Cleveland Clinic Foundation, Cleveland, OH ; 2 Neurology/Adult Epilepsy, Cleveland Clinic Foundation, Cleveland, OH ; and 3 Pediatric Neurology/Epilepsy, Cleveland Clinic Foundation, Cleveland, OH ) Rationale: Subdural grids (SDG) have been increasingly used in the evaluation of some patients with pharmacoresistant epilepsy. In a small number of patients the placement of SDG is not followed by a surgical resection. There is a paucity of data about these patients and the causes behind the lack of success in their invasive evaluation. The purpose of this study is to retrospectively review the data on patients who were evaluated at our center with SDG. Methods: Adult and pediatric patients who underwent the placement of SDG, followed by prolonged video‐EEG monitoring but did not have resective surgery were identified from the Cleveland Clinic database and their medical records were reviewed. A total of 233 patients underwent SDG placement between January 1997 and June 2005. The placement of SDG and prolonged invasive recordings did not lead to resective surgery in 29 (12%) patients who were included in this review. Mean age was 25 years (7 were < 18 years old). There were 15 males and 14 females. Average duration of seizures was 15 years. The MRI was interpreted as normal in 12 patients, other MRI findings included malformation of cortical development, schinzencephaly and focal brain atrophy. All patients had PET studies and when indicated a Wada test. The number of electrodes used varied between 56 and 160. Average monitoring duration was 8 days (between 4 and 19). Results: Reasons for NOT pursuing surgical treatment were: a) Identification of multiregional or non‐localizable epileptogenic foci (12 pts), b) presence of epileptogenic foci in eloquent cortex (8 pts), c) lack of adequate coverage of the suspected ictal onset zone (5 pts), or d) the lack of seizure recording after SDG (2 pts). In two patients the evaluation was terminated because of surgical complications. Following the SDG evaluation, 6 patients had VNS implanted and 5 others are considering this option. Three patients are considering another invasive evaluation, one had a resective surgery at a later time, and 7 patients are only considering medical treatment. Conclusions: Approximately 12% of patients were not suitable candidates to undergo surgical resection. The two most common reasons were involvement of eloquent cortex and evidence of diffuse epileptogenic zone. It is important to refine or develop better ways to identify this group of patients prior to implantation, such as the use of fMRI, MEG and sophisticated EEG to avoid costly and risky procedures. 1 Kostas N. Fountas, and 1 Joseph R. Smith ( 1 Department of Neurosurgery, Medical College of Georgia, Macon, GA ) Rationale: Implantation of subdural strip and grid electrodes represents a common methodology in the invasive evaluation of patients with medically refractory epilepsy. Although in the vast majority of cases their implantation is safe, the occurrence of implantation‐associated complications can be occasionally troublesome. The purpose of our current communication is to identify the most commonly occurred complications and calculate their frequency in our institution. Methods: In a retrospective study, 185 patients (104 males and 81 females) undergoing subdural strip/grid electrode implantation during a 20‐year period, were examined. Their age ranged between 16 and 48 years, while their mean age was 23.6. AdTech (AdTech, Racine,WI,USA) strip and grid electrodes were used in all our cases. The anatomical location of implantation were:26 implants in orbito‐frontal area, 41 in mesial frontal,14 in mesial occipital, 13 in antero‐basal area, 44 in interhemispheric areas while 180 in convexity areas (4 temporo‐polar, 83 parietal, 17 occipital and 76 temporal). The monitoring period ranged from 2–25 days (mean 10.8 and median 7 days). Results: The most commonly occurred complication in our series was the development of post‐operative epidural hematoma in 3/185 (1.6%) patients, requiring emergent surgical evacuation. In 2/185 (1.1%) cases transient aphasia occurred, which was spontaneously resolved. A surgical subdural hematoma developed in 2/185 (1.1%) while in another 2/185 (1.1%) an infection occurred. Interestingly, in 2 patients (1.1%) significant brain edema was post‐operatively evident; one of these patients was conservatively managed with no further consequences while the other one succumbed despite aggressive management of his intracranial hypertension. Finally, one patient developed post‐operatively stiff lung syndrome secondary to aspiration and eventually expired. Another interesting observation in our series was the recording of non‐habitual seizures in 5/185(2.7%) patients. Conclusions: Subdural electrode complications although rare can be troublesome and occasionally disastrous. Knowledge, early identification and prompt management of potential complications can minimize the risks associated with the surgical implantation of subdural electrodes. 1,2 William O. Tatum IV, 2 Kumar Anthony, 2 Leanne Heriaud, 1,2 Selim R. Benbadis, 2 Adele Haber, and 1,3 Fernando L. Vale ( 1 Neurology, University of South Florida, Tampa, FL ; 2 Neurology, Tampa General Hospital, Tampa, FL ; and 3 Neurosurgery, University of South Florida, Tampa, FL ) Rationale: Epilepsy Surgery is a standard of care in the treatment of intractable epilepsy. Intracranial electrodes can help identify surgical candidates when discordant information is present from the extracranial evaluation. We analyzed our experience with intracranial electrodes over 6 years. Methods: Two‐hundred and fifty‐eight resective epilepsy surgeries were performed from 2000–2006, with a mean follow‐up of 2.6 years (range 2–65 months). Twenty‐nine females and 21 males with a mean age of 29.4 years were implanted using an electrode array based upon the results of a standardized, non‐invasive, pre‐surgical evaluation. All patients were implanted at a single tertiary care epilepsy center between 2000–2006 by one neurosurgeon. Discordant information following a non‐invasive evaluation including scalp‐based video‐EEG monitoring was obtained in each case. Resective surgical procedures, implanted array, and outcome were assessed. Outcome was reported as seizure‐free, seizure improved (>50% of baseline), and no improvement, and was based upon information obtained at the time of last contact. Results: Fifty epilepsy patients (19.4%) were implanted with intracranial electrodes. Twenty‐four patients (48%) had bitemporal placement, 15/50 (30%) were unilateral, and 11/50 were bilateral‐extratemporal (22%). Thirty‐seven (74%) underwent resective epilepsy surgery, and 13/50 (26%) were rejected as surgical candidates. Seizure freedom or improvement was noted in 9/14 (64.3%) implanted with unilateral electrodes, 14/24 (58.3%) with bitemporal electrodes, and 4/10 (40.0%) with bilateral‐extratemporal electrodes. Resections were temporal in 27/37 (73.0%), and extratemporal in 10/37 (27%) including frontal in 6/37 (12.5%), multi‐lobar in 2/37 (5.4%), parietal in one (2.7%), and occipital in one. At last follow‐up, 21/27 (81.4%) of the temporal resections were either seizure free (33.3%) or seizure improved (48.1%). Only 5/27 (13.5%) failed temporal surgery, compared with 3/8 (37.5%) extratemporal resections. MRI demonstrated an abnormality in 19/50 (38%) with neuronal migrational disorders accounting for the most common cause. The presence of a lesion did not have a more favorable outcome (8/19; 42.9%) than those with cryptogenic causes (18/29; 62.1%) (p = 0.24, Fisher's exact test). Conclusions: Most epilepsy patients requiring intracranial electrodes will be candidates for resective surgery. Bitemporal electrode placement for lateralization, and temporal resections were the most frequent surgeries performed. Patients implanted only with bitemporal electrodes, were not more likely than those with a lateralized array to benefit from resective surgery. Those with lesional epilepsy had no more favorable outcome than those with a cryptogenic etiology in our series. 1 Samuel Wiebe, 1 Walter Hader, 1 Neelan Pillay, 1 Terry Myles, 1 Lorie Hamiwka, 1 Elaine Wirrell, 1 Paolo Federico, 1 Nathalie Jette, and 1 William Murphy ( 1 Clinical Neurosciences, University of Calgary, Calgary, AB, Canada ) Rationale: There is a notion that patients whose epilepsy is associated with a putative lesion have better surgical outcomes than those without such lesions. However, cases are often selected and other prognostic often remain unexplored. We explore seizure outcomes in unselected cases of lesional versus non‐lesional epilepsy surgery, as well as the variables that impact outcomes in the Calgary Comprehensive Epilepsy Program. Methods: A retrospective cohort study of consecutive, unselected cases assessed post‐surgical seizure freedom at the time of last follow‐up in patients whose MRI showed and in those in whom it did not show putative lesions. We used survival analyses for the main comparison and single and multivariate analyses to explore factors that may influence outcome, such as age, duration of epilepsy, etiology, type of surgery and location of surgery. Results: Of 113 consecutive patients, 91(81%) had putative MRI lesions. An equal proportion (70%) of lesional and non‐lesional cases had temporal lobe surgery. Patients without lesions were significantly more likely to have a family history of epilepsy (p = 0.03) and they had a trend towards a younger age at surgery. Two years after surgery 57% of lesional and 36% of non‐lesional cases were seizure free (p = 0.07). As compared with temporal lesional cases, extratemporal lesional cases were younger at presentation (20 vs 38 years) and at surgery (17 vs 35 years) (p < 0.01), and their seizure free rate was lower (not significant). Preoperative deficits were more frequent in non‐lesional extratemporal than in temporal cases. Of patients with frontal resections, most were non‐lesional (p = 0.02). Conclusions: In unselected patients undergoing epilepsy surgery those without MRI lesions have overall poorer seizure outcomes, are younger (especially extratemporal cases), more often have a family history of epilepsy, and more often have frontal lobe epilepsy. 