A Randomised Open-label Comparison of the Ef- Ficacy, Tolerability, and Hormonal Effects of Sodium Valproate and Lamotrigine Monotherapy in Newly-diagnosed Epilepsy

presents the findings of initial proof of the principle stage of this study. Method: Epilepsy patients on monotherapy with carbamazepine (CBZ) and other AEDs were recruited. A package of eye movement examinations was performed, related to serum drug levels measured before and after taking the morning dose. The responsiveness of eye movement to changes in serum level (trough and peak) was correlated to clinical responsiveness. Results: Saccadic velocities were presented in a series of patients showing the marked variation within individuals to changing drug doses. 30% differences were noted in some patients between different serum levels, with conversely little change in others. For carbamazepine, epoxide levels also influence velocity and are included in carbamazepine measures. Latency of saccades is less sensitive to the changes of serum concentration. The relation of saccadic changes to clinical effect is shown in terms of effectiveness. Conclusion: The computerised measurement of eye movements is robust and easily applied to routine clinical settings. In this study, the reliability and variance of changes in eye movement measures in relation to drug responsiveness was explored. As the molecular and neuronal mechanisms which underpin the effect of drugs on eye movements is similar to those involved in the antiepileptic effect, it seems possible that this simple clinical test will be a useful predictor of the therapeutic effect. p416 SERIAL EVOLUTION OF PLEDS 1J. Kim, 2S. Lee, and 1Y. Kim (1Department of Neurology, Chungnam National University Hospital, South Korea, 2Department of Neurology, College of Medicine, Keimyung University, South Korea) Purpose: Periodic lateralised epileptiform discharges (PLEDs) are defined as spikes or sharp waves occurring at approximately regular intervals. According to Reiher’sclassification, PLEDs can be subdivided into benign PLEDs proper (including class 1, 2, and 3) and malignant PEDs plus (including class 4 and 5). This study was performed to investigate 1) the serial pattern of PLEDs. 2) the evolution of malignant PLEDs plus to benign PLEDs proper, and 3) to suggest a more benign form in the same PLEDs class. Method: Thirty-seven patients who had two or more available EEGs were included in this study. All patients had structural brain lesions identifiedin the brain CT/MRIs. Patients with hypoxic brain damage or uncertain diagnosis were excluded. 237 EEGs from 37 patients were retrospectively analysed and the pattern of PLEDs classifiedaccording to Reiher’s classification. Results: One hundred and two PLEDs were identified out of 237 EEGs. (PLEDs proper: 100 EEGs, PLEDs plus: 12 EEGs). Seventy one EEGs belonged to class 3. We subdivided them into 1) simple group, 2) benign group, 3) vigorous group, and 4) suppressed group. Most PLEDs started with a vigorous or suppressed pattern of class 3 evolving into class 3 of simple or benign pattern, and then changing to class 1 or class 2. During the course, medium to high amplitude, complex sharp waves became lower amplitude and broader basis, simple sharp waves (class1/2). And finally they buried into theneighboring background waves. PLEDs started with class 3 of benign or simple pattern rapidly changing into class 1 or 2 and patients whose EEG started with class 1 or 2 showed the most rapid improvement in the EEGs and better prognosis. Conclusion: PLEDs have five distinguished classes. As time goes by, with proper treatment, they evolved from a malignant class to a benign class and finally disappeared. Class 3 had more diverse patterns and we suggest a vigorous and suppression pattern in more malignant forms of PLEDs in this class. p417 DIFFERENCES BETWEEN INTRACEREBRAL ACTIVITIES OF PATIENTS WITH IDIOPATHIC GENERALISED EPILEPSY BEFORE AND AFTER CHRONIC VALPROATE THERAPY 1M. Toth, 1I. Kondakor, and 2B. Clemens (1Department of Neurology, University of