1 Charles M. Zaroff, 1 Deepak Madhavan, 2 Alexis Arzimanoglou, 2 Florence Renaldo, 1 Josiane LaJoie, 1 Howard L. Weiner, 3 Eva Andermann, 4 Frederick Andermann, 4 Francois Dubeau, 4 Andre Olivier, 4 Alexei Yankovsky, 5 David N. Franz, 5 Jennifer Leonard, 6 Mary Connolly, 7 Greg D. Cascino, and 1 Orrin Devinsky ( 1 Department of Neurology, NYU Medical Center, New York, NY ; 2 Child Neurology & Metabolic Diseases Department, Hôpital Robert Debré, Paris, France ; 3 Department of Neurology and Neurosurgery and Human Genetics, Montreal Neurological Institute, Montreal, QC, Canada ; 4 Department of Neurology and Neurosurgery, Montreal Neurological Institute, Montreal, QC, Canada ; 5 Department of Neurology, Cincinnati Children's Hospital Medical Center, Cincinnati, OH ; 6 Division of Neurology, Vancouver BC Children's Hospital, Vancouver, BC, Canada ; and 7 Department of Neurology, Mayo Clinic, Rochester, MN ) Rationale: Tuberous Sclerosis Complex (TSC) is associated with medically refractory epilepsy. Epilepsy surgery, with the goal of resecting epileptogenic brain tissue for seizure relief, has become a viable therapeutic option in TSC. The purpose of the current study is to determine factors associated with seizure relief postoperatively in TSC patients utilizing multicenter data. Methods: A retrospective analysis of patients with TSC following surgery for relief of epilepsy was performed from multiple centers. Seventy subjects were analyzed. We studied patient characteristics (presence/absence of mental retardation, age at TSC diagnosis), electrophysiological characteristics (seizure onset age, ictal/interictal focality), seizure type (i.e., presence/absence of infantile spasms), tuber number, electrophysiological/neuroimaging concordance, surgical variables (age at surgery, surgical method, and number of stages), and postoperative residual dominant tuber as determined by neuroimaging. Data were analyzed for all 70 subjects as a whole and for subjects assessed two years or less postoperatively and more than 2 years postoperatively. Results: 27/70 subjects were seizure free (Engel Class I). Subjects with ictal multifocality, mental retardation, and discordant EEG/MRI data were less likely to achieve seizure freedom. However, only associations between seizure freedom and age at seizure onset, history of infantile spasms, and interictal focality were significant at all postoperative durations. The mean age at seizure onset was 2 years in seizure free subjects compared to 7 months in those with persistent seizures. Only 5/32 subjects with a history of infantile spasms were seizure free postoperatively. Focal interictal discharges were seen in 85% of seizure free subjects versus 32% of subjects with persistent seizures. The number of tubers was significantly associated with poor outcome when measured less than 2 years postoperatively. Conclusions: In patients experiencing seizure freedom following surgery for epilepsy in TSC, ictal focality, normal development, and concordant electrophysiological and neuroimaging data are common findings. However, similar features can also be observed in subjects with postoperative seizures. The current study suggests that later age at seizure onset, no history of infantile spasms, interictal focality, and fewer cortical tubers may be determining factors in seizure freedom. 1 Carolina R. Almeida, 2,3 Kette D. Valente, 2 Silvia Vincentiis, 4 Juan Castro, 4 Hung T. Wen, 1 Renato L. Marchetti, 1 Wagner F. Gattaz, and 1 Elida B. Ojopi ( 1 Laboratory of Neuroscience, Psychiatry Institute, USP, Sao Paulo, SP, Brazil ; 2 Laboratory of Neurophysiology, Psychiatry Institute, USP, Sao Paulo, SP, Brazil ; 3 LIM 21, Psychiatry Institute, USP, Sao Paulo, SP, Brazil ; and 4 Neurosurgery, USP, Sao Paulo, SP, Brazil ) Rationale: Mesial temporal lobe epilepsy with hippocampal sclerosis is one of the most common medically intractable epilepsy syndromes and its pathogenesis remains obscure. The polymorphisms of single nucleotide (SNPs), alterations in a single nucleotide of DNA sequence, when non‐synonymous, are capable of cause changes in protein's structure and have been studied in association studies. Recent studies demonstrated controversial results about the relationship between gene polymorphism and epilepsy. Several association studies involving more than 30 genes involved in Mesial Temporal Sclerosis (MTS) are described. However, it remains the uncertainty in associations and the imperfection in replications. The objective of this study is the molecular characterization of MTS. Methods: Thirty patients with MTS, well characterized by MRI, were pared by sex and age with 30 controls without family history of neurological or psychiatric disorders to be genotyped. The authors investigated, through search of new DNA SNPs, susceptibility genes to developmental of MTS. Initially, more than 100 genes coding to ionic channels and other genes related to neuronal migration and neurogenesis were evaluated. From these, 82 genes were selected. These genes were previously described in association studies of epilepsy or studies of families. To the moment, we found 20 non‐synonymous polymorphisms in 11 different genes. Results: In the initial experimental analysis, 14 SNPs find in 5 different genes were evaluated. The data show that for the first 30 patients, 8 SNPs were confirmed in 5 different genes (EFHC1, ME2, BRD2, BDNF and TMEM1). Conclusions: The number of patients and controls limited the statistical power of the ongoing study. Knowledge of these SNPs and their effect on genes involved with MTS may be important to help a better understanding of complexes pathways that characterize the neuronal hyperexcitability and the development of new strategies for treatment and diagnosis. (Supported by FAPESP 05/50135–3.) 1 Danielle M. Andrade, 1 Stephen W. Scherer, and 1,2 Berge A. Minassian ( 1 Department of Medical and Molecular Genetics, The Hospital for Sick Children, Toronto, ON, Canada ; and 2 Department of Paediatrics, The Hospital for Sick Children, Toronto, ON, Canada ) Rationale: Protein transduction domains (PTDs) may be capable of transducing full‐length fusion proteins across plasma cell membranes and the blood‐brain‐barrier (BBB). They are therefore considered a promising tool for the development of protein therapy aimed at the central nervous system. In this study we analysed the characteristics of TAT‐PTD genetically fused to cystatin B (CSTB), the protein responsible for the Progressive Myoclonus Epilepsy of Unverricht‐Lundborg disease (ULD). Methods: A PCR product containing the EPM1 coding region was cloned into the pRSET vector containing the TAT‐PTD sequence. A control protein was assembled by insertion of EPM1 cDNA into this vector, followed by removal of TAT‐PTD. TAT‐PTD‐CSTB and CSTB were expressed in bacterial cells, purified and transduced into COS‐7 cells and human lymphoblasts. Whole cell lysates and cellular fractions were obtained for study by polyacrylamide gel electrophoresis and immunoblotting. Transduced (and control) cells were also used for analysis through immunofluorescence microscopy. Results: Initial results pointed to an apparent time and concentration‐dependent transduction of TAT‐PTD‐CSTB. However, careful investigation demonstrated that the nuclear localization of TAT‐PTD‐CSTB was an artifact of fixation and that the fusion protein was adhered to the plasma membrane. Conclusions: TAT‐PTD‐CSTB does not penetrate the cells. For this reason, even if TAT‐PTD is likely able to cross the BBB, the fusion of this protein to CSTB cannot be used as a form of replacement of the intracytoplasmic protein missing in ULD. Importantly, we discuss precautions to avoid false‐positive results when working with TAT‐PTD for protein therapy of neurological diseases. 1,2 Anna‐Kaisa Anttonen, 2 Erika Ahonen, 3 Eija Gaily, 3 Elina Liukkonen, 4 Marjo‐Riitta Aine, 5 Eila Herrgård, 3 Liisa Metsähonkala, 6 Heikki Rantala, 2 Auli Sirén, and 2 Anna‐Elina Lehesjoki ( 1 Department of Medical Genetics, University of Helsinki, Helsinki, Finland ; 2 Folkhälsan Institute of Genetics and Neuroscience Center, University of Helsinki, Helsinki, Finland ; 3 Department of Pediatric Neurology, Hospital for Children and Adolescents, Helsinki University Central Hospital, Helsinki, Finland ; 4 Department of Child Neurology, Lapland Central Hospital, Rovaniemi, Finland ; 5 Department of Pediatrics, Division of Child Neurology, Kuopio University Hospital, Kuopio, Finland ; and 6 Department of Pediatrics, University of Oulu, Oulu, Finland ) Rationale: Mutations in SCN1A are associated with many epilepsy syndromes such as severe myoclonic epilepsy of infancy (SMEI) and generalized epilepsy with febrile seizures plus (GEFS+). Mutations in SCN2A2 have been described in benign familial neonatal‐infantile seizures and GEFS+ but to date only one SCN2A2 mutation has been associated with intractable epilepsy. We evaluated the contribution of SCN1A and SCN2A2 mutations to infantile onset intractable epilepsy in Finnish patients. Methods: SCN1A and SCN2A2 were screened for mutations in 33 patients, 21 of which had a clinical suspicion of SMEI. We amplified all exons and exon‐intron boundaries of SCN1A and SCN2A2 from genomic DNA and sequenced the PCR products with an ABI 3730 DNA Analyzer. Sequences were analyzed using the Sequencher 4.5 program. Results: We found six different de novo mutations in SCN1A: One nonsense, two splice site, one frameshift‐causing and two missense mutations that change highly conserved amino acids. Two of the mutations were previously described. The clinical presentation in three of the mutation positive patients is SMEI. Two patients have borderline SMEI and one has a milder phenotype with normal cognitive development. We also found one previously undescribed de novo missense mutation of a conserved amino acid in SCN2A2 in a patient with infantile onset epilepsy and episodic ataxia. None of the mutations were observed in 192 control chromosomes. Conclusions: Our findings are compatible with previous findings showing that de novo SCN1A mutations are a major cause of SMEI. SCN2A2 mutations are found only in a small subset of infantile onset intractable epilepsy. 1 Christie Barker‐Cummings, 1 W. Allen Hauser, and 1 Ruth Ottman ( 1 Sergievsky Center and Epidemiology Department, Columbia University, New York, NY ) Rationale: Epilepsy and febrile seizures are complex disorders with both distinct and overlapping genetic influences. Febrile seizures are associated with an increased risk of epilepsy (individual comorbidity). In addition, each disorder aggregates in families, and an increased risk of each disorder has been reported in the relatives of probands with the other. No single study has comprehensively examined the contributions of shared and distinct genetic influences to these observations. We aimed to evaluate the overlapping genetic contributions to epilepsy and febrile seizures by studying patterns of co‐occurrence, within individuals and within families, in a large genetic epidemiologic study. Methods: The study population comprised 2279 full siblings of 879 probands with epilepsy from the Epilepsy Family Study of Columbia University. Data on febrile seizures were obtained by interview of at least one of the proband's parents. We evaluated the risks of epilepsy in siblings without febrile seizures, and febrile seizures in siblings without epilepsy, in relation to the proband's febrile seizure history and family history of febrile seizures and epilepsy in other first‐degree relatives. We considered four types of family history: epilepsy alone, febrile seizures alone, both epilepsy and febrile seizures, and neither. We used Cox proportional hazards models to compute rate ratios (RRs) and 95% confidence intervals (CI) for epilepsy, and logistic regression to compute odds ratios (ORs) for febrile seizures. Results: Epilepsy risk in siblings was not associated with a history of febrile seizures in the proband (RR = 0.5, CI 0.17–1.35). As expected, epilepsy risk was associated with a family history of epilepsy in other relatives (RR = 3.6, CI 1.78–7.47). Epilepsy risk was not associated with a family history of febrile seizures alone (RR = 1.4, CI 0.47–3.85), but was strongly associated with a family history of both febrile seizures and epilepsy (RR = 8.9, CI = 3.92–20.05). The risk of febrile seizures in siblings was associated with a history of recurrent febrile seizures in the proband (OR = 2.3, CI 1.17–4.60), but not with a history of a single febrile seizure in the proband. As expected, febrile seizure risk was associated with a family history of febrile seizures (OR = 4.3, CI 2.54–7.39). Febrile seizure risk was also associated with a family history of epilepsy alone (OR = 2.3, CI 0.99–5.16). As in the analysis of epilepsy, febrile seizure risk was strongly associated with a family history of both epilepsy and febrile seizures (OR = 6.1, CI 2.46–15.18). Conclusions: These results provide evidence for familial clustering of epilepsy and febrile seizures that is not attributed to individual comorbidity of the two disorders. Familial clustering was strongest in families containing multiple individuals with epilepsy. The findings are reminiscent of the syndrome GEFS+, which has previously been described only in highly selected pedigrees. (Supported by NIH R01 NS43472 and an Epilepsy Foundation predoctoral fellowship.) 1 Eylert Brodtkorb, and 2 Fabienne Picard ( 1 Department of Neurology and Clinical Neurophysiology, St. Olav's Hospital, Trondheim, Norway ; and 2 Department of Neurology, University Hospital and Medical School of Geneva, Geneva, Switzerland ) Rationale: Autosomal dominant nocturnal frontal lobe epilepsy (ADNFLE) has been linked to mutations in nicotinic acetylcholine receptors (nAChR) in many families. A beneficial effect of transdermal nicotine has been reported in one single case. We aimed to investigate the influence of tobacco habits on seizure activity in two families with ADNFLE. Methods: Interviews focusing on the pattern of nicotine intake and seizure activity were performed in 22 adult mutation carriers belonging to two ADNFLE pedigrees with different mutations in the gene coding for the nAChR α4 subunit. Results: Ten of 14 tobacco consumers were seizure free, whereas all seven non‐smokers with manifest ADNFLE had persistent seizures. In several patients, remissions and fluctuations of epileptic symptoms clearly appeared to correspond to changes in tobacco habits. Furthermore, one pharmacoresistant patient who had recently begun treatment with transdermal nicotine had experienced improvement. Conclusions: Tobacco appears to be an environmental factor which influences the course of ADNFLE. An inactivation of the mutant nAChRs by desensitization during chronic nicotine exposure may explain the beneficial effect. In pharmacoresistant patients with ADNFLE, treatment with transdermal nicotine should be considered, weighing efficacy against potential long‐term harmful effects. Prospective studies are needed. (Supported by Swiss National Foundation n° 3100A0–104190/1 (Fabienne Picard).) 1 Russell J. Buono, 2 Michael R. Sperling, 3 Dennis J. Dlugos, 1 Michael D. Privitera, 4 Jacqueline A. French, 5 Warren Lo, 6 Steven C. Schachter, 7 Patrick Cossette, 8 Hongyu Zhao, 8 Zeny Feng, 1 Nancy J. Collins, 4 Theresa Scattergood, 9 Wade H. Berrettini, and 9 Thomas N. Ferraro ( 1 Neurology, University of Cincinnati, Cincinnati, OH ; 2 Neurology, Thomas Jefferson University, Philadelphia, PA ; 3 Neurology, Children's Hospital of Philadelphia, Philadelphia, PA ; 4 Neurology, University of Pennsylvania, Philadelphia, PA ; 5 Neurology, Columbus Children's Research Institute, Columbus, OH ; 6 Neurology, Beth Israel Deaconess Medical Center, Boston, MA ; 7 Neurology, CHU Hospital Notre Dame, Montreal, QC, Canada ; 8 School of Public Health, Yale University, New Haven, CT ; and 9 Psychiatry, University of Pennsylvania, Philadelphia, PA ) Rationale: The ABCB1 gene encodes P‐Glycoprotein, a membrane transporter of xenobiotics and natural molecules in many tissues. Of seven studies reporting association tests between variation in ABCB1 and response to anti epileptic drugs (AEDs), three showed positive association, while four did not. Two papers showed association in a case control design, suggesting ABCB1 variation is related to seizure susceptibility. We report the first association study on a cohort of patients and controls from the US and Canada. Methods: Genotyping single nucleotide polymorphisms (SNPs) c1236t and c3435t was performed using TaqMan assays. Samples were collected at 7 sites and all subjects (ages 3–70) were recruited according to local institutional review board policies. Blood samples were obtained and most are available as cell lines from the Coriell Institute. Both markers were typed in 297 generalized patients (70 AED resistant), 253 focal patients (88 AED resistant) and 184 controls, all of European ancestry. Hardy‐Weinberg equilibrium (HWE) was checked and alleles tested for association using chi square tests. Linkage Disequilibrium (LD) and haplotype effects were calculated using likelihood ratio and CHAPLIN respectively. AED response was defined as one year of seizure freedom since last checkup. Results: All markers were in HWE and the SNPs were in strong LD, p = 1 × 10−3. Single marker analyses showed a lack of association for drug response or seizure susceptibility. However, an association was observed between AED response and the CC haplotype in generalized patients (p = 0.007). Haplotypes CC and TT were strongly associated with seizure susceptibility comparing all cases to controls p = 1 × 10−7. Conclusions: We conclude that variation in ABCB1 is weakly associated with AED response in our generalized patients and is strongly associated with seizure susceptibility in all patients. The same haplotpye that confers drug resistance also increases risk for seizures suggesting a relationship between susceptibility and AED response. (Supported by NIH R01NS40396 to RJB, The University of Cincinnati Department of Neurology and The Neuroscience Institute, and The Center for Neurobiology and Behavior at the University of Pennsylvania.) 1 Gianpiero L. Cavalleri, 1 Mike E. Weale, 1 Kevin Shianna, 2 John Lynch, 2 Rinki Singh, 3 Kevin Murphy, 1 Roderick Radtke, 1 James McNamara, 4 Samuel Berkovic, 4 Ingrid Scheffer, 4 John Mulley, 2 Nicholas Wood, 5 Chantal Depondt, 5 Kristl Claeys, 3 Colin Doherty, 3 Norman Delanty, 2 Sanjay Sisodiya, and 1 David Goldstein ( 1 Institute of Genome Sciences and Policy, Duke University, Durham, NC ; 2 The Department of Clinical and Experimental Epilepsy, Institute of Neurology, University College London, London, United Kingdom ; 3 The Division of Neurology, Beaumont Hospital, Dublin, Ireland ; 4 Department of Medicine, The University of Melbourne, Melbourne, Australia ; and 5 Department of Neurology, Universite Libre du Bruxelles, Bruxelles, Belgium ) Rationale: Although great success has been experienced in mapping genes that harbour causative variants for familial forms of epilepsy, little progress has been made in identifying variants that contribute to the development or treatment of more complex forms of the condition. Several key shortcomings in the study design of previous studies have been identified such as insufficient sample sizes, insufficient coverage of genes and lack of replication of detected associations. Here we attempt to integrate these factors in to the largest genetic mapping study to date in epilepsy. Methods: Through the establishment of a multinational collaborative effort, a collection of 3107 epilepsy patients phenotyped for the principal seizure and syndrome types and 1797 healthy controls was assembled. Two hundred and seventy nine candidate genes were selected. For each of these genes bioinformatically‐inferred functional variants were identified. Tagging SNPs were also selected for 198 of the 297 genes. In total 4608 SNPs were genotyped using the Illumina Golden Gate platform across all 4904 participants. Results: Preliminary analyses show an enrichment for significant p values (p < 0.05) but no one p value is strong enough to withstand correction for multiple testing. Conclusions: This result suggests that the data contain a number of risk factors but the effect of any one of these risk factors is marginal. Data analyses will be enhanced by additional genotyping currently ongoing. (Supported by: This study was funded by the Medical Research Council grant G0400126 and the Programme for Human Genomics funded by the Higher Education Authority of Ireland.) 1 Fabio F. Conte, 1 Patricia A.O. Ribeiro, 1 Lourenco Sbragia, 1 Rovilson Gilioli, 1 Fernando Cendes, and 1 Iscia L. Cendes ( 1 Department of Medical Genetics, Universidade Estadual de Campinas, Campinas, São Paulo, Brazil ) Rationale: Missense mutations in the mRib72–1/Efhc1 gene have been identified to co‐segregate with juvenile myoclonic epilepsy (JME) patients. In addition, functional studies have demonstrated that expression of mRib72–1/Efhc1 gene induces apoptosis in neurons in culture. However, the exact relationship between mRib72–1/Efhc1 function and epileptogenesis is still unclear. The objective of this study was to determine the expression pattern of mRib72–1/Efhc1 gene in mouse brain in order to investigate its possible role in programmed cell death during development. Methods: Balb/c mice were submitted to programmed mating to obtain animals in different developmental stages. The brains of embryos aged 15, 17 and 18 days old and of neonates aged 1, 7, 14 and 28 days old were subsequently removed. We used brain of three animals for each age studied. After RNA extraction, the relative gene expression of mRib72–1/Efhc1 was determined by the real‐time PCR technique using TaqMan™ assay. GAPD gene expression was used as an endogenous control. Results: Preliminary results demonstrate that the expression of mRib72–1/Efhc1 is higher in embryos than in neonates (p = 0.001), with a progressive decrease from embryos aged 18 days to neonates. Conclusions: The highest expression of the mRib72–1/Efhc1 gene in embryo brains in comparison to neonates suggests that the putative pro‐apoptotic role of this gene may be necessary for the correct organization of the central nervous system during its initial phases of development. (Supported by CAPES – Coordenacao de Aperfeicoamento de Pessoal de Ensino Superior.) 1 Yaman Z. Eksioglu and 1 Masanori Takeoka ( 1 Neurology/Epilepsy, Children's Hospital Boston, Harvard Medical School, Boston, MA ) Rationale: The Aristaless‐related homeobox gene, ARX, is a transcription factor with a crucial role in the development of forebrain, pancreas and testes. Located in the X chromosome (Xp22.13), ARX mutations cause a pleiomorphic array of phenotypes such as mental retardation, lissencephaly, infantile spasms, and abnormal genitalia in males. Based on human genetics and loss‐of‐function studies in mice, ARX appears to be crucial in the regulation of GABAergic interneuron development and maintenance of circuitry. Ohtahara syndrome consists of epileptic encephalopathy, suppression‐burst EEG‐pattern and refractory epilepsy with tonic spasms in young infants. Despite reports of neonatal epilepsy, Ohtahara syndrome has not been previously described in infants with ARX mutations. Methods: We describe a boy with Ohtahara syndrome who presents with refractory seizures since day‐1‐of‐life caused by a novel mutation in the ARX‐gene. His initial presentation with startle and jerking movements, evolve into brief tonic seizures, refractory to multiple antiepileptic drugs including phenobarbital, zonisamide and levetiracetam. Despite a normal MRI, EEG reveals suppression‐burst‐pattern. He is normocephalic without dysmorphism, except for micropenis and bilateral cryptorchidism, in addition to diffuse hypotonia without weakness. Later, he develops infantile spasms with partial response to vigabatrin. Results: ARX nucleotide sequence analysis, by Athena diagnostics (Worcester, MA), including entire coding regions and intronic sequences of exon‐intron splice junctions, revealed that the proband was hemizygous for a novel mutation comprising 1 base pair insertion of C after nucleotide position 1471, resulting in a frameshift at codon 491. Analysis of mother's DNA revealed that she was heterozygous for the same mutation in ARX‐gene, hence, demonstrating her as a carrier(City of Hope Clinical Molecular Diagnostic Laboratory, Duarte, CA). Conclusions: This is the first reported case of Ohtahara syndrome as consequence of a novel mutation in exon 5 of the ARX‐gene. Insertion of a C after position 1471 in the ARX gene leads to a frameshift that completely changes the structure of the ARX protein from amino acid 491, affecting the composition of the C terminal end. This leads to the disruption of the aristaless‐domain, which is known to play crucial role in central nervous system regionalization with respect to its potential role in transactivation activity. The molecular pathology observed in this novel mutation reflects the complexity associated with ARX protein structure and function. The aristaless domain, localized to exon 5 in the C terminal end of the ARX protein, stands out as an important region in ARX protein signaling. Hence, further genetic/molecular studies are necessary to better understand the function of the aristaless‐domain and its association with the Ohtahara syndrome in our patient with this novel mutation. 