Pécs, Pécs, Hungary, 2Epilepsy Center, Department of Neurology, Kenézy Gyula Memorial Hospital, Debrecen, Hungary) Purpose: EEG background activity was localised by LORETA to test the possible differences between brain electrical activities of untreated and valproate-treated patients with idiopathic generalised epilepsy (IGE). Method: 19-channel EEG data of 15 IGE patients before and after chronic valproate therapy were studied. For source localisation of brain electrical activity LORETA, a new tomographic method was applied, where localisation is based on images of standardised current density. The method is able to localise electrical generators in 3 dimensions in four frequency domains: delta (1.5–3.5 Hz), theta (3.5–7.5 Hz), alpha (7.5–12.5 Hz) and beta (12.5–25 Hz). For the analysis 42 (S.D.: 1.7) artifact-free, 2 sec epochs were selected from the filtered raw EEG (1.5–25.0 Hz) and were processed by LORETA. Statistical differences between pairs of conditions were computed as images of t-values. To assess the possible differences in background EEG activity, voxel-by-voxel dependent t-tests were computed. The t-statistic images were examined to localise regions showing statistically significant effects thresholded at 1% probability level. Results: All patients became seizure-free. Valproate caused a statistically significant (p < 0.01) decrease of delta and theta band EEG activity in the frontal cortex, and in part of the temporal cortex. Valproate did not alter the activity of the parietal and occipital cortex, and in the remaining areas of the temporal cortex. Conclusion: The selective effect of valproate is related to areas of the cortex that might be responsible for the initiation of the seizures in IGE. The relationship between clinical and EEG effect of valproate might be suggested. p418 ELECTRICAL SOURCE IMAGING OF SPATIAL RELATIONSHIPS BETWEEN LESIONAL AND IRRITATIVE ZONES IN MEDICALLY INTRACTABLE POSTERIOR PARTIAL EPILEPSIES 1L. Maillard, 1,2L. Koessler, 1J. Vignal, 3A. Thiriaux, 2M. Braun, and 1H. Vespignani (1Service de Neurologie, Centre Hospitalier Universitaire de Nancy, France, 2IADI, Eri 13, Inserm, Nancy, France, 3Service De Neurologie, Centre Hospitalier Universitaire De Reims, France) Purpose: Medically intractable posterior partial epilepsies (MIPPE: occipital, parietal, TPOJ) remain a challenge in term localising the irritative zone (which generates interictal spikes) and the epileptogenic zone (which generates seizures). This may directly explain the low rate of success of surgery compared with temporal or frontal lobe epilepsies. Recent development of high resolution EEG (HR-EEG) which allows electrical source imaging of generators of spikes could help to localise the irritative zone and to study its relationships with MRI lesion. Method: Prospective, multicentric (Nancy, Reims) study of patients with MIPPE. Each patient had a comprehensive presurgical evaluation comprising video-EEG recordings of seizures, HR-EEG (64 channels and dipole modelling with ASA, ANT) and cranial MRI. Results: Since the beginning of the study, 3 patients have been included. In patient EN, the generator of interictal spikes (IIS) is centered on the left lateral occipitotemporal sulcus, overlapping the lesional zone (regional subcortical atrophy). In the patient PN, IIS are localised in Epilepsia, Vol. 47, S3, 2006 110 7TH EUROPEAN CONGRESS ON EPILEPTOLOGY the right inferior parietal gyrus outside but next to the lesional zone (periventricular nodular heterotopia) suggesting a pathological connection between these two distant neuronal assemblies. In the last patient, IIS have a right temporal pole generator related to the ispilateral hippocampal atrophy but distant from the occipital dysplasia. Conclusion: In MIPPE, relationships between the lesions and the irritative networks are complex. HR-EEG is a promising tool to study these relationships and take a place in presurgical investigations. p419 FRACTAL DYNAMICS OF THE THETA RHYTHM IN EPILEPSY E. Korsakova, V. Slezin, and S. Khorshev (V.M.Bekhterev Psychoneurological Research Institute, Saint-Petersburg, Russia) Purpose: The purpose of the present study consisted of in investigation of fractal dynamics of the theta rhythm at initial and later stages of epileptogenesis during the absence of paroxysmal discharge (PD). Method: The examination was carried out with 30 healthy subjects, 32 patients with a preclinical stage of epilepsy who had a single unprovoked seizure and no PD, 29 epilepsy patients with a clinical stage of epilepsy (CSE), but without PD on the routine EEG, 34 patients with CSE and PD. The power spectrum of the theta rhythm envelope were calculated for monopolar EEG on the 3 min recoding devoid of PD. Power spectrum form had the appearance of 1/fwhere f = frequency, β = fractal index. Fractal indexes were calculated within the range of low-frequency (β1) and medium-frequency (β2) fluctuations of the theta rhythm amplitude. Results: In case of discovery of paroxysms on the EEG in epilepsy, a sharp change of the theta activity fractal characteristics more marked with β2 takes place. The appearance of PD in epilepsy leads to that in the range of the theta rhythm. The values of β1 reach the greatest magnitudes in F3, F4, C3, F7, F8 and T3 as compared with all other groups of examinees (p < 0.05). The same changes occur with the values of β2, but already over all the cortical areas (p < 0.01). Conclusion: Thus, increase in stability, and regularity and decrease in complexity of the theta rhythm represent an indicator of a high probability of paroxysms appearance, a precursor of the development of PD on the EEG. p420 ACHIEVING SLEEP STATE DURING EEG IN CHILDREN: SEQUENCE OF ACTIVATION PROCEDURES J. Kaleyias, S. Kothare, M. Pelkey, H. Geofrey, A. Legido, and D. Khurana (Division of Child Neurology, Department of Pediatrics, St Christopher’s Hospital For Children, Philadelphia, USA) Purpose: Hyperventilation (HV) during an EEG is often followed by drowsiness. The effect of intermittent photic stimulations (IPS) on awake/sleep state is unclear. The current study was conducted to determine if the sequence of activation procedures have an impact on acquisition of sleep during a routine, non–sleep deprived EEG. Method: Children, scheduled between 9 to noon, who were not sleep deprived and able to hyperventilate, were included in a 3-month prospective study undertaken at the Neurophysiology Laboratory of a tertiary children’s hospital. In the first 48 children evaluated, IPS was started 5 minutes into the EEG recording and HV was elicited at the end of the session (group I). For the next 48 children, HV was started 5 minutes into the EEG recording and IPS were presented at the end of the EEG (group II). Both groups were assessed for the presence or absence of sleep during their EEG recording. Results: Only 3 of the 48 (6.2%) children in group I versus 17/48 (35.5%) in group II attained sleep (p < .0001, OR = 8.68, 95% CI 2.3432.22). Therefore, the children who hyperventilated at the beginning of the session had an 8-fold increase chance to attain sleep in comparison with children who received IPS at the beginning of the recording followed by HV at the end. Conclusion: We recommend that when performing routine non-sleep deprived EEGs in children, HV be performed first with IPS at the end of the EEG in order to maximize the yield of attaining sleep recording during the study. p421 HYPNOGENIC PAROXYSMAL DYSTONIA: DOES IT EXIST? M. Cenusa, I. Mothersill, A. Bauerfeind, and G. Krämer (Swiss Epilepsy Centre, Zürich, Switzerland) Purpose: Nocturnal or hypnogenic paroxysmal dystonia (HPD) are rarely reported complex motor attacks occurring abruptly during sleep, especially NREM sleep. The existence of HPD as a distinct nosological entity has been the cause of much discussion. Many authors consider these nocturnal motor attacks to be frontal lobe epileptic seizures (FLS). In many of the patients presented to support the hypothesis that HPDs are in fact FLS the dystonia is mostly only a minor part of what is obviously an epileptic seizure and should not be a point of discussion. The majority also have in addition to their nocturnal “dystonic” attacks, seizures of obvious epileptic origin or epileptiform activity in the interictal EEG. However, patients do exist who experience pure HPD that show no other semiology which would relate them to FLS. In the EEG during such episodes the typical delta arousals pattern that is otherwise associated with NREM parasomnias is seen. Method: A retrospective review was made of all patients who were referred for unclear nocturnal events. All night polysomnographic-video EEG recordings were performed utilising all 10–20 electrodes. Results: We found a total of 12 patients who had pure nocturnal dystonic attacks, during which the EEG showed delta arousals. On complete awakening the semiology abruptly ceased. There was no additional semiology that would support a diagnosis of epilepsy. Conclusion: Our findings support the fact that HPD as a distinct nosological entity does exist. We suggest that it be considered, in analogy to confusional arousals, an incomplete motor awaking with associated delta arousal in the EEG. p422 POSTLESION FOCAL NEOCORTICAL EPILEPSY IN HUMANS: THE ROLE OF THE BLOOD–BRAIN BARRIER 1O. Tomkins, 1,4A. Korn, 1,2A. Friedman, 2A. Cohen, and 3I. Shelef (1Department of Physiology, Ben Gurion University of the Negev, BeerSheva, Israel, 2Department of Neurosurgery, Soroka University Medical Center, Beer-Sheva, Israel, 3Department of Neuroradiology, Soroka University Medical Center, Beer-Sheva, Israel, 4Department of Pediatric Neurosurgery, Tel Aviv-Sourasky Medical Center, Tel Aviv, Israel) Purpose: Recent studies in experimental animals suggest that lesions of the blood–brain barrier (BBB) play a key role in the pathogenesis of focal neocortical epilepsy (Seiffert et al., J Neurosci 2004;24:7829–36). The purpose of the present study was to evaluate possible correlations between areas of enhanced BBB permeability and areas with abnormal neocortical electrical activity in human patients suffering from lesionrelated seizures. Method: Included in this study were patients with mild-moderate blunt head trauma. Patients underwent standard magnetic resonance brain scan sequences, as well as dynamic studies aimed at evaluating the integrity of the BBB. In addition, a postgadolinium T1 sequence was used to evaluate statistically significant changes in cortical signal, suggesting diffusion of the contrast material across an altered BBB. EEG was recorded from subjects using a clinical EEG unit (CEEGRAPH IV, Bio-logic Systems Corp., Mundelein, Illinois). Areas of aberrant power spectra slowing and interictal like events were localised using low resolution electrotomography (LORETA) and statistical parametric mapping. Results: Increased BBB permeability following focal brain trauma can be measured and localised in the neocortex. In some cases BBB permeability may last up to several months following the traumatic event. In ca. 30% of the patients (n = 21) examined under this study, spatial correlation was found between regions with enhanced BBB permeability and regions with localised focal cortical slowing or inter-ictal activity. Conclusion: Our data suggest that a prolonged increase in BBB permeability following head trauma may be associated with abnormal, hypersynchronised neocortical activity, perhaps predisposing to epileptic seizures. Epilepsia, Vol. 47, S3, 2006 7TH EUROPEAN CONGRESS ON EPILEPTOLOGY 111 p423 INTEROBSERVER RELIABILITY OF VIDEOPOLYSOMNOGRAPHIC DIAGNOSIS OF NOCTURNAL FRONTAL LOBE SEIZURES F. Bisulli, L. Vignatelli, F. Provini, F. Pittau, I. Naldi, A. Zaniboni, P. Montagna, and P. Tinuper (Department of Neurological Sciences, Bolgna, Italy) Purpose: The diagnosis of nocturnal frontal lobe epilepsy (NFLE) is mainly based on video-polysomnographic examination. Paroxysmal sleep-related disturbances distinguishing NFLE are characterised by one or all of the following movement-related subtypes: 1) paroxysmal arousals (PA–brief and sudden recurrent motor paroxysmal behaviour), 