1,2,3 Marco Fedi, 1,4 Leon Bach, 1,3 Samuel F. Berkovic, 1,3 Ingrid Scheffer, and 1,5 David C. Reutens ( 1 Department of Medicine, The University of Melbourne, Austin Health, Heidelberg, Victoria, Australia ; 2 Clinical School, The University of Melbourne, Austin Health, Heidelberg, Victoria, Australia ; 3 Department of Neurology, Austin Health, Heidelberg, Victoria, Australia ; 4 Department of Endocrinology, Alfred Hospital, Monash University, Melbourne, Victoria, Australia ; and 5 Southern Clinical School, Monash University, Clayton, Victoria, Australia ) Rationale: The hypothalamo‐pituitary system is innervated by cholinergic fibers and activation of the neuronal nicotinic acetylcholine receptors (nAChRs) modulates the release of anterior pituitary hormones. Mutations of nAChRs subunits are linked to autosomal dominant nocturnal frontal lobe epilepsy (ADNFLE), a syndrome characterized by clusters of nocturnal frontal seizures with no overt endocrinological manifestations. Here we assessed the effect of central cholinergic activation on the release of anterior pituitary hormones in individuals with ADNFLE. Methods: Thirteen subjects with the α4‐Ser248Phe mutation (4 male, mean age 43.2 ± 16.8) and 41 healthy volunteers (24 male, mean age 36.2 ± 12.2) were studied. Serum levels of GH, LH, FSH, PRL, TSH, thyroxine and cortisol were measured at baseline, 30 and 60 minutes after infusion of physostigmine. Results: In controls, cholinergic activation increased the serum levels of GH (mean change + 993%, p < 0.001), PRL (169%, p < 0.001) and reduced the levels of FSH (−7.9%, p < 0.002). In contrast, in subjects with the a4‐Ser248Phe mutation no changes in the release of these hormones were observed (GH 104% p > 0.2; PRL 5.2%; p > 0.8, FSH – 0.8% p > 0.2). Physostigmine increased cortisol levels both in controls (p < 0.01) and in subjects with the mutation (p < 0.03). Between subjects repeated measure ANOVA showed a selective difference in the release of GH and PRL (time*group interaction GH p < 0.04; PRL p < 0.01) following administration of physostigmine. Conclusions: The α4‐Ser248Phe mutation selectively affects cholinergic‐mediated GH and PRL secretion, a finding that may be of pathophysiological significance given the role of somatostatin in experimental models of focal epilepsy. 1 Tracy A. Glauser, 1 Katarzyna Bryc, 1 Todd G. Nick, 1 Donald L. Gilbert, 1 Ericka Holt, 1 Steven Fordyce, and 2 Frank R. Sharp ( 1 Pediatric Neurology, Cincinnati Children's Hospital, Cincinnati, OH ; and 2 Neurology, University of California at Davis, Sacramento, CA ) Rationale: Specific genomic profiles have been noted in the blood of patients with various neurological disorders including Down syndrome, Neurofibromatosis, Tuberous Sclerosis, and Tourette syndrome. This study aimed to determine if specific genomic profiles existed in the blood of children with newly diagnosed and untreated idiopathic epilepsy compared to normal children. Methods: Patients with untreated idiopathic generalized epilepsy (IGE), untreated idiopathic partial epilepsy (IPE) or healthy children with no history of epilepsy (NML) were enrolled in this IRB approved study. RNA was extracted from whole blood and gene expression was determined using Affymetrix U133 2.0 PLUS microarray chips. Using microarray expression data, an analysis of variance (ANOVA) model with epilepsy type (IGE v. IPE v. NML), gender and age (where age was dichotomized over the median of 10 years) was used to identify genes (with a 90% confidence of having false discovery rate, FDR < 0.1) that differentiated between the three groups (IGE, IPE and NML). Pairwise comparisons were performed to identify significantly differentiated genes between groups. Gene ontology (GO) analysis was performed on the IGE and IPE groups from the genes that differentiated between the three groups. GO classes and parent classes with at least 5 observations in the selected subset and with an Observed vs. Expected ratio of at least 2 were reported. Results: A total of 79 children were enrolled; 21 with IPE, 30 with IGE and 28 NML patients. After adjustment for age and gender, 460 genes (with a 90% confidence of having FDR < 0.1) were differentially expressed among the epilepsy groups. Pairwise comparisons showed: IGE v. NML had 301 significantly differentiated genes, (overall p = 0.002), IPE v. NML had 308 significantly differentiated genes (overall p = 0.003) and IPE v. IGE had 26 significantly differentiated genes (overall p < 0.001). From the 460 genes which differentiated the IGE v. IPE v. NML with FDR < 0.1, two gene ontologies were found in the comparison of IGE v. IPE, where the 61 most significant genes contained no more than 5 false discoveries with 90% confidence. The two gene ontologies were intracellular non‐membrane‐bound organelle (GO id 0043232) and non‐membrane‐bound organelle (GO id 0043228) each with Observed vs. Expected ratios of 2.67. Conclusions: Specific genomic profiles exist in children with idiopathic generalized epilepsy and idiopathic partial epilepsy that differentiate from normal controls and from each other. These profiles may provide insight into the underlying process and may represent an additional tool for proper diagnosis and classification of pediatric epilepsy. (Supported by NIH‐R21‐NS044956.) 1 Alica M. Goldman, 2 Daniel L. Burgess, 3 Richard A. Gibbs, 1,4 Kevin E. Chapman, 2,4 Angus A. Wilfong, 1,5 Richard A. Hrachovy, 1,5 Perry J. Foreman, 1 Ian L. Goldsmith, 1 Amit Verma, 2 Marcia J. Hill, and 2 Jeffrey L. Noebels ( 1 Peter Kellaway Section of Neurophysiology, Department of Neurology, Baylor College of Medicine, Houston, TX ; 2 Department of Neurology, Baylor College of Medicne, Houston, TX ; 3 Human Genome Sequencing Center, Baylor College of Medicne, Houston, TX ; 4 Department of Pediatrics, Baylor College of Medicine, Houston, TX ; and 5 Michael E. Debakey VA Medical Center, Houston, TX ) Rationale: Mutations of ion channel genes are recognized as an important cause of defined inherited syndromes in epilepsy and cardiac arrhythmias. We are reporting preliminary data on the following; (1) the prevalence of cardiac co‐morbidities and (2) the frequency and spectrum of genetic variation in the five major genes associated with long QT syndrome in patients with idiopathic epilepsies (IE). Methods: 153 patients with IE were systematically recruited and detailed phenotypic data regarding epilepsy and cardiac arrhythmias were collected. DNA samples were analyzed by sequencing across five major LQT genes (SCN5A, KCNQ1, KCNH2, KCNE1, and KCNE2). Results: Routine ECG was performed in 41/153 individuals. The study was abnormal in 13/41 (32%) subjects. 9/41 (22%) patients had both an abnormal ECG and a medical history positive for syncope, cardiac palpitations or skipped beats. 2/153 (4%) patients were lost due to sudden unexplained death in epilepsy (SUDEP) within one year of follow‐up. Interim DNA sequence analysis of the five major LQT genes in the IE population showed that 27% of all the LQT single nucleotide polymorphisms (SNPs) were exon bound variants of which 45% was represented by protein altering/non‐synonymous SNPs (NS‐SNPs). NS‐SNPs were most commonly found in the SCN5A (33%) and KCNH2 (43%) genes. SCN5A and KCNH2 NS‐SNPs were also the most frequent genetic variants found in patients with dual phenotype of cardiac arrhythmias and epilepsy. Conclusions: The prevalence of inter‐ictal cardiac arrhythmias in our cohort of idiopathic epilepsy patients was similar to the prevalence of ictal arrhythmias cited in the literature ((39%) by Nei et al.Epilepsia.2004;45(4):338–345, and (35%) by Opherk et al. Epilepsy Res.2002;52(2):117–27). The presence of relatively frequent non‐synonymous coding variants in the SCN5A and KCNH2 genes in IE patients with a cardiac phenotype suggests their involvement in the comorbid dual clinical phenotype of cardiac arrhythmias and seizures. (Supported by NINDS 5K08NS47304–3 and NS049130.) 1 An Jansen, 2 Marijke Bauters, 3 Marjan De Rademaeker, 1 Linda De Meirleir, 2 Guy Froyen, 4 Jean‐Pierre Fryns, and 4 Hilde Van Esch ( 1 Department of Child Neurology, AZ Vrije Universiteit Brussel, Brussels, Belgium ; 2 Human Genome Laboratory, Centre for Human Genetics, VIB, Leuven, Belgium ; 3 Medical Genetics, AZ Vrije Universiteit Brussel, Brussels, Belgium ; and 4 Centre for Human Genetics, University Hospital Gasthuisberg, Leuven, Belgium ) Rationale: Mutations in the cyclin‐dependent kinase‐like 5 gene (CDKL5) have been detected in girls with the atypical Rett syndrome variant (MIM 312750) that is characterized by early‐onset infantile spasms and severe neurodevelopmental retardation. We present a young boy with early‐onset infantile spasms, severe encephalopathy, tetralogy of Fallot, and bilateral cataract, in whom a deletion at Xp22 comprising the Nance Horan Syndrome (NHS) gene and CDKL5 gene was detected by array‐CGH. Methods: Brain MRI, video EEG, metabolic and infectious work‐up, karyotype, subtelomeric screening, and array‐CGH were carried out in the proband. Results: The boy was diagnosed shortly after birth with a Tetralogy of Fallot, mild facial dysmorphism, extreme hypotonia, and bilateral microphthalmia with microcornea and congenital nucleocortical cataract. At age 7 weeks he developed myoclonic seizures, occurring most frequently post‐prandial or during drowsiness. EEGs during wakefulness and sleep were normal. At age 4 months, the seizure pattern was consistent with infantile spasms, but no hypsarrhythmia was seen on EEG. At age 5 months, he developed tonic spasms, sometimes preceded by crying or accompanied by laughter or oral automatisms. A video EEG at age 6 months showed multiple myoclonic jerks without EEG correlate, as well as several tonic spasms characterized by sudden onset low voltage rhythmic diffuse 15 Hz activity followed by generalized slowing. No typical epileptiform activity was seen. Seizures were resistant to treatment with valproic acid, vigabatrine, topiramate, phenobarbital and levetiracetam. At age 10 months, he is severely retarded with no reaction to light stimuli, severe axial hypotonia, and absent midline organization. Additional investigations were normal, except for a de novo deletion of approximately 2.8 Mb at Xp22 including 16 genes of which the only known disease‐associated genes are the NHS gene that causes the ophthalmologic abnormalities and the CDKL5 gene that is responsible for the severe epileptic encephalopathy. Conclusions: This is the first description of a male patient with a deletion of the NHS and CDKL5 genes, showing the involvement of CDKL5 in severe epileptic encephalopathy in males. The epilepsy phenotype was characterized by drug‐resistant early‐onset infantile spasms, tonic spasms and myoclonic jerks, without clear epileptiform EEG correlate. (Supported by: H.V.E. is a post‐doctoral researcher of the FWO‐Vlaanderen. M.B. is a doctoral student of the IWT‐Vlaanderen. This work was supported by a research grant G‐0229–01 of the FWO‐Vlaanderen, Belgium.) 