2) supplementary motor area seizures (SMAs–motor attacks with dystonic features), 3) hypermotor seizures (HMs–motor attacks with complex dyskinetic features) 4) episodic nocturnal wanderings (ENW–stereotyped, agitated somnambulism). The aim of this study was to estimate interobserver reliability of video-polysomnographic diagnosis in patients with suspected NFLE among sleep medicine experts, experts in epileptic disorders and trainees in sleep medicine. Method: Sixty-six patients referred to our sleep centre for suspected NFLE were included. All underwent nocturnal video-polysomnographic recording. Six doctors (3 residents and 3 trainees) independently classified each case as “NFLE ascertained” (according to the above specified subtypes: PA, SMAs, HMs, ENW) or “NFLE excluded.” Interobserver reliability was calculated by means of Kappa statistics, and interpreted according to standard classification (0.0–0.20 = slight agreement; 0.21– 0.40 = fair; 0.41–0.60 = moderate; 0.61–0.80 = substantial; 0.81–1.00 = almost perfect). Results: The observed raw agreement on the diagnosis of NFLE ranged from 63% to 79% between each pair of raters; the interobserver reliability ranged from “moderate” (kappa = 0.50) to “substantial” (kappa = 0.72). The major source of variance was the disagreement in interpreting arousal movements as PA or not. There were no differences in the level of agreement between residents and trainees. Conclusion: Among sleep medicine/epileptologist residents and trainees, interobserver reliability of diagnosis of NFLE on the basis of videotaped observation of sleep motor phenomena is not satisfactory. One of the sources of variance is probably the interpretation of motor phenomena resembling simple arousals or PA. Explicit videopolysomnographic criteria for the classification of paroxysmal sleep motor phenomena are needed. p424 CLINICAL, MRI AND DIPOLE LOCALISATION ANALYSIS IN DIFFERENT SUBTYPES OF MALFORMATIONS OF CORTICAL DEVELOPMENT (MCD) 1,2S. Kochen, 1,3C. Papayannis, 1,3H. Laforcada, 1D. Consalvo, 1W. Silva, 4J. San Martin, 4N. Von Elleniueder, and 4,2C. Muravchik (1Epilepsy Center, Hosp “R.Mejia,” School Medicine, Univ Buenos Aires, Argentina, 2CONICET, Buenos Aires, Argentina, 3Health Secretary Buenos Aires City, Argentina, 4School Engineering, University La Plata, Argentina) Purpose: MCD represent the second cause of refractory epilepsy in adults. Clinical, epidemiological and localisation of the epileptogenic zone, through dipole localisation of the different subtypes, were compared. Method: For dipole localisation, our work group developed software that allowed optimisation of the application of Polhemus and BESA. The EEG (10-10 system) were obtained for 3 hours with 64 channels (Stellate). The patients were divided in 3 groups: G1: abnormal proliferation (cortical focal dysplasia and neoplasias), G2: abnormal migration (heterotopias), and G3: abnormal organization (polymicrogyria). We analysed different variables, particularly the occurrence of gradient of divergence (GD) clinical-MRI-dipole localisation. Results: We included 80 patients. The significance data, in G1 (n = 40): women 42.5%, neurological deficit (ND) 7.5%, GD 12.5%. G2 (n=24): women 75%, ND 8.3%, GD 16.7%. G3 (n=16): women 43.7%, ND 75%, GD 31.2%. Conclusion: We found between the subtypes of MCD a significant difference in relation to sex, women’s majority in G2, neurological deficit in G3. And we observed a gradient of divergence minor in G1 and most important in G3, corresponding with the worst postsurgical prognostic in the last group. The clinical information, MRI and the source localisation methods proved useful in distinguishing the differences and the prognosis in the patients with MCD. p425 DYNAMIC CHANGES OF ICTAL HIGH FREQUENCY OSCILLATIONS IN NEOCORTICAL EPILEPSY: USING MULTIPLE BAND FREQUENCY ANALYSIS 1A. Ochi, 1H. Otsubo, 1K. Imai, 1E. Donner, 2J. Rutka, and 1C. Snead (1Division of Neurology, The Hospital for Sick Children, Toronto, Canada, 2Division of Neurosurgery, The Hospital for Sick Children,