1 Ruth Ottman, 1 Kay Kamberakis, 2 Edyth Wiggs, 2 Pat Reeves‐Tyer, 2 Eva H. Baker, 2 John A. Butman, 2 Sadat Shamin, and 2 William H. Theodore ( 1 G.H. Sergievsky Center, Columbia University, New York, NY ; and 2 Clinical Epilepsy Section and Diagnostic Radiology Department, National Institutes of Health, Bethesda, MD ) Rationale: Autosomal dominant partial epilepsy with auditory features (ADPEAF) is a form of idiopathic lateral temporal lobe epilepsy frequently caused by mutations in the leucine‐rich, glioma inactivated 1 (LGI1) gene. Subjects with epilepsy in families with LGI1 mutations have not generally been found to have mesial temporal sclerosis or other structural brain abnormalities detected on MRI. However, in one previously reported family with an LGI1 mutation, a lateral temporal malformation was identified in 10 individuals. This result was important because it strongly supported the concept that LGI1 influences risk through a neurodevelopmental mechanism. We attempted to replicate these findings in other families. Methods: We studied 15 subjects (10 affected, 2 clinically unaffected mutation carriers, 3 unaffected non‐carriers) from six ADPEAF families, each of which had a different LGI1 mutation. Subjects underwent high‐resolution structural MRI on a Philips 3‐tesla scanner, including MPRAGe, FLAIR, t1 and t2 sequences, and neuropsychological testing with a special emphasis on language measures. MRIs were analyzed by radiologists blinded to clinical and genetic data. Results: Subjects ranged in age from 20–81 years; nine were male and six female; and education ranged from 13–21 years. These factors were similar in affected and unaffected subjects. We found no evidence of mesial temporal sclerosis, malformations of cortical development, or abnormal gyral patterns in any of the 15 subjects. One affected individual (age 63) had bilateral T2 increases, and two unaffected subjects, neither of whom carried a mutation, had incidental findings thought to be unrelated to epilepsy (a bifrontal white matter abnormality, and marginally lower right temporal neocortical volume). Verbal IQ scores were significantly lower in subjects with epilepsy than in those without epilepsy (ave. 108 vs. 124, p = 0.012). Among subjects without epilepsy, however, verbal IQ was similar in subjects who carried a mutation (ave. = 124) and those who did not (ave. = 125), suggesting that the difference in verbal IQ was more likely to be due to epilepsy or its treatment than to the mutation per se. In contrast, average scores for the oral word association test were lower both in subjects with epilepsy (94) and clinically unaffected mutation carriers (95) than in noncarriers (120) (p = 0.048). Conclusions: The results of this study do not support the previously reported association of LGI1 mutations with a structural brain abnormality detectable on high‐resolution MRI. Although the number of subjects tested is small, findings from the oral word association test may reflect an impact of mutations on language function. (Supported by NIH R01 NS36319 and the NINDS Division of Intramural Research.) 1 Marcelo A. Kauffman, 2 Dolores Gonzalez Moron, 1 Damian Consalvo, and 1 Silvia Kochen ( 1 Centro de Epilepsia, Hospital Ramos Mejia, Buenos Aires, Argentina ; and 2 Neurology Residency, Hospital Ramos Mejia, Buenos Aires, Argentina ) Rationale: Previous studies have examined the role of a single nucleotide polymorphism (SNP) at the promoter region of Interleukin 1b gene (IL1b −511T) on Temporal Lobe Epilepsy (TLE) and Febrile Seizures (FS) susceptibility, but they have been inconclusive. The aims of the present study were to define the role of this polymorphism on TLE and FS susceptibility and to obtain more precise estimates of the risk associated with this genetic variant. Methods: Published studies up to Febuary 2006 of TLE, FS and IL1b −511T SNP were identified by searches of PubMed, Biosis, Web of Science and Embase databases. Meta‐analysis of TLE and FS case‐control data were performed to assess the association of IL1b −511T SNP with TLE, TLE with Hippocampal Sclerosis (TLEHS) and FS. Pooled Odds Ratios were estimated by means of a genetic‐model free approach. Quality of Studies included was assessed by a score. Publication Bias was investigated by Egger and Begg tests. Results: A total of 12 studies were included where 7 examined the association with TLE (with a total of 1798 subjects) and 5 examined the association with FS (with a total of 1299 subjects). The data analyzed by the genetic model‐free approach indicated that the best model of inheritance is recessive. This model for allele T produced significant results in the TLE sample (fixed effects odds ratio 1.30 (1.01–1.69)) and in the TLEHS sample (fixed effects odds ratio 1.69 (1.26–2.27)) (Figure below), whereas for FS the results were non‐significant (fixed effects odds ratio 1.33 (0.95–1.84)). There was not evidence of Publication bias (Egger and Begg tests non‐significant for TLE, TLEHS and FS samples). Quality was heterogeneous between studies. Conclusions: Overall, these findings support a role for IL1b −511T gene variant in TLE under a recessive mode of inheritance. Further explorations of the involvement of the IL1b gene in the susceptibility to FS, with a greater number of studies with larger sample sizes, are needed to fully establish the role of the IL1b gene in this condition. (figure 1) (Supported by: Dr. Kauffman has a PhD scholarship from CONICET.) 1 Ok Joon Kim and 2 Nam Keun Kim ( 1 Neurology, Pochon CHA University, Sungnam, Kyunggi‐do, Republic of Korea ; and 2 Clinical Research Institute, Pochon CHA University, Sungnam, Kyunggi‐do, Republic of Korea ) Rationale: Elevated plasma homocysteine level (hyperhomocysteinemia) is associated with various vascular diseases, abortion, and cancer. In epilepsy, prolonged antiepileptic drugs induce hyperhomocysteinemia, especially in the methylenetetrahydrofolate reductase (MTHFR) TT type, which induce cerebrovascular disease and congenital fetal abnormalities such as neural tube defects. Because folic acid decreases homocysteine level, folic acid is useful treatment in patients with hyperhomocysteinemia.We studied to assess factors that have affects on the decrease of homocysteine levels in epilepsy patients, especially according to single nucleotide polymorphisms. Methods: We prospectively included 75 epilepsy patients admitted at Pundang CHA Hospital between 2002 and 2005. All patients treated with antiepileptic drugs during one year and examined homocysteine levels at two times: initial and last time (one year later). When hyperhomocysteinema (homocysitne > 12μmol/L) was found at initial time, patients were supplemented with daily 400μg folic acid for one year. All patients were divided into two groups; increase group (IG) and decrease group (DG) according to the change of homocysteine levels between initial and last time. Also, all patients examined various polymorphisms, such as MTHFR C677T, MTHFR A1298C, methionine synthase A2756G, and thymidylate synthase gene polymorphisms. Results: Among 75 patients, IG was 31 patients and DG was 44 patients. Decrease of homocysteine levels during one year was observed in young age, initial high level of homocysteine, use of folic acid, presence of family history, and semiology of complex partial seizure. There were no significant differences of homocysteine levels according to polymorphisms. Conclusions: Our results suggest that folic acid can cheaply decrease homocysteine levels in epilepsy patients with hyperhomocysteinemia. We recommend administration of folic acid in old epilepsy patients with generalized and symptomatic seizures since homocyteine level did not increase in the young age patients with complex partial seizure and family history. 1 Patrick Kwan, 2 Virginia Wong, 3 Ping W. Ng, 4 Colin H.T. Lui, 5 Ngai C. Sin, 1 Andrew C.F. Hui, 1 Lawrence K.S. Wong, 1 Yu Evelyn, and 1 Larry Baum ( 1 Division of Neurology, Chinese University of Hong Kong, Hong Kong, Hong Kong ; 2 Department of Paediatrics, University of Hong Kong, Hong Kong, Hong Kong ; 3 Department of Medicine, United Christian Hospital, Hong Kong, Hong Kong ; 4 Department of Medicine, Queen Elizabeth Hospital, Hong Kong, Hong Kong ; and 5 Department of Paediatrics, Prince of Wales Hospital, Hong Kong, Hong Kong ) Rationale: There is conflicting data whether CC genotype of the C3435T single nucleotide polymorphism (SNP) of the ABCB1 gene is associated with multi‐drug resistance in Caucasian epilepsy patients. There is wide ethnic variation in the genotype frequencies of this SNP. We investigated whether the association might be observed in ethnic Chinese. Methods: ABCB1 3435 was genotyped in 518 Chinese patients with epilepsy, classified as drug resistant in 221 and drug responsive in 279, and 179 control subjects without epilepsy. Phenotypic definitions were similar to previous studies in Caucasian subjects. Drug resistance was defined as at least 1 seizure per month despite treatment with 2 or more antiepileptic drugs, and drug responsive as seizure free for at least 1 year on treatment. Genotype was carried out by polymerase chain reaction‐restriction fragment length polymorphism assays. Results: As compared with patients with drug responsive epilepsy, patients with drug resistant epilepsy were more likely to have the TT genotype at ABCB1 3435 than CT or CC genotypes (odds ratio 2.2, 95% confidence interval 1.2 – 4.1, p = 0.004). The frequency of CC genotype was similar between patients with drug resistant and drug responsive epilepsy. There was no difference in genotype frequencies between the control subjects and the two patient groups. Conclusions: In Chinese patients with epilepsy, ABCB1 3435TT was found to be associated with drug resistance. The results contradict those reported in Caucasian subjects, and highlight the complexity of the potential role of ABCB1 3435 SNP in the pathogenesis of drug resistant epilepsy. (Supported by: Partially supported by the HKU Small Project Funding.) 1 Walter J. Lukiw, 1 Victor L. Marcheselli, and 1 Nicolas G. Bazan ( 1 LSU Neuroscience Center, Louisiana State University Health Sciences Center, New Orleans, LA ) Rationale: Experimentally induced epileptogenesis elicits profound changes in early response gene (ERG) expression in the mammalian brain. In these studies we analyzed early (0–6 hr), medium (1–7 da) and long term (2–8 wk) changes in the DNA‐binding of transcription factors (TFs) AP1, AP2, Egr1, HIF‐1, STAT1, NF‐kB, NFIL‐6, SP1 and TFIID to target DNA sequences in adult rat hippocampus after a single kainic acid (KA) injection. These TFs are enriched in the immediate promoters of the cFOS and cyclooxygenase‐2 (COX‐2) ERGs. Methods: To trigger epileptogenesis, male albino Wistar rats were injected i.p. with 10 mg/kg KA using saline as a vehicle. At indicated time points (0, 1, 3, 6 hr; 1, 3, 7 da and 2, 4, 8 wk) rats were sacrificed and hippocampal and cortical cellular and nuclear proteins and total RNA were co‐isolated from the same samples. TF‐DNA binding was studied using gel shift and super‐shift assay; RNA message levels were determined using DNA array analysis, RT‐PCR and Northern blots. cFOS and COX‐2 protein levels were quantified using Western immunochemistry. Results: Of the TFs examined, AP1, HIF‐1, NF‐kB and STAT1 binding were found to display oscillatory DNA binding profiles over the short, medium and long term. AP1‐DNA binding kinetics strongly correlated with COX‐2 gene activation over the short term while NF‐kB‐DNA binding paralleled COX‐2 gene expression at longer time points. Sustained increases in TFIID‐DNA binding suggested prolonged de novo induction of transcription from TATA‐containing rat hippocampal genes. Conclusions: Taken together, these results suggest that induction of the pro‐inflammatory and pro‐apoptotic TFs AP1, HIF‐1, NF‐kB and STAT1, and their binding to specific target DNAs in ERG promoters, drives inflammatory and apoptotic gene expression in a phasic fashion. These data also support the notion of a global re‐programming of hippocampal gene expression patterns long after a single KA‐induced triggering of epileptogenic events. (Supported by NIH AG18031 and NIH NS23002.) 1 Carla Marini, 2 Paolo Aridon, 3 Chiara Di Resta, 4 Elisa Brilli, 2 Maurizio De Fusco, 3 Fausta Politi, 1 Elena Parrini, 2 Irene Manfredi, 5 Tiziana Pisano, 5 Dario Pruna, 3 Giulia Curia, 5 Carlo Cianchetti, 4 Massimo Pasqualetti, 3 Andrea Becchetti, 1,6 Renzo Guerrini, and 2 Giorgio Casari ( 1 Division of Child Neurology and Psychiatry, IRCCS Stella Maris Foundation, Pisa, Italy ; 2 Human Molecular Genetics Unit, Dibit San Raffaele Scientific Institute, Milan, Italy ; 3 Dipartimento di Biotecnologie E Bioscienze, Università di Milano‐Bicocca, Milan, Italy ; 4 Unità di Biologia Cellulare E Dello Sviluppo, Dipartimento di Biologia, Università di Pisa, Pisa, Italy ; 5 Child and Adolescent Neuropsychiatric Clinic, University of Cagliari, Cagliari, Italy ; and 6 University of Pisa, Pisa, Italy ) Rationale: Autosomal dominant frontal lobe epilepsy is characterized by clusters of sleep‐related hypermotor seizures, and has been associated with mutations of the alpha4 and beta2 subunits of the neuronal nicotinic acetylcholine receptor (nAChRs) genes (CHRNA4 and CHRNB2). An Italian family from Sardinia with sleep‐related seizures with prominent fear and wandering is presented. Methods: Clinical information, EEG studies and blood samples were obtained. A genome‐wide linkage analysis was performed, followed by electrophysiological studies of transfected HEK293 cells expressing either the newly identified nAChR alpha2 mutant subunit or the wild type receptor. Expression pattern of the genes coding for the alpha2 and 4, and beta2 nAChRs in areas of autoptic human brain was also performed. Results: Ten family members had clusters of seizures during sleep. Mean age at seizure onset was 10 years. The proband had since 6 years of age, clusters of sleep‐related brief episodes with awakening, vocalization, fearful expression and movements of the tongue. The remaining individuals had nocturnal clusters of similar episodes including complex motor behaviors such as walking outside the room. The genome‐wide linkage analysis showed a locus on chromosome 8p12.3–8q12.3, and a heterozygous missense mutation in the nAChR alpha2 subunit gene (CHRNA2) was detected in all 10 affected individuals. Whole‐cell recordings of transfected HEK293 cells showed that the CHRNA2 mutation markedly increased the receptor sensitivity to acetylcholine. The expression pattern of alpha2 mRNA in autoptic human brain exhibited the highest level in the thalamus, but a diffuse expression was detected in the investigated samples. beta2 and alpha4 mRNA were uniformly distributed throughout the samples. Conclusions: CHRNA2 is the third nAChRs gene to be associated with familial sleep‐related epilepsies. CHRNA2 mutation causes a complex ictal behavior with prominent fear and wandering. The alpha2 subunit showed the highest expression in the thalamus, hence confirming that mutations in nAChRs genes are involved in the epileptogenesis related to sleep‐arousal mechanisms. (Supported by the Italian Minister of Health; Telethon Foundation; Italian Ministry of Education, University and Research (MIUR).) 1 Claudia V. Maurer‐Morelli, 1 Rodrigo Secolin, 1 Romenia R. Domingues, 1 Rafael B. Marchesini, 1 Neide F. Santos, 2 Eliane Kobayashi, 2 Fernando Cendes, and 1 Iscia Lopes‐Cendes ( 1 Medical Genetics, UNICAMP, Campinas, São Paulo, Brazil ; and 2 Neurology, UNICAMP, Campinas, São Paulo, Brazil ) Rationale: The association between temporal lobe epilepsy and mesial temporal sclerosis (MTS) has been well established; as well as the use of hippocampal atrophy (HA) on magnetic resonance imaging (MRI) as an in vivo surrogate marker of MTS. One of the risk factors associated to MTS is childhood prolonged febrile seizures. We have described a type of mesial temporal lobe epilepsy with evident familial recurrence associate with HA but low frequency of febrile seizures. Previous pedigree and complex segregation analyses provided evidence for the presence of a major gene predisposing to HA in familial mesial temporal lobe epilepsy (FMTLE). The objective of this study was to identify the region harboring the main gene associated with HA in FMTLE by a genome‐wide linkage study. Methods: Genome‐wide scan was performed using a total of 332 microsatellite markers at ∼12cM intervals. An additional 14 markers were genotyped in the candidate region. Two‐point and multipoint LOD scores were calculated with the LINKAGE computer package. We assumed an autosomal dominant inheritance with incomplete penetrance. Results: We identified linkage to chromosome 18p11.3–11.2, with a maximum LOD score of 3.60 at θ= 0.0 for the D18S976 marker in a single family (F‐10) with 11 affected individuals with HA. Multipoint and haplotype analyses localized the locus within a 6 cM interval flanked by markers D18S976 and D18S452. No other significantly positive LOD score was detected for this family in the entire genome. Conclusions: This is the first conclusive evidence that HA may be caused by genetic factors, which can have major implications in the study of the pathophysiological mechanisms underlying MTS and its relationship with temporal lobe epilepsy. (Supported by FAPESP.) 1 Marcelo J. Murai, 1 Ricardo Horiuchi, 2 Daniel Martins, 1 Cláudia Maurer‐Morelli, 2 José Camillo Novello, 3 Fernando Cendes, and 1 Iscia Lopes‐Cendes ( 1 Medical Genetics Department, Unicamp, Campinas, São Paulo, Brazil ; 2 Biochemical Department, Unicamp, Campinas, São Paulo, Brazil ; and 3 Neurology Department, Unicamp, Campinas, São Paulo, Brazil ) Rationale: Mesial temporal lobe epilepsy (MTLE) is the most common and severe type of partial epilepsy, representing ∼50% of all adult epilepsy patients and frequently associated with pharmaco‐resistance. The relationship between MTLE and hippocampal sclerosis (HS) is well established. However, the precise pathogenesis of HS and its relationship with MTLE is not completely clarified. Two‐dimensional electrophoresis (2DE) is a powerful fractionation method for complex protein mixtures. In difference gel electrophoresis (DIGE) based proteomics, the experimental and control samples are labeled with different fluorophores and are run in the same gel, thereby minimizing gel preparation variation. DIGE is one of the few techniques that is capable to perform quantitative proteomics, generating statistical data to differences in protein abundances. Methods: We analyzed the proteome of three hippocampus removed from patients with refratory MTLE who underwent epilepsy surgery. Results: 2DE‐DIGE identified 2 up‐regulated and 10 down‐regulated proteins as determined by Student's T‐test (p ≤ 0.01). The observed molecular weight of the spots ranged from 28 to 93 kDa. Conclusions: The identity of these spots will be elucidated by mass spectrometry in order to gain additional information in a global scale about the mechanism of epileptogenesis in MTLE. (Supported by Fapesp, CAPES and CNPq.) 1,2 Rima Nabbout, 2 Stéphanie Baulac, 1 Nadia Bahi‐Buisson, 1 Catherine Chiron, 1 Olivier Dulac, and 2 Eric Leguern ( 1 Neuropediatrics Department, INSERM U663, Hopital Necker‐Enfants Malades, Paris, France ; and 2 INSERM U679, Hôpital de La Pitié Salpêtrière, Paris, France ) Rationale: We report the clinical and genetic study of a large French family with febrile seizures (FS) and childhood absence epilepsy (CAE). Methods: This family was identified through a national French campaign for familial epilepsy. It spans 4 generations and consists of 51 members with 13 affected. The medical history of all members was obtained by personal information and by consulting the medical files of each affected member. All family members gave a written consent to participate and 26 DNA were available for genetic study. Results: Clinical study: All affected members presented FS with CAE in 5 members and TLE in one. All FS stopped before the age of 6 and they recurred less than 4 times. Patients presenting CAE had recorded absences and characteristic EEG with 3 Hz spike waves. FS were simple except in one patient who had a long lasting complex FS (45 minutes) that occurred at 8 months of age. He presented later pharmaco‐resistant TLE. He had a left hippocampal sclerosis on brain MRI. All family members have a normal psychomotor development.Genetic study: The FS trait segregates as autosomal dominant trait. The genetic study allowed the exclusion of any linkage with reported loci for FS and FS plus (FS+), especially with the locus of GABRG2 gene on chromosome 5q reported in a family presenting FS+ and absences. A genome wide search with 380 markers allowed us to localize a new gene responsible for FS on 3p. We could not exclude another genomic segment with positive Lodscore (Z = 1.94 at θ= 0.00). All patients presenting further epilepsy (CAE and TLE) shared a common haplotype on this locus in addition to the FS haplotype on 3p. Conclusions: These findings emphasize the genetic heterogeneity of FS. Epilepsy in association with FS could result in this family from an interaction between at least 2 genes: the gene on 3p with a modifier gene. (Supported by ARGE, FFRE.) 1 Ruth Ottman, 1 Christie Barker‐Cummings, 2 Vincent M. Vasoli, 2 James P. Burke, 3 Janet L. Sobell, 1 W. Allen Hauser, and 2 Jeffrey R. Buchhalter ( 1 G.H. Sergievsky Center, Columbia University, New York, NY ; 2 Mayo Clinic, Rochester, MN ; and 3 Zilkha Neurogenetic Institute, University of Southern California, Los Angeles, CA ) Rationale: Valid screening methods to identify people with a history of seizures or epilepsy are crucial for epidemiologic studies. They are also essential to identify affected family members in genetic studies. Screening relies on careful history‐taking, but it is not clear what questions should be asked, or how accurately they identify affected individuals. Methods: We designed an 8‐question screening interview to identify people with a history of seizures. The first question asks, “Have you ever had, or has anyone ever told you you had, epilepsy or a seizure disorder?” If this question is answered “no,” it is followed by a second question that asks, “Have you ever had a seizure, convulsion, fit or spell under any circumstances?” This is followed by six additional questions that screen for symptoms possibly related to seizures (e.g., uncontrolled body movements, unexplained change in mental state, etc.). Subjects who screen positive are given a detailed diagnostic interview to confirm seizure occurrence and obtain additional clinical details. The interview was administered by telephone to 148 subjects with medical record documented incidence of unprovoked seizures from 1935–1994 while residing in Rochester, Minnesota, and 168 age‐ and sex‐matched controls. Interviewers were blinded to case‐control status. Results: Sensitivity was defined as the proportion who screened positive (“yes” or “possible” to any screening question) among those with either epilepsy (n = 124) or an isolated unprovoked seizure (n = 24); specificity was defined as the proportion who screened negative among the controls. Sensitivity was 86% for epilepsy and 75% for isolated unprovoked seizures. Specificity was 92%. Sensitivity was significantly higher for subjects with incidence in 1955 or later than for those with earlier incidence (epilepsy: 89% vs. 71%, isolated unprovoked seizure: 82% vs 0%(0/2)). Among subjects with epilepsy who screened positive, 87% answered “yes” or “possible” to the first question (asking about epilepsy or a seizure disorder), 9% to the second question (asking about seizures under any circumstances), and 4% to another question. Among subjects with isolated unprovoked seizures who screened positive, 56% answered “yes” or “possible” to the first question and 44% to the second question. Conclusions: This screening instrument is highly effective (89% sensitivity) for identifying people with onset of epilepsy up to 50 years before interview. Subjects are more likely to be missed if they had onset of epilepsy even earlier, or had an isolated unprovoked seizure rather than epilepsy. Further analyses should address the relation of sensitivity to clinical epilepsy features, so that sensitivity can be improved by targeting new questions to particular subgroups. (Supported by NIH R01 NS043472.) 1 Sumit Parikh, 2 Deepak K. Lachhwani, 2 Ajay Gupta, 2 Elaine Wyllie, and 2 Prakash Kotagal ( 1 Neurology, Section of Pediatric Neurology, Cleveland Clinic Foundation, Cleveland, OH ; and 2 Neurology, Section of Pediatric Epilepsy, Cleveland Clinic Foundation, Cleveland, OH ) Rationale: Up to 30% of children with mental retardation (MR) have epilepsy and up to 45% of children with epilepsy have MR (Epilepsia. 2000 Sep;41(9):1214–20). The report of the Amalfi group raised concern about missing treatable causes of epilepsy; as a result diagnostic metabolic testing in children with developmental delay and epilepsy is commonly performed . (J Child Neurol. 2002 Dec;17 Suppl 3:3S98–102) A standardized approach to such testing does not yet exist, and the prevalence of metabolic disease in this population is not known. Methods: Records of the 429 children admitted to the pediatric epilepsy monitoring unit during 2005 were reviewed. We noted the presence or absence of developmental delay, type of EEG discharges (focal‐mixed or generalized), and presence or absence of metabolic testing. Metabolic studies included: a) Plasma amino acids, biotinidase, lactate/pyruvate, acylcarnitines, ammonia levels, transferring isoelectric focusing, peroxisomal testing, and uric acid; b) CSF amino acids, glucose and lactate/pyruvate levels; and c) Urine organic acids, uric acid and creatine/guanidinoacetate levels. Results: 20% of our patients received some metabolic testing (85/429). Metabolic abnormalities were found in 48% of these patients (40/83), with 31% of these individuals having findings of mitochondrial dysfunction (27/85) with elevations in a) plasma alanine/proline/sarcosine (amino acids); b) plasma lactic acid; and c) urine lactic acid, 3‐methyl‐glutaconic acid and Kreb cycle intermediates (organic acids). Metabolic abnormalities were most common in those with focal discharges (70%; 21/30), when compared to multi‐focal (53%; 8/15) and generalized (40%; 13/32) discharges. Secondary carnitine deficiency (26%, 22/85) and valproate related isolated elevations in plasma glycine (63%, 5/8) were also commonly seen. A single patient with a disorder of creatine metabolism was identified. Only 11% of the patients had CSF analysis done (10/85) with a 30% yield of abnormalities (3/10), including elevations in alanine, glutamate or glutamine. Of 9 patients with seizures but normal development, 6 had metabolic abnormalities identified (66%). Only 17 individuals received their first MRI at our institiution, with 86% (12/14) having structural abnormalities found, and one of these individuals, with a cortical heterotopia, having mitochondrial disease diagnosed. Conclusions: Metabolic abnormalities may represent a significant finding in individuals with epilepsy, especially in those with developmental delays, even despite findings of a focal structural brain abnormality. A systematic approach to the diagnoses of these individuals is needed. A consult with a metabolic specialist is encouraged in these patients if a diagnosis is not known. Disorders of mitochondrial metabolism should be placed high on the differential list. 1 Patricia A.O. Ribeiro, 2 Lourenço Sbragia, 3 Rovilson Gilioli, 4 Fernando Cendes, and 1 Iscia Lopes‐Cendes ( 1 Medical Genetics, UNICAMP, Campinas, São Paulo, Brazil ; 2 Pediatric Surgery, UNICAMP, Campinas, São Paulo, Brazil ; 3 CEMIB, UNICAMP, Campinas, São Paulo, Brazil ; and 4 Neurology, UNICAMP, Campinas, São Paulo, Brazil ) Rationale: Mutations in LGI1 gene were described in patients with autosomal dominant lateral temporal lobe epilepsy and preliminary functional studies point to a possible involvement of LGI1 with migration and/or neuronal proliferation. However, the precise function LGI1 remains unknown. The objective of the present study was to determine the expression pattern of the Lgi1 gene in mice brain during development and in adult animals. Methods: Programmed mating was carried with Balb/c mice in order to obtain embryos of different ages. The brains of three animals at the following ages were removed: E15, E17, E18 days (E: embryo), P1, P7, P14, P28, P42 and P56 days (P: post‐natal). Gene expression assays were carried out using real time PCR with the TaqMan™ system. In addition we used an endogenous control (Gapdh gene) and all experiments were performed in duplicates. Results: Lgi1 gene expression was significantly low during the intra‐uterine ages increasing gradually until P56 (adult animal). Samples from P28, P42 and P56 presented a seven fold increase in expression as compared to E15 samples. Conclusions: The pattern of Lgi1 gene expression that we observed suggests a predominantly inhibitory function during development of the central nervous system (low expression during embryonic stages). In addition, we may speculate that in more advanced ages, when neurons are already differentiated, its inhibitory function could be also essential, which can be suggested by the high expression in adult mouse brain. (Supported by CNPQ and FAPESP.) 1 Graeme J. Sills, 1 Mohd Makmor‐Bakry, 1 Nikolas Hitiris, 1 Elaine Butler, and 1 Martin J. Brodie ( 1 Epilepsy Unit, Division of Cardiovascular & Medical Sciences, University of Glasgow, Glasgow, Scotland, United Kingdom ) Rationale: The dose of carbamazepine (CBZ) required to achieve optimal seizure control varies widely from patient to patient. Genetically‐determined variability in the activity and/or capacity of various drug metabolising enzymes (DMEs) is believed to influence the pharmacokinetics, and thereby dosage requirements, of a number of therapeutic agents. We have investigated polymorphic variants in the genes encoding DMEs involved in the metabolism of CBZ in an effort to identify predictors of CBZ maintenance dose. Methods: A total of 70 epilepsy patients (49% male; median age 34 years, range 14 – 72 years) who had benefited (≥50% reduction in seizure frequency) from treatment with CBZ monotherapy were included in the analysis. Individuals who did not tolerate the drug and/or those who did not experience significant efficacy were specifically excluded. Common genetic variants in CBZ‐related DMEs, including specific cytochrome P450s (CYP3A4, CYP3A5 and CYP1A2) and microsomal epoxide hydrolase (EPHX1), were screened by conventional polymerase chain reaction – restriction fragment length polymorphism (PCRRFLP) or direct sequencing. Associations were identified and characterised by multiple logistic and linear regression analyses, respectively. Results: All genotype frequencies were consistent with Hardy‐Weinberg equilibrium (p>0.05). No single clinical characteristic or genetic variant was sufficient to predict CBZ maintenance dose. However, a multiple logistic regression model incorporating patient age at time of optimal seizure control (OR = 1.03, 95%CI 1.00–1.07, p = 0.024), genotype of EPHX1 c.337T>C (OR = 0.44, 95%CI 0.22–0.87, p = 0.018) and genotype of EPHX1 c.416A>G (OR = 0.46, 95%CI 0.22–0.98, p = 0.044) demonstrated a significant association with maintenance dose (r2= 0.16; p = 0.009). Conclusions: This analysis suggests that a combination of age and genetic variants of the EPHX1 gene can be used to predict optimal maintenance doses of CBZ. Further investigations are required in an effort to strengthen the predictive value of this observation before its clinical utility can be realistically assessed. (Supported by: MMB is supported by a studentship from the Government of Malaysia and Universiti Kebangsaan Malaysia.) 1 Lisa J. Strug, 1 Martina Durner, 3 Effie Cayanis, 1 Fengli Zhang, 1 Dana Politis, 1 Irene Klotz, 1 Elisa Dicker, and 1,2,3 David A. Greenberg ( 1 Biostatistics, Division of Statistical Genetics, Columbia University, New York, NY ; 2 Psychiatry, Columbia University, New York, NY ; and 3 Genome Center, Columbia University, New York, NY ) Rationale: In linkage genome scans of Idiopathic Generalized Epilepsy (IGE) we found evidence for linkage on chromosome 5p with the juvenile absence phenotype. There are at least two candidate genes in this region: Succinate dehydrogenase complex subunit A (SDHA) and Programmed cell death 6 (PDCD6). Methods: We examined this region for association with Juvenile Absence Epilepsy (JAE) and Childhood Absence Epilepsy (CAE), genotyping single nucleotide polymorphisms in and around SDHA and PDCD6. We used case‐control and family‐based association methods to determine the degree of association. Results: We found that CAE is strongly associated with all SNPs tested in the PDCD6 gene (maximum chi‐squared = 21.22, p < 0.0001 at rs4957014). JAE, in contrast, was not significantly associated with PDCD6 or any SNPs in the region studied. These results were confirmed in the family‐based association analysis. The odds of having CAE if one carries at least one copy of the associated allele at the rs4957014 locus is 3.2 times the odds for an individual without a copy of the allele at this locus (95%CI: 1.91 – 5.37). Conclusions: Our results suggest that not only does PDCD6 play a role in CAE susceptibility, but that different genes contribute to susceptibility in CAE and JAE; no association was seen in the JAE group. This implies that CAE and JAE cannot necessarily be treated as having the same genetic basis. Although there may be susceptibility genes common to the two (or more) syndromes, treating them as one genetic entity could mask the existence of genes for only one of the syndromes. It is especially interesting that this is at least the second apoptosis gene that appears to be related to a form of IGE. We wish to acknowledge our many collaborators who were critical to this study. Space limitations make it impossible to name them here. (Supported by NIH RO1 NS37466 (to MD) and NIH RO1 NS027941 (to DAG).) 1,6 Miyabi Tanaka, 2 Marco T. Medina, 2 Reyna M. Duron, 3 Ramon Castro, 4 Iris J. Martinez, 1 Machado J. Salas, 4 Maria Elisa Alonso, 1,5 Julia Bailey, 6 Richard W. Olsen, and 1 Antonio V. Delgado‐Escueta ( 1 Comprehensive Epilepsy Center, UCLA & VA GLAHS, Los Angeles, CA ; 2 Neurology, National Autonomous University, Tegucigalpa, Honduras ; 3 University of Sonora, Hermosillo, Mexico ; 4 Neurology, National Instiutute of Neurology & Neurosurgery, Mexico City, Mexico ; 5 Neuropsychiatrics Institute, UCLA, Los Angeles, CA ; and 6 Pharmacology, UCLA, Los Angeles, CA ) Rationale: Because GABRB3 deficient mice shows absence like features including EEG characteristics and pharmacological response because atypical absences in Angelman syndrome correlates with deletions which includes GABRB3, and because Transmission Disequilibrium Test showed possible genetic association between GABRB3 and CAE, we screened for mutations in GABRB3 in families with CAE. We detected two heterozygous missense mutations (P11S, S15F) in the alternative signal peptide, exon1a of GABRB3 segregating with CAE affected members of three Hispanic families. The same missense mutations were not present in 34 other probands with CAE and 440 Hispanic healthy controls Here, we report on the functional consequences of these missense mutations. Methods: Expression constructs: We used mammalian expression vectors, pCMV‐SPORT 6 into which was cloned exon1–9 of GABRB3 and pFP‐N1 (GFP fusion vector). We amplified the mutation containing Exon1a with designed primers with restriction enzyme sites, and replaced Exon1 in pCMV‐SPORT6 with Exon1a with/without mutations. The construct which has the region from exon1a to exon9 without stop codon was amplified from the replaced vector. After digestion, each construct was cloned into pFP‐N1. Immunoblot analysis. HeLa cells were transfected with expression constructs and examined 48 hours post‐transfection and homogenized. Protein samples were run on a 10% Tris‐HCl gel and transferred onto a nitrocellulose filter. After blocking, the filter was processed through sequential incubations with the primary antibody for 1 h, and then with horseradish secondary antibody for 1h. Immunoreactive proteins on the filter were visualized using Typhoon software 9410. Results: Immunoblot studies revealed that both P11S and S15F mutations in exon1a decreased the expression of the protein level when compared with controls. The average densitometric image quantitation was 33%(±0.014) of control for P11S and 69%(±0.14) of control for S15F, normalized to actin density(n = 3). Conclusions: Our results indicate that both P11S and S15F mutations in exon1a decrease expression of protein levels and suggest less trafficking of the nascent peptide from the endoplasmic reticulum to cell membrane. GABRB3 is a main component of GABAR in reticular nucleus of thalamus and expressed transiently in other thalamic nuclei in the perinatal period, decreasing in adult. We speculate that the decreased expression of GABRB3 in developmental brain leads to absence seizures. 1 Simone S. Tsuneda, 1 Fabio R. Torres, 2 Maria A. Montenegro, 2 Marilisa Guerreiro, 2 Fernando Cendes, and 1 Iscia T. Lopes‐Cendes ( 1 Medical Genetics, UNICAMP, Campinas, SP, Brazil ; and 2 Neurology, UNICAMP, Campinas, SP, Brazil ) Rationale: Recent studies have demonstrated that mutations in FLN1 gene are responsible for bilateral periventricular nodular heterotopia (BPNH). We identified a novel mutation (1159G>C), in a family segregating BPNH. However, the exact molecular mechanism by which this mutation lead to abnormal FLN1 protein was unclear. The purpose of this study was to investigate the molecular mechanism of this new mutation. Methods: Total RNA was obtained from peripheral blood samples, from 2 BPNH patients (mother and daughter) and control individuals. RT‐PCR was performed by standard techniques and the cDNA was amplified using specific primers spanning the region containing exon 6 and exon7. Amplicons were cloned into pGEM‐T vector, gel‐purified and subsequently sequenced, using SP6 and T7 primers. Results: Analysis of the cDNA amplicon demonstrated a different pattern of electrophoretic migration between patients and controls. The sequencing of these fragments showed that amplicons from individuals with BPNH kept the intronic sequence between exons 6 and 7. Conclusions: Our data clearly showed that the molecular mechanism of the mutation 1159G>C is the abolishment of the exon 6 donor splicing, resulting in an alternative stop codon and, possibly, in a truncated protein. Thus an aminoacid substitution, as suggested previously, is not the mechanism involved in the etiology of BPNH sydrome in the patients analysed. (Supported by FAPESP.) 1 V. Udani, 2 R. Shukla, 1 P. Munot, 1 L. Iyer, 2 S. Prakash, 1 S. Pujar, and 2 G. Chandak ( 1 Child Neurology & Epilepsy, PD Hinduja National Hospital, Mumbai, India ; and 2 Genome Research Group, Centre for Cellular and Molecular Biology (CCMB), Hyderabad, India ) Rationale: Cohen syndrome (MIM 216550) is an autosomal recessive disorder with features of psychomotor retardation, microcephaly, characteristic facial features, hypotonia, joint laxity, progressive retinochoroidal dystrophy, myopia and neutropenia. Characteristic facial features include high‐arched or wave‐shaped eyelids, a short philtrum, thick hair, and prominent incisors. Clinically homogenous in the Finnish population, non‐Finnish patients show considerable variability. Seizures have never been described in Cohen's syndrome. The genetic locus for Cohen syndrome was recently mapped to chromosome 8q and truncating mutations in the COH1 gene were identified. In Finnish patients a founder mutation, C1117fsI1124X has been consistently found while other mutations have been associated with the atypical non‐Finnish Cohen's syndrome. One of us (VPU) observed several clinical features of Cohen's syndrome in Indian patients with cryptogenic epilepsy. A pilot study confirmed eight Cohen‐like features in a group of children with cryptogenic epilepsy when compared with controls. This prompted us to study the COH1 gene in such patients. Methods: We systematically studied patients with cryptogenic epilepsy and a Cohen‐like phenotype for all features of Cohen's syndrome and included only those with at least 5/8 clinical features that we had found to be significant in our previous study. We initially screened for all the reported common mutations and later sequenced all 62 exons of the COH1 gene in this cohort of 24 patients. We also collected DNA from age, sex and ethnically matched controls. Results: We found several novel mutations such as Y413X, A829T and G3407R in the COH1 gene in 8 of 24 patients (33%). Interestingly, we did not find any of the earlier reported mutations including the most common mutation, C1117fsI1142X. The commonest mutation Y413X was identified in 4 patients, of which one was a compound heterozygote with the G3407R mutation. These mutations were not observed in any of the 100 controls. The mutations were mostly observed in the heterozygous condition except in one patient who was homozygous for the G3407R mutation. We could not find any clinical differences in the patients with and without mutations. However the patients with mutated COH1 had a significantly higher number of Cohen's syndome features than those without the mutations. Conclusions: We conclude that novel COH1 mutations are strongly associated with Indian patients of cryptogenic epilepsy with a Cohen – like phenotype. There is however a possibility of additional mutations in the non‐coding regions of the COH1 gene or in other genes.Functional analysis will throw more light on the mechanisms by which these mutations may be causally related to the epilepsy which is a characteristic feature in these patients. (Supported by Research Society, PD Hinduja National Hospital.) 1,3 Kette D. Valente, 2 Monica Varela, 2 Cintia Fridman, 2 Celia Koiffmann, and 4 Maria J. Marques‐Dias ( 1 Neurophysiology Laboratory‐Psychiatry, USP, Sao Paulo, SP, Brazil ; 2 Biology, USP, Sao Paulo, SP, Brazil ; 3 LIM 21 – Psychiatry, USP, Sao Paulo, SP, Brazil ; and 4 Pediatrics, USP, Sao Paulo, SP, Brazil ) Rationale: Angelman syndrome (AS) results from deletion (DEL), uniparental disomy,imprinting anomalies or UBE3A mutations. Patients determined by DEL have a severer phenotype. In 95% of AS patients caused by DEL,two main classes are found: Class I with breakpoints at BP1 (proximal) and BP3 (distal), and Class II with breakpoints at BP2 (proximal) and BP3 (distal).The remaining 5% have the distal breakpoint at BP4 and BP5.We formerly described that BP1‐BP3 patients are more severely affected,on clinical grounds,than BP1‐BP2 patients. The study of electroclinical data in these distinct groups of DEL patients has not been done. Therefore, we aimed to evaluate the importance of breakpoint on the electroclinical phenotype in patients with DEL. Methods: The authors evaluated 18 DEL patients (6 were Class I; 10 were Class II and 2 were Class III and IV). Diagnosis was made by methylation pattern analysis of exon 1 of the SNRPN‐SNURF gene and by microsatellite profiling of loci within and outside the 15q11‐q13 region. We characterized epilepsy by history obtained with a questionnaire including:occurrence of epilepsy;age of onset;seizure type;epilepsy aggravated by fever;severity of epilepsy determined by:daily seizures; disabling/injurious seizures;more than 3 seizure types;status epilepticus (SE) and;history of refractory epilepsy. These data were corroborated by medical records, personal contact with previous physicians, and video‐EEG monitoring. Patients underwent a mean of 2.6 EEGs. Suggestive EEG patterns for AS were classified according to Boyd et al.and were studied as to morphology, duration, occurrence, frequency, amplitude and distribution. Results: Class I had more daily (p 0.0357) and disabling seizures (p 0.0350). They also presented a higher frequency of SE(p 0.0357). Both groups had similar rates of epilepsy aggravated by fever (p 0.5879); however, in Class I these events evolved to SE(p 0.0357). Polytherapy was more frequent in Class I (p 0.0350), associated with the refractoriness observed in this group.No statistical difference was observed as to age of onset, age of seizure control, seizure type and diversity of seizures presented by these patients. Background abnormalities, interictal and ictal epileptiform discharges were similar in both groups. Suggestive EEG patterns of AS were high in both groups. Occurrence of prolonged and frequent runs of these patterns did not show statistical differences. Conclusions: Epilepsy was more severe and refractory to treatment in patients with large deletions. Therefore, deletion is not a homogeneous group and breakpoint is predictive for epilepsy severity in AS, representing an important factor in parents' counseling. On the other hand, EEG patterns of AS seem to be reliable for diagnosis, despite the breakpoint. (Supported by FAPESP.) 1 Patrick Van Bogaert, 2 Regis Azizieh, 1 Alec Aeby, 3 Linda De Meirleir, 4 Florence Christiaens, 2 Julie Desir, and 2 Marc J. Abramowicz ( 1 Pediatric Neurology, ULB‐Erasme Hospital, Brussels, Belgium ; 2 Medical Genetics, ULB‐Erasme Hospital, Brussels, Belgium ; 3 Pediatric Neurology, AZ‐VUB, Brussels, Belgium ; and 4 Pediatric Neurology, UCL‐Saint Luc, Brussels, Belgium ) Rationale: Rare, mendelian forms of epilepsy amenable to positional cloning help understand the molecular mechanisms of epilepsy. Methods: Three female patients from two consanguinity loops in a large inbred Morroccan family were investigated for progressive myoclonic epilepsy (PME). After exclusion of known causes, linkage study and analysis of candidate genes were performed. Results: Epilepsy started between 18 and 24 months of age after normal psychomotor development. Seizure types were multifocal myoclonus affecting limbs and face and aggravated by movements, and febrile generalized tonic‐clonic in one patient. EEG showed slow dysrythmia, multifocal epileptiform discharges without reproducible temporal relationship with myoclonus and occasional generalized epileptiform discharges. Photosensibility was present in one patient. Giant somatosensory evoked potentials were not found. Cerebral MR imaging and fundoscopy were normal. In 2 patients, seizures were refractory to anti‐epileptic drugs. Evolution was unfavourable in both of them: one patient died from septic shock in the course of a corticosteroids trial, and the other one, now aged 9 years, showed progressive dysarthria, dementia and ataxia with lost independent walking at age 7. Epileptic seizures were controlled under lamotrigine and clonazepam in the third patient, now aged 16 years, who presented profound mental retardation and dysarthria without ataxia. GeneChip analysis of 11K SNPs showed a 15 Mb homozygous haplotype common to the 3 patients. Further analysis using microsatellite markers confirmed linkage of the disease to a 5 cM pericentromeric region of chromosome 7, with a maximum multipoint LOD of 4.2 at D7S663. In silico inspection showed the gene for a Potassium Channel Tetramerization Domain homolog, KCTD7, close to D7S663 within the segment. KCTD7 contains a BTB/POZ‐type domain and has strong homology with the T1 domain of the voltage‐gated potassium channels. Direct sequencing of KCTD7 revealed a C to T transition creating a stop codon in exon 2, homozygous in patients and heterozygous in parents. Conclusions: PME in our patients was associated with a mutation of the KCTD7 gene encoding a putative potassium channel tetramerization domain, consistent with a defect of membrane repolarization.

Journal

EpilepsiaWiley

Published: Oct 1, 2006

There are no references for this article.